**FUP**

Old Church Slavic

Anna Polivanova

51

# Anna Polivanova **Old Church Slavic**

Grammar and Dictionaries

*Translated by* Lev Blumenfeld

*Edited by* Artemij Keidan

#### BIBLIOTECA DI STUDI SLAVISTICI

ISSN 2612-7687 (PRINT) - ISSN 2612-7679 (ONLINE)


### BIBLIOTECA DI STUDI SLAVISTICI

### *Editor-in-Chief*

Laura Salmon, University of Genoa, Italy

#### *Associate editor*

Maria Bidovec, University of Naples L'Orientale, Italy

### *Scientific Board*

Noemi Albanese, University of Rome Tor Vergata, Italy Rosanna Benacchio, University of Padua, Italy Maria Cristina Bragone, University of Pavia, Italy Giuseppe Dell'Agata, University of Pisa, Italy Claudia Olivieri, University of Catania, Italy Francesca Romoli, University of Pisa, Italy Laura Rossi, University of Milan, Italy Marco Sabbatini, University of Pisa, Italy

### *International Scientific Board*

Giovanna Brogi Bercoff, University of Milan, Italy Maria Giovanna Di Salvo, University of Milan, Italy Alexander Etkind, European University Institute, Italy Lazar Fleishman, Stanford University, United States Marcello Garzaniti, University of Florence, Italy Harvey Goldblatt, Yale University, United States Mark Lipoveckij, University of Colorado-Boulder , United States Jordan Ljuckanov, Bulgarian Academy of Sciences, Bulgaria Roland Marti, Saarland University, Germany Michael Moser, University of Vienna, Austria Ivo Pospíšil, Masaryk University, Czech Republic

Anna Polivanova

# Old Church Slavic

# Grammar and Dictionaries

translated by Lev Blumenfeld edited by Artemij Keidan

FIRENZE UNIVERSITY PRESS 2023

Old Church Slavic : grammar and Dictionaries / Anna Polivanova, translated by Lev Blumenfeld, edited by Artemij Keidan. – Firenze : Firenze University Press, 2023. (Biblioteca di Studi Slavistici ; 51)

https://books.fupress.com/isbn/9791221501049

ISSN 2612-7687 (print) ISSN 2612-7679 (online) ISBN 979-12-215-0103-2 (Print) ISBN 979-12-215-0104-9 (PDF) ISBN 979-12-215-0105-6 (XML) DOI 10.36253/979-12-215-0104-9

Graphic design: Alberto Pizarro Fernández, Lettera Meccanica SRLs

Front cover image: Fragment of the *Codex Zographensis*, f. 225r, © Russian National Library. Electronic Library of the Manuscripts Department of the Russian National Library: <https://expositions.nlr.ru/ ex\_manus/Zograph\_Gospel>.

Volume published with the support of Sapienza University of Rome, Department of Istituto Italiano di Studi Orientali.

Data from the dictionaries are freely available as a searchable database online here: <https://integral. github.io/osd/>.

### *Peer Review Policy*

Peer-review is the cornerstone of the scientific evaluation of a book. All FUP's publications undergo a peer-review process by external experts under the responsibility of the Editorial Board and the Scientific Boards of each series (DOI 10.36253/fup\_best\_practice.3).

#### *Referee List*

In order to strengthen the network of researchers supporting FUP's evaluation process, and to recognise the valuable contribution of referees, a Referee List is published and constantly updated on FUP's website (DOI 10.36253/fup\_referee\_list).

#### *Firenze University Press Editorial Board*

M. Garzaniti (Editor-in-Chief), M.E. Alberti, F. Vittorio Arrigoni, E. Castellani, F. Ciampi, D. D'Andrea, A. Dolfi, R. Ferrise, A. Lambertini, R. Lanfredini, D. Lippi, G. Mari, A. Mariani, P.M. Mariano, S. Marinai, R. Minuti, P. Nanni, A. Orlandi, I. Palchetti, A. Perulli, G. Pratesi, S. Scaramuzzi, I. Stolzi.

*FUP Best Practice in Scholarly Publishing* (DOI 10.36253/fup\_best\_practice)

The online digital edition is published in Open Access on www.fupress.com.

Content license: except where otherwise noted, the present work is released under Creative Commons Attribution 4.0 International license (CC BY 4.0: http://creativecommons.org/licenses/by/4.0/legalcode). This license allows you to share any part of the work by any means and format, modify it for any purpose, including commercial, as long as appropriate credit is given to the author, any changes made to the work are indicated and a URL link is provided to the license.

Metadata license: all the metadata are released under the Public Domain Dedication license (CC0 1.0 Universal: https://creativecommons.org/publicdomain/zero/1.0/legalcode).

© 2023 Author(s)

Published by Firenze University Press Firenze University Press Università degli Studi di Firenze via Cittadella, 7, 50144 Firenze, Italy www.fupress.com

*This book is printed on acid-free paper Printed in Italy*

# **Contents**


Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

#### CONTENTS



### CONTENTS


*contamination and competition between paradigmatic classes. Excursus on* 

# **Preface to the English edition**

Artemij Keidan

A few words to explain the genesis of the English edition of this monumental grammar of Old Church Slavic are in order. The book had been originally published in Russian in 2013 by the Institute of Slavic Studies of the Russian Academy of Sciences in Moscow.1 Lev Blumenfeld undertook the English translation of the book, working closely with the author and editor for several years. The translation has been both a titanic and an easy task. On the one hand, the author's style is terse, almost formulaic, with no space for rhetorical digressions, which simplified the translation process. On the other hand, the innovative character of this grammar led to innovative terminology; this represented an additional challenge for the translator, requiring close collaboration with the author to find apt English equivalents for novel terms (see the *Translator's note* containing a commented index of newly coined terms, p. XI).

I wish to advise the reader about some formal conventions adopted in this book which could appear unusual to the English-speaking audience.

The peculiar continuous numbering of the sections (paragraphs) has been maintained in order to preserve consistent citability with regards to the original. This was not possible with the numbering of the footnotes, since a few footnotes that were too Russian-oriented were omitted from the translation.

One most important descriptive tool used by the author are tables, found in many paragraphs. Tables in this book are not simply a nice visual representa-

1 See A. K. Polivanova, *Staroslavjanskij jazyk. Grammatika. Slovari*. Moskva: Universitet Dmitrija Požarskogo; Institut Slavjanovedenija RAN, 2013.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) First name Last name, affiliation, affiliation country, email, ORCID

Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com),

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0

tion of data; rather, tables represent a scientific statement whereby one variable is mapped onto another. Note that, except for the Preface, tables are numbered with the number of the paragraph they appear in. However, some tables are left unnumbered, on purpose rather than inadvertently: these are merely used as visual aids without making a theoretical claim.

Bibliographical references are quoted by the last name of the author only, followed by the page or section number (cf.: Vaillant, § 2). The year is only added to avoid ambiguity (cf.: Lunt 1974 and Lunt 2001).

Regarding the representation of the data in Slavic, the decision in this translation has been to keep all such data as in the original, i.e. in the Cyrillic *poluustav* (a medieval semi-uncial script), rather than in Latin transliteration. This decision follows the tradition of grammars of other ancient Indo-European languages intended for a general linguistics audience (Whitney, Smyth, Sihler), including grammars of OCS (Lunt, Vaillant).

While this decision leads to a certain complexity for some readers—though, arguably, the complexity of the script is far surpassed by that of the grammatical apparatus itself—it also leads to greater clarity. The main issue with transliteration is that most reasonable transliteration systems are not segmentally bijective (e.g. *ю* is rendered by *ju*, and this is not the sole use of either *j* or *u*). Furthermore, commonly used transliteration rules coincide in large part with the rules of the phonology that generate phonological representations, which, here as in the original, are given in Latin-based transcription (with the addition of the letter *ъ* for what is conventionally also transliterated as *ŭ*, likewise *ь* for *ĭ*). Thus, maintaining the Cyrillic-Latin distinction of the original version of this grammar leads to a clear typographic difference between phonological representations and other kids of strings—graphic or morphophonological.

Note also that OCS forms and examples are not glossed in English, except for the small number of cases where a Russian gloss was given in the original (only done for disambiguation and clarity for Russian readers). It did not seem practical to supply English glosses for all of the OCS material in the grammar.

With respect to the Russian edition, only a few minor details and typos have been corrected during the translation.

Also, a searchable online database containing the lexical and grammatical data from this grammar is available at the book's internet page.

I have to express my personal gratitude to Firenze University Press for having accepted this book for publication and a special thanks to Laura Salmon, the editor of the *Collana di Studi Slavistici*, who kindly agreed to include this book in the series.

I am also grateful to the Grant Office of the Sapienza University of Rome for providing financial aid in support of publication of this book.

# **Translator's note**

Lev Blumenfeld

### I. Introduction

This is a difficult book. For specialists in Slavic linguistics, the familiar material of Old Church Slavic appears under shockingly novel light. For theoretical linguists, this work presents a construct unlike any other with the word grammar in the title. For all readers, the *grammar* is difficult because the complex formal apparatus is stated declaratively, not procedurally, and thus its validity can be evaluated—or even its content grasped—only after overviewing the entire assembly.

The general mindset of this grammar is a radical one: it is not bound by either Slavist tradition or by any particular theoretical framework, but only by its own internal consistency, the goals it sets for itself, and some common axiomatic assumptions. In the remainder of this preface, as an aid to the reader of the English translation, I will give an overview of the general properties of the grammar, of the architecture of the construction, and explain some non-obvious terminology.

### II. Explicitness and its consequences

The most salient feature of the grammar is its explicitness. While explicitness is a commonly declared goal of grammatical description, an ostensibly formal presentation often conceals an imprecise grammatical construct. Not so in Polivanova's grammar, whose explicitness is nearly absolute. The data under analysis, the tasks of the grammar, the tools available to the linguist, and the solutions adopted by her are nearly always fully laid out.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) First name Last name, affiliation, affiliation country, email, ORCID

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0

Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com),

First, the data covered by the grammar are entirely contained in the *benchmark corpus*, which is the basis for the *benchmark lists* of wordforms and lexemes. The grammar is responsible for all of the contents of these benchmark lists.

This benchmarking not only creates an explicit domain of responsibility for the grammar, but also acts as one of the layers of abstraction between the raw data and the grammatical description, more fully detailed below. Also explicit is the list of grammatical categories and category values which are available in descriptions of any wordform under analysis.

Second, the tasks of the grammar are stated overtly. Subparts of the tasks are given at the beginning of the syntagmatic and paradigmatic parts of the grammar. Unifying the various sets of benchmark tasks, the grammar specifies the following algorithm: for any name of a lexeme and a set of category values called a *paradigmatic call*—it provides a wordform.

The usually fraught relationship between linguistic theory and messy empirical reality poses a challenge to any explicit grammatical description of a variable corpus. In this grammar, the strategy of utmost explicitness extends to the process of scientific idealization itself, allowing an unstable corpus to be handled by a precise grammatical apparatus.

At the core of Polivanova's strategy is the notion of *canon*, an overtly given idealized form of the language. It is the canon that is the subject of the grammatical analysis, not the unfiltered corpus data. The choice of the canon is not empirically determinate, but remains under the grammarian's control, just as any other analytical choice made in the grammar. The benchmark lists of lexemes and wordforms are part of the canon, and thus the grammatical algorithm that relates paradigmatic calls and wordforms operates with data from the canon.

Once the grammar is able to handle the canonical forms, it may also depart from the canon in various ways. The description of the variability of the data in the corpus is handled in terms of departures, or *aberrations*, of the grammatical algorithm from canonical derivations. These alternative derivations usually do not require separate mechanisms, but result from alternative applications of mechanisms which form part of the grammar itself.

The canonical—aberrant dimension describes irregularity in terms of its distance from the declared ideal. This dimension is one of two orthogonal dimensions of irregularity used in the grammar. The other expresses irregularity within the canon itself.

More generally, the high degree of explicitness highlights the grammar's distinctive theoretical stance. It is not empiricist, in the sense that linguistics does not deal with raw data, and the grammarian does not produce an analysis using some discovery procedure applied to directly observable empirical facts. The canon walls off the grammar from raw and highly variable empirical material. Neither is there a cognitivist or mentalist orientation of the grammar: it does not claim to describe speakers' competence or psychologically realistic derivational steps. In this sense, the grammar in this book is not a theory that is directly empirically testable against data, the way that one normally conceives of e.g. generative theories.

Instead, Polivanova's grammar is conceived as a coherent algebraic object. Its axioms are the architecture of the grammar, such as the notions like segment and formative, availability of mechanisms such as alternations, and so forth (see below in this introduction for an overview of the grammatical architecture). Within that axiomatic structure of abstract grammar, language-particular choices are made in constructing a grammar of this language.1 In a sense, the topic of the book is not so much OCS grammar per se, as the grammatical construction itself. As such, it can be applied as a tool to the description of other languages of a similar inflectional-paradigmatic type.

Less obviously, the grammar makes no attempt to characterize the *class* of languages that its grammatical axioms predict. This is simply not one of the stated benchmark goals of the grammar, and thus the grammar cannot be held responsible to it. This represents another unconventional aspect of the present grammar for readers familiar with generative-style grammatical description, where so-called "typological predictions of the theory" tend to dominate the grammarian's choices.

For readers accustomed to historical/comparative descriptions of dead languages, the grammar presents a distinct set of unconventionalities. The author here is not bound by diachrony in her synchronic choices. For example, aberrations are not characterized as archaic or innovative, and neither are other variant choices provided by the grammar. The classification of e.g. verbs and nouns into inflectional classes does not follow diachronically determined divisions. While the resulting system is often strikingly unusual from a historical point of view, the analyst's synchronic orientation frees her from long-established truths, leading to highly ingenious and economical classifications. An example may be provided by the apparently simple but deeply innovative declensional classification in Table 299, which does away with familiar notions like *a*-stems, *u*-stems, etc. Another example is the verbal classification in Tables 432 and 440, where the author found an apparently unique arrangement of columns that allows an economical description of stems used in various verb classes and categories.

I will not note here all differences between Polivanova's treatment and a more traditional diachronically-oriented treatment of the same facts, but a few examples are worth highlighting.

There is no notion of first, second, third palatalization, and no distinction between two kinds of *yat* morphophonemes, which would be necessary to keep first and second palatalizations apart. This move is made necessary by the author's assumed principle that phoneme and morphophoneme alphabets must coincide. As a result, the alternations corresponding to traditonal palatalizations are not segmentally decidable.

The principle here named Jakobson's Law is the familiar law of open syllables, reformulated without reference to any notion of syllable, but instead regulating the adjacencies of formatives of various shapes and the terminal selection rules.

<sup>1</sup> See the author's "Principy postroenija segmentnoj grammatiki" (Polivanova 2008), and "Formal'naja paradigmatika i klassy slov v russkom jazyke" (Polivanova 2001).

Another example of an innovation is the letter Ж, the voiced counterpart of щ, postulated for the so-called standard Cyrillic alphabet.

### III. Architecture of the grammar

Because of its declarative statement, the architecture of the grammatical construction is not immediately apparent from the text of the grammar itself. It is given in a brief procedural summary in Ch. 25, and here in this preface.

A distinctive feature of the grammar is the use of alternative derivational pathways to describe various kinds of irregularity. The "main road" of the derivation, used by fully regular forms, is shown in Figure I. It starts with the paradigmatic call, or a combination of a lexeme with a set of grammatical categories and category values, and proceeds by the successive (serial) application of various selection and replacement rules provided by the grammar.

Figure I. Main derivational pathway

The operations at each step draw from a library of adjustments available not only to these operations on the main derivational pathway, but also to alternative paths of the derivation, in particular to aberrations. For example, there is a set of of rules, called *alternations*, that replace particular segments in particular locations in formatives. Alternations are used by *paradigmatic effects* and by *boundary adjustment rules*, which are two separate grammatical mechanisms that modify strings. One of those mechanisms, *paradigmatic effects*, is itself available to various parts of the grammar. It not only forms verbal workstems in the standard paradigmatic classes, but is also used in constructing paradigms of irregular verbs (within-canon irregularity), and in aberrant forms. To use a programming analogy, mechanisms like alternation and paradigmatic effects are subroutines that may be called on by different parts of the program.

As mentioned above, there are two dimensions of regularity: regularity within canon (primary/secondary, anomalies) and in sources (aberrations). The grammar treats both of these types as departures from the main derivational pathway, and both of them draw on the same library of available mechanisms such as alternations and paradigmatic effects.

### IV. On terminology

I supply a list of explanations for some key terms used in this grammar that are unusual or otherwise non-self-explanatory. I draw the reader's attention especially to the terms *spellout* and *variant*, because they are used in an unfamiliar way in this grammar. The original Russian term or terms are given in parentheses

**Aberrant, aberration (девиантный, девиация).** Any form is either canonical, aberrant, or erroneous. Aberrant forms, resulting from aberrant application of grammatical processes, are departures from canonical derivations. The aberrant-canonical dimension is the formal tool used to handle the variability of the corpus data that coexists with a well-behaved idealized form of the language that is subject to the grammatical analysis.

# **Adjacency (смежность).** See *alternation*.

# **Allotment rules (правила назначения).** See *basic stem*.

**Alternation (чередование).** Alternations are a mechanism responsible for describing polyformy (multiple different shapes) of formatives. It is a general mechanism, in that it is used by different parts of the grammar: it is a library of possible segmental *replacements* that is called upon by both canonical and aberrant derivations, and by different parts of those derivations. Formally, alternations are tables whose rows are *grades* and whose columns are *series*, and whose cells are segments or sequences of segments. Replacement rules by an alternation substitute a member of a series belonging to one grade by the member belonging to an *adjacent* grade. Such replacement partners are called *pairings*. Key alternations are *velar palatalization*, *substitutive softening*, and the *fundamental vowel alternations*. See also *terminal*.

# **Ambivalence (амбивалентность).** See *CVC ambivalence*.

**Anomaly, anomalous (аномалия, аномальный).** The notion of *anomaly* is a formal tool to describe degrees of regularity *within the canon*. It is orthogonal to the canonical—aberrant dimension that describes variability of the corpus. Anomalous forms are those that are part of the canon but either their structure or their derivational pathway departs from the standard in some way. Anomalies can be syntagmatic (e.g. containing otherwise prohibited clusters), or paradigmatic (e.g. showing unusual application of paradigmatic effects).

**Basic component (базовый компонент).** Part of a form's morphological structure that excludes suffixes. This term loosely corresponds to the English term *base*. **Basic correspondence (базовое соответствие).** In a *writing* or *graphic system*, the rules that relate phonemes to letters are its *basic correspondence*. Each of the OCS *sources* has its own basic correspondence and writing system.

**Basic stem (базовая основа).** The basic stem, along with the *workstem*, are intermediate stages in the derivation of a wordform. Basic stems result from the application of *basic stem allotment rules*, which select either the *expanded* or *truncated* version of the stem based on the *paradigmatic call*. The result is then subject to *paradigmatic effects*, which output the *workstem*.

**Benchmark (контрольный).** The *benchmark* corpus is the set of data for which the grammar is responsible. From it are derived the benchmark list of lexemes and wordforms. The grammar is responsible for relating the wordforms, their paradigmatic properties, and their parent lexemes.

**Bundle (комплект).** Subsets of forms within finite subparadigms that are opposed to each other in tense/aspect are organized into four bundles: present, imperfect, imperative, and aorist.

**Canon, canonical (эталон, эталонный).** The grammar describes an explicitly given idealized form of the Old Church Slavic language, called the *canon*. Forms which are part of the canon are *canonical*. Other forms may be erroneous, if they have no grammatical description at all, or aberrant, if they result from alternative application of grammatical processes. The aberrant—canonical dimension is the formal tool used to handle the variability of the corpus data that coexists with a well-behaved idealized form of the language that is subject to the grammatical analysis. The term *canon* and *canonical* in this grammar should not be confused with two other senses of the word: (a) the set of texts given some official status by a church, and (b) the set of texts considered standard by a scholarly tradition. Here, the canon is an idealization, not an observable corpus. The original Russian term *эталон* (< Fr. *étalon*) does not easily yield to translation; although the expression *étalon language* was used, in a very similar sense, in the English translation of B. Uspenskij's *Principles of structural typology* (see Uspenskij), we have preferred not to overload this book with unusual terms.

**CVC ambivalence, agreement (CVC-амбивалентность, согласование).**  This is one of the two key alloformy mechanisms (along with the *twofold rule*) that governs the selection of terminals based on phonological shape of the stem: by CVC ambivalence, generally, V-initial terminals are used after C-final stems, and vice versa. The principle that this alloformy enforces is called *CVC agreement*, or *Jakobson's Law*.

**Effect (эффект).** See *paradigmatic effect*.

**Expanded stem (расширенная, развернутая основа).** See *basic stem*.

**Formative, alloform, alloformy (форматив, аллоформ, аллоформия).**  *Formatives* are strings of segments. Thus, they are units of a rank immediately higher than segment. The paradigmatic part of the grammar operates on strings

of formatives. The notion of formative loosely corresponds to the more familiar morpheme, but unlike morpheme, is a purely formal notion, i.e. a one-sided object containing only the signifier, not the Saussurean two-sided pairing of sound and meaning. The notions of *alloform* and *alloformy* refer to sets of formatives belonging to a single family and distributed according to some grammatical principle. One innovative consequence of divorcing the notion of formative from its meaning is that the grammarian becomes more free in setting up formatives where convenient for the analysis, for example by unifying homophonous morphemes with different senses under one formative, or splitting a morpheme like тел҄ь into formatives т and ел that occur elsewhere.

**Gloss (глосса).** This term is not used in its usual sense of 'translation into the language of the grammar', but refers to individual occurrences of wordforms in specific sources.

**Grade (ступень).** See *alternation*.

**Graphics, graphic (графика, графический).** The texts in both their raw and their canonical form are given in *graphic* representation, as sequences of letters and spaces. While sounds, not writing, are the usual object of analysis in linguistics, the sound—letter distinction is immaterial in the case of OCS, a dead language where the only evidence for sound is a writing system.

**Index (индекс).** See *paradigmatic index*.

**Jakobson's Law (закон Якобсона).** See *CVC ambivalence*.

**Monovariate (моновариантный).** See *twofold*.

**Normalization (нормализация).** The notion of *normalization* refers to the set of writing conventions used to represent canonical forms. This includes both the alphabet—the use of the narrower *standard Cyrillic* compared to the *expanded Cyrillic* used to represent forms from sources—as well as various other conventions that form the writing system of the canonical OCS.

**Pairing (соотношение).** See *alternation*.

**Paradigmatic call, address (парадигматический заказ, адрес).** Each well-behaved wordform is endowed with grammatical properties: a set of category-value pairs appropriate to a particular class (e.g. gender for nouns and adjectives, person for verbs, etc.). A *paradigmatic call* is an expression of the form *K*(*L*), where *K* is a set of category-value pairs and *L* is a lexeme. It thus "calls" a particular cell in the paradigm of a particular lexeme. A *paradigmatic address* is a property of a wordform, and can be thought of as the set of all paradigmatic calls to which that wordform answers.

**Paradigmatic effect (парадигматический эффект).** *Paradigmatic effects* are a library of changes that apply to create both regular and irregular forms within the canon, and aberrant forms outside the canon. In the regular derivation, paradigmatic effects apply to basic stems to generate workstems. The application

of paradigmatic effects is determined by the paradigmatic index of the lexeme.

**Paradigmatic index (парадигматический индекс).** A lexical diacritic mark that carries information about paradigmatic properties of a form—the set of terminals it uses, the set of paradigmatic effects that are applied to it, etc.—is called the *paradigmatic index*.

**Profile (анкета).** A summary of key forms in the paradigm of a word is called its *profile*. Profiles are not formal objects in the grammar, but practical tools that help the reader build full paradigms from a subset of the forms.

**Pure (чистый).** See *sonant*.

**Replacement (замена).** See *alternation*.

**Representation, in paradigmatics (репрезентация).** Not to be confused with the use of *representation* synonymous with *spellout*, this term refers to the largest division within the forms of a lexeme. For verbs, this is the division between finite and nominal representations, the latter containing all the participles. For adjectives, the two representations contain the so-called long and short forms.

# **Series (ряд).** See *alternation*.

**Sonant (сонант).** This term (a traditional one in Slavic and Indo-European linguistics) refers to some of the series of the fundamental vowel alternations, namely those where the undergoing segment sequence contains a son(or)ant along with a vowel. The series that contains only vowels is called *pure*.

**Source (памятник).** The corpus of texts that this grammar describes consists of seven sources: *Kiev Missal*, *Codex Zographensis*, *Codex Marianus*, *Codex Assemanius*, *Psalterium Sinaiticum*, *Sava's book*, and *Codex Suprasliensis*. This is a strictly bounded corpus; the grammar is not responsible for information outside of it.

**Spellout (запись).** This unusual term is used as a general cover word for any linguistic object regardless of its analytical status. It is thus a more general term than *representation*, which presupposes that something is being represented in terms of something else. Forms found in corpora are *spellouts*, as are canonical words, formatives, etc. This term is a convenient alternative to the over-loaded word *form*. The term *spellout* is not related to "spellout" used in various generative theories of syntax and morphology. Note also the use of the term *representation* in a narrow sense in the paradigmatics.

**Starting form (исходная форма).** The form of a lexeme given in the Paradigmatic Dictionary. In generating the output from a paradigmatic call, the starting form serves as the initial step.

**Stem allotment rules (правила назначения основы).** See *basic stem*.

**Stem (основа).** See *basic stem*; *workstem*.

**Subparadigm (субпарадигма).** Large sets of wordforms in the paradigm of

a lexeme are broken into subparadigms in such a way that all forms of a single subparadigm share a workstem.

**Terminal (окончания, флексии).** Inflectional formatives are called *terminals*, distinct from *suffixes*, which are derivational formatives. Terminal selection laws enforce basic phonotactic principles such as avoidance of clusters and hiatus. Terminal selection laws are distinct from *alternations* in two ways: alternation rules are replacements of segments by other segments, while terminal selection does not replace a formative with another formative, but selects from a list of available choices. Second, terminal selection is segmentally determinate, while alternations need not be. See also *alternations*.

# **Truncated stem (усеченная основа).** See *basic stem*.

**Twofold (двухвариантный).** This term, borrowed from Lunt 1976 and greatly expanded in its use, refers to the pervasive pattern of alloformy of formatives that depends on the segmental content of the immediately adjacent formatives. The traditional analog of twofold formatives are "soft" and "hard" declensions. Here the distribution is handled by the *twofold rule* (§ 86), which determines the choice of twofold formatives in appropriate contexts. Declension types which are not twofold are called *monovariate*.

**Unique (уникальный).** *Unique* lexemes are those whose oblique forms are too irregular to be formed according to any rules. Strictly speaking, such forms are simply listed and not given any mophological structure.

**Variant, variance, -variate (вариант, -вариантность).** The term *variant* and related words are used in this book in the sense of 'alternative forms of the same formative'. For example, the pairs of terminals that participate in alloformy patterns like CVC agreement or the twofold rule may be referred as 'variants'. Formatives that do not participate in such patterns are called *monovariate*. Importantly, the term carries no meaning related to dialectal variation, or optionality, as may be more familiar to some readers.

**Workstem (рабочая основа).** The *workstem*, along with the *basic stem*, is an intermediate stage in the derivation of a wordform. Workstems result from applying *paradigmatic effects* to basic stems, but prior to the selection of terminals. The arrangement of the rules of this grammar ensures that the workstem is the same for all forms that belong to a subparadigm.

# V. Acknowledgements

I am grateful to Artemij Keidan, who has participated in all aspects of the work on this translation. He has given countless hours of his time both to substantive discussion of the grammar's content and to the translation issues, as well as the technical side of typesetting the work. I also wish to express my sincere gratitute to the author for the opportunity to work on this translation and for the many fruitful hours of discussion.

# **From the author**

When opening a new book, readers often want to figure out what it has and what it doesn't have, and to what kind of reader the author is primarily addressing it. Answering these questions directly is the goal of this preface.

This book contains the grammar and grammatical dictionaries of the Old Church Slavic (OCS) language. The following are its important features. First, in its genre this is not a textbook, but a scientific monograph. Second, in its content it is a purely synchronic grammar, and not a synchronic description in light of diachrony. Third, the theoretical framework of the grammar is classical, as opposed to some narrower framework (e.g. generative grammar, optimality theory, etc.). From the beginning I approached this work as an attempt to improve on the classical monographs of André Vaillant (1948), Paul Diels (1963), and Horace Lunt (1974). The improvements are in the direction of utmost completeness, explicitness, and deliberate consistency with the corpus of texts and the grammatical dictionaries. Explicitness, as it turns out, necessitates discussion and revision of almost the entire grammatical apparatus. This is the reason why the book may seem brimming with innovative terminology and formal constructs.1

1 I should note especially the profusion of tables in this book. Perhaps there are too many; almost every paragraph contains one; some contain several. Of course, the data from tables can be linearized, and I am aware that some readers resent tables. It is also no secret that making tables is not always easy. When building my tables, I recalled the ironic remark of my teacher Alexander Wentzell: «I think I got it: linguistics is the art of table-making». For me, this observation is filled with deep scientific meaning, which uncovers an answer to the persistent question of the connections between linguistics and mathematics.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) First name Last name, affiliation, affiliation country, email, ORCID

Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com),

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0

Of course, this grammar treats some facts differently from predecessors (for example, verbs are classified in a different way compared to the authors mentioned above). It is useful, however, to note the principal novelty in content compared to previous works. The novelty is that this grammar consistently examines morphophonological spellouts of wordforms, where their external shape is represented by strings of formatives. For all starting forms, that spellout is given in the grammatical dictionary, cf. for example безакон҄ѥниѥ — без.за.(кон҄).ен.ьј.е, врѣменьнъ — (врѣт).м.ен.ьн.ъ, достоинъ — до.(сто*i̯*).ьн.ъ. Accordingly, a root dictionary is also included, where each root is matched to all lexemes that contain it. Thus, the grammar gives rules that ensure the transformation between morphophonological and phonological/graphic spellout.

The goal of the book is to give a complete description of the synchronic grammar of OCS (graphics, phonology, morphophonology, and paradigmatics). The grammar is responsible for a strictly bounded set of sources (*Kiev Missal*, *Codex Zographensis*, *Codex Marianus*, *Codex Assemanius*, *Psalterium Sinaiticum*, *Sava's book*, and *Codex Suprasliensis*). Due to the significant heterogeneity of sources, peculiarities in the language of the sources are described as systematic departures from canonical OCS, which is the language determined by the present grammar. Principles of the grammatical description that underly the construction of canonical OCS and of the rules that compute the transition from the canon to the language of specific sources are discussed in detail in the introduction (*Differences across idiolects and instability within idiolects in OCS sources*). It is useful to note that in most textbooks and monographs on OCS, it is the so-called common Slavic language that functions as this canon, i.e. the result of a comparative-historical reconstruction, following an established tradition in OCS studies that goes back to the classical grammar by Leskien (1871). In the few grammars whose goal is synchronic description (Vaillant; Lunt 1974), the canon is introduced implicitly, as if it were an obvious empirical given. Accordingly, linguistic problems encountered in constructing a canonical grammar remain mostly hidden from the reader. Building a complete, synchronically-oriented description of the OCS grammatical system makes it possible to pose, and sometimes to solve, a series of nontrivial theoretical problems relevant for grammars of inflectional languages, as well as problems particular to grammars of dead languages represented by limited corpora.

The goal of a complete description of sources is not posed in this book. A selected list of parameters along which sources more or less systematically depart from canonical grammar is explicitly given. Observed departures outside of that list are noted sporadically, without any attempt at completeness. Given that the old literature is hard to access, many well-known facts are repeated and reinterpreted here, regardless of how they are treated in the literature (for example, almost all facts given in Diels' grammar are included here). And yet, this book should not be seen as a complete handbook on the sources.

The book consists of four parts. The first two contain the grammar; the third part contains some addenda, among them a small selection of texts; the fourth contains the dictionaries. The material of the first two parts, along with the dictionaries, makes it possible to construct, for each word of the starting corpus, its entire paradigm, and for each wordform, to determine its grammatical characterization, and its phonological and morphophonological spellout.

The book is addressed primarily to specialists—linguists and philologists working in Slavic studies. However, it is also written in such a way that it can be successfully used as a practical guide by those who are first starting to study OCS and reading original texts. From 1970, when A. A. Zaliznjak first taught a course on OCS for linguists, and I was responsible for the exercises, until 2011, I taught OCS to linguistics students every year. This book largely reflects this long pedagogical experience.

In my work on this book I benefited from advice and direct help of A. V. Dybo, V. A. Dybo, Ju. M. Gizatullina, A. S. Kasjan, B. V. Kravcov, E. S. Logunova, I. S. Pekunova, A. V. Ter-Avanesova, M. N. Tolstaja, M. V. Vinogradova, A. A. Zaliznjak, and M. V. Živova. I would like to especially note the invaluable help of I. S. Pekunova, who not only gave feedback on the many drafts of the text, but also proofread the great number of quotes and illustrations, and in many cases selected them. To all of these colleagues I express my heartfelt gratitude.

I also wish to express my sincere gratitude to Lev Blumenfeld, the brilliant English translator of this book. Working with Lev was both business and pleasure. It turned out that there were passages that remained obscure even to myself, and which I was able to clarify only after the translator assailed me with requests of elucidations about them.

I am happy that the English edition will finally appear. I am grateful to Laura Salmon who made this publication possible, and to A. Keidan, who undertook the labor of coordinating the editorial process.

INTRODUCTION

# **Differences across idiolects and instability within idiolects in OCS sources**

§I. The Old Church Slavic language and its sources

Old Church Slavic (OCS) is defined, on the one hand, as the language of the first Slavic translations of liturgical books by Sts. Cyril and Methodius in the second half of the IXth century, and, on the other hand, as the language represented by a small corpus of the oldest surviving texts (hereafter *sources*), created in the X–XII centuries.

There is no agreement among researchers on which sources constitute the OCS canon and which are outside of it.1 For example, the dictionaries of Sadnik and Večerka differ in the list of sources. The benchmark corpus in this book includes only the following seven sources: *Kiev Missal*, *Codex Zographensis*, *Codex Marianus*, *Codex Assemanius*, *Psalterium Sinaiticum*, *Sava's book*, and *Codex Suprasliensis*.

Data from other sources are admitted unsystematically.

### §II. Glagolitic and Cyrillic writing

The most salient difference between the sources is the use of two scripts, or alphabets, the *Glagolitic* and *Cyrillic* ones. Some sources use the Glagolitic script

1 For example, some authors are inclined to consider the Ostromir Gospel (the oldest dated manuscript, from 1056) as an OCS source. The oldest sources of Church Slavic should also be distinguished from OCS, differing from it in both place and time of composition. The language used today in orthodox services is also called Church Slavic; see details in e.g. Kraveckij.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

(KIEV, ZOGR, MAR, AS, PS SIN), while others use the Cyrillic script (SAV and SUPR). This difference, however, is trivial, and can be eliminated using very simple transliteration rules that replace Glagolitic letters with their Cyrillic analogs according to very simple replacement patterns.2

The following fragment from Codex Assemanianus, shown in its original Glagolitic script and its Cyrillic transliteration, serves as an illustration (Mk 8, 34).

In most editions, Glagolitic texts are represented in transliteration. Below, all texts from sources are taken from published versions, in particular the Glagolitic ones from published Cyrillic transliterations.

§ III. Differences across idiolects and instability within idiolects

Apart from the differences in the alphabets, sources also differ in more or less significant details, showing both differences across idiolects (i.e. between the sources) and instability within idiolects (i.e. variation within a single source).



Table I shows the variant distribution for the spellout of the stem गणुभ्तुь in seven sources. Only ZOGR and SUPR show instability within an idiolect in this case-there are two variants in each of these sources. Differences across idio-

These rules are identical across editions of different Glagolitic sources. See details in § 132.

lects reveal the pairwise opposition between four groups of sources: (Kiev) ~ (Zogr) ~ (Supr) ~ (Mar, As, Ps Sin, Sav).

# §IV. Canonical OCS

The simplest way of describing differences between idiolects and instability within an idiolect assumes the definition of a certain arbitrary canon; the observed diversity is described as a departure from that arbitrary canon. This grammatical fiction is referred to here and below as *canonical OCS language*, or simply *the canon*. As a rule, this book describes the grammar of that canonical OCS. Luckily, there are no significant disagreements between researchers on the definition of various grammatical features of this language.3

The question is simply of selecting a convenient baseline for describing all observed facts. The selection of a canon as a necessary descriptive tool should not be confused with the historical and philological questions of the existence of an "original" OCS (see paragraph XI below).4

# §V. The grammar of canonical OCS and the description of sources

The goal of describing the grammar of canonical OCS is distinct from the goal of describing the sources as such.

The present book aims to construct such a canonical OCS that makes it possible to obtain the data observed in sources using some conversion rules, and to offer these conversion rules.5 Data from sources are admitted only out of necessity to show the reader that the observed diversity of the sources can really be represented as declared conversions from the canonical language.6

# §VI. Parallel texts

Below are parallel fragments from four sources, Zogr, Mar, As, and Sav.7 Verse numbers are shown using Arabic numbers at the beginning of the verse. The


number in parentheses refers to the commentary to Table II, which pertains to the section of the text that precedes the reference.

## *Codex Zographenis*, Mt 8, 28–34

28ꙇ пришьдъшюму на онъ полъ· въ странѫ ћерꙿћесинꙿскѫ· сърѣтосте (1) и дъва бѣсъна (2)· отъ гребищь (3) ꙇхъ· ꙇсходѧща (4) л҄юта ѕѣло· ѣко не можааше (5) никꙿтоже· минѫти пѫтемь (6) тѣмь· 29ꙇ се възъписте (7) глѭща· чьто естъ нама ꙇ тебѣ исе· сне бжіи· пришелъ ли еси сѣмо· прѣЖе врѣмене мѫчитъ насъ· 30бѣ же далече отъ н҄ею· стадо свин҄иі (8) много (9) пасомо· 31бѣси же мол҄ѣахѫ (10) и глѭще· аще ꙇзгониши ны· повели намъ ꙇти въ стадо свиное· 32ꙇ рече ꙇмъ· ꙇдѣте· они же шьдъше вънидѫ въ свиниѩ· ꙇ абие устръми (11) сѧ стадо все по брѣгу (12) въ море· ꙇ умрѣшѧ· ꙇ утопошѧ (13) въ водахъ· 33ꙇ пасѫщеи бѣжашѧ· ꙇ шьдъше въ градъ възвѣстишѧ всѣ (14)· ꙇ о бѣсъную· 34ꙇ се всь (15) градъ ꙇзиде противѫ ісви· ꙇ видѣвъше и молишѧ· да би прѣшьлъ (16) отъ прѣдѣлъ ꙇхъ·

### *Codex Marianus*, Mt 8, 28–34

28ꙇ пришедъшу ему исви на онъ полъ· въ странѫ ћерћесинскѫ· сърѣтете (1) и дъва бѣсъна (2) отъ жалии (3) ꙇсходѧща (4) лютѣ ѕѣло· ѣко не можааше (5) никтоже минѫти пѫтемь (6) тѣмь· 29ꙇ се възъписте (7) глща· что естъ нама и тебѣ иссе сне бжии· пришелъ еси сѣмо прѣЖе врѣмене мѫчитъ насъ· 30бѣ же далече отъ неѭ стадо свинии (8)· мъного (9) пасомо· 31бѣси же молѣхѫ (10) и глѭще· аще изгониши ны· повели намъ ити въ стадо свиное· 32ꙇ рече имъ идѣте· они же ишедъше идѫ въ свиниѩ· ꙇ абье устръми (11) сѧ стадо вьсе по брѣгу (12) въ море· ꙇ утопѫ (13) въ водахъ· 33а пасѫщеи бѣжашѧ· ꙇ шедъше въ градъ възвѣстишѧ вьсѣ (14)· ꙇ о бѣсъную· 34ꙇ се весь (15) градъ изидѫ противѫ исви· ꙇ видѣвъше и молишѧ· да би прѣшелъ (16) отъ прѣдѣлъ ихъ·

### *Codex Assemanius*, Mt 8, 28–34

28прішедъшу ису въ странѫ ћерћесиньскѫ· сърѣтосте (1) и дъва бѣсъна (2)· отъ гробищъ (3) исходѧща (4)· люта ѕѣло· ѣко не можааше (5) мінѫті ніктоже пѫтемъ (6) тѣмь· 29и се възьпісте (7) глща· чьто естъ нама и тебѣ исе сне бжіи· прішелъ еси сѣмо· прѣЖе врѣмене мѫчитъ насъ· 30 бѣ же далече от нею стадо свиніи (8) много (9) пасомо· 31 бѣси же молѣахѫ (10) и глще· аще изгониші нъи· повели намъ иті въ стадо свиное· 32и рече имъ идѣте· они же ишедъше вьнидѫ въ свиниѩ· и абіе устръми (11) сѧ стадо въсе по брѣгу (12) въ море· и утопѫ (13) въ водахъ· 33а пасѫщеи бѣжашѧ· и шедъше въ градъ· възвѣстішѧ вьсѣ (14) о бѣсъную· 34и се весъ (15) градъ изіде протівѫ исви· и видѣвъше молишѧ и да бі прѣшелъ (16) ѡтъ прѣдѣлъ их·

Cyrillic letters denoting pages of the manuscript are replaced with roman letters. Texts are shown as they appear in the editions used; in particular, the editor has broken up the text into wordforms (inserting spaces and possibly other dividers), removed corruptions (or supplied emendations), etc. Note that the researcher, when constructing a grammar, operates with a corpus of (at least partially) interpreted texts, knowing not only the contents of the text (its translation, so to speak), but also at least some of its grammatical features.

### Sava's book, Mt 8, 28-34

28 пришъдъ ic въ земля геръгеснномъ сърътоста (1) и в въсъюстия (2) са отъ грова (3) исходациа (4) лють зъло гако не могжцю (5) никольбже прити пятьмъ (6) тъвль: 2°н авнє възъписта (7) гажціа: что є нама и тєє в сбор вжи' придє прѣждє врђмєнє: насъ мжчить: 39″в жє далече: оттъ нею стадо свинни (8) много: (9) пасомо: 33 всен же его молъух (10) глิжще аще изгоннши насъ: повели намъ ити въ стадо свиное. 32 рече имъ идъте: они же идж въ свинна: и авне оустрьми (11) са все стадо по врегроу (12) въ море н истопия (13) въ водахъ 33 пасящен же въжаша: и шъдъше въ градъ повъдаша вса (14). о въськосно 34 и авне всь (15) градъ изнде въ сърътение ičoy. и видѣвъше его молишь: тако да прендетъ (16) отъ пределъ ихъ

### § VII. Analysis of selected examples

The forms of different sources are not compared with each other, but each is compared with the canonical form. As long as the canon is fixed, each form of the source is easily identified as canonical or non-canonical, or alternative.

Table II on p. XXVI shows canonical forms with their grammatical address and their analogs in the four passages from different sources, for 16 wordforms.

### § VIII. Source-to-source and source-to-canon comparison

In some cases, sources differ lexically and syntactically in parallel passages. For example, at the end of the fragment above we see in ZOGR да Ем пръшьмъ отъ пръдъ сув, but in SAV гако да преидетъ отъ пръдълъ пхъ (the construction in MAR and As is the same as in ZOGR). Accordingly, the canon for the fragments in ZOGR, MAR and AS (да ви првшылъ отъ пръдълъ ихъ) differs from the canon for the corresponding fragment in Sav (пко да пръидетъ отъ повдѣлъ ихъ). Such differences, while philologically quite interesting, are outside of the scope of this book, which deals only with features of segmental grammar and paradigmatics. Thus, when considering sources, the subject of analysis is the following pair: (actual spellout of wordform in text, its canonical analog). Members of such pairs are eponymous wordforms, e.g. (cъptreve, coptirocre) 3Du2PIAor (съръсти); (пжтємъ, пятьмь) ISg (пять) etc. The actual spellout of a wordform that differs from the canon is called an alternative spellout. The so-called doublet wordforms and doublet lexemes constitute a separate case. Two distinct wordforms, identical in their grammatical characterization and both belonging to the canon, are called doublets. Such are the forms BECA and BBC's, in the passages above, as well as, e.g., GLSg wordforms словесе//словесн.


# Table II. Analysis of compared wordforms in four sources

Notes to Table II


# §IX. Transition from Glagolitic to Cyrillic script

When considering actual material from sources, Glagolitic forms are examined in their Cyrillic transliteration, which is then compared with the corresponding canonical form. For example, in (1°), first the Zogr form ⱄⱏⱃⱑⱅⱁⱄⱅⰵ is converted into its Cyrillic transliteration сърѣтосте, which is then compared against the canonical сърѣтосте; Mar ⱄⱏⱃⱑⱅⰵⱅⰵ (сърѣтете) is likewise compared with the same canonical сърѣтосте, where a partial mismatch is observed. We compared Zogr ⰺⱄⱈⱁⰴⱔⱎⱅⰰ (ꙇсходѧща) (4°), As ⰻⱄⱈⱁⰴⱔⱋⰰ (исходѧща), Sav исходѧща with the canonical исходѧща, and also identify partial mismatches. In (10°), Zogrⰿⱁⰾ҄ⱑⰰⱈⱘ (мол҄ѣахѫ), Marⰿⱁⰾⱑⱈⱘ (молѣхѫ), Asⰿⱁⰾⱑⰰⱈⱘ (молѣахѫ) is compared with the canonical мол҄ꙗахѫ, and in all three pairs partial mismatches are found. In (16°), Zogr ⱂⱃⱑⱎⱐⰾⱏ (прѣшьлъ) and Mar ⱂⱃⱑⱎⰵⰾⱏ (прѣшелъ) is compared with the canonical прѣшьлъ, where Zogr shows a complete match and Mar a partial mismatch.

## §X. Aberrant spellouts in sources

As these illustrations show, alternative forms in sources live alongside canonical ones. It may be the case that the same form in the same source in some of its occurrences acts as canonical, while in others as alternative (cf. in Sav the spellout of the preposition въ and вь). The observed diversity of aberrant forms is induced by a limited number of *aberrations*. A complete list of aberrations must make available a suitable *aberrant derivation* for any observed aberrant form. In the majority of cases, aberrant derivations have the modality of permission rather than requirement. Because each aberrant form corresponds to a single definite canonical form, the transition from the text of a source to its canonical analog is determinate, but the converse is not the case: a canonical text cannot be converted into its prototype in a source using determinate rules, at least because the selection of a form as canonical or aberrant at any point in the text does not follow a rational pattern: their distribution is random.

The share of alternative spellouts in the passages shown remains below 20%. This ratio is stable across the general corpus of the texts under consideration, although in some sources there are more aberrant spellouts than in others (e.g. in Supr in some places the share is 25–30%).

Differences across idiolects and instability within idiolects are described using the same set of aberrations. The same instability within idiolects, established separately for each source, forms the differences between idiolects, because sources differ not so much in their assortment of aberrant forms, as much as by the character of competition between different canonical and aberrant forms.

Table III shows, for example, the distribution of the variant spellouts of the words къто and мъногъ in two sources. This shows that for мъног-, canonical forms predominate in Mar and aberrant ones in Zogr. For къто, canonical forms predominate in Zogr and aberrant ones in Mar.


Table III. Spellouts of the forms мъногъ and къто in Zogr and Mar

Cf. the spellout кьто Lk 10, 36, As.

The distribution of aberrant spellouts generated by the same aberration can differ in the same source in different lexemes (or even forms). Thus, we can suppose that in different sources different aberrant spellouts were treated as acceptable alternative variants. As far as segmental aberrations are concerned, the character of the competition can be understood as a certain scribal regimen, rather than a phonetic law. It is also not а graphic rule, because such rules admit no exceptions and use no information on units of higher rank than segment, while scribal regimen concerns individual wordforms.

At the same time, differences between sources and their within-idiolectal instability can be described using a simplified schema that estimates only the proportion of a given segmental aberration or graphic peculiarity in a given source. Such a schematic table is given below (Table IV) for the seven basic sources; such ratings as "no", "rare", "occasional", "present", "often" are meant to reflect the increasing proportion of spellouts that reflect a given segmental property of the source.


Table IV. Overview of the main segmental peculiarities of sources

### §XI. A note on the interpretive substance of the differences across idiolects and instability within idiolects

The purely linguistic problem of describing a group of idiolects and instability within an idiolect amounts to creating such mechanisms as described above. However, from a historical-philological point of view, both the status of the canon and the status of differences between idiolects are of paramount importantce.

Manuscripts were being created at different times, in different places, and by different scribes. In the general case idiolectal differences can be interpreted as temporal (more vs. less archaic), regional (reflecting dialectal differences), or as social and register differences (e.g. as the opposition between oral and written, formal and informal, and the like). The sociolinguistic situation can be quite complex when such oppositions are interlinked.

As far as the status of the canon, the question can be posed as follows. First, can the canon be interpreted as the "original" OCS, i.e. as the language of translation of the first Greek liturgical books, the language of the first Slavic apostles? Second, can the canon be considered the genetic prototype of the idiolects observed in the sources? For the specific canonical OCS that is usually considered as such and described in this book, the answers are obviously negative: it is neither "Cyril's original" language, nor a genetic prototype. We should note that while the latter question of the genetic prototype is fairly clear, the meaning of the question of "Cyril's original" is not well-defined as posed: it is not obvious a priori how to verify a proposed reconstruction of such a language. Durnovo (1929) carefully considers the question of "Cyril's original" language, using the oldest data on alphabets. Taking his conclusions on the traits of that language, we must admit that it does not coincide with the canon described here.8

8 Interested readers are referred to Durnovo's works, which contain a comprehensive analysis of these questions both in their grammatical and interpretive aspects. Here we limit ourselves to one quote, which clarifies the main direction of Durnovo's thinking: «Sts. Cyril and Methodius, with their translations, originated the Slavic literary language that is known to us in its oldest attested form as Old Church Slavic. Since it is defined as a literary language, we understand under the term a certain norm that authors, translators, and scribes who were writing in this language attempted to follow, and which cannot be identified with their individual languages or a living dialect. Only those linguistic traits that were perceived by the writer as the norm form part of OCS, more or less consistently according to the writer's level of literacy. Traits not consistently maintained in these sources and amounting to departures from the scribe's adopted literary norms, are not part of OCS as a literary language, and should be regarded as reflecting various living dialects, or another literary language. It also follows from the same definition that, even though a living dialect lay at the basis of OCS, and that it is possible for a literary language to coincide with a living dialect in all its traits, without including traits of other dialects or languages, we may not assume without sufficient reason that OCS as we know it coincided with another Slavic language or dialect» (Durnovo 1929, quoted from Durnovo 2000, p. 567).

PART I

Segmental grammar

# CHAPTER 1 Acquaintance

The part of the book entitled Segmental grammar deals with the segmental systems: graphics, phonology, and morphophonology. The goal of this auxiliary chapter is to fix the terminology and to mention some particular issues which would be confusing without explanation. These clarifications concern only the sense of the notions under discussion, while their content is treated in the rest of this Part, and in the dictionary.

### § 1. Samples of segmental representations

This Grammar uses three types of segmental representations, or spellouts:4 graphic representation, phonological representation (ph), and morphophonological representation (mph). The graphic representation, or spellout, is also called normalization (norm).

Below are examples of these three spellouts, for a short passage (Mt 8, 28-30).3

### Normalization

2° И ПОНШЬДЪШОУ ЕМОУ НА ОНЪ ПОЛЪ' ВЪ СТРАНЖ ГЕОБГЕСННЬСКЖ' СЪРЪТТОСТЕ И ДЪВА БЪСЬНА' ОТЪ ГРОБИШТЪ ИХЪ' НЕХОДАШТА ЛЮТА БЪЛО' ГАКО НЕ МОЖААШЕ НИКЪТОЖЕ линяти пжтьмь тѣмь: 2% се възъписте глагодняшта: чьто естъ нала и тєбъ нсоусе: съще пожии: пришьлъ ли есн съмо: пръжде връмене мжчить насъ 30 гв ЖЕ ДАЛЕЧЕ ОТ В НЕЮ СТАДО СВИНИИ МЪНОГО ПАСОМО


Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup referee list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215-0104-9, DOI 10.36253/979-12-215-0104-9

Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

# Phonological spellout

28i prišьdъšu jemu na onъ polъ vъ stranǫ gerьgesinьskǫ sъrětoste i dъva běsьna otъ grobištь ixъ isxodęšta l̕uta dzělo jako ne možaaše nikъtože minǫti pǫtьmь těmь 29i se vъzъpiste glagol̕ǫšta čьto jestъ nama i tebě isuse syne božii prišьlъ li jesi sěmo prěžde vrěmene mǫčitъ nasъ 30bě že daleče otъ n̕eju stado svinii mъnogo pasomo

# Morphophonological spellout

28(и) при.шьд.ъш.у j.ему (на) он.ъ пол.ъ (въ) стра.н.ѫ (герьгесиньскѫ) съ.рѣт.осте j.ь дъв.а бѣс.ьн.а (отъ) гроб.ищ.ь j.ихъ из.ход.ѧщ.а л҄ут.а ѕѣл.о (ꙗко) (не) мож.ѣ.аше ни.к.ъ.т.о.же ми.н.ѫ.т.и пѫт.ьмь т.ѣмь 29(и) (се) въз.ъп.и.сте гла.гол҄.ѫщ.а ч.ь.т.о [ѥстъ] [нама] (и) [тебѣ] (исусе) сын.е бож.ьј.ь при.шьд.л.ъ (ли) [ѥси] (сѣмо) (прѣЖе) врѣт.м.ен.е мѫч.и.т.ъ [насъ] 30[бѣ] (же) (далече) (отъ) н҄.еј.у ста.д.о свин.ьј.ь мъног.о пас.ом.о

# §2. Wordforms

In all three types of spellouts, spaces delimit *wordforms* (in their graphic, phonological, or morphophonological spellouts). A raised dot in the graphics delimits certain strings of wordforms. Note that manuscripts have no spaces, and thus it is the editor's job to segment the text into wordforms. Punctuation is variable both in manuscripts and in editions of texts.4

Examples and notes


<sup>4</sup> In the graphics, as well as in quotes from sources, a raised dot stands for all punctuation.

expressions like the following: "the lexeme вьсь is represented by 12 glosses in Sav, including six Sg forms and six Pl forms"; "the spellout вси of the lexeme вьсь has four glosses in Sav"; "Ps Sin contains no glosses of the lexeme вьсь", etc.

### §3. Benchmark list of wordforms

The set of wordforms in canonical OCS is fixed. Each wordform in the list corresponds to, first, its asegmental name (its paradigmatic address; see § 245); second, to the three segmental spellouts—the graphic, phonological, and morphophonological ones. In this book, these representations are fully defined only for the *benchmark list of wordforms*, and may be undefined to some extent outside of that list.

The boundaries of the benchmark list of wordforms are defined by the *benchmark list of lexemes*, which is given by the *paradigmatic dictionary* (hereafter PD) together with the rules of paradigmatic synthesis, which describe the formation of all oblique wordforms.

The benchmark list does not include: 1) extraparadigmatic forms,5 2) compounds; 3) proper names (toponyms and anthroponyms), and 4) unassimilated loans and their derivatives.

In the text samples above all wordforms that are not in the benchmark list are parenthesized; герьгесиньскѫ and исусе are examples of a toponym and an anthroponym, respectively.

#### §4. Graphics

The normalization, or graphics, is a writing system for canonical OCS, as well as the alphabetic spellout of concrete canonical wordforms. By default, it is used as the basic segmental representations of wordforms, and of passages comprising several wordforms.

The atomic units of the graphics are the letters of the *standard Cyrillic* (see § 16 below). From now on, whenever we talk about graphics, unless otherwise noted, we are referring to the normalization.6

In the graphics, a wordform is a string of letters.

### §5. The phonological representation

The atomic units of the phonological representation are *phonemes* of the so-called standard phonology, which are defined by their inventories (see § 23–27 below;

<sup>5</sup> See Part II, Ch. 8 on paradigmatic and extraparadigmatic forms.

<sup>6</sup> Note that normalizations, with varying degrees of systematicity, are used in all grammars and dictionaries of OCS. In this sense we can talk about different normalizations: e.g. Vaillant's normalization, Lunt's normalization, Večerka's normalization, etc. For example, глагол҄ѥши, глагол҄еши and глаголѥши are spellouts of the same wordform in different normalizations.

for nonstandard phonology, see § 866). Either Cyrillic or Roman letters can be used to represent phonological segments. The choice of the Roman or Cyrillic alphabet is only a matter of convenience for each particular context. Thus, spellout pairs like *prišьlъ* and пришьлъ, *jesi* and jеси, *l'uta* and л҄ута, are equivalent phonological spellouts of the corresponding wordforms.

In the phonology, a wordform is a string of phonemes.

### Examples and notes

The spellout герьгесиньскѫ is aberrant both in graphics and in phonology (*gerьgesinьskǫ*), as well as in morphology (герьгес.ин.ьск.ѫ): the sequences of velars and front vowels are prohibited everywhere. As we already noted, this word does not belong to the benchmark wordlist. Phonological interpretation of such forms, generally speaking, goes beyond the scope of the present grammar.

### §6. The morphophonological representation

The atomic units of morphophonological representations are standard phonological phonemes.7 Periods separate *formatives*, which are the linear units of the rank immediately above the phoneme.8 Thus, a formative is a string of phonemes. Morphophonological spellouts, just as phonological ones, can be written in either Cyrillic or Roman script: *pri.šьd.l.ъ* or при.шьд.л.ъ, *l'ut.a* or л҄ут.а.

Morphophonological spellouts make use of a special auxiliary symbol *i̯*, called *epenthetic i̯* (or *epenthetic yod*). This symbol is not a segment. It always precedes a vowel, and corresponds either to the phoneme /j/ or to zero in the phonological spellout.9

In two cases, the morphophonological representation is not defined: first, for forms outside of the benchmark list of wordforms (such forms are parenthesized in the sample), and second, for *morphologically anomalous* wordforms of paradigmatic lexemes (they are placed in square brackets in the sample; see details in §261).

In the morphophonology, a wordform is a string of formatives.


<sup>7</sup> It is commonly assumed that morphophonological spellouts are built from distinct units called the *morphonemes*. In this grammar the identity of phonological and morphophonological atomic units is assumed as a postulate in the construction of the segmental grammar.

Examples and notes

1°. Formatives are classified by position class: *prefixes* (p), *roots* (R), *suffixes* (s), and *terminals* (t). Membership in position class is a lexical property of a formative. In the paradigmatic dictionary root formatives are parenthesized, cf. (пол).ъ, (ста).д.о, (бож).ьj.ь, из.(ход).и.т.и, etc. All formatives preceding the root are prefixes; all formatives following the root but preceding the terminal are suffixes. In the rest of the book, parentheses around roots can be left out, as in the sample above. Periods separating formatives can also be dropped, and some spellouts can be mixed, cf. въз.л҄юбити for the morphophonological spellout въз.л҄уб.и.т.и and the graphic spellout възюбити, or из.жити, for the morphophonological spellout из.жи.т.и and the graphic spellout иЖити.

According to each formative's position class identity, each wordform has a *pRs schema*. Thus, (ста).д.о has the pRs schema R.s.t; из.(ход).и.т.и has p.R.s.s.t.


### §7. Sounding the text

In reading OCS texts out loud, academic practice allows certain phonetic conventions that bring it closer to the language of the reader.10

<sup>10</sup> Common Russian practice, for example, does not distinguish е and ѣ, and almost does not distinguish ъ and ь.

### §8. Segmental systems

The graphic, phonological, and morphophonological representations form three *segmental systems* of the OCS grammar: the graphics, phonology, and morphophonology, respectively. Atomic units in each system are called *segments*. The phonetics of OCS as a separate system is not treated in this book due to the absence of sufficient data.11

Segmental grammar must fully describe both the segmental systems themselves and the *mapping* rules which link the systems with each other.

### §9. Setting up a segmental system

In order to set up a segmental system, one must provide 1) an *inventory* of its atomic units, and 2) *dictionaries* or *inventories* of derivative units—that is, the units of the next-highest rank—strings of atomic units, strings of strings, etc.12

A grammatical description of a segmental system includes 1) determining and investigating paradigmatic and syntagmatic relationships between atomic units, and 2) investigating the formation of derivative units, which are strings of basic units.

### §10. Basic and derived units: segments, formatives, and wordforms

The primary, basic units of all three systems are segments, and they are atomic. In the graphics these are *letters*; in the phonology and morphophonology, these are *phonemes*. Derivative units in the graphics are graphic words, or graphic wordforms, that is, strings of letters separated by spaces. In the phonology, derivative units are phonological words or wordforms, that is, strings of phonemes separated by spaces.

The morphophonology includes derivative strings of two types. First, they are *formatives*, which are strings of phonemes separated by periods. Second, they are *morphophonological words*, or strings of formatives separated by spaces. Accordingly, the morphophonology gives a formative inventory, and treats the morphological composition of two types of derivative units: the construction of formatives as strings of phonemes, and the construction of morphophonological words as strings of formatives. See more details in Ch. 24, § 876, *On the morphological composition of stems and wordforms*.

### §11. Syntagmatics

A description of well-formed derivative units in the segmental grammar is called *syntagmatics*. It defines allowed and prohibited adjacencies of units of a lower


rank in a unit of a higher rank. The graphics defines allowed and prohibited adjacencies of letters within a graphic word (see § 20). The phonology—or, the phonological syntagmatics of phonemes—describes the allowed and prohibited combinations of phonemes in a phonological word. The morphophonology describes allowed and prohibited combinations of phonemes both within formatives and within morphophonological words—the morphophonological syntagmatics of phonemes, as well as allowed and prohibited combinations of formatives in morphophonological words—the syntagmatics of formatives. Below the phonological and morphophonological syntagmatics of phonemes are treated in parallel (see Ch. 4). We consider formative syntagmatics only in describing segmental combinations that are allowed or prohibited at formative boundaries. See more details in Ch. 24, § 900–909.

### §12. Mapping rules between systems

The mapping between graphics and phonology supplies a phonological representation of a wordform given its graphic representation, and vice versa. For example, for л҄юбл҄ѭ, such rules give *l'ubl'ǫ*, and for *l'ubl'ǫ*, they give л҄юбл҄ѭ. These rules are given below in § 32 and ff. The mapping between morphophonology and phonology is unidirectional, from morphophonology to phonology, but not vice versa (see rules of the type mph ⇒ ph/norm, § 63–77). So, the spellout пас.т.и gives пасти, but the same phonological spellout can also correspond to mph пад.т.и.

### §13. Alloformy

Some formatives of the same position class are considered alloforms of each other. Such are, for example, осте and сте (3Du2PlAor) in съ.рѣт.осте and въз.ъп.и.сте. Such are the root formatives j and н҄ in j.ь, j.ихъ, and н҄.еj.у. This grammar defines a particular *alloformy relation*: that is, for each formative it is given whether it has alloforms. Formatives with no distinct alloforms are *monoformemic* or *monovariate*, while those that have two or more distinct alloforms are *polyformemic* or *polyvariate*.13

The most important types of segmental alloformy are described using a system of *alternations* (see § 98 and ff.).

### §14. Aberrant spellouts and segmental aberrations

Segmental grammar also deals with segmental aberrations and aberrant forms generated by them.14 Most of the material dealing with segmental aberrations is in Chs. 6 and 7. However, the main chapters also include some comments on aberrations. Ch. 4, § 77 deals with aberrant mph ⇒ ph/norm rules; Ch. 5, § 111, 117–118 deal with aberrant versions of alternations. Ch. 23 consideres aberrant

<sup>13</sup> The alloforms of a formative may also be called variants.

<sup>14</sup> Paradigmatic aberrations are treated in Part II.

versions of corresponding formatives. A general overview of segmental aberrations and their classifications is given in § 129; see also Ch. 24, § 886–897.

### §15. The benchmark task of Part I of the book

The benchmark task15 of the segmental grammar in this book is to provide a *segmental analysis* of any wordform of the benchmark list of wordforms.16

More specifically, given the paradigmatic address of the target wordform (given by the paradigmatic grammar; see Part II), the segmental grammar must:


It may be the case in the analysis of a manuscript text (from the benchmark list of texts) that the target wordform does not have a paradigmatic derivation. In this case we must first find the canonical analog of the target wordform. In order to do this, it is necessary to (1) make a hypothesis about the paradigmatic address of the target wordform, and (2) build a canonical wordform with that paradigmatic address. Then, comparing the target wordform with the canonical wordform, we can attempt to understand the observed differences as the effect of some *aberrations* (for paradigmatic aberrations, see Ch. 13 for nominals and Ch. 22 for verbs; for segmental aberrations, see Chs. 6–7). In case of success, the target form is treated as an aberrant spellout of the corresponding canonical one. If the form is not representable as an effect of aberrations, it must be treated as corrupt. Hypotheses about its supposed prototype in this case are not controlled by grammar (cf. брѣгру for брѣгу, Mt 8, 32 Sav).

<sup>15</sup> A benchmark task is the content that (1) must be provided by the relevant part of the grammar, and (2) where success can be evaluated relatively formally. A grammar may pursue other aims that are more weakly controlled and lie outside of the benchmark task.

<sup>16</sup> The benchmark list of wordforms gives the graphic spellout.

<sup>17</sup> Morphophonological representations of starting forms are given explicitly in the paradigmatic dictionary (PD). For oblique forms they are produced by paradigmatic synthesis.

# CHAPTER 2 **Graphics and phonology**

# **Alphabet**

### §16. Segment inventory

The alphabet of normalization (i.e. the graphics of canonical OCS) contains the following 40 letters:


The following 16 letters are vowels: а, ꙗ, е, ѥ, и, о, у, ъ, ы, ь, ѣ, ю, ѧ, ѩ, ѫ, ѭ; the rest (24) are consonants. Note that the spellouts Ж, щ, у, ы, as well as л҄, н҄, р҄ each represent a single letter.

### §17. Letter names

Some letters have special names:


First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

The letters л, н, р differ from л, н, р with a diacritic called *kamora*. In manuscripts kamoras sometimes occur over other letters (see details in § 132, 167).

# §18. Simple and iotated vowel letters

Some vowels letters form simple ~ iotated pairs, see Table 18.

Table 18. Simple ~ iotated vowel letter pairs


### §19. Classes of letters

Tables 19.1 and 19.2 show classificatory features that can be used to refer to letters and their groups.1

Table 19.1. Groupings of vowel letters


Table 19.2. Groupings of consonant letters


# **Letter combinations**

### §20. Contextual conditions for vowel letters

The use of vowel letters word-initially, after vowels, and after consonants is regulated in part by the corresponding morphophonological and phonological con-

<sup>1</sup> Here and below tables are numbered as follows: if there is only one table in a paragraph, its number coincides with the paragraph number; if there is more than one, then the table number consists of the paragraphs number and the table's serial number within the paragraph, e.g. 19.1, 19.2.

straints, and in part by purely graphic rules.2 Table 20 shows a general overview of prohibitions in these positions.3


Table 20. Use of vowel letters by preceding context

§21. The distribution of simple and iotated letters

As can be seen from the Table 20, simple and iotated letters stand in partially complementary distribution. This is summarised in Table 21 on p. 14.

In this table, tilde indicates that the letters in the corresponding position stand in contrast, reflecting different phonological strings. Examples:

```
у~ю — утро, югъ, научити, краю (DSg);
```


Table 21. Simple and iotated vowel letters: contextual distribution

# **Phoneme inventories**

§22. Vowel membership

Figure 1 represents all the vowel phonemes of OCS.

Figure 1. Vowels of OCS

The phonological independence of all 11 vowel segments can be seen after a simple consonant; see § 51 below.

# §23. Vowel inventory

The feature set for the vowel inventory contains the following features.


The system of the phonological oppositions is shown in Table 23.


# Table 23. Phonological vowel inventory

# §24. Vowel phoneme classes

The following groups can be distinguished among vowels: front (/ě, e, ь, i, ę/), back (/a, o, ъ, y, ǫ/), *reduced* (/ь, ъ/), also called *yers*, and nasal (/ę, ǫ/). In the inventory, nasal vowels are undefined for height. One may suppose that they do not differ in height from their pure counterparts /e/ and /o/. See details in Ch. 24, *Supplement*, § 866, *Trubeckoj's nonstandard phonology*.

# §25. Phonological vowel pairings

The most important phonological pairings of vowels are shown below in Table 25.

Table 25. Pairings of vowel phonemes


Note that nasal vowels do not form pairings with pure vowels. Between themselves, they form an isolated phonological pairing *ę*‖*ǫ*. Likewise, the phoneme /u/ only forms a pairing with the phoneme /y/.

# §26. A terminological note: long and short vowels

The pairings by the feature inner/outer corresponds to the opposition of *etymologically short* (central) and *etymologically long* (peripheral) vowels, adopted in grammars of Slavic languages. The acoustic and articulatory interpretation of the feature inner/outer is difficult. In particular, in treating alternations, the grades represented by inner vowels are called "short" (the short grade), while those represented by outer vowels are called "long" (lengthened grade). In this grammar, there is no phonological opposition of vowels by length.

# §27. The set and inventory of consonants

The feature set of the consonant inventory contains the following features.


Labial Dental Dentopalatal 1 2 3 Velar Plosive Voiceless Voiced *p b t d št žd c dz č k g* Fricative Voiceless Voiced *s z š ž x* Approximant Nasal Glide *m v n n̕ j* Lateral Trill *l r l ̕ r̕*

Table 27. Phonological consonant inventory

The phonological independence of all 25 consonant phonemes is observed before the vowels /a/, /u/, and /ǫ/; see § 51 below.

# §28. A note on dentopalatals

The acoustic and articulatory interpretation of the ternary feature *affrication*, defined for dental plosives, can be as follows. 1: dentoalveolars; 2: alveopalatals; 3: palatoalveolars.

Of course, the OCS data can hardly speak in favor of one or another acoustic/articulatory interpretation of this feature, but can give evidence of the organization of the system of opposition. The system adopted here ensures relative consistency with the system of pairings (see § 31 below). Those pairings, in turn, agree as much as possible with the observed alternations.

Dentopalatals, which are below referred to as *morphophonologically soft*, form a compact group of consonant phonemes, in opposition to all other consonants. This opposition is widely used in paradigmatics, defining the so-called twofold rule.

# §29. A note on the phonemes č and *g*

Note, on the one hand, the absence of a voiced counterpart of *č* in the system, and on the other, the phonological indeterminacy of the only voiced velar *g*, which can both be a plosive (as in this inventory and grammar), as well as fricative. The first interpretation is consistent with the morphophonological pairings *k*‖*c*, *g*‖*dz*, *x*‖*s*; the second with the morphophonological pairings *k*‖*č*, *g*‖*ž*, *x*‖*š*. See details in Ch. 24, *Supplement*, § 873, *On the types of segmental pairings*.

# §30. Classes of consonant phonemes

Table 30 shows the classes of consonant phonemes and their names that may be used below, but which, perhaps, do not follow directly from the inventory data.


Table 30. Names of some classes of consonant phonemes

# §31. Phonological consonant pairings

The most important phonological pairings of consonant phonemes are shown in Tables 31.1–3.

Table 31.1. The velar~shibilant pairing Table 31.2. The velar~sibilant pairing




### Table 31.3. The dental~dentopalatal pairing

# **Transformation from graphics to phonology and vice versa**

### §32. General considerations

As already noted, phonemes in this book are notated using two equivalent systems, Roman and Cyrillic. The choice is not substantively significant; the dichotomy only serves illustrative purposes.

In some cases the correspondence between phonological and graphic spellouts of an expression is trivial. Cf. graphic дома, phonogical дома or *doma* (or the same in slashes: /дома/, /doma/). Different notations are distinguished by fonts (see Table 33.2 below), while within the same notation, only the meaning of the atomic symbols changes: they are letters in the graphics and phonemes in the phonology.

However, the correspondence between graphics and phonology is not always trivial. Indeed, on the one hand, the standard Cyrillic alphabet has no single letter for the phoneme /j/; this phoneme is rendered using iotated letters. Thus, the phoneme string /moja/ corresponds to the graphic string моꙗ, the phoneme string /mojǫ/ corresponds to the graphic string моѭ etc. On the other hand, iotated letters in the graphics are used after kamorated letters to notate the same phonemes which in other occurrences are transcribed by non-iotated letters. Thus, the phoneme string /l̕u/ corresponds to the graphic spellout л҄ю (cf.л҄юбити), while the phoneme string /lu/ corresponds to the spellout лу (cf. прилучаи) etc.

Thus, for example, the letter ꙗ corresponds either to the single phoneme /a/ (cf. кон҄ꙗ — *kona̕* ), or the phoneme string /ja/ (cf. моꙗ — *moja*). Likewise, the phoneme /a/ corresponds either to the letter а (cf. дома), or the letter ꙗ (cf. кон҄ꙗ).

# §33. Graphics ⇔ phonology transformation rules

These rules are determinate in the sense that a given phonological string corresponds to one and only one graphic string (regardless of the string's content, i.e. which word it represents), and conversely, a given graphic string corresponds to one and only one phonological string.

The rules are shown in two tables: the first (Table 33.1, special cases) is responsible for the treatment of complex strings, i.e. strings that contain iotated letters (in the graphics) or the phonemes /j/ and kamorated phonemes (in the phonology). The second table (Table 33.2, trivial cases, simple alphabetic correspondence) is responsible for the transformation between notations (within a notation the same symbol is interpreted as a letter in the graphics and a phoneme in the phonology).



Table 33.2. Graphics ⇔ phonology alphabetic correspondence: trivial cases

Table 33.1. Graphics ⇔ phonology transformation rules: special cases


§34. From phonological spellout to normalization

All occurrences of the phoneme /j/ must be eliminated; all combinations of the form j+V are replaced with the corresponding iotated letters (cf. *jako*⇒ꙗко, *znaješi*⇒знаѥши, *zmijǫ*⇒змиѭ, *šija*⇒шиꙗ).4 In other cases the vowel phonemes *a*, *e*, *ę*, *u* and *ǫ* are rendered either by the corresponding iotated letter after kamorated Cs (cf. *zemla̕* ⇒земл҄ꙗ), or by a simple letter elsewhere (cf. *duša*⇒душа).

In those cases the phoneme /j/ is simply removed, while the kamorated consonant, of course, remains. Thus, we have: *jako*⇒ꙗко, *zemla̕* ⇒земл҄ꙗ, *duša*⇒душа, *nǫžda*⇒нѫЖа; *jego*⇒ѥго, *pole̕* ⇒пол҄ѥ, *lice*⇒ лице; *jęti*⇒ ѩти, *molę*⇒ молѧ; *jugъ* ⇒ югъ, *n̕eju* ⇒ н҄ѥю, *mǫžu* ⇒ мѫжу, *voždu* ⇒ воЖу,*l ̕ubl ̕ǫ* ⇒ л҄юбл҄ѭ; *mojǫ*⇒моѭ, *zeml ̕ǫ*⇒земл҄ѭ, *dušǫ*⇒душѫ, *rǫkǫ*⇒рѫкѫ.

<sup>4</sup> The phoneme /j/ does not occur in any other context in phonological spellouts; in *j*+V combinations, only the five vowels *a*, *e*, *ę*, *ǫ*, *u* are possible.

# §35. From normalization to phonological spellout

In order to transform a normalization into a phonological spellout, all occurrences of iotated letters must be removed and replaced with the corresponding j+V sequences, if the iotated letter follows a vowel or is word-initial. Otherwise, they are replaced by the corresponding vowel phoneme. For example, for ꙗко we have *jako*, for моꙗ we have *moja*, for земл҄ꙗ we have *zemla̕* , for душа we have *duša*, for кон҄ѥмь we have *konemь ̕* , for знаѥши we have *znaješi*, for ꙗзва we have *jazva*, for мытар҄ь we have *mytarь̕* .

# CHAPTER 3 **Formatives and morphophonological representations of words**

# **Positional classes of formatives**

## §36. Classes of formatives

Positional classes of formatives, on the one hand, determine the possible positions of the formative in the wordform, and on the other, are constrained by the so-called *CVC norm*. The main positional classes are roots (R), prefixes (p), suffixes (s), and terminals (t), or inflections.1 Table 36 shows the positional classes of formatives and their CVC norms.


Table 36. Positional classes of formatives and their CVC norm

\* The count in the Root column is limited to the benchmark corpus of wordforms, i.e. excluding proper names, borrowings, extraparadigmatic items, and compounds.

1 Outside of the benchmark corpus of wordforms there are some formatives that do not belong to these main positional classes. For example, the *connective vowels* о/е in compounds (cf. бог.о.род.иц.а). A separate class comprises so-called *autonomous* formatives, which can both form their own extraparadigmatic wordforms, and combine with each other or with formatives from main positional classes. Such are, for example, the formatives не, ни, же, бо, ли, нъ (cf. such forms as ли, же, ни; не.бо.нъ, ни.к.ого.же). Some other types of auxiliary formatives are possible.

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

Note that, according the CVC norm, all OCS formatives are monosyllabic except the terminals.2

In the CVC norm, as in all other CVC schemata, C stands for a single consonant or a cluster, while V stands only for a single vowel.

The attribution of a formative to a positional class is, generally speaking, a lexical property.3

### §37. Segmental positions of formatives

For each formative, the CVC norm determines its *segmental positions*. So, each segment or sequence of segments is associated with a particular segmental position. Prefixal formatives have two segmental positions, initial C and final V. Root formatives have three, initial C, medial V, and final C. Suffixal ones have two, initial V and final C. Terminals have three, initial V, medial C, and final V. Examples of segmental positions of some formatives are shown in Tables 37.1 and 37.2.

Table 37.1: An example of the segmental positions of root formatives


Table 37.2. An example of the segmental positions of suffixal formatives


Statements such as the following are used below: "The initial C position of the formative твор is replaced with тв, or the initial C position of the formative твор is тв", etc.

Segmental positions of formatives are computed only on the morphophonological representation. In the graphics and phonology they are not observable in the general case. From the graphic spellout of the wordform ичезнѫти one cannot tell that the prefix has a final C (cf. из.чез.н.ѫ.т.и); likewise in обидѣти

<sup>2</sup> See § 295–298 on the decomposition of disyllabic terminals for nominals, and § 611 for verbs.

<sup>3</sup> See a list of the formatives of the basic position classes in Ch. 23 (prefixes, roots, and suffixes), § 289 (nominal terminals), and § 455 (verbal terminals).

for the initial C of the root (cf. об.вид.ѣ.т.и); in издрекѫ, the segment д does not belong to any segmental position (cf. из.рек.ѫ).4

# §38. Nonstandard formatives

CVC norms come with *edge requirements* (or *edge conditions*), which are determined by the oppositions between Initial Cs and Vs, and final Cs and Vs, as shown in Table 38.1.

### Table 38.1. Edge requirements for CVC norms


The CVC norm determines the standard which the majority of formatives of a given position class follow, but a small number of formatives violate. Some violate the edge requirements (such as the prefixes без and из, roots ѣд, зна, дѣ, и, suffixes ѫ, ѣ, а, and terminals хъ, ста). Others violate the length requirements, e.g. чловѣк ‹1077›.5 Others violate both conditions, e.g. the roots алък ‹9›, зм ‹318›, and the suffixes т, н, д.

Formatives which violate the CVC norm of their class are called *nonstandard*. Such formatives occur in all positional classes.

Tables 38.2–38.4 (p. 24) show the computation of segmental positions for nonstandard formatives.


Table 38.2. Nonstandard root formatives

Table 38.3. Nonstandard prefixal formatives


4 Note that in the case of the so-called sonant vocalism, the segmental positions cannot correspond to the CVC schema. See more details in § 123.

5 Numbers in angled brackets, as ‹1077› here, show the root's number in the paradigmatic dictionary and the root dictionary.


### Table 38.4. Nonstandard suffixal formatives

It is important that a formative is treated as C- or V-initial, or as C- or V-final based on its CVC schema, not its CVC norm. For example, a prefixal formative is called V-initial and C-final; root formatives зм ‹318› and j ‹343› are C-initial and C-final.

Of course, in degenerate cases, the notion of segmental position is meaningless. For example, there is no sense in asking which of the C positions is filled and which is empty in the roots j ‹343›, or с ‹276›.6

For roots, it makes sense to distinguish *closed* and *open* roots. Closed roots have a non-empty final C in all alloforms; open roots have an empty final C in all alloforms. *Finally ambivalent* roots (see below) cannot be classified as open or closed in the general case; such a classification can only apply to individual alloforms.

### §39. Ambivalence

Formatives with alloforms that can be either C- or V-initial are called *initially ambivalent*. Such are, for example, the root formatives въп and ъп, suffixes н and ен, and terminals охъ and хъ. Formatives with C- and V-final alloforms are *finally ambivalent*. Such are prefixal formatives съ and сън, root formatives да and дад, клѧ and кльн, and suffixes ов and у. Both types can be referred to as *CVC*-*ambivalent*.

Alloforms of CVC-ambivalent formatives are sometimes morphophonological variants (i.e. they show standard alloformy), and sometimes not, depending on whether the distribution of alloforms follows the CVC agreement rule. If the rule is followed, the alloforms are morphophonological variants; otherwise, they are not morphophonological variants. (See § 92–93 on CVC ambivalence as morphophonological variation).

<sup>6</sup> There is also no need in computing the segmental positions for polysyllabic opaque roots (see § 628), although their final C (if it is present) is determined as usual. For example, in чловѣк, the final C is к; in мъног, it is г.

# Formative sequences in the word: morphological composition of stems

## §40. Morphological composition of stems

The morphological composition of a word or a stem comprises those properties which can be described in terms of presence or absence of formatives or sequences of formatives. For example, simple and prefixed verbs are distinguished by their morphological composition. While these properties can be observed in morphophonological representations, they are properties of the wordform itself, and sometimes of the lexeme (as is the case with the difference between simple and prefixed verbs). The morphological composition of a word or a stem in terms of the positional classes of the formatives that make it up is described by its pRs schema. On constituent structure and on determining the basic components see Ch. 24, § 876, On the morphological composition of stems and wordforms.

### §41. An overview of possible pRs schemata

In a paradigmatic wordform, the possible arrangement of formatives is determined, first, by the natural order of positional classes, viz. p < R < s < t, and second, by limitations on the number of formatives of a positional class, as shown in Table 41.1.

Table 41.1 Possible pRs schemata: a general overview



Table 41.2 Possible pRs schemata: examples

The following are possible departures for the standard shapes of paradigmatic word- forms described above. First, a single wordform may contain more than one root. This is standard in compounds, cf. ( Бог ).о. ( род ).нц.а. In addition, there are several cases of reduplication, e.g. въ.( тръ. тър)).а.т.н <975>, (гла.гол).ъ <159>, из.((мрь.мръ).т.н <539>, and(попел).ъ, (пепел).ъ, оч.(пепел).н.т.н <654>.

Among these, въ.⸨тръ.тър⸩.а.т.и is a borrowing (cf. Gk. τάρταρος); the root does not occur in either of the shapes тръ or тър. On the other hand, гла.гол is related to the root глас ‹162› only etymologically; cf. (глас).ъ and its derivatives. Thus, the single root does not occur in either the shape гла or гол. For измрьмрѣти and попелъ we have forms with the standard shape of the root, cf. мрѣ.т.и and пал.и.т.и. In all these cases, the stem is opaque (on opaque roots and stems, see § 628, *Parsing problems and roots: opaque stems*).

Second, a wordform of a paradigmatic lexeme may lack a terminal in some special cases. This is possible with syncopated щ- and ш-Part forms.7

Third, some terminals, called *bicomponential*, are comprised of two formatives in a row. See § 611 for verbs and § 295 for nominals.

# §42. CVC norm for words

Conditions on the arrangement of formatives by position class (the pRs conditions) and the requirements of the CVC norms of position classes are set up in such a way that formative boundaries in a wordform can only be of the type C.V or V.C, but not C.C or V.V. Thus, the formatives' CVC norm determines the CVC norm for words, as seen in Table 42.

Table 42. CVC norms of words


A word may begin with a vowel if its first formative is nonstandard, such as a V-initial prefix (cf. из.бити, у.мрѣти), or a V-initial root (cf. агн.ьць). However, a word may not end in a consonant, regardless of whether its final formative is a terminal or not.

# §43. Jakobson's law

The law determines the behavior of formative boundaries. As Table 43 shows, the law evaluates boundaries on the basis of their CVC schemata, only allowing contrasting C.V and V.C boundaries.

Prohibited boundaries of the type C.C and V.V are called *heteroformemic clusters* and *heteroformemic hiatuses*, respectively. Clusters and hiatuses within a formative are called *tautoformemic*.

<sup>7</sup> Note that for many wordforms of unique lexemes, their inflectional spellout (i.e., the division into stem and terminal, cf. § 251) is not analyzed.

### Table 43. Jakobson's law


Here, as usual, V stands for a single vowel, and C for either a single consonant or a cluster (see § 36).

### §44. Jakobson's law and CVC norms

Jakobson's law and CVC norm requirements are related in such a way that pRs strings without nonstandard formatives automatically obey the law. In such cases, the morphological composition of a word is for the most part transparent, and observable even in its phonological and graphic spellouts.

For example, consider the form непричѧстьникъ:


Likewise for the word възакониѥ:


### §45. Violations of Jakobson's law

Violations of the law are only possible in case a nonstandard formative intrudes into a licit pRs string, that is, a formative which violates the CVC norm of its positional class—or, more precisely, its edge conditions. For example, нес.т.и, contains a nonstandard C-initial suffix which creates a heteroformemic cluster; зна.еши contains a nonstandard V-final root which creates a heteroformemic hiatus.

However, nonstandard edges can compensate each other. For example, a V-initial root can be compensated by a C-final prefix, cf. из+ити (изити), сън+ѣсти (сънѣсти), etc. Likewise, a V-final suffix can be compensated by a C-initial suffix. Cf. дѣл.а+въш.и (дѣлавъши), etc.

### §46. Hiatuses

The CVC norms of formatives preclude tautoformemic hiatus, while Jakobson's law prohibits heteroformemic hiatus. Nonetheless, OCS occasionally does allow hiatus. Tautoformemic hiatus is encountered in syntagmatic anomalies, e.g. the prefix наи; cf. also borrowings аеръ, геона, архиереи, уалъ, etc. Such forms are not subject to Jakobson's law. The intrusion of a nonstandard formative in a pRs string threatens the possibility of a heteroformemic hiatus.

If nonstandard edges are not compensated, hiatus may occur in the morphophonological representation. In the phonological representation, hiatus may or may not be eliminated, and may occur in such representations. For example, we have да+а.т.и⇒*dajati* (graphic даꙗти), зна+еши⇒*znaješi* (graphic знаѥши). Here, the insertion of the phoneme *j* eliminates heteroformemic hiatus. On the other hand, на+уч.и.т.и⇒*naučiti* (graphic научити), трьп.ѣ+аше⇒*trьpěaše* (graphic трьпѣаше), where heteroformemic hiatus is preserved.8 Figure 2 shows the overview of the cases of hiatus.

A separate class of cases includes those where the phonological representation acquires a hiatus absent in the morphophonological representation. This is possible where a *j*-final formative encounters an initial и or ь. Cf. строј.и.т.и⇒*stroiti*, крај.ь⇒*krai*. See more details in § 63–77.

Figure 2. Overview of heteroformemic hiatus cases and their sources in mph and ph spellouts

# §47. Clusters

Jakobson's law prohibits heteroformemic clusters, but the CVC norm of the formatives does not exclude clusters inside formatives. Tautoformemic clusters are found in formatives of all positional classes: cf. the prefix прѣдъ, roots страд, близ, мѣст, the suffixes ьск, ост, the terminals оста, осте. Some tauto-

<sup>8</sup> See § 63–77 for more details on the treatment of hiatus in the mapping from the morphophonological representation to the phonological one; cf. rules of the form mph⇒ph/norm.

formemic clusters are syntagmatically allowed (e.g. пр, стр, бл, ст, ск, as above), while others are syntagmatically prohibited. Such prohibited clusters are found in *syntagmatic anomalies*, e.g. седморъ. Tautoformemic clusters, however, are not subject to Jakobson's law. Nonstandard formatives that intrude in a pRs string create the possibility of a heteroformemic cluster, but nonstandard edges can be compensated. For example, a C-final prefix may be compensated by a V-initial root, cf. въз+ъп.и.т.и (възъпити); из+и.т.и (изити). A C-initial terminal may be compensated by a V-final root or suffix, e.g. зна+хъ (знахъ), трьп.ѣ+хъ (трьпѣхъ).

If nonstandard edge conditions are not compensated, the morphophonological representation will contain a heteroformemic cluster. However, in the phonological representation, some heteroformemic clusters, namely the prohibited ones, are eliminated. For example, без+цѣн.ьн.ъ⇒*becěnьnъ* (graphic бецѣньнъ), рек+т.и⇒*rešti* (graphic рещи), сла.д+т.ь⇒*slastь* (graphic сласть). The clusters are either eliminated or mapped into allowed clusters.9 Figure 3 shows an overview of cluster cases.

Figure 3. Overview of heteroformemic cluster cases and their sources in mph and ph spellouts

<sup>9</sup> See § 63–77 for more details on the treatment of clusters in the mapping from morphophonological to phonological representations. See rules of the type mph⇒ph/norm.

### CHAPTER 4

# **Phoneme syntagmatics and the mapping rules**

### §48. General

Phoneme syntagmatics considers allowed adjacencies between phonemes inside a unit of a higher rank: formatives and wordforms.1 Phoneme combinations that are allowed in the morphophonology but prohibited in the phonology are eliminated in the mapping between morphophonological and phonological representations, i.e. the mph⇒ph/norm rules that are described in §63–77. Note that syntagmatic statements of the form "the combination is allowed" or "the combination is prohibited" are supplemented by information on *syntagmatic anomalies*, or wordforms that contain prohibited phoneme combinations.

Phonological syntagmatics of OCS can be given by examining phonemes combinations of length 2, because the following recursive rule curtails the *syntagmatic depth* of the observed combinations:

*xyz* — is allowed if and only if *xy* is allowed and *yz* is allowed; *xyz* — is prohibited if and only if *xy* is prohibited or *yz* is prohibited.

### §49. Outline of the presentation

Allowed and prohibited phoneme combinations are described in two steps: we consider first syntagmatic restrictions up to the CVC schema, and second, CVC

1 See more details on the technique of description of phoneme syntagmatics adopted in this grammar in Ch. 24, § 900–909.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

types of combinations that are not generally prohibited up to specific segments or their classes.

# §50. Phoneme sequences in terms of the CVC schema

Table 50.1 shows the syntagmatic properties of CVC combinations. The schemata contain periods in parentheses, which means that the statements apply both to internal combinations with a boundary (*x*.*y*) and to internal combinations without a boundary (*xy*). For edge conditions, the conditions are internal to the wordform; in other cases, the conditions in the wordform and in the formative are identical.


Table 50.1. Phoneme combinations in terms of the CVC schema

Here, not only V stands for a single vowel, but C also stands for a single consonant and not a cluster.

"Allowed" in the table above means that among allowed combinations there are those meeting the corresponding schema. "Prohibited" means that all combinations meeting the corresponding CVC schema are prohibited.

Classes of combinations with additional restrictions, allowances, or anomalies, are treated in more detail below, in the order given in Table 50.2 on p. 33.


Table 50.2. Syntagmatic restrictions

Vowel syntagmatics

# *Sequences of the type xV*

### §51. General table

Complete data on *x*V-type combinations, where *x* is a consonant, vowel, or left word boundary (#), is shown in Table 51 (p. 34). Rows are numbered and correspond to the context (*x*); columns show vowel phonemes (V). Cells show syntagmatic conditions. Each row is divided into two subrows, for mph and ph conditions, respectively.

Row 2, in addition to the phoneme *j*, shows the epenthetic *i̯* for clarity. Epenthetic *i̯*is possible in intermediate working expressions. The syntagmatic possibilities of *j* and *i̯* are identical. Note that just as for other consonants, word-initial combinations with *j* (i.e. combinations such as #jа, #*i̯*а, #jѣ, #*i̯*ѣ, etc.) are accounted for in line 2.

Here plus (+) and minus (–) signs mean that the combinations are respectively allowed and prohibited. A minus with a star (–\*) means that the combination is prohibited but checked; plus with a circle (+°) means that the combination is allowed, although in isolated cases is possible not only at boundaries, but also inside formatives (see more details in § 54).

Combinations with a plus in both subrows are called *absolutely allowed*, e.g. the (6,1) cell. Those with a minus in both subrows are *absolutely prohibited*, e.g. the (6,8) cell. Of the remaining combinations, those with a minus in the ph subrow are called *repairable*, e.g. cell (2,2), (1,2), (3,2). Those with a –\* in the ph row are *unrepairable*, e.g. the (1,1) cell.


Table 51. Allowed and prohibited phoneme combinations of the *x*V type

# §52. Word-initial and postvocalic vowels

All vowels except и are prohibited in both mph and ph (line 1 in Table 51.1). However, all back vowels except ъ and ы are allowed in these positions in anomalous cases both in mph and in ph (cf. cases like агньць), while front vowels are allowed there only in mph (cf. cases like ѧ.т.и [ѩти]). See more details in Ch. 24, § 867.

Below are examples of word-initial vowels.

Back vowels: #а: агньць, азъ, #о: овьца, отрокъ, #у: ухо, убити, #ѫ: ѫтроба, ѫродъ; front vowels: (ѣд).ьц.а [*i̯*ѣд.ьц.а⇒*jadьca*/ꙗдьца], (ес).т.ьств.о [*i̯*ес.т.ьств.о⇒*jestьstvo*/ѥстьство], (ьн).ок.ъ [*i̯*ьн.ок.ъ⇒*inokъ*/инокъ], (ѧ).т.и [*i̯*ѧ.т.и⇒*jęti*/ѩти].

Inside formatives, hiatuses violating their CVC norms are possible only in loans, cf. аеръ from Gk. ἀήρ, and in the nonstandard formative наи, e.g. наитрѣбл҄ии. Heteroformemic hiatuses are possible at boundaries in case a V-final formative meets a V-initial one. Such combinations violate Jakobson's law. They are repaired in some cases (cf. зна.еши⇒*znaješi*), and are preserved in other cases (cf. на.уч.и.т.и⇒научити; нес.ѣ.аше⇒несѣаше). In the latter case the combination is a syntagmatic anomaly. However, V(.)и clusters are not anomalous, since they are absolutely allowed. On the special status of the phoneme и see below in § 57.

# §53. A note on loose formative adjacency

Heteroformemic hiatus is preserved in cases of *loose formative adjacency*.2 Loose adjacencies occur in two cases. First, prefixes form loose adjacencies with the following formative (cf. на#.учити, прѣ#.об#.разити, на#.ѧ.т.и). Second, a loose adjacency occurs where an а-initial imperfect terminal attaches to a V-final formative (cf. плака+ахъ — плак.а#ахъ [плакаахъ], трьпѣ+ахъ — трьп.ѣ#ахъ [трьпѣахъ]). Likewise, a loose adjacency is found in bicomponentialimperfect terminals, such as ѣ.ахъ, ѣ.аше, etc., cf. нес.ѣ#ахъ [несѣахъ]. There is also a loose adjacency in compounds before a root: ⸨мъног⸩.о.(оч).ит.ъ → [мъног.о.#оч.ит.ъ] (for мъногоочитъ); (рѫк).о.(ѧ).т.ь → [рѫк.о.#ѧ.т.ь → рѫк.о.#*i̯*ѧ.т.ь] (for рѫкоѩть).

See an overview of the examples in Table 53.


Table 53. Hiatus examples: tight and loose adjacencies

§54. Vowel phonemes after morphophonologically soft consonants

The rows for morphophonologically soft consonants in Table 51 (rows 2, 3, and 4), while differing from each other, show significant similaries. Let us examine the cells containing +°. We use C\* for the consonants in row 3 (that is, shibilants and kamorated consonants).

2 Initial cycle rules, see § 69.

(1) C\*ѣ. All eight combinations in this cell are morphophonologically allowed and all are checked at boundaries. These combinations are prohibited in the phonology, and are repaired by mph ⇒ ph/norm rules (cf. мол҄.ѣаше⇒ мол҄ꙗаше, мож.ѣаше⇒можааше).

### Special cases of the жахъ type

In some cases, the C\*ѣ combination occurs inside a formative rather than across a boundary. These are forms with the е‖ѣ alternation, forms of the old Aor (жещи): жѣг.хъ⇒жахъ, and secondary imperfectives ичазати [из.чѣз.а.т.и], пожагати [по.жѣг.а.т.и]. In the derivation of the ph representations the combination C\*ѣ is repaired by the same rules as the rules repairing the combinations across boundaries.

(2) C\*ъ and C\*ы. These combinations are morphophonologically allowed, but represented by a few isolated cases (see below). They are prohibited in the phonology and repaired by the mph⇒ph/norm rules (cf. шы.т.и⇒шити, шъв. ен.ъ⇒шьвенъ). See § 72, rule block B2°.

### Special cases

Formative-internal шы and шъ are found in the root ‹1105› (шити). This root has the H(*u*) sonant vocalism with ы‖ъ alternation grades: шы.т.и [шити] and шъ*i̯*.ѫ [шиѭ], шъв.ен.ъ [шьвенъ]. The verb шити in this representation has the same vocalism as the verbs of the group крыти 4h\*⩨⤸.3 (See also Ch. 24, § 874).4 Note that the combination щ.ъ (i.e. at formative boundaries) occurs only in the inflectional spellouts of supines of velar-final 4c class verbs (see § 88).

#### §55. On the phonological independence of phonemes

The position after simple Cs (row 5 in Table 51) is the maximally free position for vowels. All combinations of the form simple C + V are absolutely allowed. This establishes the phonological independence of all 11 vowel phonemes.

The maximally free position for consonants is before the vowels а, у and ѫ (columns 1, 9, and 10 in Table 51). All C+а, C+у, and C+ѫ are absolutely allowed. This establishes the phonological independence of all 25 consonant phonemes.

<sup>3</sup> The notation 4h\*⩨⤸ represents the lexeme's paradigmatic index; see § 257.

<sup>4</sup> See more details in Ch. 24, § 900–909. Cf. also the morphophonological spellouts земл҄.а (and not земл҄.ѣ) with spellouts such as ног.а or жен.а. Whenever there are no paradigmatic or morphophonological grounds to the contrary, mph representations by default contain the unmarked vowel (cf. жаб.а and not жѣб.а, час.ъ and not чѣс.ъ, жити and not жыти, etc.). Note also that the combination jь, while morphophonologically allowed and repairable, is checked at boundaries and not checked inside formatives. It is, however, only represented by the opaque stem оjьм ‹633›.

# §56. The law of the velars

All velar+front vowel combinations are absolutely prohibited (see Table 56). This combination, whenever it appears in the morphophonological representation (e.g. in borrowings such as герьгесиньскъ), cannot be repaired by mph⇒ph/ norm rules, and remains in the phonological representations, yielding a syntagmatic anomaly (see more details in Ch. 24, § 869).

# Table 56. The law of the velars


§57. On the partial complementary distribution of vowels along the front-back dimension

The four vowel pairs, а~ѣ, о~е, ъ~ь, and ы~и enter into opposition by advancement. Vowels in all these pairs are in complementary distribution in most positions. In particular, they are in complementary distribution after velars (see row 6 in Table 51.1), and after morphophonologically soft consonants except ц, ѕ, and j (see row 3 in Table 51.1). The pair а~ѣ has special behavior: these two vowels contrast after ц and ѕ (cf. отьца vs. тѧжьцѣ, but are neutralized word-initially and postvocalically, and after other morphophonologically soft consonants (see rows 1–3 in Table 51.1). (See more details in Ch. 24, § 870).

The fact that the phoneme и, unlike other front vowels, is absolutely allowed wordinitially and postvocalically, but prohibited in the second (ph) subrow after j, is due to the fact that OCS has no phonological opposition between и and jи. Forms without j (и) and forms with j (jи) are chosen in morphophonological representations based on paradigmatic or morphophonological considerations. Cf. j.ихъ [ихъ] (root j ‹343›), but иг.о [иго] (root иг ‹331›); or краj.и [краи], but бу.и [буи]; строj.и.т.и [строити], but сто.иши [стоиши] for стоꙗти, etc.

# *Sequences of the type Vx*

### §58. Absolutely allowed and repairable sequences

Among combinations of the type V*x*, where *x* is a consonant or a right word boundary, all are absolutely allowed except two: the combinations ъj(*i̯*) and ьj(*i̯*) represent a special case in that they are prohibited in the phonology but allowed in the morphophonology, not only across boundaries but within formatives. These combinations are eliminated in the mapping from morphophonological to phonological representations. (See more details in Ch. 24, § 900).

## Special case

The combination ьj occurs widely within formatives in morphophonological representations. First, some roots contain it, cf. бити, бьj.еши ‹23›, вьj.а.л.иц.а ‹95›. Second, it is found in the frequent suffix ьj. Cf. бож.ьj.ь, въз.нес.ен.ьj.е, etc. The suffix ьj also occurs as the first part of bicomponential nominal terminals, cf. л҄юд.ьj.е (NPl).

The combination ъ(*i̯*) (but not ъj) occurs in irregular verbs of class 4 of the крыти 4h\*°⩨⤸ group (cf. кръ*i̯*.еши, шъ*i̯*.еши), as well as in the combined declension (terminals 2-combi), cf. нов.ъ*i̯*.ь [новыи].

# Consonant syntagmatics (clusters)

## §59. General table

The general Table 59 (p. 39) shows combinations of the type c1c2, where c1 and c2 are single consonants. Rows and columns are numbered; rows correspond to c1 and columns to c2.

The table is separated into two subtables, the upper rows 1–4 and the bottom rows 5–9. The upper subtable's rows have two subrows for mph and ph. The lower subtable does not show separate subrows for ph, since what is prohibited in mph cannot be allowed in ph. The lower subtables contains absolutely prohibited clusters.

A minus with an asterisk means that the combination is prohibited but found in syntagmatic anomalies.5 Plus and minus in a cell mean that the combinations are allowed or prohibited, respectively. All cases with stars are commented upon in § 62 below.

The symbol S\* (found in the 5th row and 5th column) abbreviates the class of shibilant and sibilant affricates—that is, ч, ж, ш, щ, Ж, ц and ѕ.

### §60. Data abbreviations in the general table

Table 59 does not include the following information.6

First, the table does not include the consonant phonemes j, л҄, н҄, and р҄. The situation with these cases is as follows:

1) all combinations with j—both Cj and јC—are absolutely prohibited;

6 Recall that clusters are subject to a recursive rule that allows us to limit the description to binary clusters; see above in § 48.

<sup>5</sup> The minus in the (2,3) cell does not take into account the unique formative гда (cf. къгда, тъгда, вьсегда, etc.). Since these forms are extraparadigmatic, they are not part of the benchmark list of wordforms.


Second, the table does reflect the phonological prohibition against binary consonant clusters disagreeing in voicing. Phoneme combinations that disagree in voice but are allowed in the first subrow (mph), are repairable.7


Table 59. General table of c1c2-type clusters

# §61. On allowed and prohibited clusters

As is clear from the Table 59, the vast majority of clusters are absolutely prohibited, as is the case with the entire lower subtable, rows 5–9. In the upper subtable all clusters are allowed morphophonologically (there is a plus in the mph subrow—36 such groups of clusters). Of these, 21 groups are prohibited in the phonology, and are repairable, or partly repairable—that is, are preserved in *potentially correctable anomalies*; such cells are marked by a minus with an asterisk.8 Note that *geminates* are prohibited in the phonology but allowed in the morphophonology, cf. по.мет.т.и [помести].

7 Note that sometimes such combinations are repaired by rules that do not refer to voicing. For example, сла.д.т.ь⇒сласть (see below § 74, rules of type c1c2⇒c0c2).

8 Anomalies occur in tautoformemic clusters—i.e. those that occur within a formative, and not by heteroformemic ones that are created by a C-final formative meeting a C-initial one. For example, the cluster дм is not repaired in седм.ор.ъ [седморъ], and represents a potentially

All clusters that are allowed in the phonology are checked. Among clusters that are allowed in the morphophonology but prohibited in the phonology, there may be unchecked ones, that is, those that are not represented in any wordform from the benchmark list. Such are the clusters мн and хт. The cluster мн is absolutely allowed; cf. the potentially possible охръмнѫти. The cluster хт is allowed in the first (mph) subrow, but prohibited in ph representations (хт is repaired by the rule хт⇒щ; cf. the potentially possible врѣх.т.и [врѣщи], cf. Church Slavic врѣщи, врьхѫ 'thresh'). On the other hand, among absolutely prohibited clusters there are those that occur in a small number of uncorrectable anomalies. Such is the cluster цв in the anomalous root цвѣт ‹1050› (cf. цвѣтъ, цвисти). All anomalous items are root formatives.

Note also that some aberrant spellouts may contain some clusters prohibited in ph representations. Such is the aberrant infinitive блисцати; its canonical analog is блистати. Such cases arise also as a result of aberrant alternations (cf. л҄юдьсции NPlmPlen for canonical л҄юдьстии), or aberrant boundary adjustment rules (cf. бесчьстиѥ for canonical бечьстиѥ). See more details in § 76 and Ch. 24, § 872.

### §62. Lists of syntagmatic anomalies

Syntagmatic anomalies contain prohibited tautoformemic clusters. The formative—more specifically, the root formative—is the carrier of such anomalies. The complete list of roots with prohibited clusters is given here.

Most of the roots in the list contain absolutely prohibited clusters (such as, for example, жр, which is prohibited both in mph and in ph). However, others contain clusters that are allowed in mph but prohibited in ph, such as тл, дл, and ср, зр. Such groups of roots are marked with the sign ‡. The anomaly of these roots consists of the fact that ph-prohibited clusters are not repaired, i.e. are not subject to mph⇒ph/norm rules. Such clusters are *potentially correctable anomalies*. That is, before applying mph⇒ph/norm rules, it is necessary to check whether a root belongs to the list of roots not subject to mph⇒ph/norm rules (that is, one of the groups marked with ‡).

correctable anomaly, while the cluster д.м is repaired in руд.м.ѣн.ъ [румѣнъ]. Likewise, the cluster т.л is repaired in мет.л.ъ [мелъ] (съмести), but not in тлъст.ъ [тлъстъ] (отлъстѣти). All formatives that contain such potentially correctable clusters are listed below in § 62. (See more on potentially correctable and uncorrectable phoneme combinations in Ch. 24, § 900–909).

Group 1: Cл, Cр combinations


Group 2: isolated combinations

This group contains the root седм ‹814› (potentially correctable), and the roots цвѣт ‹1050›, ѕвѣзд ‹305›9 (both uncorrectable).

It is noteworthy that the well-populated group 1 (combinations of the Cл, Cр type) contains roots without an etymological cluster, since the р and л in the OCS cluster is the C-initial version of the sonant.10 From a synchronic point of view, sonant vocalism is set up for a limited number of roots, namely those that show CVC ambivalence, e.g. зрь ‹309›, cf. зьрѣти but зрьцало, or those that show alternations particular to sonant series, such as тлѣк ‹962›, cf. тлѣщи, тлькѫ (see details in § 677).

# The mapping from the morphophonological to the phonological representation

### §63. General

Morphophonological representations of wordforms are mapped to phonological or graphic ones by the mph⇒ph/norm rules. These rules are called

<sup>9</sup> The roots цвѣт ‹1050› and ѕвѣзд ‹305› etymologically descend from roots with the combinations \**kv*, \**gv*: \**květ*, \**gvězd*. The sound changes *kv* > *cv* and *gv* > *dzv* is part of the so-called second palatalization (the change from velars to sibilants).

<sup>10</sup> Of course, for opaque roots this etymological interpretation is uncertain.

mph ⇒ ph/norm because, in practice, it is usually necessary to deal with mph~graphics pairs, and because the rules relating phonology to graphics are trivial. However, the analytical burden is on the mph⇒ph component of the rules; mph⇒ph/norm rules can be understood as the composition of mph⇒ph rules (see below in § 69–77) and ph⇒norm rules (see above § 34). A ph spellout can be derived with only the mph⇒ph rules (cf. *zna.n.ьj.a* ⇒ *znanija*). To derive the graphic spellout (norm), one must first apply the mph⇒ph rules, and then apply ph⇒norm rules to the output (*znanija* ⇒ знаниꙗ).

Note that mph ⇒ ph/norm rules are related to syntagmatics in that they repair combinations that are allowed in the morphophonology but prohibited in the phonology.

### §64. A note on internal and external expressions

The input to mph⇒ph/norm rules is the mph representation of a wordform, and produce as output its ph representation. The mapping of a particular mph representation into a ph representation is shown by its *derivation* of the form *A* → *K*1 → *K*2 → … → *K*<sup>n</sup> → *B*, where *A* is the input expression, *B* is the output expression, and *K*<sup>i</sup> are intermediate results. Each transition between adjacent expressions in this sequence is effected by one of the rewrite rules. The beginning of the derivation (the *input expression A*) and the end of the derivation (the *output expression B*) are *external expressions* which are tightly controlled by the grammar. Inputs are morphophonological representations of wordforms, which are determined by the dictionary and the rules of paradigmatic synthesis. Outputs are phonological representations of those wordforms. The syntagmatics of inputs and outputs is described in terms of allowed and prohibited combinations (§ 51, 59).

The intermediate results—the expressions *K*1, *K*2, …, *K*n—are internal expressions. These are strings that mix morphophonological and phonological objects with auxiliary symbols. The structure of internal expressions is not controlled by grammar directly. Any expression that can be derived from a mph representation using mph ⇒ ph rules is considered to be a valid internal expression. Such expressions need to be cited in describing the application of mph⇒ph rules. To avoid confusion, such expressions are given in angled brackets, cf. ⟨#отъ#ѣд.ъ⟩. Mappings between intermediate espressions *K*i and *K*j are shown with a single arrow → (the double arrow links the input and the output). Note that the expression *K*<sup>i</sup> → *K*<sup>j</sup> does not necessarily mean that *K*<sup>i</sup> and *K*<sup>j</sup> are immediate neighbors in the derivation.

### §65. Cycles and blocks

The rule system contains three cycles: the initial cycle (blocks A.I and A.II), the main cycle (block B), and the final cycle (block C). The rules of the initial cycle examine the boundaries between formatives and in some cases introduce auxiliary symbols, namely # (showing loose formative adjacency), and epenthetic *i̯*. Rules of the main block B effect various segmental changes. Rules of the final block C remove any remaining auxiliary symbols. All rules except those of the initial cycle are *segmentally decidable*. This means that the rules require no information outside of the string, and the context depth does not exceed 2, that is, they only require information about the symbol being worked on and either the immediately preceding or the immediately following one.

Rules of the initial cycle use information on the membership of formatives in the class of prefixes (given by a list, see § 640), or to a certain set of terminals (namely, а-initial imperfect terminals, see § 455). Information about formative boundaries (the period) is only necessary on the initial cycle, to the extent that the formatives of the string being examined are compared against a list.11

A rewrite rule can (1) replace one or two symbols with a symbol, (2) introduce a symbol, or (3) remove a symbol.

### §66. How to apply the rules

Rules apply in order by block. Only when the rules of a block are exhausted can the rules of the next block apply. The order of the blocks is A.I, A.II, B, C. The procedure is as follows.


In other words, each rule applies to a given string as long as it has substrings to which the rule is applicable. For example, in the string зна.ѫщ.а.его we apply the rule of epenthetic *i̯*-insertion, namely (VV)⇒(V*i̯*V) (see below § 70, rule A.II.2°), which produces ⟨зна*i̯*ѫщ.а.его⟩. This rule then applies to the string ⟨зна*i̯*ѫщ.а.его⟩, giving ⟨зна*i̯*ѫщ.а*i̯*его⟩. The resulting string has no substring to which the rule is applicable. We then move on to other rules. In block B, we find the rule of epenthetic *i̯*-removal (§ 77), and apply it twice: ⟨зна*i̯*ѫщ.а*i̯*его⟩<sup>→</sup> ⟨знаjѫщ.а*i̯*его⟩ → ⟨знаjѫщ.аjего⟩.

If the symbol # introduced by #-insertion rules is non-initial, cutting the string into two substrings, all subsequent rules apply as if the string contains two independent wordforms. For example, from the string отъ.(ѣд).ъ we get

<sup>11</sup> Parentheses and periods can be removed at any point of the derivations; the symbol # is removed last; see below §77. In the examples below periods are removed in an affected substring after the first successful application of a rule to the input string. However, the algorithm can be organized differently. In some cases, periods are preserved to make the examples more readable.

<sup>12</sup> At the first step it is an external string; however, in subsequent iterations, step 1° can apply to internal strings as well.

⟨#отъ#ѣд.ъ⟩. The substring ⟨#отъ#⟩ remains intact—it contains no substring to which any rules apply. In the substring ⟨#ѣд.ъ⟩ we apply the rule of epenthetic *i̯*-insertion (see below, rule A.II.1°), giving ⟨#ѣд.ъ⟩ → ⟨*#i̯*ѣд.ъ⟩. Next, ⟨*#i̯*ѣд.ъ⟩ → ⟨*#i̯*ад.ъ⟩ (rule B.1°). Next, ⟨*#i̯*ад.ъ⟩ → ⟨*#*jад.ъ⟩ (epenthetic *i̯*-removal, block C). The resulting string remains intact, as it has no substrings that match any rule. Only now all instances of the auxiliary symbol # can be removed, and we get the final result: the form *otъjadъ*, or normalized отъꙗдъ.

# §67. Unchecked segment sequences and anomalies

Unchecked segment strings (which is the majority of strings) belong to two non-intersecting classes: unchecked and allowed in mph representations, and unchecked and prohibited in mph representations. In the first class are strings such as кк, пж, or кпк. In the second class are strings such as л҄о, цы, or пжв. Unmarked strings that are allowed in mph representations but prohibited in ph representations must be repaired by rewrite rules along with checked strings. In other words, the left part of rules contains strings that are never found in OCS wordforms. For example, we have rules that repair the clusters кк or пж: кк⇒к; пж⇒ж. (see the rule c1c2⇒c2 below, § 74).

# A technical note

Potentially correctable anomalies are not subject to rewrite rules. In other words, before applying the rules to a wordform, we must determine whether it belongs to the class of anomalies. The list of anomalies by default contains not wordforms, but lexemes or even roots that are carriers of anomalies. For example, the root седм ‹814› is represented by five lexemes: седмь, седмъ, седморъ, седморица, седмица. All wordforms of all five lexemes preserve the anomalous cluster дм.

# §68. Rewrite rule format

A symbol or a string of symbols in parentheses is subject to replacement. The context symbol is in curly braces. A single symbol is a phoneme or one of the auxiliary symbols (# or *i̯*).13 If a rule has no expression in curly braces, the symbol in parentheses is rewritten in all its occurrences.

An expression of the form (*x*){*y*}⇒*z* means: replace *x* with *z* if *x* is immediately followed by *y*.

An expression of the form {*x*}(*y*)⇒*z* means: replace *y* with *z* if *y* immediately follows *x*.

An expression of the form (*x*)(*y*)⇒*z* means: replace immediately adjacent *xy* with *z*.

13 A period is not a symbol in this sense: it cannot be part of segmental rewrite rules.

A single expression may contain several symbols separated by commas in parentheses or braces. This means that several rules are abbreviated by a single expression. Thus:

The expression (*a*, *b*){*y*}⇒*z* is understood as two independent rules: (*a*){*y*}⇒*z* and (*b*){*y*}⇒*z*; The expression (*x*){*m*, *n*}⇒*z* is understood as two independent rules: (*x*){*m*}⇒*z*  and (*x*){*n*}⇒*z*; The expression {*m*, *n*}(*x*)⇒*z* is understood as two independent rules: {*m*}(*x*)⇒*z* and {*n*}(*x*)⇒*z*. Finally, an expression like (*a*/*b*){*y*}⇒*q*/*r* should be understood as the following two rules: (*a*){*y*}⇒*q* and (*b*){*y*}⇒*r*.

Rule format given by the cluster table is explained in § 74.

# Rules of the first cycle: the A.I and A.II blocks

# §69. Block A.I: insertion of the # symbol


### §70. Block A.II: epenthetic i ̯-insertion


### Note

Epenthetic *i̯* can be thought of as a symbol that improves the CVC schema of the formative, bringing it into line with the CVC norm at the beginning of a nonstandard V-initial root, at the end of a V-final root, suffix, or terminal (in the latter case, as part of bicomponential terminals of the adjective declension, but not as part of the bicomponential terminals of the imperfect). In the dictionary, in the morphophonological spellout of starting wordforms, *i̯* is included, even though it does not belong to the morphophonological representation. This is done for convenience.

# §71. Examples of rule application of the initial cycle

In Table 71, the input is the morphophonological representation; the output is an internal expression that results from the application of rule of the initial cycle A. In the examples, for the reader's convenience external expressions preserve parentheses that mark roots.


Table 71. Examples of initial cycle rules

# Rules of the main cycle: the B block

# Segmental rewrite rules applying to combinations with vowels

# §72. Block B: partial neutralization of vowel contrasts and absorption of *j* and epenthetic i ̯


B.2°. Partial neutralization of the ы~и and ъ~ь contrasts

{л҄, н҄, р҄, ч, ж, ш, щ, Ж}(ы/ъ)⇒и/ь ⟨#шъ*i̯*.еши⟩→⟨#шь*i̯*.еши⟩, ⟨#шъв.ен.ъ⟩→⟨#шьвенъ⟩, ⟨#шы.т.и⟩→⟨#шити⟩ The phonemes ы and ъ appear in mph representations after morphophonologically soft con-

sonants only in special cases. See above § 54; cf. also the supine forms, see § 88.



# *Segmental rewrite rules applying to clusters*

# §73. Cluster rewrite rules: rule B.5°

Two groups of rules apply to clusters in order. First, rules removing voiced+voiceless sequences apply (see B.5° below), and then all other rewrite rules (see rule B.6°, § 74, and Table 74.2).


## §74. Cluster rewrite rules, main table: B.6° rules

All clusters in the mph representation are subject to inspection against the main Table 74.2. Some are evaluated as allowed (marked as +); these undergo the rules vacuously, cf. зб⇒зб, as in ⟨#из.би.т.и⟩→⟨#из.би.т.и⟩. Others are prohibited in ph representations. For these clusters, the table shows the type of rule that the clusters undergo in the mapping between mph and ph representations, or between internal expressions. The rules are classified according to the following types, shown in Table 74.1.


Table 74.1 Cluster rewrite rule types

Table 74.2 (p. 49) is the main table. Rows show the initial members (c1) of the clusters in the mph representations, and columns show final members (c2). The cell for row *x* and column *y* shows the rule type that applies to the cluster *xy*, if *xy* is allowed in mph representations but prohibited in ph representations. Otherwise—that is, if *xy* is allowed both in mph and ph representations, the cell contains a plus sign. For example, for об.вид.ѣ.т.и, the cluster is бв. In the corresponding cell (1,9) we find c1, which means that the cluster (allowed in mph but prohibited in ph) must be replaced by its first member, in this case б. Thus, we have бв⇒б; ⟨#об.вид.ѣ.т.и⟩→⟨#обид.ѣ.т.и⟩. For rules that insert a phoneme differing from both c1 and c2, marked as c0, the content of c0 is explained in the comments below the table.14

Note that the rules shown in the main Table 74.2 do not take into account some special cases discussed separately in § 75.

<sup>14</sup> Note that some rules are inert, because the corresponding clusters, being allowed in the morphophonology, are not checked or checked only among anomalies which are not subject to mph ⇒ ph/norm rules. For example, there are no examples subject to ср⇒стр rules, since срѣда, срамъ, etc. are in the list of anomalies; see above § 62. Likewise, there are no examples subject to the rule гд⇒Ж, or to the rules кк⇒к, пб⇒б, and many others. See data on clusters attested in OCS in Lunt 2001, § 2.522.


### Table 74.2. Main cluster table: rule B.6°

In the cell (2,3): c0 = щ/Ж (г+д⇒Ж, elsewhere щ). In the cell (3,3): c0 = с.

In the cell (4,8): c0 = т/д (с+р⇒стр, з+р⇒здр).

S\* in column 5 stands for shibilant and sibilant affricates (ч, ж, ш, щ, Ж, ц, ѕ).

#### §75. Special cases of cluster tranformation rules

Here we state some necessary and optional addenda to the rules in Table 74.2.

Cell (1,6) of the main table

All three clusters in this cell (пн, бн, мн) are absolutely allowed, and all are checked, cf. об.новити, гыб.нѫти, истоп.нѫти, утоп.нѫти, тим.н.а. No such cluster is found inside any formative. The peculiarity of these clusters is that they are sometimes preserved and sometimes eliminated in the mapping between mph and ph, not being prohibited. With prefixes, the clusters are eliminated in two lexemes, об.(нѣм).ѣ.т.и⇒онѣмѣти, and об.(нѣмл҄).а.т.и⇒онѣмл҄ꙗти. It is preserved in other cases. With the suffix н.ѫ the situation is more complex (see details in Ch. 24, § 871). Clusters outside of these two situations (stems of нѫ verbs and prefixes) are eliminated. These other cases are: (1) пн with the root съп ‹918› (без.съп.н.ьj.е [бесъниѥ], съп.н.ьj.е [съниѥ], съп.н.ъ [сънъ]); (2) мн with the root тим ‹952› (тим.н.ьн.ъ [тиньнъ], тим.н.ав.ъ [тинавъ], тим.н.а [тина], cf. тимѣно); (3) combinations with the root усм ‹1017› (усм.н.ьj.ѣн.ъ [усниꙗнъ], cf. усмѣнъ).

#### Cell (3,1) of the main table

Here, all six combinations are allowed in mph and prohibited in ph. Of the six theoretically possible combinations, only three are found: тм (врѣт.м.ѧ⇒врѣмѧ), дм (плед.м.ѧ⇒племѧ, руд.м.ѣн.ъ⇒румѣнъ), and тп (от.про.(врѣг).т.и⇒опроврѣщи). The rule does not apply to the potentially correctable anomalous root formative седм ‹814› (see the list in § 62).

Cell (4,5) of the main table

Here all 14 combinations are allowed in mph and prohibited in ph. Of these, only four are found: зч (из.чит.т.и), зж (из.жи.т.и), зш (раз.шир.и.т.и), and зц (без.цѣн.ьн.ъ). The rule eliminates the first consonant: из.чит.т.и⇒ичисти, раз.шир.и.т.и⇒раширити, без.цѣн.ьн.ъ⇒бецѣньнъ. This rule does not apply to зж, where the special rule of the type c1c2⇒c0 applies, namely зж⇒Ж: из.жи.т.и⇒иЖити, раз.жег.т.и⇒раЖещи, въз.жел.а.т.и⇒въЖелати.

### Cells (3,7) and (4,8) of the main table

These cells contain four combinations: TA, AA, cp, and 3p. Of these, three are found across boundaries: TA (oct.n.a), дл (раз.пад.л.нн.а), 3р (въз.раст.т.и). Rewrite rules remove the first consonant in the combinations from the (3,7) се11: ост.л.а >осла, раз.пад.л.нн.а = распалина. In the cell (4,8), а consonant is inserted, viz. co = crp and 30 > 3 go. Thus, we get BB3.pact.r.H > BB3 gactry. These rules do not apply to roots from the list of potentially correctable anomalies (see the list in § 62).

### \$ 76. Aberrant cluster rewrite rules

Aberrant spellouts may show results of alternative, aberrant cluster rewrite rules. They are the following.

1°. Alternative rewrite rules for clusters with affricates


E.g. I привъсъ I къ оученикомъ твоимъ и не могж его исцелити Mt 17, 16 MAR (cf. H Привъс и кь оученикомъ твоммъ и не взмогоша ицелити и Mt 17, 16 AS); БЕС'ЧЕСТИГА ВО ИСПЛЬ(НЕНЕ); НЕ СТЪРДИШИ (Л)И СА ПОВИСПОДЪЯНИ ТЕМЬНУВИ ДНГАВОЛЕ SUPR 74, 25-27 (cf. СЪЕРАША СЪБОРЪ ДА НЕ БОГА СЛАВАТЪ" НЪ ДА христосоу вечьстию сътворат' supr 385, 11-12); кде свъшта и мечи' и говори БЕШТИСЛЬНИИ' КДЕ ЛЮДИЕ И ГИЕВАНИЮ SUPR 448, 12-14 (сf. ВИДЪАХЖ БО ЕГО БЕЧІСЛЬНОЮ БЪШЕНЬЮ SUPR 565,26-27); С ПОЖТЪ ВЪТООЗІ ЖЕНЖ' Г ТЪ ФУМОВТЪ EEUTAAA Lk 20, 30 ZOGR. Note also the spellouts истълитъ SUPR 115, 6 and icertaenut ps sin 37, 4 and 37, 8.

See details on combinations with prefixes in Ch. 24, § 872.

2°. Alternative rewrite rules for clusters with velars


Cf. 3anob Egbmi TBolmi приспы намъ помоць твов (KIEV 6b, 14-16) for canonical помошть [по.мог.т.ь].

3°. Alternative rewrite rules for combinations with kamorated Cs in second position1s


С. просити милости ог тебе оумръцивьшааго съмръть: авльшааго свътъ омраченъ кусн 39b, 4-7 (for canonical оумрьтвльшають from оумрьтвити);

<sup>15</sup> The symbol C stands for the consonant resulting from the substitutive softening alternation. See details in § 112-118.

ДЮБАІ МА: ВЬЗЛЮБЛЕНЪ БЯДЕТЪ ОЦЕМЬ МОСМЬ И АЗЪ ВЪЖЛЮБЛЯ І' І ВВАНЯ СА EMOY CAMB IO 14, 21 ZOGR (Cf. BL3AKBENER in the same verse in AS, BB3AKOENO in MAR); TBOM CI CANYKLE'SH BENNENEHIM KIEV Sa, 2-4; likewise for canonical EE3 HIEro, H3 METO We have such aberrant spellouts as EEER TTEML БЪЩА: С БЕЖНЕГО НИЧЬТОЖЕ НЕ БЪКТЪ: ЕЖЕ БЪСТЪ IO 1, 3 ZOGR, AS; 1 ПОНДЖ КЪ СВИ' I ОБОЕТЪ СЪДАШТА ЧКА' IЖНЕГОЖЕ БЪСИ L3ИДЖ' ОБЛЬЧЕНА L СЪМЪКЛАШТА' при ногог своу Lc 8, 35 ZOGR (Cf. the same verse in MAR 13 мегоже въси изндж).

Because combinations with kamorated Cs most often result from the substitutive softening alternation, the corresponding aberrant forms can be understood also as resulting from substitutive softening in its alternative, stronger version (for double substitutive softening mph effect see § 118).

Note that aberrant rule application may generate prohibited clusters. These are: сч, сц, што, ждо, шл, жд, шн, жн, штвл.

Rules of the final cycle: the C block

§ 77. Deletion of auxiliary symbols


# CHAPTER 5 **Polyformy of formatives**

# **Alloformy of formatives and types of alloformy**

### §78. Monoformemic and polyformemic formatives. families of formatives

Some formatives of a single position class are set up as *alloforms* of each other. The alloformy relation partitions the set of formatives into non-intersecting classes, which are called *families of formatives*, or their *assortments*. A family may contain one, two, three, etc. formatives.1 Formatives of families with more than one member are called *polyformemic* (or *polyvariate*); otherwise they are *monoformemic* (or *monovariate*). For example, among prefixal formatives, без is monoformemic, while от is polyformemic, with two alloforms, от and отъ. Among root formatives, зна ‹319› is monoformemic, while рек ‹766› is polyformemic, with alloforms рок, рьц, and others.

The partitioning of formatives among families is given by list, not controled by any formal procedure.2


FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

# §79. A note on intersections between formative families

Generally speaking, families of formatives do not intersect, so one formative (or one string of phonemes) in most cases represents only one family of formatives. There are relatively few violations of this principle. For example, there are two root formatives вар: вар1 in варити1 'anticipate', вар2 in варити2 'boil, cook'. A similar example occurs with the terminal и ⟨NPl⟩, e.g. in гради, and the terminal и ⟨2–3SgImv⟩ e.g. in неси. Among prefixal and suffixal formatives, there are almost no intersections of this kind (see isolated cases in § 843).

# §80. The internal organization of families

Families of polyvariate formatives are quite diverse. Here are families of three root formatives.


Clearly, the internal organization of the ‹535› and ‹537› families are more similar to each other than each of them is to the organization of ‹766›.

§81. A note on the goals of investigating varieties of formative families3

Investigation of the variety of formative families can be reduced to the following goals: (1) classification of families into particular types; (2) discovery of the arsenal of devices generating the observed families. From the point of view of grammar proper, the most important goal is to uncover mechanisms that govern the distribution of different alloforms of the same formatives, that is, to find the correlation between, on the one hand, oppositions between forms in terms of alloformy, and on the other, oppositions between forms in terms of some other grammatical characteristic. Traditionally, this task is distributed among *inflection* (oppositions of forms within a lexeme), and *derivation* (oppositions of forms of different lexemes). Thus, we can talk about two sub-goals: (3) investigation of polyformy in inflection (i.e. in the paradigmatics), and (4) investigation of polyformy in derivation.

<sup>3</sup> The general remarks below are limited to roots. Extending these ideas to other position classes of formatives requires some technical adjustments. For roots, the partitioning into families is controled by factors external to morphophonology, such as semantics and etymology, much more strictly than for other position classes.

In this grammar we limit ourselves to the goals (2) and (3). Excluding the goal (1) allows us to avoid the difficulties that result from the random incompleteness of the benchmark list of lexemes. Excluding the goal (4), in turn, allows us to avoid distinguishing active rules from random facts that are due to tradition. Limiting the goals in this way is consonant with the main tool for representing the variety of polyformy, namely alternations. Alternations both serve the needs of paradigmatics, and allow an overview of polyformy, which is manifested both in the variety of the generated paradigmatic forms of one lexeme, and in the variety of fossilized derivational forms.

# §82. Standard types of alloformy

Any two formatives of one family are alloforms of each other. However, certain pairs of alloforms have tighter connections than others. For example, рек and реч are more tightly connected than рок and рьц. In the family ‹844› (слу, слов, слав, славл҄, слы), the alloforms слу and слов, слав and славл҄ are more tightly connected than, for example, the alloforms слов and славл҄.

Some pairs of alloforms derive their tight connections from comparisons of the segmental content of equivalent positions (such are the pairs рек and реч, слав and славл҄ given above), while others do so due to their particular distributions.

Those types of alloformy that are observed in a significant number of pairs are treated as standard.

Table 82 shows the *standard types of alloformy* and their names.


Table 82. Standard alloformy types

Segmental alloformy is treated as standard only if it satisfies one of the established alternations. For examle, the alloformy of the root alloforms жьд‖жид is standard, because the pairing ь‖и is among the pairings of fundamental vowel alternations. On the other hand, the root alloformy сѣд‖сѧд is nonstandard: there is no alternation containing the pairing ѣ‖ѧ. See more details in Ch. 24, *Supplement*, § 873, *On the types of segmental pairings*.

A single formative family may show different standard types of alloformy, and may combine standard and nonstandard ones. For example, in ‹844› слу/слов shows CVC ambivalence, while слав‖славл҄ shows consonantal instability. Each of these standard types of alloformy is discussed in more detail below.4

### §83. A note on the distribution of alloforms

Polyformemic formatives raise the question of distribution. The term *distribution* is used in two senses. On the one hand, a distribution is the totality of data on the occurrences of *one single* alloform; let us call this *particular distribution*. On the other hand, a distribution is the result of comparing the particular distributions of *two* (or more) alloforms of a polyformemic formative; let us call this *comparative distribution*.

Usually we distinguish three types of comparative distributions: 1) complementary distribution, 2) free variation, and 3) contrast. For comparative distributions, we establish the condition of distribution: it might be the segmental context, the grammatical properties of the wordforms that realize a particular formative, or some other properties of the observed occurrences of the alloforms in question. Of course, classifying a particular comparative distribution according to one of these three types does not entail that each particular distribution should be subject to the same conditions.

Although such a partion would be desirable, the three distribution types are not mutually exclusive. The set of all occurrences of a pair of alloforms may be divided into two subsets in such a way that one of the subsets shows one type of distribution (e.g. complementary), and the other shows a different type (e.g. contrast). For example, рѫк and рѫч contrast by segmental context (cf. рѫк.ѫ and по.рѫч.ѫ), but the alloform рѫк is not found before front vowels.

Luckily, this grammar does not aim to describe comparative distributions of alloforms of polyvariate formatives. However, the typology of the comparative distribution is the basis for distinguishing types of standard alloformy, viz. morphophonological variation vs. standard segmental alloformy. However, as can be seen from Table 82, possible formal imprecision of the distinction between comparative distribution vs. contrast (see the last column) does not preclude a precise distinction between the resulting classes of the partion; these classes and their names are given in the first column of Table 82. Note that contrast is defined as follows: two linear units *A* and *B* are in *contrast* if (1) *A* and *B* are not free variants of each other5 and (2) there exists at least one context where both *A* and *B* are possible. There is no requirement that the set of contexts of *A*

<sup>4</sup> In order to determine whether a pair of alloforms are linked by standard alloformy, it is sufficient to test all cases of standard alloformy discussed below. If the pair does not satisfy any of the four cases, its members are not linked by standard alloformy.

<sup>5</sup> Formally, in this grammar no pair of alloforms are in free variation. Factually such a relation could be found in such pairs as нѫд and нуд ‹621›, ав and ꙗв ‹1›, or with suffixes ен and ѣн. In this grammar, such facts are treated lexically: in some cases we establish lexical doublets, e.g. for авити//ꙗвити. In other cases, one of the lexical variants is treated as aberrant, e.g.

and *B* be identical. For example, in Russian, the phonemes /n/ and /r/ contrast, because there is *nos* 'nose' vs. *ros* 'grew', *nad* 'above' vs. *rad* 'happy', even thought there is *drug* 'friend' but not \**dnug*.

# **Morphophonological variants**

§84. A terminological note on morphophonological variants

Two alloforms in complementary distribution by segmental context are called *morphophonological variants* of each other. These are the alloforms of the first two standard types: twofold rule (see the members of this set below in § 90), and CVC ambivalence (see members of this set below in § 95).

Twofold rule

# §85. Properties of the variants

Twofold rule applies to pairs of formatives that are alloforms and that are distinguished only by their initial V. Thus, twofold rule is only possible for suffixes and terminals. Alloforms of twofold formatives are called alloforms (or variants) of hard and soft subtypes,6 as shown in Table 85.

Table 85. Initial vowels of hard and soft subtypes of twofold formatives


Such are, for example, the terminals ъ/ь (NSgm, cf. град.ъ and кон҄.ь); ѣхъ/ихъ (LPlmn, cf. град.ѣхъ and кон҄.ихъ); the suffixes ы/ѧ (the suffixes of щ-participles, cf. нес.ы and плач.ѧ); от/ет (cf. наг.от.а and тъщ.ет.а). In the spellout of twofold formatives, the overslash variants are *hard subtypes*, while the underslash variants are *soft subtypes*.

for нѫдити and нудити. In other cases, several lexical variants are considered to be equally possible derivational alternatives, as for the suffixes ен and ѣн see § 896.

<sup>6</sup> The conventional names of these variants have nothing to do with the phonetic and phonological opposition between "hard" and "soft" sounds, which applies to consonants. Recall that this grammar does not contain the feature hard ~ soft in the inventory of consonant phonemes.

### §86. The twofold rule

This rule examines the segmental content of the preceding formative as shown in Table 86.1 (p. 58). Examples are shown in Table 86.2.

Table 86.1. The twofold rule


Table 86.2. Examples of the application of the twofold rule


### §87. A technical note on the application of the twofold rule

A formative preceding the one to which a rule is being applied can itself be polyformemic. For example, one of its alloforms can end in a morphophonologically soft consonant and the other end otherwise. In the examples above, this is the case in {враг, враж, враѕ} before the terminals ъ/ь and ѣхъ/ихъ, and for {плак, плач} before the suffixes ы/ѧ. Here, order of application matters.

1. The wordform враѕ.ѣхъ is derived from a morphological skeleton of the shape "workstem + terminal", in this case враг+ѣхъ. The rule that builds the morphological skeleton consists of two parts: creating the workstem (in this case враг, as in the input), and selecting the appropriate terminal (in this case, the twofold rule determines the choice of the overslash ѣхъ given ѣхъ/ихъ). The resulting morphological skeleton (враг+ѣхъ) is mapped into враѕ.ѣхъ by boundary adjustment rules (replacement by velar palatalizaton; see below in § 105).

2. Cases like плач.ѧ are different. The wordform плач.ѧ is derived from a morphological skeleton of the shape "workstem + suffix", in this case плач+ѧ. The rule that builds the morphological skeleton consists of two parts: creating the workstem (in this case плач, as in other forms of the PRAE system from the base stem плак by substitutive softening; see below § 112), and selecting the appropriate terminal or suffix (in this case, the twofold rule selects the underslash ѧ given ы/ѧ). The resulting morphological skeleton (плач+ѧ) is mapped into плач.ѧ.

Generally, in paradigmatics (as well as in stem formation, which is not treated in detail in this book), there is never need to apply rules to expressions where formatives are represented by their families, even though sometimes one needs to replace an input alloform by an alternative one, as in workstems (e.g. from the basic плак to плач), or to select from several morphophonological variants, as for terminals or suffixes, e.g. ѣхъ/ихъ, or ы/ѧ.

# §88. A note on the supine forms of velar‑final verbs

Consider the supine of the verb рещи 4. Supine forms, as other forms of nominal subparadigms of the verb, have the morphological skeleton of the form [workstem INF-AOR + nominal suffix] + nominal terminal. In the case of the supine, the nominal terminal is the twofold ъ/ь. For рещи we get [рек.т]+ъ: by the twofold rule, we select the overslash variant, because the stem ends in a т, which is not a morphophonologically soft consonant. We get the morphophonological representation рек.т.ъ. By the mph⇒norm/ph rules, we get рек.т.ъ→рещ.ъ→рещь. Note that the final ь in the graphics does not indicate a twofold terminal, and yet there is no monovariate terminal ъ. On the other hand, setting up a monovariate terminal ь (from the 1-simplex set) for the same form would contradict such supine forms as нестъ, сѣꙗтъ, etc. Cf. облещь Lk 24, 29 Mar and As; въврѣшть Mt 10, 34 Mar (въврѣщъ in the text).

# §89. Special cases

The twofold rule is violated by terminals beginning with о/е in the declension of loans. Vowel-final workstems get the hard subtype of the terminal, cf. архиереови (i.e. архиере+ови) instead of архиере+еви, which is expected by the twofold rule. Cf. also архиереомъ (i.e. архиере+омъ) instead of архиере+емъ. Cf. also the suffix ов in архиереовъ, etc. However, there may be aberrant forms where this correction does not apply (cf. мосеемъ Mk 9, 4 Mar).

# §90. The list of cases

Twofold rule applies only to suffixes and terminals. Almost all nominal terminals of twofold declension types, and the terminals of secondary personal dative (D2), are twofold; cf. the catalog of nominal terminals in § 289. In the verb, imperative terminals and the suffixes of м- and щ-Part in the *e*-conjugations are twofold (cf. the catalog of verbal terminals and suffixes in § 455). Among the remaining suffixes, the following are twofold:7 {10} ој/еј, {16} ък/ьк, {20} ъл/ьл, {24} ом/ем, {29} ън/ьн, {32} ор/ер, {36} ост/ест, {38} ъств/ьств, {42} от/ет, {43} ът/ьт, {43} ъш/ьш, {55} ов/ев.

# §91. A note on new twofold rule

In aberrant forms there may occur terminals from the nonstandard *new twofold* set (2-duplex), which contains new twofold terminals that result from the union of two monovariate ones, e.g. ъмь/ьмь (from ъмь and ьмь). See more details in Ch. 13, *Aberrant nominal forms in sources*, § 388.

Some twofold suffixes show special departures from classical distributions in their canonical forms. This is the case, for example, for the suffix ък (льг.ък.ъ,

<sup>7</sup> The number in curly braces refers to the suffix's number in the suffix catalog; see § 841.

but тѧж.ьк.ъ). These departures are called *secondary twofold rule* below (see Ch. 23, *Formative inventories: prefixes, roots, suffixes*, § 863, *A note on secondary twofold rule*).

# CVC-ambivalence

# §92. Properties of CVC-ambivalent variants

CVC ambivalence as morphophonological variation is defined as follows: two formatives that are alloforms are CVC-ambivalent if, first, they differ either in the initial or final position of their CVC schema (initial and final ambivalence, respectively), and second, if they are distributed according to the CVC agreement rule.

In a small number of roots and suffixes there exist pairs of CVC-ambivalent alloforms that are not distributed by CVC agreement, i.e. are not morphophonological variants. Such are, for example, the root alloforms да and дад ‹216› (cf. даꙗти [да.а.т.и] and дадѧтъ [дад.ѧтъ]), выч and уч ‹132› (cf. навыкнѫти [на.вык.н.ѫ.т.и] and научити [на.уч.и.т.и]); the suffixes ств and ъств/ьств (cf. бѣство [бѣг.ств.о] and божьство [бож.ьств.о]).

# §93. The CVC agreement rule

This rule examines the CVC schema of adjacent formatives as shown in Table 93.1. Examples of corresponding ambivalent formatives are shown in Table 93.2 (p. 61).

Table 93.1. The CVC agreement rule


§94. A note on the application of the CVC agreement rule

When a finally ambivalent formative meets an initially ambivalent one, there are two ways to achieve agreement, and the choice between them is regulated not only by the CVC agreement rules, but by the rules of paradigmatic synthesis as well. Cf. клѧ/кльн+охъ/хъ (1SgAor) gives клѧ.хъ, not кльн+охъ, but клѧ/кльн+ъш/въш gives кльн.ъш, not клѧ.въш (ш-Part is кльнъши), because the paradigmatic synthesis rules prescribe a V-final variant in the aorist system, but a C-final variant in the imperfect system (see § 427).


### Table 93.2 Examples of the CVC agreement rule application

### §95. The list of cases

Morphophonological variation of the CVC ambivalence type is found in all position classes.

The following prefixes have final CVC ambivalence: въ/вън, съ/сън. The alloforms от and отъ show final ambivalence that is not distributed by CVC agreement; cf. отъврѣщи, отврѣсти.

Among roots, final CVC ambivalence is found first of all in the so-called *labile roots*, i.e. roots with sonant vocalism whose CVC alloforms are distributed by the CVC agreement rule (see § 128, *A terminological note: labile roots*). A CVC-ambivalent root is called labile if it has sonant vocalism (see § 128). Such are, for example, the roots кла/кол ‹364› (кла.т.и, кол.ъ); слу/слов ‹844› (слу.т.и, слов.ѫ).8 Initial CVC ambivalence is found in a relatively small number of formatives, e.g. ъп/въп ‹129› (cf. въз.ъп.ити, въп.ити). (See details in the lists in § 682–784).

Among suffixes, initial CVC ambivalence is found in the participial suffixes of the imperfect system (н/ен for н-Part, (ъш/ьш)/въш for ш-Part; see §455). Among other suffixes, the following are initially ambivalent: {35} с/ес, {44} х/ъх; the following are finally ambivalent: {55} (ов/ев)/у, {56} ъв/ы, {57} ѧ/ен, {58} ѧ/ѧт.

Finally, among terminals, standard aorist ones have initial CVC ambivalence (хъ/охъ, cf. зна.хъ, нес.охъ, etc.); cf. § 455.

Examples of CVC ambivalence not distributed by the CVC agreement rule in root and suffixal formatives are found above in § 92.

<sup>8</sup> The only exception is the anomalous root (и, ид) ‹329›. Its alloforms are distributed by the CVC agreement rule, but the vocalism is pure.

# §96. Special cases


# §97. A note on new CVC ambivalence

Newly developing CVC ambivalence is found in aberrant forms of prefixes, which in canonical OCS are monoformemic, e.g. with the prefix из. Cf. aberrant forms such as изъчитати Supr 179, 18, represented as изъчитаѧи; изъчезеSupr 487, 3 (represented as изꙿчезе), but ищезошꙙ Supr 450, 2, etc.

# **Segmental alloformy**

# Alternations as a tool for describing segmental alloformy

# §98. General

Alternations are a tool for describing segmental alloformy. It is important that the same system of alternations is responsible both for *intraparadigmatic* alloformy (i.e. different alloforms as part of the same paradigm, e.g. богъ, боѕѣLSg; чисти, чьтѫ1SgPrae), as well as *interlexemic* alloformy (e.g. different lexemes with the same root, e.g. льгъкъ and польѕа ‹514›). An alternation is an object made of segmental pairings10 that ensure representation of families of formatives as ordered structures that can be easily compared with each other. Certain formal connections are established between alternations and the formatives that carry those alternations, which make it possible to represent alloforms (members of a family of formatives) in a table defined by an alternation (see below Tables 102.1 and 102.2).

## §99. Alternations

An *alternation* is a table whose columns are *series* (see below the columns *q*1−*q*4) and whose rows are *grades* (see below the rows *d*1−*d*3). Cells are filled with segments or strings of segments, which are the *undergoers* of the alternation. Strings of segments that are undergoers in an alternation are called *quasisegments*.

Table 99 shows a fragment of the OCS velar palatalization alternation.

<sup>9</sup> Non-uniform behavior of this suffix can have an etymological explanation, but within OCS it is more appropriate to unify all suffixal н and ен under one family of formatives.

<sup>10</sup> See details in Ch. 24, § 873.


Table 99. An alternation table

### §100. An order on the set of grades: the adjacency relation

Grades of an alternation are partially ordered: some grades are defined as adjacent (in this case, *d*1 and *d*2 are adjacent, as are *d*1 and *d*3). The adjacency relation is directional: in two adjacent grades, one is the *leading* and the other the *following* grade. In this case, *d*1 leads *d*2 (written ad *d*1>*d*2), and *d*1 leads *d*3 (*d*1>*d*3). Some pairs of grades may not be comparable in terms of adjacency.

Paradigmatic *grade replacement rules* apply in terms of the adjacency relation. For example the rule *d*1→*d*2 defines the replacement к→ч, ск→щ, г→ж, and х→ш.

### §101. Segmental position tied to an alternation

Each alternation corresponds to a particular *segmental position* that is *tied* to that particular alternation. Grade replacement rules prescribe replacements only in the segmental position tied to that alternation. For example, the replacements к→ч, ск→щ, г→ж are only applied to the final C segmental position of the corresponding formative.

The OCS alternation system contains two consonantal alternations with the final C segmental position, namely velar palatalization (see § 105–111), and substitutive softening (see § 112–118), and one vocalic alternation with the medial V segmental position, namely fundamental vowel alternations (see § 119–128).

### §102. Alternations and alloformy

Let us assume as obvious the following wording, setting aside formal definitions.

"Alloforms *G* and *H* represent the alternation *A*" (by the series *q*n in the segmental position *t*). For example, alloforms рок‖роч represent the velar palatalization alternation, as do the alloforms рок‖реч, рѫк‖рѫц, мог‖мож, мог‖моѕ, etc.

"Some family of formatives represents the alternation *A*" (by the series *q*n in segmental position *t*), or "The alternation *A* is represented in some family of formatives", or more briefly, "in a formative" (by its series *q*n in segmental position *t*). For example, in the formative family of the root ‹766› (рек), the velar palatalization alternation is represented (by its series *q*1 in the final C segmental position).

If two formatives represent the same alternations, their families can be shown in comparative tables, as in the following examples in Table 102.1.

Table 102.1 compares four roots that show the same alternation. For the root ‹69› (бѣг), the alloform бѣѕ is not attested.


Table 102.1. Comparative tables 1

Table 102.2 compares two roots, ‹535› (мог), and ‹537› (мок), which shows two alternations: velar palatalization and fundamental vowel alternation. Alloforms in square brackets are not attested.

Table 102.2. Comparative tables 2


The *domain* of an alternation is the set of formative families that represent it. An alternation is called *free* if its domain can be defined in terms of the segmental properties of formatives; otherwise it is *limited*.

In OCS, all consonantal alternations are free, and all vocalic ones (the fundamental alternations) are limited.

### §103. Alternations and intraparadigmatic alloformy

Paradigmatic synthesis rules include grade replacement rules by a particular alternation. Rules of the form *d*1→*d*2, where *d*1 and *d*2 are grades of some alternation *A*, require that in some selected formative (i.e. the last one) of some expression, the segment *x* of the grade *d*1, which is found in the segmental position *t* tied to alternation *A*, be replaced with the segment *y* of the grade *d*2, which is the correlate of the segment *x* by alternation *A*. For example, in deriving nominal forms, the replacement rule of the velar palatalization alternation requires the grade к to be replaced with the adjacent grade ц (cf. ног+ѣ⇒ноѕ.ѣ; рѫк+ѣ⇒рѫц.ѣ; дъск+ѣ⇒дъст.ѣ, etc.). In the derivation of the PRAE forms, the replacement rule of the velar palatalization alternation requires the replacement of the grade к with the adjacent grade ч (cf. рек+еши⇒реч.еши, мог+еши⇒мож.еши; иск+еши⇒ищ.еши, etc.). Replacement rules apply both in building the workstems and in boundary adjustment rules. They cause the transition from a given alloform of a formative to some other alloform.

### §104. Alternations and interlexemic alloformy

Because there is no algorithm for synthesizing families of lexemes sharing a root, or for word-formation synthesis, the role of alternations outside paradigmatics is simply that of a tool which, on the one hand, allows a comparative view of root formative families, and on the other hand, controls the boundaries of such families. This control is achieved by requiring that a formative family only contains those formatives that represent standard alloformy types (i.e. are morphophonological variants, or represent some alternation).11 An important filter is, on the one hand, isolating the segmental position tied to each alternation, and on the other, the series isolation condition.

Restricting the domain of an alternation to a particular segmental position prevents the inclusion in one family of such etymologically related formatives as, for example, град ‹192› and жрьд ‹293› (in Sadnik's dictionary these formatives are alloforms; see «253»). The replacement of grades in these two formatives is found in the initial C position, while velar palatalization is tied to the final C position.

### Series isolation condition

If an alternation *A* is represented in a formative family by a series, e.g. the series *q*i, then the alternation *A* cannot be represented in the same family by a different series *q*<sup>j</sup> distinct from *q*i.

In other words, the series isolation condition prohibits formatives from being in the same family if they have a segmental position occupied by undergoers of the same alternation belonging to different series. For example, the formatives рож and рок cannot be alloforms; likewise, ду ‹259› (дунѫти) and дѫ ‹268› (дѫти) cannot be alloforms, since у and ѫ belong to different fundamental vowel alternation series. The only violation of this condition is found in the root ‹496›: the formatives лих and лѣк, treated in this grammar as belonging to one family, do not show standard segmental alloformy. Indeed, the segments х and к are the same grade of different series of the velar palatalization alternation. The lexemes отълѣкъ and лихъ are etymologically related (cf. Vasmer, *lixoj* and *olek*).

11 Recall that the alloformy relation is given a priori; in this book, this is done with the list of roots and the catalogs of functional formatives. Therefore, the statements discussed here are simply conditions on the correctness and completeness of the grammar; in some cases we must admit violations of these conditions. For example, the formatives гън and жен are alloforms (root ‹153›), even though they do not show standard alloformy. Likewise, there is no standard alloformy in terminals sharing a name, e.g. ⟨3SgPrae, итъ⟩ and ⟨3SgPrae, етъ⟩, and some others. The terminals итъ and етъ are paradigmatic variants; cf. § 259. The question of whether paradigmatic variants are alloforms is generally scholastic, at least for a grammar that does not use the notion of morpheme. See discussion on alloformy types in Plungjan, p. 53 and ff.

# Velar palatalization

§105. The alternation table

# This alternation has 6 series and 3 grades (Table 105).



\* One form found: дрѧздѣ Supr 12, 8 (as дрꙙздѣ).

\*\* The only forms are NPl влъсви and Voc влъшвеSupr 111, 21 (a single gloss).

# §106. The tied segmental position and the domain

The segmental position is final C. The domain includes all roots and suffixal formative families that contain a formative whose final C position is occupied by a к undergoer of this alternation. This means that the alternation is free.

# §107. Names and notation for series and grades

The series are named after the participants of the first grade: к, г, х, ск, зг, хв. Likewise for grades: к, ч, ц.

# §108. Adjacent grades and replacement rules

Adjacent grades are к>ч and к>ц. Two replacement rules apply in the paradigmatics: к→ч (replace the к grade segment with the ч grade segment), and к→ц (replace the к grade segment with the ц grade segment).

A note on the ц‖ч and ѕ‖ж pairings

The pairings ц‖ч and ѕ‖ж, being morphophonological, are not phonological and are not adjacent by the velar palatalization alternation. However, these pairings form an anomalous replacement ц→ч and ѕ→ж, which applies in the same environment as the standard replacement by adjacent grades in the velar palatalization alternation, such as к→ч. This anomalous replacement is added to the standard replacement in boundary adjustment rules of vocative forms. The forms Voc отьче for отьць, Voc кънѧже for кънѧѕь contain anomalous replacements, while Voc вльче for влькъ contains a standard replacement. Anomalous replacements are sometimes responsible for interlexemic alloformy; see examples in § 109. Note that the replacements ц→ч and ѕ→ж are also among the standard replacements of the substitutive softening alternation.

## §109. Alloformy

Here are some examples.

Intraparadigmatic alloformy: рѫк‖рѫц: рѫка, рѫцѣ (LSg for рѫка); ик‖иц: мѫченикъ, мѫченици (NPl for мѫченикъ); дъск‖дъст: дъска, дъстѣ (LSg for дъска); мог‖мож‖моѕ: могѫ (1SgPrae), можетъ (3SgPrae), моѕѣте (1PlImv), which are forms of the verb мощи (мог.т.и).

Interlexemic alloformy: мрък‖мрьц: помръкнѫти, помрьцати; х‖ш: грѣхъ, грѣшьнъ.

Many formatives show all three grades: дъск‖дъщ‖дъст: дъска, дъстѣ, дъщица; блиск‖бльщ‖блист: блискати, бльщати, блистати; ик‖ич‖иц: мѫченикъ, мѫченичьскъ, мѫченици.

In interlexemic alloformy, there can be derivationally related pairs whose basic and derived alloforms are not adjacent, e.g. отьць and отьчьство (the alloforms ьц and ьч).

# §110. A note on velar palatalization

In the tradition of Slavic linguistics, replacements by the к→ч rule are usually treated as the so-called *first palatalization*, while replacements by the к→ц rule are called *second* or *third palatalization*. These palatalizations are conceived as historical sound changes that deposit synchronic alternations, to the extent that the discussion focuses on historical or comparative issues. Accordingly, such grammars formulate segmental conditions that distinguish these three palatalizations (see more details, e.g. in Lunt 2001, § 29, p. 193 and ff.).

# §111. Alternative pairings in the velar palatalization alternation

Along with the ск→ст replacement by the standard ск‖ст pairing, a frequent aberration is the ск→сц replacement by the alternative ск‖сц pairing. Note that сц is a prohibited cluster. See the distribution in sources in Vaillant, § 41, p. 82.

## Illustrations

ютру же бывъшу· съвѣтъ сътворишѧ вьси архиереи и старьци людьсции на иса· ѣко убити и Mt 27, 1 Mar (cf. утру же бывъшу· съвѣтъ сътворішѧ вьсі архиереи и старци людьстіи на иса· ѣко убіті и Mat 27, 1 As); кънѩѕи людъсциі събърашѩ сѩ съ бмъ авраамлемъ Ps Sin 46, 10 (cf. ѣко бъ спсетъ сіона· ї съзіЖѫтъ сѩ граді їюдѣісті Ps Sin 68, 36); въ црствии небсцѣемь Mt 8, 11 Mar (cf. въ цѣсарьствѣ небесьстѣѣмъ Supr 67, 19–20); cf. likewise for паска: бѣ же параскевьћи пасцѣ· година бѣ ѣко третиѣѣ· ꙇ гла ꙇмъ· се црь вашь· они же въпиѣхѫ· възьми възъми пропьни і· Io 19, 14–15 Zogr (cf. бѣ же параскевћіи пастѣ· годіна же бѣ ѣко третіѣа Io 19, 14 As).

# Substitutive softening

# §112. The alternation table

This alternation has 20 series and 2 grades (Table 112).


Table 112. Substitutive softening


# §113. The tied segmental position

The segmental position is the final C. The domain of the alternation includes all root and suffixal formative families that contain C-final formatives ending in undergoers of the C° grade of the alternation. Thus, this alternation is free.

# §114. Names and notations for series and grades

Series are named after the undergoers of the first grade: п, б, etc. Grades are called C° (simple), and C• (substitutively soft).

# §115. Adjacent grades and replacement rules

The adjacency relation is as follows: C°>C•. In the paradigmatics, there is a replacement rule C°→C•.12

A note on the ц‖ч and ѕ‖ж pairings

The pairings ц‖ч and ѕ‖ж, being morphophonological, are not phonological, even though they are adjacent by the substitutive softening alternation. If we remove from the velar series of substitutive softening the ц‖ч and ѕ‖ж pairings, what remains is velar paltalization without the ц grade. In paradigmatics, the ц‖ч and ѕ‖ж pairings by substitutive softening are necessary for paradigms of

<sup>12</sup> Note that in case the paradigmatic synthesis rules require replacement by substitutive softening of a formative-final consonant that is not in the C° grade, the consonant remains unchanged, i.e. the rules apply vacuously. So, by boundary adjustment rules, we have куп+ѫ⇒купл҄.ѫ, but луч+ѫ⇒луч.ѫ (for 1SgPrae of купити and лучити, respectively).

verbs like °рицати (Prae °ричѫ), but the functional load of these pairings (and the corresponding standard replacements) is limited to several dozen forms, of which only a few are actually found.13 In interlexemic alloformy, the ц‖ч pairing is widely represented: cf. коньць and бесконьчьнъ; троица and троичьнъ, троичьскъ, etc. (the ѕ‖ж pairing is represented by isolated cases: кънѧѕь and кънѧжь; пѣнѧѕь and пѣнѧжьникъ; помиѕати and помѣжити). However, interlexemic alloformy is not generated by the ц‖ч and ѕ‖ж pairings by substitutive softening, but rather by the anomalous ц→ч and ѕ→ж replacements by velar palatalization. Indeed, when the starting forms end in simple consonants, there are no replacements, cf. зѫбьнъ, водьнъ, квасьнъ, рабьскъ, градьскъ, пьсьскъ, etc.14

### §116. Alloformy

Here are some examples.

Intraparadigmatic alloformy: топ‖топл҄: утопити, утопл҄ѭ (1SgPrae for утопити), съл‖съл҄: сълати, съл҄ѥши (2SgPrae for сълати); выс‖выш: възвысити, възвышааше (2–3SgImf for възвысити).

Interlexemic alloformy: въп‖въпл҄: въпити, въпл҄ь; зем‖земл҄: подъземиѥ, земл҄ꙗ; от‖ощ: поработити, порабощати; д‖Ж: сладъкъ, услаЖати. See also § 864, *A note on segmentless substitutive softening*.

### §117. Alternative pairings in the substitutive softening alternation

In aberrant spellouts there are forms that result from replacements by alternative pairings in the substitutive softening alternation. The following are groups of aberrant pairings.



Illustrations

дазь намъ· вьсемогы бже· да ѣкоже нъи есі небесьскъиѩ \*піцѩ насъитілъ· \*такозе же ꙇ животъ нашь сілоѭ твоеѭ утврьді (Kiev 4b, 2–6); подазь намъ просімъ тѩ вьсемогъиꙇ бже· блаженъиѩ раді мѫченіцѩ твоеѩ феліцітъи въкупьнѫѭ молитвѫ· ꙇ \*тоѩзе раді \*зашчіті нъи (Kiev 2a, 12–17); да исправі

<sup>13</sup> These are class 3 verbs whose truncated stem ends in ц or ѕ. These are °лѧцати 3 ‹526›, °рицати 3 ‹766›, °трьѕати 3\* ‹973›. Alloforms ending in -ч, -ж can be expected only in the PRAE system forms, i.e. °лѧцати, лѧчѫ, °рицати, ричѫ, °трьѕати, трѣжѫ. There are only 26 such forms found in sources, of these 24 in Supr for the verb °рицати, and лѧчѫтъ Supr 328, 24 (written as лꙙчѫтъ), and трѣжетъ Supr 520, 2. Cf. 10 verbs of class 7 ending in -цати, -ѕати: °клицати ‹383›, °кльцати ‹387›, °мрьцати ‹567›, °выцати ‹132›, °двиѕати ‹222›, °жиѕати ‹279›, °миѕати ‹548›, °стриѕати ‹894›, °сѧѕати ‹933›, °тѧѕати ‹1006›.

<sup>14</sup> Although the etymological sources of these pairings are well known, their interpretation from a comparative-historical point of view is not completely clear. In any case, etymological data does not help the synchronic understanding of OCS data. Cf. Lunt 1974, § 3.30 and ff.

N'TH L ONICTI' HE HAWING ATENS DAAI' N'E OETETA TROELO DAAI LINE ECI "ORBELLEAR намъ (кіеv 3b, 8-12); въсжда твоего гі насъицені просимъ тъ»: отъ въсъхъ \* противьяция съ съя намъ съпаси нъи (клет ба, 22-6b, 2); просимъ ти дазь НАМЪ' ДА СВЕЖТ'ЪІ ТВОІ ВЪСЖДЪ "ПОЕМЛІЖЦЕ ДОСТОІНІ БЯДЕМЪ \*ОЧШЧЕНИЪ ТВОЕГО (KIEV 3a, 10-13); BB KOHELT RAONZ TECHI \*OCBIALE(NET) AOMOY JABA (PS SIN 29, 1); гръхъ юності мобы и \*невъзєства моего не поміни (рѕ SIN 24, 7). Сf. по средув ДЬНЪСЪ ЖІВЪІМЪ І МОЪТВЪІМЪ ЪВЛЕЊА СА" СОУГОУБЯ ПАКЪІ ГЛА ЖІЗНЬ' СОУГОУБО 0036 TB0 (CLOZ 13b, 39-14a, 1; here the correction жд is written above the letter 3 in the word posseTBo), and also the contaminated spellout визжь in по томь гла томъ принеси пръста твос е го съмо с визжь ряц в мон ( Jn 20, 27 мАR).

Cf. likewise in SUPR: заштицаа (352, 7 and 412, 17) alongside заштиштати SUPR 407, 29-30.

Starred spellouts are lexical aberrations; see more details in Ch. 24, Supplement. Excursus on aberrations, § 892, Lexical aberrations.

2°. Alternative pairings among labials

The following aberration is a degenerate case of alternative pairings in the substitutive softening alternation.


Usually the cases are seen as lacking the expected epenthetic l (i.e. this aberration is understood as an atomic one of the type l' = 0, see § 139-141).15

#### Illustrations

и пристяпь едниъ къмижьникъ рече емогу одчителю сдя по тевъ: ъможе колиждо гдеши Mt 8, 19 ZOGR (cf. пристяпь in the same verse in мак, but пристяплъ in AS); не оставых вась сиръ поидж к валь Io 14, 18 AS 30a (сf. не оставлия вась сирь придж кь вамъ Io 14, 18 AS 93d); любан же ма' вьзлюсенъ БЯДЕТЪ ОЦЬМЬ МОІЛЬ IO 14, 21 SAV (сf. a ЛЮБАН МА ВЬЗЛЮБЛЕНЪ БЯДЕТЪ ОТЦИЪ MOHNL IO 14, 21 MAR); DBLER ABAN SUPR 240, 14-15 (cf. MBAR CA EMOY CAN'S Io 14, 21 SAV); w ofe rabbeno mn E'bi' tako hanagehing whiche our possenning a ПОСТРАДАВЪ: ХОШТЕШИ СЪ СТАВПА СЪЛЪСТИ SUPR 558, 8-10.

Cf. also: и препловъше: придж въ земых генисарефь Mt 14, 34 SAV (cf. пришъдъ IC въ земля геръгєснномъ Mt 8, 28 SAV); in PS SIN: 0T' зємььк 80, 11, \* 3€ NEW 21, 30 and our 3€ 3€ MAIM 80, 6 for canonical 3€ MAM 16, 11, \*3EMM 77, 55, 3EMAX 79, 10 and BB 3EMAR 84, 10 for canonical 3EMAR; Na \*3EMH 54, 21 and въ земли 77, 40 for canonical землн.

<sup>15</sup> In the Slavic linguistic tradition, epenthetic I refers to those instances of the phoneme l' that occur after labials reflecting earlier labial+j sequences. These are, first of all, labial+l' sequences that result from substitutive softening, including asegmental substitutive softening, cf. âювити - люслях, зємла. Less commonly epenthetic I is found after root-initial consonants, cf. пльвати, клюдъ, etc.

Starred spellouts are lexical aberrations; see more details in Ch. 24, Supplement. Excursus on aberrations, § 892, Lexical aberrations.

### § 118. Double substitutive softening morphophonological effect

In some cases the substitutive softening replacement rule of the C° → C° type applies to a final C position occupied by a cluster rather than by a single consonant. In such cases, the double substitutive softening effect is possible. First, the final consonant of the cluster is replaced by standard substitutive softening: CC → CC ; we have, e.g., съклазния 1SgPraefor съелазнити; мъклія 1SgPrae for мъслити. Next, the resulting cluster, which is generally allowed, is subject to alternative cluster rewrite rules by the CC® > C°C° schema. So, from съблазняя we get съвлажня; from мъклия we get мъщаяя (see § 76 on aberrant cluster transformations; note also that such aberrant rules can apply to produce aberrant spellouts such as EEERERS for canonical EE3 Alero).

This effect is not systematic. In some cases, forms with double substitutive softening are not observed at all, cf. ₪38ΜΤΗ, Ια 3Βλίατη. In other cases, sources show alternative variants, cf. pachmatplatin. packmallmonth. In yet other cases, only forms with double substitutive softening are found: сf. мъклити, мъщийж and мъщатахъ. The double substitutive softening effect generates aberrant forms with prohibited clusters.

When only an aberrant form with double substitutive softening is found, that form is shown in the dictionary.46

### Illustrations

L БЪША ЛЮДЬЕ ЖИДЖЦЕ ЗАУДИРА: L ЧЮЖДААХХ СА ЕЖЕ КЪШНЪАШЕ ТЪ ВЪ Цръкъвє Lc 1, 21 MAR; в ЕЛАЖНЕАХЖ СА О НЕМЬ Mt 13, 57 ZOGR; ЄГДА ЛИ ПАКЪІ Въню носашта и сърътають христоса и ликър ставашта: и стельшта ризъл съмоштря: радоуск са такожде SUPR 332, 28-333, 1.

# Fundamental vowel alternations

§ 119. The alternation table

This alternation has 7 series and 8 grades (Table 119, p. 72).

### § 120. The tied segmental position and the domain

The segmental position is the medial V. The alternation is represented only in root formatives, and is limited. Root formatives that show fundamental alter-

<sup>16</sup> These are: <578> размъншлатн, размъшляния, размъншâган and others; <28> съслажнати, <705> опражнати, оупражнение, <1116> объщтовение, <542> оусъмоштриенние, <600> пръмждрати, < 539> оумрьштвлати, оумрьштвление.

nations are given by a list (fewer than 200 roots out of more than 1000 roots; see the lists in § 656, 682).


Table 119. Fundamental vowel alternations

# §121. Series and their names and notations

The first series is called *pure* (shown as P in the Table 119); the other series are *sonant*: H(*n*), H(*m*), H(*u*), H(*j*), H(*r*), H(*l*). Roots that show the P alternation series are said to have *pure vocalism*; roots that show the sonant alternation series are said to have *sonant vocalism*.

# §122. Grades and their names and notations

Grades are named and labeled by the pure series P: the ь grade, the е grade, etc. The following pairings between grades, ъ — ы, о — а, е — ѣ, and ь — и, shown in the Table 119 by vertical arrows, in the pure series represent the pairing by the inner ~ outer phonological opposition. See § 25–26.

§123. Undergoers of the alternation

The undergoers of the alternation in the pure series are single vowels.

In the sonant series, the undergoers are not only single vowels, but also *quasi-segments*, viz. CV combinations (ра, лѣ), and VC combinations (ен, ор).17 In the sonant series, each grade contains a pair of CVC-ambivalent *vocalic realizations*:18 a C-final one (above the slash), notated Hc, and a V-final one (below the slash), notated Hv. The symbol H without an index indicates that the vocalic segmental position is filled with vocalic realizations of the sonant vocalism. C-final realizations in all series are represented by V+C combinations, where V is the vowel of the corresponding grade in the pure series, and C is one of the consonants *n*, *m*, *v*, *j*, *r*, *l* from each series of the alternation. V-final realizations in the H(*r*) and H(*l*) series have the form C+V; in other series V-final realizations are single vowels. C-final (Hc) and V-final (Hv) realizations are distributed by CVC agreement. Closed roots show only V-final realizations, cf. влак, влѣк, вльк ‹105›; дъх, духъ, дых ‹261›. In other cases, a formative can show V- or C-final realizations in different alloforms, cf. мрь, мрѣ ‹539›, and мьр, мор ‹539›.

Thus, in the sonant series each grade is represented by two coupled vocalic realizations that are in complementary distribution by the CVC agreement rule. For example, the pairing between the ь and the е grades of the H(*r*) series can be represented by the following quadruplet of vocalic realizations: ьр/рь‖ер/рѣ; cf. мьр/мрь‖[мер]/мрѣ ‹539›: мьреши, измрьмрѣти.

Sonant combinations are understood to occupy a single segmental position, namely the medial V. Below are examples showing the segmental positions for two roots, влѣк ‹105› and клѧ/кльн ‹384›.


Thus, roots with sonant vocalism can be *closed* (as влѣк, with the segmental positions occupied by CHvC), or *labile* (as клѧ/кльн, with segmental positions occupied by CHv/CHc). Note that for labile roots, or, more generally, for finally ambivalent roots, it is only possible to categorize as "open" their particular alloforms (the ones ending in Hv or V). See § 128 below.

### §124. Intersections between series and grades

As the table shows, the same vocalic realization may belong to different grades of the same series, e.g. рѣ belongs to the grades е and ѣ of the H(*r*) series. Likewise, the same realization may belong to different series, e.g. ѫ belongs to the grades ъ, ы, о, and а of the H(*m*) and H(*n*) series. Some realizations, such as ь and и, ъ and ы, as well as ѣ, can represent both pure vocalism and sonant vocalism of the H(*u*) or H(*j*) type.

<sup>17</sup> See more details in Ch. 24, § 873.

<sup>18</sup> The term "vocalic realization" can refer both to a root vowel with pure vocalism in a particular alloform, as well as a root vowel with stable vocalism.

In some cases, this overlap makes it difficult to categorize a particular root formative in terms of grades or series of the alternation. For example the root трѫс‖трѧс ‹984›, which only shows these two alloforms, cannot be categorized as either H(*m*) or H(*n*), at least within the synchronic description (cf. PIE \**trem*, Latin *tremō*, Lithuanian *trìmti*, etc.). In such cases we adopt the notation H(*n*/*m*). See § 692–701.

### §125. Adjacent grades

The adjacency relation on the grades is shown by arrows in the first row of the table. This relation defines two regions: the front region (ь, е, и, ѣ), and the back region (ъ, о, ы, а). No two adjacent grades belong to different regions. The boundary between the regions is shown in Table 119 with the tilde, indicating the lack of adjacencies between grades belonging to different regions.

Adjacent pairings can be *horizontal* (ь>е, и>ѣ, ъ>о), and *vertical* (ь>и, е>ѣ, ъ>ы, о>а). In the adjacent grades and pairings, the leading grade is called *heavy* compared to the following one, and the following grade is called *light* compared to the leading one. So, ь is the light grade compared to е; и is the heavy grade compared to ь, etc. The same terms can be used to refer to alloforms: e.g. the formative чьт is light compared to чит; нѣс is heavy compared to нес. (See Ch. 24, *Supplement*, § 874.4–5, *On verbs with unstable root vocalism*).

### §126. Replacements

All adjacent grades can be subject to replacement rules in the paradigmatics. Note that horizontal replacement are typical for sonant formatives, while pure formatives more often show vertical replacements. Cf. жьдати, жидѫ (pure), and зъвати, зоветъ (sonant H(*u*)), дъхати, душетъ (sonant H(*u*)). (See more details in Ch. 24, *Supplement*, § 874.4–5, *On verbs with unstable root vocalism*).

### §127. Alloformy

Intraparadigmatic alloformy: жьд‖жид: жьдати, жидѫ (1SgPrae for жьдати); зъв‖зов: зъвати, зовѫ (1SgPrae for зъвати); бьр‖бер: бьрати, берѫ (1SgPrae for бьрати); мрѣ‖мьр: мрѣти, мьрѫ (1SgPrae for мрѣти).

Interlexemic alloformy: твор‖твар: творьць, тварь; зѧб‖зѫб: прозѧбати, зѫбъ; гни‖гној: гнити, гнои; зла‖зел: злакъ, зеленъ; зрь‖зьр: зрьно, съзьрѣти.

In interlexemic alloformy, there can be derivationally related pairs whose basic and derived alloforms are not adjacent, e.g. рещи and отрокъ (alloforms рек and рок).

### §128. A terminological note: labile roots

Finally ambivalent roots are called *labile* if the vocalic realizations of their formatives fit one of the sonant series of the fundamental alternation. Most labile roots show different grades of the series, but roots that have C- or V-final realizations of only one grade are also called labile and are treated as showing fundamental alternations. So, the root ‹780› (alloforms ру/ров) is treated as having unstable vocalism representing the fundamental alternation of the H(*u*) series. By definition, labile roots are those that are, first, finally ambivalent; second, whose C- and V-final alloforms are distributed by the CVC agreement rule, and third, whose vocalism is sonant. However, in some labile roots, the distribution of C- and V-final alloforms is obscured by the paradigmatic effect called "C-final stem arrest" (see § 433). For example, this is the case for the root ‹438› (крыти).

# CHAPTER 6 **Segmental peculiarities of sources**

§129. Graphically alternative spellouts and aberrant forms

As noted above, spellouts from sources are always placed in correspondence with the canon, i.e. a grammatically equivalent canonical wordform is made to correspond with each wordform from the source that is being inspected. In case of differences between that form and the canon, the former is called *alternative*. Among alternative forms we distinguish: (1) *graphically alternative* spellouts (such as ѣко and ꙗко, начѩти and начѧти), (2) *aberrant* spellouts (such as много and мъного, срѣтоста and сърѣтосте), and (3) *erroneous* spellouts (cf. брѣгру for брѣгу, рарабъ for рабъ). Graphically alternative spellouts, as opposed to the other types, have the same phonological shape as the corresponding canonical form.1 Among aberrant forms, we distinguish *segmentally aberrant* (cf. много for canonical мъного), and *paradigmatically aberrant* (cf. сърѣтоста for canonical сърѣтосте). Paradigmatically aberrant forms are not examined in this Part (see Part II, *Paradigmatics*).

Graphically alternative spellouts originate in different writing systems, or *graphic systems*. Segmentally aberrant forms, on the other hand, originate in *segmental aberrations*. (See details in Ch. 24, *Supplement*, § 886–897, *Excursus on aberrations*).

1 This is an arbitrary convention that represents the format of description adopted here, rather than a hypothesis about the actual state of affairs.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

# **Graphic systems of the sources**

### §130. Graphic system

A graphic system consists of a *letter inventory* and a *basic correspondence*. The inventory defines the alphabet used, or the set of letters, or graphic segments. The basic correspondence *g* is the correspondence between letters and phonemes; it is a correspondence between atomic units, or segments, of different types. The basic correspondence is reversible, i.e. it defines not only the transition from letters to phonemes, but also from phonemes to letters. Given a basic correspondence *g*, the derivative correspondence *G* can be built, which is a correspondence between strings of atomic units of different types: a string of letters and a string of phonemes. Usually, the basic correspondence *g* can be set up in such a way that the transition from letter strings to phoneme strings would be compositional: *G*(*abc*)=*g*(*a*)*g*(*b*)*g*(*c*).

Thus, the basic correspondence ensures the transition from a letter string into a phoneme string (reading rules, so to speak), and the transition from phoneme strings to letter strings (writing rules, so to speak). In more or less simple cases, these rules can be stated reversibly, i.e. the same rule system can be used as reading as well as writing rules. In any case, the basic correspondences that are found in this book are stated in such a way.2

The *principle of compositionality*, i.e. the equivalence *G*(*abc*)=*g*(*a*)*g*(*b*)*g*(*c*), sometimes is not fully satisfied. In some cases the same letter in different occurrences corresponds to different phonemes, or the same phoneme in different occurrences is rendered by different letters. Such an indeterminacy must be accounted for by rules that prescribe the choice of one and only one output segment for each input segment. Such rules are called *fork resolution rules*.3 Cf. the fork /a/⇒(а, ꙗ) in the basic correspondence of the normalization (see § 34 above). For this reason, basic correspondence rules can include, alongside simplest rules of the type *x*⇒/y/ (or /y/⇒*x*), more complex rules that can be represented as *axb*⇒/pyq/ (or /pyq/⇒*axb*). In other cases the ambiguity has no resolution rules; for example, after shibilants and ц, ѕ, in Kiev the phoneme /a/ can be rendered by the letter а or the letter ѣ (cf. срьдьцѣ 2b, 15–16, and срьдьца 5a, 21), and, accordingly, the letter ѣ can render both the phoneme /a/ and the phoneme /ě/ (see Ch. 7, § 200).

Each source has its own unique graphic system. The canon also has its own graphic system, distinct from that of the sources. The graphic system of the canon, termed *normalization*, is given in Ch. 2, *Graphics and phonology*. Recall that in treating the graphic systems of Glagolitic sources, the subject of analysis are not the primary Glagolitic spellouts but their Cyrillic transliterations.

<sup>2</sup> Some special cases where basic correspondences depart from the simple schema described here are noted in the course of presentation.

<sup>3</sup> If the fork resolution rule uses no information other than segmental neighborhood, then the fork and the relevant rules are called *segmentally decidable*.

The discussion below proceeds as follows: first the transliteration is shown, and some questions relating to letter inventories are discussed (§131–133); then the technique for describing the basic correspondences in sources is explicated (§ 134–135), as well as the construction principles for phonological representations of alternative spellouts (§ 136–138). The next subsection, *Segmental aberrations*, catalogs all segmental aberrations (§ 139–141). Ch. 7, *An overview of sources*, shows text samples of all the examined sources, and a treatment of their segmental peculiarities (writing systems and segmental aberrations).

# Alphabets and transliteration

### §131. The list of alphabets

Glagolitic sources use the *Glagolitic* alphabet. The normalization uses the alphabet known as *standard Cyrillic*. Cyrillic sources use the alphabet known as *natural Cyrillic*. Natural Cyrillic contains all the letters of standard Cyrillic, plus some additional ones; for example, the letter (*uk*). Also considered is the *secondary Cyrillic* alphabet, consisting of the Cyrillic correspondents of the Glagolitic alphabet as used in editions of Glagolitic sources in their Cyrillic transliteration. The union of all Cyrillic alphabets is called *expanded Cyrillic*. The alphabets themselves are given in Table 132.1 (p. 80).

### §132. Transliteration

Table 132.1 gives the Glagolitic ⇔ secondary Cyrillic transliteration. This table, along with rows that give the correspondence between Glagolitic and Cyrillic letters (rows 1–2), gives rows for standard Cyrillic (normalized, row 3) and natural Cyrillic (as used in Cyrillic sources, row 4).

As this table shows, transliteration gives a reversible one-to-one correspondence. In the pairs (ⰺ, ꙇ/ї) and (ⱏⰺ, ы/ы), the slashes separate letters that are in complementary distribution in sources: the letters under the slash are the secondary Cyrillic letters used in the Sever'janov's transliteration of the *Psalterium Sinaiticum*.

The table shows also the standard Cyrillic correspondents for letters of the secondary and natural Cyrillic. Below, in the discussion of the sources, secondary Cyrillic letters that correspond to the same standard Cyrillic letter are called *allographs*; same for natural Cyrillic letters.

The transliteration is perfectly compositional: α(*xyz*)=α(*x*)α(*y*)α(*z*) for any letter *x*, *y*, *z*, where α is the transliteration from the Glagolitic to the secondary Cyrillic alphabet or vice versa.




Table 132.2. Additional Glagolitic and Cyrillic letters


Table 132.3. Numerical values of Glagolitic letters (as secondary Cyrillic) and Cyrillic letters


ѭ

ⱙ

### Addendum

Some letters used in Glagolitic and Cyrillic sources have no standard Cyrillic correspondents. These are letters that are almost exclusively used in spellouts of new borrowings, mostly in personal names; they are called *additional* in Table 132.2 and below. The standard Cyrillic alphabet is designed to service the normalization. Strictly speaking, in this book normalization is only set up for wordforms from the benchmark corpus, which does not include wordforms with additional letters in Glagolitic sources. Nonetheless, in some particular cases the book uses Cyrillic spellouts that correspond to Glagolitic spellouts with additional letters. This is the case, for example, for the spellout герьгесиньскѫ for the Glagolitic ⰼⰵⱃꙿⰼⰵⱄⰹⱀꙿⱄⰽⱘ (§1, normalization sample). In spellouts of borrowings that contain the prohibited velar + front vowel combinations, sources may use the letters к҄ and г҄ (see details in Ch. 24, *Supplement*, § 869, *On the law of the velars*). In Zogr, kamora is sometimes used with labials; see § 167.

Cf., for example, ⰵⰼⱛⱂⰵⱅⱏ (ећѵпетъ) Mt 2, 13 As; ѳома Jn 14, 5 As, Sav; указаниѥ отъ свꙙтааго ѱаниꙗ Supr 304, 28–29.

Table 132.3 shows the numerical values of the letters (see details in Vaillant, § 12, 18, 47, Lunt 2001, Ch.I).4

### §133. Departures from sequential spellouts of letters

Here must be mentioned spellouts under titlo, as well as superscript and added letters. In citing forms, spellouts under titlo and superscript letters, as well as ligatures marked in editions, are indicated by *titlo* or by letters printed as superscripts. The *payerok* is rendered as an apostrophe; the ligature for от/отъ by the symbol ѿ. Other diacritics found in manuscripts and often reflected in editions are simply ignored in this book. Cf., for example:


# Basic writing systems in sources

§134. The basic correspondence and ways of characterizing it

For the normalization, the basic correspondence was given completely in the description of the graphics and phonology of the canon (see § 33–35). For sources, basic correspondences need not be described fully: it is sufficient, for each

<sup>4</sup> The letter Ϟ is the *koppa*, an obsolete Greek letter that indicates 90 in numerical use.

source, to consider the fragment where the basic correspondence of that source differs from the basic correspondence of the canon. For all sources, such a fragment gives the letter equivalent of the following six phonemes: /e/, /e/, /u/, /o/, /a/, /e/. These nontrivial fragments of basic correspondences are given by special tables that are supplied in the discussion of each source. These tables are called below basic writing systems of the source.

Outside of this fragment, the basic correspondence is trivial. However, if within the trivial region the basic correspondence between letters and phonemes is one-to-one, in sources one phoneme (and, accordingly, one standard Cyrillic letter) can correspond to several allographs. Allographs from the expanded Cyrillic alphabet used in each source are shown in the overview of sources below the basic writing system tables. The use of nonstandard allographs creates graphically alternative spellouts that are phonologically equivalent to the canon. Let us give a few examples.


#### § 135. Basic writing system tables

The format of these tables reflects a pragmatic compromise: it collapses the transition from phonology to graphics with the transition from graphics of the canon (normalization) to the graphics of the source. This combination is possible because the normalization & phonology transition is unambiguous.

Consider such a table fragment (Tables 135.1-2, p. 83) for MAR (the full table is given as Table 144). Table 135.1, as in other such tables shown in the overview of the sources (Ch. 7), the phonology is not given directly but follows from the normalization; in Table 135.2, it is given explicitly. Here and below in basic writing system tables each row corresponds to a phoneme (or a phoneme pair in the case of /a/ ~ /ě/ ), and each column corresponds to a left context. The cells give the rendition of the corresponding phoneme or phoneme string in the normalization (shaded row) and in the source's writing system (unshaded row).

Distribution rules that regulate the selection of an allograph out of the set of possible allographs for a given source are not given and not investigated in this book.


Table 135.1. Basic writing system table fragment (variant 1)

«ш» stands for shibilants, C stands for simple consonants

Table 135.2. Basic writing system table fragment (variant 2)


«ш» stands for shibilants, C stands for simple consonants

Every spellout from Mar, including alternative spellouts, can be evaluated as satisfying (+) or not satisfying (–) the basic writing system of Mar:


Aberrant spellouts may also either satisfy or fail to satisfy the basic writing system; e.g. the aberrant form ощѫтитъ Jn 11, 57 Mar (for canonical ощутитъ), from the point of view of graphics, is in perfect agreement with the writing system of Mar.6

<sup>6</sup> Note that this table, as well as the full basic correspondence, only ensures the transition from a phoneme string to the corresponding letter string and vice versa, but does not assign any segmental content to any wordform. While characteristic of the canon, stability of the phonological content of most wordforms is often broken in the language of the sources. For example, in Zogr we find ѣко alongside ако, which points to unstable phonology (/jako/ vs. /ako/) as well as unstable graphics. In other cases graphically distinct strings can correspond to identical phonology, cf. помышл҄ен- (Supr 296, 1) and помъишл҄ен- (Supr 249, 18; note

§ 136. Phonological representations of alternative spellouts

In most cases, the phonological representation of any alternative spellout can be constructed using the basic correspondence in the direction from letters to phonemes. Comparing that phonological representation of the alternative spellout with the phonological representation of the canonical spellout gives the evaluation of the spellout as graphically alternative (gr) or segmentally aberrant (ph). Table 136 shows examples.


Table 136. Evaluation of alternative spellouts: simple cases

However, not every alternative spellout can be treated as simply as in the cases above. Special problems arise, first, with kamorated phonemes (1', n', r'), second, in undecidable ambiguities in the basic correspondence itself, and third, in case the spellout observed in the source does not follow the basic writing rules.

In particular, if the source's spellout departs from its own basic writing system, the symbol-by-symbol transcription process is simply blocked. This is the case for 4tch in ZOGR: for the letter t after shibilants the basic correspondence offers no phonological form. And if the spellout contains a letter that, by the basic correspondence, leads to an undecidable fork, the ambiguity remains (cf. in KIEV срьдыць: /srbdbce/ or /srbdbca/ ).

§ 137. Phonological representations of alternative forms: special cases

In the cases considered below, grammatical information is necessary for the construction of the phonological representation of an alternative form. Remem-

allographs bland bh), or in AS pacilaT- and pacniar- (cf. paccument and paccuaca; both spellouts in Lk 23, 33; see § 173-174 below). Such graphical ambiguity is built into the writing system and is noted in the discussion of the source.

ber, that in the analysis of alternative forms we consider pairs (*A*\*, *A*), where *A*\* is the alternative form under consideration, and *A* is its canonical analog, and the grammatical properties of these forms are identical. In the simplest cases, however, phonological representations can be assigned to alternative spellouts without using grammatical properties of the form and even without any data supplied by the corresponding canonical form.

### Kamora

Most sources do not use the kamora, and those that use it do so unsystematically. The following convention is adopted in this book: the phonological opposition simple ~ kamorated that is posited for the canon is also assumed for the language of the sources, but the graphics of the sources do not reflect it or reflect it only sporadically. At the same time, the distribution of kamorated phonemes is identical to that in the canon. According to this convention, such spellouts as, e.g. Mar землѭ, землѣ, are assigned the phonological representations /zeml'ǫ/, /zeml'a/, and accordingly such a pair of spellouts as землѣ and земл҄ѣ are assigned the same phonological representation /zeml'a/. Thus, grammatically interpreting the form is necessary to construct the phonological representation of alternative spellouts (compare лѣто /lěto/ with землѣ /zeml'a/ and лѫкавъ /lǫkavъ/ with землѫ /zeml'ǫ/, etc.).

Accordingly, spellouts that differ from the canon only in absence of the kamora are treated as graphically alternative, not aberrant.

#### Undecidable ambiguities

In order to assign a phonological representation to a spellout such as Kiev срьдьцѣ, /srьdьcě/ or /srьdьca/, the grammatical interpretation must be first determined: if срьдьцѣ is GSg(n) or NAPl(n) from срьдьце, then the fork is decided in favor of /srьdьca/. The principle here is the same as with the kamora.

Departures from the basic writing system

As mentioned above, sources may occasionally depart from the basic writing system. For example, the spellout пѩть is found in As in place of canonical пѧть, despite the fact that in As the letter ѩ is not used after simple consonants. However, the possibility of a spellout like пѩ is admitted as a departure from the basic writing system. Accordingly, for As the spellout пѩть is considered phonologically equivalent to пѧть. Thus, this spellout is not aberrant, but simply graphically alternative. Another example is furnished by the following spellouts in Mar: пѣнѧзю, пѣнѧѕу, пѣнѧзу. According to the basic writing system of Mar, the letter ю must be used after ѕ, while the letter у after з. Thus пѣнѧзю /pěnęzu/ is not only phonologically aberrant (з in place of canonical ѕ), but violates the graphics required by the basic writing system of Mar. The spellout пѣнѧзу /pěnęzu/ is phonologically aberrant, but follows Mar's graphics, and the spellout пѣнѧѕу /pěnędzu/ matches the canon in phonology, but violates the graphics required by the basic writing system of Mar.

Frequently, spellouts that violate the basic writing system of a given source contain graphically prohibited letter combinations: cf. Mar тачѣе, Zogr

чѣсъ, As пѩть, As оставѭ, Mar притѧжѭ, Mar пѣнѧзю, and many others. These letter combinations are prohibited not only in the writing system of these sources, but in the normalization as well. Thus, the phonological representation cannot be obtained using the basic writing system. In these cases phonological representations can be reverse-engineered. If the form shows mentioned cases of departure from the basic writing system of the source (as in the case of чѣсъ and пѩть), the phonological representation is obtained from the phonological equivalence of those spellouts: чѣсъ is equivalent to часъ, and thus /časъ/ is the phonology; пѩть is equivalent to пѧть, and thus /pętь/ is the phonology. If, on the other hand, the spellout does not show any expected departures from the basic writing system (cf. Ps Sin земѭ) grammatical analysis is necessary, which would compare the canonical and aberrant derivations of the corresponding forms. Ps Sin земѭ for canonical земл҄ѭ /zeml'ǫ/ from земл҄+ ѫ, where the last consonant of the workstem is kamorated, and the terminal is /ǫ/; земл҄ is the substitutively softened alloform of the root ‹313› (зем, земл҄); this root has an aberrant assortment resulting from an aberrant version of the substitutive softening alternation. Thus, зем is the aberrant alloform of земл҄, and the morphological skeleton зем+ѫ is aberrant. Further, this morphological skeleton yields the phonological form /zemǫ/. Table 137 shows some examples.


Table 137. Evaluation of alternative spellouts: special cases

Consider in detail the spellout Ps Sin земъѭ (Ps Sin 16, 11) showing combined aberrations. First, it shows the aberrant pairing м‖мь in place of the canonical м‖мл҄ (see § 141 below); second, ъ in place of the canonical ь (see § 140 below).

#### § 138. A note on the phonological representations of alternative spellouts

lt is taken as a basic convention that an aberrant spellout differs from the corresponding canonical one in its phonological representation, while a graphically alternative spellout is identical to the canon in the phonology but differs only in the writing. This solution is fairly arbitrary and somewhat simplistic, but avoids vicious indeterminacy both for the grammarian and the reader.

The convention itself is debatable in two aspects: first, one may like or dislike its sense, and second, one may like or dislike the demarcation line drawn in the material.

As for the sense of the notion, graphic mismatches in the general case can reflect not only graphic differences, but also phonological ones. Distinct spellouts can correspond to phonologically identical strings (cf. Russian лиса 'fox' /lisa/ and леса 'forests' /lisa/; cf. also English need vs. knead), and identical spellouts can correspond to phonologically distinct strings (cf. Russian конечно 'finite' /kon'ečno/ and конечно 'of course' /kon'ešno/, or села 'sať' /s'ela/ and села 'villages' /s'ola/; cf. also English lead guide' /liid/ and lead 'heavy metal' /led/). Thus, alongside the convention adopted here, solutions that do not adopt it may also be linguistically acceptable. Table 138 shows the possible diversity of imagined grammatical interpretations (ph 2 and ph 3 below) of different aberrant spellouts of two forms: for canonical вьсь, aberrant forms вєсь, въсь, всь, and for canonical мъночн, aberrant мънози.'


Table 138. Possible phonological interpretations of alternative spellouts

The line ph 1 corresponds to the solution adopted in this book; ph 2 and ph 3 show alternatives, of which there are many more. Of course, we have no data that allow us to select phonological representations on the basis of their phonetic shape. Indirect data that lead outside of the facts of the sources immensely complicate the logic of the construction. It must be remembered that in the general case the same grammatical form can have different phonetic shapes (new and old, formal and informal, etc.). In any case, these problems lie outside of the aims of this book, as belonging not to the canonical OCS language but to the language of the specific source and various hypotheses about the evolution of OCS and its many variants in space and time.

The question of the demarcation line between graphically alternative and aberrant forms, in terms of the basic convention adopted here, depends entire-

Grammatical solutions that interpret these alternative spellouts as graphically alternative are not considered here, but of course not excluded as possibly acceptable solutions.

ly on the rules that assign phonological representations. These rules are given by the basic correspondence of each source and by some specific conventions. Of course it would be possible to establish somewhat different basic correspondences from those adopted in this book while maintaining the basic convention on alternative and aberrant spellouts.

# **Segmental aberrations**

# §139. General

Aberrations are divided into *segmental*, *paradigmatic*, and *lexemic* (or *lexical*). Segmental grammar deals with segmental and lexemic aberrations. Segmental aberrations are divided into atomic and modifying aberrations (or simply modifications). In atomic aberrations, the substituted entities are single segments; in modifications, the substituted entities are segmental rewrite rules (individual mph⇒ph/norm rules or individual rules of replacement by an alternation). Segmental aberrations are described below. In lexemic aberrations, the substituted entities are individual lexemes, and are noted alongside corresponding canonical lexemes (in the lists of commented lexemes in the main text of the book). See details in Ch. 24, *Supplement*, § 886–897, *Excursus on aberrations*; see overview of lexical aberrations *ibid*.

# §140. Atomic aberrations

Atomic aberrations attack the phonological or morphophonological representations of wordforms.

Atomic aberrations targeting phonological representations All such aberrations are shown in Table 140.1.


Table 140.1. Atomic aberrations of phonological representations

The aberration (V+) j (+V)⇒0 could be treated as a modification that attacks morphophonological representations, canceling the *i̯* epenthesis rule. For example, for покаꙗти we have mph representation по.ка.а.т.и, which in the canonical derivation corresponds to /pokajati/, and without *i̯* epenthesis to /pokaati/. Cf. also великаа Mk 13, 2 Mar, and outside of the benchmark word list зълодѣа (Euch 63b, 10; Lk 23, 33 Mar), олѣа (Mt 25, 8 Mar), and many others among borrowings. Note that most spellouts without intervocalic j are those where the phoneme /j/ in the phonology corresponds to epenthetic *i̯*in the morphophonol-

ogy. Forms of the lexeme божии [бож.ьј.ь] in Supr are notable here: the spellout божиꙗ occurs more than 40 times, while божиа occurs only once (Supr 294, 13), and there are no other spellouts of this wordform without j; same in Mar. These facts point to treating this aberration as a modification, but the technically simpler solution adopted in this book does not contradict the data.

The aberration *dz*⇒*z* may be treated as attacking phonological or morphophonological representations.

Only the aberration V1V1⇒V1 must necessarily attack phonological representations: so, in the example above we have прииметъ [при.ьм.етъ] /priimetъ/.8

Notes

1°. The aberration (V+) j (+V)⇒0 (intervocalic j loss) can be found in aberrant forms of long adjectives, cf. новааго for новаѥго and similar, and Prae forms of the type дѣлаатъ, which are treated in this grammar as resulting from paradigmatic aberrations (cf. § 394, 617 below). 2°. Aberrant forms of the iotated imperfect of the type исъхнѣꙗше can be treated as hypercorrections (insertion of j into a hiatus), or as departures from *i̯* epenthesis rules in the mph⇒ph/norm transition (see § 69–70 above). In this book these forms are treated as resulting from paradigmatic aberrations (see § 471 below).

Atomic aberrations targeting morphophonological representations These are the so-called *yer* aberrations. There are four of them; they are listed in Table 140.2.


Table 140.2. Atomic *yer* aberrations of morphophonological representations

### Notes

1°. All *yer* aberrations are particular deformations of mph⇒ph rules.9 Aberrant forms that only show the application of *yer* aberrations have the same mph represen-


tation as the corresponding canonical form, but their phonological representation differ: [k.b.t.o] ⇒ /koto/ > KBTO and [k.b.t.o] ⇒ /kto/ ⇒ KTO; [pbj.o] ⇒ /pijo/ ⇒ ΠΗΗΣ and [pbj.o] > /pbjo/ > nbm, cf. also BEAEH. In yer regression, the aberrant derivation simply bypasses one of the mph > ph rules, 0 while in the classical triad (fall/strengthening/confusion) the aberrant derivation adds new rules, which are given above (Table 140.2, column 1). 2°. A special group of aberrations, dealing with aberrant spellouts corresponding to the canonical combinations ин, ъи, adjoins yer aberrations. Cf. in MAR отъ тоуждикъ Mt 17, 25 and отъ тогждинуъ Mt 17, 26 (for canonical on b Toy'x gunx b); AS a.3b me BENTI H Jn 8, 55 (for canonical a35 жє въмь и); and others. The latter are separated into their own group because the boundary between graphics and phonology is hard to draw here, cf. a spellout like ΕΛΑΓΈΜ Lk 18, 18 AS, where the string τη can have two analyses.11

### § 141. Modifying aberrations

Here two types of aberrations are distinguished: first, aberrant cluster rewrite rules (i.e. aberrant mph = ph/norm); second, aberrant pairings in alternations (see Table 141).


### Table 141. Modifying aberrations

This table shows artificial spellouts of aberrant forms (and their canonical analogs), which show only the segmental changes discussed here, corresponding to spellouts cited in the indicated paragraphs.

10 See yer rules creating the phonemes n and 'ar before j (see above \$ 72, rule 3°).

11 See more details on the classification and morphophonological interpretation of yer aberrations in § 898-899, Excursus on yer aberrations. Havlík proposed the rule that regulates alternative yer aberrations: fall, strengthening, or confusion, as the case may be.

# CHAPTER 7 **Survey of the sources**

§142. Treatment scheme for segmental idiosyncrasies of a source

The typical format for treatment of the sources is as follows: text samples are given first, then a table showing the basic writing system, and then departures from that writing system. Other segmental idiosyncrasies of a source are then listed in the following order: (1) allographs (for the letters щ, и, ы, and others); (2) mixing of letters (з/ѕ, у/ѫ, ю/ѭ);1 (3) rendition of the kamorated letters; (4) alternative alternations and combinations with epenthetic *l*; (5) omission of intervocalic *j*; (6) rendition of the *yers*; (7) rendition of the combinations -ии-, -ыи-; (8) additional commentary on text samples.

# *Codex Marianus*

§143. Text samples2

Mark 4, 26–28

1ꙇ глаше тако естъ цѣсарествие бжие· 2ѣкоже члкъ въмѣтаатъ сѣмѧ вь землѭ· 3ꙇ съпитъ· 4ꙇ въстаатъ нощь и дьнь· 5ꙇ сѣмѧ прозѧбаатъ· 6ꙇ растетъ ѣкоже


First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

не вѣстъ онъ· 7о себѣ бо землѣ плодитъ сѧ· 8прѣЖе трѣвѫ по томь же класъ· 9по томъ же и пьшеницѫ въ класѣ·

### Matthew 15, 22–28

10ꙇ се жена хананеиска· 11отъ прѣдѣлъ тѣхъ ишедъши· 12възъпи глѭщи· 13помилуи мѧ ги сну давдвъ· 14дъщи моѣ зълѣ бѣсънуетъ сѧ· 15онъ же не отъвѣща еи словесе· 16ꙇ пристѫпьше ученици его молѣхѫ и глѭще· 17отъпусти ѭ ѣко въпиетъ въ слѣдъ насъ· 18онъ же отъвѣщавъ рече· 19нѣсмъ посъланъ· 20тъкмо къ овъцамъ погыбъшимъ дому излву· 21она же пришедъши поклони сѧ ему глщи· 22ги помоѕи ми· 23онъ же отъвѣщавъ рече· 24нѣстъ добро отѧти хлѣба чѧдомъ и поврѣщи псомъ· 25она же рече еи ги· 26ꙇбо и пси ѣдѧтъ отъ крупицъ падаѭщихъ съ трапезы господеи своихъ· 27тъгда отъвѣщавъ ис рече еи· 28ѡ жено вельѣ естъ вѣра твоѣ· 29бѫди тебѣ ѣкоже хощеши· 30ꙇ исцѣлѣ дъщи еѩ во тъ часъ·

### Matthew 20, 8–14

31вечеру же бы(въ)шю· 32гла гнъ винограда· къ приставънику своему· 33призови дѣлателѧ· 34ꙇ даЖь имъ мъздѫ· 35наченъ отъ послѣдьниихъ до пръвыихъ· 36пришедъше же иже въ единѫѭ на десѧте годинѫ· 37приѩсѧ по пѣнѧзю· 38ꙇ пришедъше пръвии мьнѣахѫ сѧ вѧще приѩти· 39ꙇ приѩсѧ по пѣнѧзу· 40приемъше же ръпътаахѫ на гнъ глѫще· 41како сьѩ послѣдьнѧѩ· 42единъ часъ сътворьшѧ· 43ꙇ равъны намъ сътворилъ ѩ еси· 44понесъшеимъ тѧготѫ дьне и варъ· 45онъ же отъвѣщавъ рече единому ихъ· 46друже не обиЖѫ тебе· 47не по пѣнѧѕу ли съвѣщахъ съ тобоѭ· 48вьзьми свое и иди· 49хощѫ же сему послѣдьнюму дати ѣко и тебѣ·

#### Mark 2, 8–17

50что тако помышлѣате въ срдцихъ вашихъ· 51что естъ удобѣе рещи ослабленуму· 52отъпущаѭтъ ти сѧ грѣси· 53ли рещи въстани и вьзъми одръ твои и ходи· 54нъ да вѣсте ѣко власть иматъ снъ члвчскы· 55отъпущати на земи грѣхы ꙇ гла ослабленуму· 56тебѣ глѭ въстани и вьзьми одръ· твои и иди въ домъ твои· 57и въста абье и вьзѧтъ одръ· 58ꙇ изиде прѣдъ вьсѣми· 59ѣко дивлѣахѫ сѧ вьси и славлѣхѫ ба глѭще· 60ѣко николиже тако видѣхомъ· 61ꙇ изиде пакы къ морю· 62ꙇ весь народъ идѣаше къ [къ] нему и учааше ѩ· 63ꙇ мимо грѧды исъ видѣ левьћиѭ альфеова· 64сѣдѧща на мытьници· 65и гла ему по мьнѣ грѣди· 66ꙇ въставъ въ слѣдъ его иде· 67ꙇ быстъ [же] възлежѧщю ему въ дому его· 68ꙇ мъноѕи мытаре и грѣшъници възлежахѫ съ исмь и съ ученикы его· 69бѣахѫ бо мьноѕи· 70ꙇ по немь идѫ 71и кънижъници фарисеи· 72видѣвъше и ѣдѫщъ съ мытари и грѣшъникы· 73глаахѫ ученикомъ его· 74что ѣко съ грѣшъникы ѣстъ и пьетъ· 75ꙇ слышавъ исъ гла имъ· 76не трѣбуѭтъ съдравии балиѩ нъ болѧщеи· 77не придъ призъватъ праведьникъ· 78нъ грѣшъникы въ покаание·

### § 144. Basic writing system

The basic writing system of the Codex Marianus is given in Table 144.


Table 144. Basic writing system: Codex Marianus

«ш» stands for shibilants, C stands for simple consonants.

§ 145. Departures from the basic writing system

Rare departures are possible. Here are some examples:


### § 146. Allographs of the letter யா

MAR uses as free variants two allographs: шт and ц. Thus, in the fragments shown above we have or be burars (27) and or BB tuaBb (45).

### \$ 147. Allographs of the letter и

MAR uses three allographs: n, i and, less frequently, i. Cf. BB icTHHHHMML LC 16, 11; въ ктиня Lc 12, 44; извавление Mt 20, 28, and likewise in the fragments shown above, where only n and t are used.

### \$ 148. Allographs of the letter TI

Normally, the shape ъı is used for the letter ъг, but сf. татъвъ: обидън ЛЖКАВЪСТВО: ЛЬСТЬ СТОУДОДЕАНИЕ МК 7, 22; ВЪ ТОМЬ ЖЕ ДОМОУ ПОВЕЗІВАНТЕ 1.

# §149. A special spellout for the letter ѧ

In some occurrences we find a special version for the letter ⱔ (rendered as ѧ in the norm), for instance (transliterated as ꙙ by Jagić). Such are: горꙙ Lk 24, 32 and Jn 5, 35; ѣдꙙи Jn 6, 58; сꙙи Jn 19, 38. See § 622.

# §150. Mixing of the letters з and ѕ

The letter з can occur in place of ѕ. Cf. in the framgents shown above: по пѣнѧзю (37), по пѣнѧзу (39) and по пѣнѧѕу (47), and elsewhere.

# §151. Mixing of the letters ѫ and у, ѭ and ю

Sometimes we find у in place of the expected ѫ, cf.: вьметаему (ASgf) Lk 12, 28; възлюблю Jn 14, 21; while sometimes we find ѫ in place of the expected у, cf.: бѫрѣ Lk 8, 23; мѫжѫ буѭ Mt 7, 26, къ немѫ Mt 8, 5; ощѫтитъ Jn 11, 57, шѫѭѭ Lk 23, 33.

# §152. Rendition of kamorated letters

Kamorated letters occur in isolated cases. Such are, for example, аврамл҄е Lk 16, 22; възл҄юбл҄ены Lk 20, 13, магдалын҄и Mt 28, 1 and Mt 27, 61, бол҄еи Mt 12, 6, земл҄ѧ Jn 12, 32, земл҄ѭ Mt 25, 18.

There are no spellouts with kamora in the fragments shown above. Cf.: землѭ (2), землѣ (7), молѣхѫ (16), послѣдьнѧѩ (41), сътворьшѧ (42) and others. Departures from the basic writing system are possible, cf.: куплѫ Lk 19, 13 alongside with куплѭ Lk 19, 15, приклучьшю сѧ Mk 6, 21 alongside with ключи сѧ Lk 1, 9, and elsewhere.

# §153. Substitutive softening

Epenthetic *l* is often absent. Cf. in the fragments shown above: пристѫпьше (16), на земи (55); but землѭ (2), землѣ (7), ослабленуму (51), дивлѣахѫ (59) and elsewhere.; cf. also оставь Mk 8, 13 alongside with оставль Mt 13, 36, and elsewhere. See § 117, *Aberrant versions of replacements by the substitutive softening alternation*, § 141, *Modifying aberrations*.

# §154. Omission of intervocalic *j*

This occurs fairly often. In the fragments above: покаание (78); cf. also великаа Mk 13, 2, but пропѧтаѣ Mk 15, 32; въздаание Lk 14, 12, дааниѣ Mt 7, 11, нечааниѣ Lk 21, 25, and many others.

Here note also cases with deformed terminals: въстаатъ (4), прозѧбаатъ (5) and others.

§155. Rendition of the *yers*

All aberrations are possible. Cf. the following in the fragments shown above.

Strengthening: цѣсарествие (1), ишедъши (11), господеи (26), во тъ часъ (30), наченъ (35), приемъше (40), весь (62); cf. also понесъшеимъ (44) (for понесъшиимъ with the terminal ь*i̯*.имъ).

Fall: псомъ (24), пси (26), что (50) and others.

Confusion: вьзьми (48), вьзъми (53), вьзьми (56), вьзѧтъ (57); cf. also по томь (8) and по томъ (9), and many others.

Regression: вельѣ (28), сьѩ (41), абье (57), пьетъ (74).

§156. On spellouts for normalized combinations -ии- and -ыи-

Standard spellouts in the fragments shown above: пръвыихъ (35); послѣдьниихъ (35), пръвии (38), съдравии (76). Aberrant spellouts in the fragments shown above: погыбъшимъ (20), падаѭщихъ (26) and others.

Likewise праведъныхъ Lk 14, 14, etc.; отъ кыхъ приемлѭтъ дани Mt 17, 25 and ꙇ рече има кыхъ Lk 24, 19; cf.: отъ туЖихъ Mt 17, 25 and отъ туЖиихъ Mt 17, 26; въ неправедьнѣемь жити Lk 16, 11. Likewise къто убо естъ вѣръны приставьникъ и мѫдры Lk 12, 42; выистинѫ Lk 20, 21; выиюдеихъ Jn 11, 54; аще оставимы-и тако· вьси вѣрѫ имѫтъ въ нъ Jn 11, 48; видѣлы-и еси Jn 9, 37; cf. ꙇ протешетъ і Lk 12, 46. Aberrant spellouts with the strengthening of the *yers*: завѣтъ стои (свѧтои) свои Lk 1, 72 and others.

§157. Additional commentary on samples


# *Codex Zographensis*

§158. Text samples

Luke 5, 29–35

1ꙇ бѣ народъ многъ мытар҄ь ꙇ инѣхъ ꙇже бѣахѫ съ н҄имь вьзлежѧще· 2ꙇ рьпътаахѫ кънижьн҄ици ꙇ фарисѣі· 3къ ученикомъ его глѭще· 4по чьто съ мытари ꙇ грѣшьникы ѣстъ ꙇ пьетъ· 5ꙇ отъвѣщавъ ис рече къ н҄имъ· 6не трѣбуѭтъ съдравиі врачевъ· 7нъ болѧщеі· 8не придохъ призъватъ правьдьникъ· 9нъ грѣшьникы въ покаанье· 10они же рѣшѧ къ н҄ему· 11по чьто ученици ꙇоанови постѧтъ сѧ чѧсто· 12ꙇ молитвы творѧтъ· 13такоЖе

и фарисъссии 14 а твої фдать и пынуть 15 онъ же рече къ нилъ 16 еда можете сбъе врачьнъым: 17догдеже женнуъ есть съ нили. 18 сътворити постити сах 19 ПОИДЖТЪ ЖЕ ДЬНЬЕ 20 ЕГДА ОТЪРУТЪ БЯДЕТЪ ОТЪ ЯНУЪ ЖЕНИХЪ 2 " ТЪГДА ПОСТАТЪ СА ВЪ ТЪ ТЪІ ДНИ

#### Matthew 7, 1-8

22 не осжжданте: 23 да не осяждени вждете: 24 мм же во сждомь сядите: 25 сядать вамъ: 20 вь нья же мъря мърите: 27 възмератъ вализъ 20 чьто же видиши сячьць сжь єсть въ оці вратра твоєго. 22 врьвна єже єсть въ оць твоємь не чюєши: 30ли како речеши пратоу своємоу: 31 остави и нзъмж схчець 13 очесе твоего: 32 се връвъно въ оцъ твоемь лицемъре: 33138ли пръвъе пръвъю 1.3 очесе твоего г тъгда оузьриши сзати с случесе врата твоего. 35 не дадите стаго псоль: 36ни получаите внееръ вашихъ предъ свиниками: 3 да не попержтъ дуъ ногами своими: 38 враштьша сл растръгнятъ във: 39просите: 40 гдастъ са вамъ: 4 сштъте 42 обраштете: 43 тлъцъвте с отвръзетъ са вамъ: 44 Въстькъ во просан приемлетъ: 45 нштан обретаєтъ: 40 таъкжшторимог ОТВОЪЗЕТЪ СА:

#### John 8, 2-11

47 ютро же пакъ приде въ цръковь: 40 вси людие гдѣахъ къ немоч: 40 съдъ оучааше м. 50 привъсл же кънижьници и фарисъи женя въ пръдюводувани тата: 51 поставльше ся по средъ глаша емор. 500чителью си жена ката естъ НЪИЧЕ ВЪ ПОВАЮЕОДЪСИНИ: 33 ВЪ ЗАКОНЪ НАМЪ МОСИ ПОВЕЛЪ: ТАКОВЪНА КАМЕННЕМЬ повивати. 54 тъ же чьто глеши: 5 се же овша искоушаеяште и 5 да слубли на нь чьто глати. 5°нс же низъ поклонь сл. 5°прьстомь писаше на земн. 5° ко же прилежахж въпрашатиште г. "Въсклони сл. с рече глу" " иже васъ без гръха естъ: 62 пръжде връзн калень на неж: 63 пакът поклонь са писаше на земли: 64 слъщавъше ссуождауж: 65 гдниз по еднюмоу. 66 начьнъше отъ старьць 97 оста единъ 68 жена стовашті по средъ 69 въсклонь же са не рече ег 7 жено. 7 къде сятъ иже на та важдаух: 72никъсже ли теве осжди: 73 ма же рече: 74никътоже ги 75 рече же ис. 7 ни азъ теве осжждами. 7 1 ди. 7 с отъ сель не съгръшая къ томочъ

#### § 159. Basic writing system

The basic writing system of the Codex Zographensis is given in Table 159 (p. 97).

§ 160. Departures from the basic writing system

Rare departures are possible. Here are some examples:

1) 04 ~ ю: тлъкжштоумог (46);



Table 159. Basic writing system: *Codex Zographensis*

«ш» stands for shibilants, C stands for simple consonants.

# §161. Allographs of the letter щ

Only щ.

# §162. Allographs of the letter и

Zogr uses three allographs: и, ꙇ and, less frequently, і. In the samples shown above note several spellouts with і: твоі (14), въпрашаѭще і (59), съгрѣшаі (78).

## §163. Allographs of the letter ы

The shape ы is usually used for the letter ы, but cf. малыхъ Mt 10, 42, отыметъ Mk 4, 15 (see also § 170 below).

# §164. A special spellout for the letter ѧ

In some occurrences we find a special version for the letter ⱔ (rendered as ѧ in the standard transliteration), for instance (rendered as ꙙ in the standard transliteration). Such are: несꙙ Mk 14, 13; ѣдꙙи Jn 6, 54; сꙙи Jn 1, 18 and Jn 6, 46; живꙙи Jn 6, 57; грѧдꙙи Mt 11, 3. See § 622.

# §165. Mixing of the letters з and ѕ

Occurs fairly often. Thus, for example: помози Mk 9, 22, мнози Mk 9, 26. Cf. the form връѕи (62) in the text sample above.

## §166. Rendition of kamorated letters

The use of kamora is usual but inconsistent. The cases where kamora is missing are relatively rare; in the text sample above, see мытари (4) despite мытар҄ь (1), поставльше (51), поклонь (57), and others. Noteworthy are the occurrences of kamorated letters for phonemes н, л or р (hypercorrection effect); see кънижьн҄ици (2), жен҄ихъ (20), брьвн҄а (29) and свин҄иѣми (36) in the text sample above.

# §167. Substitutive softening

In a number of spellouts the expected *l-epentheticum* is missing. Cf. на земи (58) and наземл҄и (63) in the text sample above; likewise пристѫпь Mt 8, 19, корабь Mt 9, 1. Note also spellouts with kamorated labials: на зем҄и Mt 9, 6; пристѫп҄ь Mt 25, 24, despite пристѫпл҄ь Lk 10, 34; пристав҄енье Lk 5, 36, despite приставл҄еньѣ and приставл҄ѣетъ (Lk 5, 36). See § 117, *Alternative pairings in the substitutive softening alternation*, § 141, *Modifying aberrations*.

# §168. Omission of intervocalic *j*

Occurs fairly often. In the text samples above: покаанье (9); cf. also дааниѣ Mt 7, 11, but даѣти (in the same verse), сѣаниѣ Mt 12, 1, and others.

### §169. Rendition of the *yers*

All aberrations are possible.

Strengthening: сѫчець (31), but сѫчьць (33); цръковь (47). Cf. also образось Mk 12, 16; родось Mt 11, 16; ꙇзо облака Lk 9, 35.

Fall: многъ (1), дни (21), брьвн҄а (29), псомъ (35), вси (48). Confusion: вьзлежѧще (1), вь н҄ѭ (26), бръвъно (33), цръковь (47), etc. Regression: пьетъ (4), покаанье (9), дьнье (19), etc.

§170. On spellouts for normalized combinations -ии- and -ыи-

The text samples above show the following aberrant spellouts: въ прѣл҄юбодѣани (50) despite въ прѣл҄юбодѣаниі (52), likewise съдравиі (6); никыже (72).

Likewise for ыи: дше нечисты Mk 5, 8, бѣсъновавы сѧ Mk 5, 18, отъ малыхъ сихъ Mt 10, 42, ѡбрѣты Mt 10, 39, въ днь сѫдъны Mt 11, 24; for ии: орѫжьіми Mk 14, 48; нъ бол҄ьі васъ· да бѫдетъ ѣко ꙇ мьн҄иі Lk 22, 26. Cf. отыметъ Mk 4, 15 (for canonical отьметъ3SgPrae from отѧти). Aberrant spellouts with *yers* strengthening: завѣтъ стоі (свѧтоі) своі Lk 1, 72; крѣпл҄еꙇ Mk 1, 7.

§171. Additional commentary on samples


# Codex Assemanius

#### ട്ട് 172. lext samples

#### Luke 16, 24-31 (As 55b, 10-55d, 11)

'оче аврааме пом'оун ма' 2 и посьлі лазара' да омочіт'ї конець пр'вста своєго въ водъ: 4 и оустоудит възкь мон: 5 вко страждж въ пламени сельъ брече же емоу авраал: 74-адо помъні: 15ко въсприналъ есі благаа твоъ въ животъ твоемъ: °н лазаръ такожде зьлаа: "нънъ же сьде оутъшаєтъ са а тъи страждеши: 1°н надъ въсъмн сими междю вали и нами: "Пропасть вели в оствръди сл: 17-вко да хотации мінжті не възмаганятъ: 13ни сже отъ тждѣ къ намъ пръходатъ: 4 рече же молих та: фуво оче: да и посьлешь въ дом оца моего: имамь во пьють вратия: 15 кв да съвъдътельствотетъ ниъ да не и ти приджтъъ на мъсто се лижчьное: 10гла же авраамъ: 17имяттъ мосеа и прокъи: 18 да послоушаютъ нуъ. 190нъ же рече ни обе авраале: 20 гъ аще кто о" мрътвънихъ идетъ къ нимъ: 21 покажть са: 22рече же ємоу: 23 аще мосеа и прокь не послоушаннув: 24 ни аште кто отъ мрътвъннуъ въскръснетъ: 25не имятъ върън:

#### Luke 10, 33-37 (As 59b, 9-59c, 13)

26 самартьным же етеръ грьды: 27 приде надь нь 20 и видѣвъ и мосрдова: 22 и пристяпль облаза строупът его: 30 възливаю ольн. 31 в вино: 32 въсаждъ же и на свои скотъ: 33приведе и въ гостиньниця: 34и прилежа елъ: 35и на оутриа ншедъ 36 ВЬЗЕМЪ ДЪВА ПЪНАВА: 37 дастъ гостинником. 38 рече емоу прилежи емъ: 3 н еже аште приждівеши: 40азь егда възвраштът сл въздамъ тт: 41квто оуво отъ тъх три мънитъ ті са бънти искрьни въпадшоумог въ развонникън: 40мъ же рече сътворен мость съ нимъ: 43 рече же елюу исъ- 4 иди и тъи твори такожде

#### Matthew 13, 24-42 (As 126a, 12-126b, 28)

45 подовъно есть црство несное: 40 чкоу свавьшоу довроє съмл на селъ своємъ 47 съплинемъ же члемъ 48 пріде врагь его. 49 въства плъвелъ по средъ пшеница И ОТІДЕ 50 ЕГДА ЖЕ ПООБАВЕ ТОВВА И ПЛОДЪ СЪТВООГ 5 ТОГДА ЪВША СА И ПЛЪВЕЛІ 5° ПОНШЕДЪШЕ ЖЕ РАБІ ГДНА ОВША ЕЛЮУ ГИ· 53 НЕ ДОБОЕ ЛИ СЕМАЛ СЪАЛЪ ЕСІ НА сель своемъ: 54 отъ кждоу оусо имать плъвелъ: 50 мъ же (ре)че имъ: 56 врагъ члкъ то сътвори 570нг же ръша: 58хоштеши лг 99да шедъще възверємъ њу 60 мъ же р че бънн. 62 еда како въстръганяште плъвелъ. Въстръгнете сь ними и пшеннця: 64 оставите коупно расти обоє до жентвъ: 65 въ връма житвъ: 6 рекл жытельнемы: 6 шедъше изверъте прежде плъвелъ: 8 съвъжате ма въ снопъ тко съжешти м. 6 а пшениця съверъте въ житьниця можъ 7 пристжплъше же оучениці его 7 рубша ємор гі: 7 съкажі намъ притчя плъввлъ селънъ ихъ 3 3 Wвъщавъ же гъ рече имъ: 7 въсъавъ! доврое съмля естъ снъ чачь: 75 а село естъ весь лиръ: 7 доброе же съма син сять сбіве цротвиъ: 7 а плъввли случаснепривании 7 а врагъ въсть въсть длаволъ: "а жнатва кончание въка єсть: 8°а жьятельне антал сять: 81 вкоже оуво плъвелі съвіраннуть слі 8° н огнемъ съжаганятъ са. 8 тако бидетъ и въ съконьчание въка сего. 9 посьлетъ гь анћлъи своѩ· 85и сьберѫтъ ѿ конець землѧ въсѧ съблазны· 86и творѩщѧѩ безаконие· 87и въвръгѫт ѩ въ пещь огньнѫ·

# §173. Basic writing system

The basic writing system of the *Codex Assemanius* is given in Table 173.


Table 173. Basic writing system: *Codex Assemanius*

«ш» stands for shibilants, C stands for simple consonants.

# §174. Departures from the basic writing system

Rare departures are possible. Here are some examples.


# §175. Allographs of the letter щ

The allographs щ and щ are used in As as free variants. Cf. in the text samples above: аще (23), аще (24), etc.

# §176. Allographs of the letter и

Three allographs are used in As: и, і, and ꙇ. Cf. in the text samples above: ни ꙇже (13), гостіньницѫ (33), and гостиннику (37), etc.

# §177. Allographs of the letter ы

The shapes ы and ъи are used for the letter ы. Cf. in the text samples above: тъи (9), вѣръи (25), бъити (41), but грѩды (26), etc. See § 183 below.

# §178. Mixing of the letters з and ѕ

Occurs infrequently. Cf. in the text samples above: проѕѧбе (50) alongside the regular пѣнѧѕа (36).

# §179. Rendition of kamorated letters

The kamora is not used. Cf. in the text samples above: молѭ (14), посьлеші (14), землѧ (85), likewise за нѭже Lk 8, 47 (57d, 24); cf. also разорѭ Lk 12, 18. There is a small number of aberrant spellouts, cf. створѫ Jn 14, 13 (29d, 11), but cf. the standard spellout in the preceding verse творѭ Jn 14, 12 (29d, 3).

# §180. Substitutive softening

The expected *l-epentheticum* is absent in some spellouts: cf. оставѭ Jn 14, 18 (30a, 5–6), вьзлюбень Jn 14, 21 (30a, 25–26) and вьзлюбѫ Jn 14, 21 (30a, 28–29), but soon afterwards ѣвлѫ (30a, 29). The spellout земи occurs constantly (e.g., Mk 9, 20 [77c, 25]), cf. also the spellout землѧ (85) with the ligature м+л, likewise лоно авраамле Lk 16, 22 (55b, 1). Spellouts with *l-epentheticum* predominate, cf. in the text samples above: прістѫпль (29), пристѫплъше (70), cf. also иаковлѣ Mk 16, 1 (11d, 24), оставлениние Mk 1, 4 (137b, 23, for оставлѥние). Imperfects like любѣаше Jn 11, 5 (79d, 4–5) can be treated as imperfects with a PRAE workstem (see details in § 469, *Present imperfect*), or as resulting from aberrant substututive softening. See § 117, *Alternative pairings in the substitutive softening alternation*, § 141, *Modifying aberrations*.

### § 181. Omission of intervocalic j

Occurs fairly frequently. In the text samples above: ΕλλΓΔΑ (7), 3ьΛΔΑ (8), ςΈΑΛ'Σ (53), въсъавън (74), etc. Cf. however: вътора в Mk 12, 31 (115d, 3), дакахя Mk 15, 23 (104a, 15), дроуга в Mk 3, 5 (75a, 29).

### § 182. Rendition of the yers

All aberrations are possible, except for yer regression.

Strengthening: конець (3), сътворен (42), вєсь (75), etc .; cf. also тько оученикотъ не оумьретъ Jn 21, 23 (31c, 1-2), мирос (for миръ сь) Jn 9, 39 (22c, 13); иноплеменьникось Lk 17, 18 (63c, 5-7).

Fall: авраам (6), въ дом (14), кто (20), гостинникоу (37), въвръгжт (87), etc. Confusion: посьлі (2), пръста (3), зьлаа (8), октвръди (11), надь нь (27), кьто (41), врагь (56), пристяплъшє (70), въсл (85), etc.

\$ 183. On spellouts for normalized combinations -ии- and -тый-

Aberrant spellouts in the text samples above: въставъ! (74) and въставъ! (78); forms that follow the canon are e.g.: χοταμιά (12), мрътвъникъ (20) and (24), приждівеши (39), трин (41, GPI), искрьнии (41), селънъпиуъ (72), син (76).

Other aberrant spellouts: да не видации відать Jn 9, 39 (22c, 14), and immediately below и видации слеп кждя" (22с, 15-16); пастъирь довръи Jn 10, 11 (114a, 19), and soon afterwards пастъирь довръин Jn 10, 14 (114b, 8); cf. also азь же въмъ и Jn 8, 55 (21c, 23); оукнъмо и Mt 21, 38 (46d, 5-6). Note here also the hypercorrect spellout въ срдциихъ вашкъ Mk 2, 8 (76a, 18).

### § 184. Additional commentary on samples


# Psalterium Sinaiticum

§ 185. Text samples

```
Psalm 12
```

```
1 доколъ гі завждешь мых до конъца 2 доколь отъвраштаеши лще твое от мене
3 доколь положя съветъ въ дши моен. Теользувъ въ создьщи моемъ друг п
ноштъ доколъ възнесетъ съл врагъ мог на мна: "призъри оуслъши мю гі вже мог
```
7просвѣті очі моі еда когда усънѫ во съмръті· 8еда когда речетъ врагъ моі (о) укрѣпіхъ сѩ на нъ· 9сътѫжаѭщеі мі въздрадуѭтъ сѩ· 10аще сѩ подвіжѫ· 11азъ же на мілостъ твоѭ надѣахъ сѩ· 12въздрадуетъ сѩ сръдъце мое о спі твоемъ· 13поѭ гю благодѣавъшюму мнѣ· 14і вьспоѭ імені гю въішьнюму·

#### Psalm 50

15помилуі мѩ бже по веліцѣі милості твоеі· 16ї по многыимъ щедротамъ твоімъ оцѣсті безаконньѣ моѣ· 17наїпаче омыі мѩ отъ безаконеньѣ моего· 18ї отъ грѣха моего очісті мѩ· 19ѣко безаконнье мое азъ знаѭ· 20ї грѣхъ моі прѣдъ мноѭ естъ вынѫ· 21тебѣ едіному съгрѣшіхъ ї зълое прѣдъ тобоѭ створихъ· 22ѣко да оправьдіші сѩ въ словесе твоіхъ· 23ї прѣпьріши вънегъда осѫдиті сѩ· 24се бо въ безаконены зачѩтъ есмъ· 25ї въ грѣсѣхъ роді мѩ маті моѣ· 26се бо рѣснотѫ възлюбилъ есі безвѣстьнаа и таинаа прѣмѫдростъ твоѭ ѣвілъ ми есі· 27окропіші мѩ ософомь очищѫ сѩ· 28омыеші мѩ паче снѣга убѣлѭ сѩ· 29слуху моему дасі радостъ и веселье· 30въздрадуѭтъ сѩ кості съмѣреныѩ· 31ѡтъвраті ліце твое отъ грѣхъ моіхъ· 32ї вьсѣ безаконньѣ моѣ оцѣсті· 33срдце чісто съзіЖі въ мнѣ бже· 34ї дхъ правъ обнові въ ѫтробѣ моеі· 35не отъвръѕі мене отъ ліцѣ твоего· 36ї дха стаго твоего не отыми отъ мне· 37въздаЖь ми радостъ спнѣ твоего· 38ї дхмъ вдадычънемъ утвръді мѩ· 39научѭ безаконныѩ пѫтемъ твоімъ· 40ї нечьстівії къ тебѣ обратѩтъ сѩ· 41їзбави мѩ отъ кръві бже бже спнѣ моего· 42въздрадуетъ сѩ ѩзыкъ моі правъдѣ твоеі· 43гі устънѣ моі отвръзеші· ї уста моѣ възвѣстѩтъ хвалѫ твѫѭ· 44ѣко аще бі вьсхотѣлъ жрътвѣ далъ бімъ убо· 45вьсесъжагаемъїхъ же не благоволіші· 46жрътва бу дхъ съкрушенъ· 47срдца съкрушена и съмѣрена бъ не учъжітъ· 48ублажі гі благоволеньемъ твоімъ сиона· 49ї да съзіЖѫтъ сѩ стѣны· їимъскы· 50тогда благоволіші жрътвѫ правьдѫ· 51възношеньѣ въсесъжагаемаа· 52тогъда възложетъ на олтръ твои тельцѩ·

#### Psalm 71

53сънідетъ ѣко доЖь на руно· 54ї ѣко каплѣ капѫщіѣ на землѫ· 55вьсіѣетъ въ дьні его правъда· 56і мъножъство міру доіЖеже отыметъ сѩ луна· 57ї обладаетъ отъ морѣ і до морѣ· 58і отъ рѣкы до конеца вьселеныѩ· 59прѣдо нь пріпадѫтъ етіопѣні· 60ї враѕі его пръсть поліжѫтъ· 61цѣсарі тарьсьсіції отьці дары прінесѫтъ· 62цѣрі аравьстїі саво дары прінесѫтъ· 63ї поклонѩтъ сѩ ему цѣрі земльстії· 64вьсі ѩзыці поработаѭтъ ему· 65ѣко ізбавілъ еестъ ніща отъ сілъна· 66і убога емуже бѣ помощьніка· 67пощедітъ ніща і убога· 68ї дшѩ убогыхъ спетъ· 69ѡтъ ліхвы и отъ неправъды ізбавітъ дшѩ іхъ· 70і честьно їмѩ его прѣдъ німі· 71ї жівъ бѫдетъ дастъ сѩ ему отъ золъта аравііска· 72і помолѩтъ сѩ ему вынѫ· 73вьсь день благоствітъ і·

#### §186. Basic writing system

The basic writing system of the *Psalterium Sinaiticum* is given in Table 186, p. 104.

§187. Departures from the basic writing system

Rare departures are possible. Here are some examples.



Table 186. Basic writing system: *Psalterium Sinaiticum*

«ш» stands for shibilants, C stands for simple consonants.

# §188. On the spellouts of the word ангелъ

The following spellouts are attested in Ps Sin: аѧг҄ел- (6×), e.g. прѣдъ аѧг҄елы Ps 137, 1; анг҄ел- (3×), e.g. вьсі анг҄лі Ps 102, 20, and аг҄ел- (2×), e.g. аг҄елы люты Ps 77, 49. The letter ⱔ (transliterated as ѧ) is not attested outside the aforementioned spellouts.

# §189. Allographs of the letter щ

Two allographs are used in Ps Sin: щ and щ. Cf. щедръ Ps 85, 15 and щедръ Ps 102, 8; see also the examples in the text samples above отъвращаеші (2), нощъ (4), ніща (65), аще (10), щедротамъ (16), очищѫ (27), etc.

§190. Allographs of the letter и

Three allographs are used in Ps Sin: и, і, and ї. See in the text samples above: наїпаче (17) and others. See § 132.

# §191. Allographs of the letter ы

The allograph mainly attested in Ps Sin is ы, but cf. вышьнѣго Ps 90, 9.

# §192. Mixing of the letters з and ѕ

The occurrences of з in place of the expected ѕ are infrequent. Cf. отъвръзѣмъ (Imv for отъврѣщи) Ps 2, 3. In the text samples above we have не отъвръѕі (35), враѕі (60).

# §193. Rendition of kamorated letters

Kamora is not used. In the samples above, according to the basic writing system: убѣлѭ (28) and землѫ (54), въішьнюму (14), and many others. However, vacillation is possible, cf: вышьнуму Ps 7, 18 and 81, 6; пріклонѫ Ps 48, 5; възвеселѭ сѩ Ps 103, 34 and вьзвеселѫ сѩ Ps 30, 8; похвалѭ Ps 55, 5 and похвалѫ Ps 55, 11.

### §194. Substitutive softening

In a number of spellouts the expected *l-epentheticum* is missing. In the text samples above we have капѫщіѣ (54), but каплѣ (54). Cf. the spellouts of the wordforms of земл҄ꙗ: земъѣ (cf. Ps 32, 5) and землѣ (cf. Ps 32, 8); земъѩ (Ps 21, 30), земьѩ (cf. Ps 80, 11) and землѩ (cf. Ps 1, 4); на земъѭ (Ps 16, 11), земѭ (Ps 77, 55) and землѭ (cf. Ps 84, 10), землѫ (cf. Ps 79, 10); земі (cf. Ps 43, 26) and землі (cf. Ps 40, 3).

See § 117, *Alternative pairings in the substitutive softening alternation*, § 141, *Modifying aberrations*.

# §195. Omission of intervocalic *j*

Occurs fairly often. Cf. in the text samples above: надѣахъ (11), безвѣстьнаа (26), таинаа (26), въсесъжагаемаа (51).

### §196. Rendition of the *yers*

All aberrations are possible.

Strengthening: во съмръті (7), прѣдо нь (59), честьно (70), день (73), etc. Cf. also во съмрът-і-хъ Ps 72, 4 (for въ съмрьти ихъ); нечестіѭ Ps 72, 6; прішьлецъ Ps 38, 13 and прішельці Ps 104, 12; деньніцѩ Ps 109, 3; день Ps 36, 13; денъ Ps 18, 3; likewise весь Ps 36, 26 and весъ Ps 19, 5; врабеі Ps 10, 1, cf. also сѫдобъ (for сѫдьбъ) Ps 47, 13; ложь Ps 115, 2, крѣпокъ Ps 7, 12, собьра Ps 40, 7, возъвахъ Ps 65, 17; во-чьсті Ps 48, 21, etc.

Fall: от (2), мнѣ (13), створихъ (21), etc.

Confusion: конъца (1), вь (4), на нъ (8), вьспоѭ (14), правъдѣ (42), мъножъство (56), сілъна (65), etc.

Regression: веселье (29), безаконеньѣ (17), безаконнье (19), безаконньѣ (32), възношеньѣ (51), etc. Cf. also безаконнъе Ps 54, 11, велъѣ (for canonical велиꙗ) Ps 49, 3, прѣрѣканъѣ Ps 30, 21.

Note cases of *yer* insertion: вънегъда (23), cf. also вънегда Ps 4, 2; тогъда (52), cf. тогда (50); cf. also когъда Ps 37, 17 and когда (7).

§197. On spellouts for normalized combinations -ии- and -ыи-

Aberrant spellouts in the text samples above are the following: безаконены (24) (for безаконьнии LSg), вьсесъжагаемъїхъ (45), убогыхъ (68); forms that follow the canon are: многыимъ (16), омыі (17), нечьстівії (40), земльстії (63), аравііска (71), etc.

Other aberrant spellouts: прѣмѫдроі (for прѣмѫдрыи NSgmPlen) Ps 48, 11, пожрѣхомои (for пожрѣхомъ и) Ps 34, 25; змъи Ps 103, 26; выистълѣнъе (for въ истьлѣниѥ) Ps 29, 10, etc.

§198. Additional commentary on samples


# *Kiev Missal*

§199. Text samples

2b, 8–17

1мьшѣ на вьсѩ дьнꙇ вьсего лѣта обідѫцѣ· 2бъ ꙇже тварь своѭ вельмі помілова· ꙇ по гнѣвѣ своемь· 3изволі въплътиті сѩ съпасениѣ раді чловѣчьска· 4ꙇ въсхотѣвъ намъ утврьді срьдьцѣ нашѣ· 5ꙇ милостиѭ твоеѭ просвѣті нъи·

3a, 9–15 6по въсѫдѣ· 7просімъ тѩ гі дазь намъ· 8да свѩтъі твоі въсѫдъ пріемлѭце достоіні бѫдемъ очішчениѣ твоего· 9ꙇ вѣра твоѣ въ насъ да въздрастетъ·

4b, 1–6 10по въсѫдѣ· 11дазь намъ· вьсемогы бже· 12да ѣкоже нъи есі небесьскъиѩ піцѩ насъитілъ· 13такозе же ꙇ животъ нашь сілоѭ твоеѭ утврьді·

6a, 1–8 14надъ оплатъмь· 15прінесены тебѣ гі сьі даръ ꙇже тъи есі далъ ꙇ сътворꙇлъ· 16цꙇръкъве раді твоеѩ· 17ꙇ жівота ꙇ прѣставлениѣ нашего раді· 18ꙇ съвѣстуемъ нъи· 19ѣко бальство естъ то живота вѣчьнаго·

# §200. Basic writing system

The basic writing system of the Kiev Missal is given in Table 200. The limited size of the source does not allow us to establish the basic writing system unambiguously for all positions shown in the table. Thus, there are no examples with /a/ after kamorated and some other consonants.


Table 200. Basic writing system: *Kiev Missal*

«ш» stands for shibilants, C stands for simple consonants.

# §201. Departures from the basic writing system

# Here are some examples.

ѣ, а after shibilants and ц, ѕ, in the text samples above: мьшѣ (1), обідѫцѣ (1), etc., cf. съгрѣшаті 4a, 21; cf. also срьдьцѣ нашѣ APl (4) and срьдьца нашѣ APl 5a, 21.

# §202. Rendition of the letter щ

In place of the letter щ of the normalization Kiev shows either ц, or шч (see § 207 below).

# §203. Allographs of the letter и

Three allographs are used in Kiev: и, і and ꙇ.

# §204. Allographs of the letter ы

Two allographs are used in Kiev: ы and ъи. Cf. in the text samples above: свѩтъі (8), вьсемогы (11), прінесены (15), and many others for NSgmPlen; otherwise: нъи (5), нъи (12), небесьскъиѩ (12), насъитілъ (12), тъи (15), нъи (18); likewise небъитіѣ 4a, 17; даръи 5b, 2; свѩтъихъ 2a, 20, etc.; however, in one occurrence вьсемогъиꙇ бже 2a, 13 and in one occurrence даръ сь принесенъи 7a, 20. See § 210 below.

### §205. Mixing of the letters з and ѕ

Cf. дазь (7), (11) for даѕь; даѕь, in turn, by the aberrant pairing д‖ѕ instead of canonical д‖Ж. Cf. also такозе (13), тузꙇмъ 4b, 10–11. Kiev shows no examples of expected ѕ other than due to the application of this mechanism.

### §206. Rendition of kamorated letters

Kamora is not used. Thus, according to the basic writing system, in the samples above we have: пріемлѭце (8); likewise въишьнімі 3a, 4; беж негоже 6a, 14–15, любъвь 2b, 23, людиꙇ 2a, 18.

# §207. Substitutive softening

The *l-epentheticum* occurs as expected: пріемлѭце (8), прѣставлениѣ (17); cf. also земльскаѣ 4a, 22–23; въжлюбленъиѩ 5a, 3–4.

In place of canonical т‖щ, д‖Ж, and ст‖щ in Kiev we find т‖ц, д‖з, and ст‖шч; likewise ц in the suffixes of the *št*-participles and ц as the result of the simplification of the clusters к.т and г.т. In the samples above: обідѫцѣ (1), дазь (7), пріемлѭце (8), очішчениѣ (8), дазь (11), піцѩ (12). Likewise: насъицені

6a, 22–23; обѣцѣниѣ 2a, 8–9; обѣцѣлъ 3b, 11; зашчіті 2a, 17; помоць 6b, 15; помоцьѭ 5a, 17. Note here also такозе (13) and тузꙇмъ 4b, 10–11.

See § 117, *Alternative pairings in the substitutive softening alternation*, § 141, *Modifying aberrations*.

§208. Omission of intervocalic *j*

Not observed. However, contracted forms in the adjectival declension are usual, as in вѣчьнаго (19).

§209. Rendition of the *yers*

Overall, *yer* aberrations do not occur. See however the spellout оплатъмь (14), and also 2b, 18; 3a, 24; 4a, 7; 4b, 16; 5b, 1, as opposed to оплатмь 1b, 9; 6b, 10; 7a, 19; and the spellout въсѣхъ 7a, 22–23.

Cases of regression in the text samples above: сьі (15), likewise сьі NSgm 3b, 1 and помоцьѭ 5a, 17.

§210. On spellouts for normalized combinations -ии- and -ыи-

In the text samples above such are the spellouts свѩтъі (8), вьсемогы (11), прінесены (15); although вьсемогъиꙇ NSgm 2a, 13, cf. also принесенъи NSgm 7a, 20. See also § 204.

§211. Additional commentary on samples


# **Sava's Book**

§212. Text samples

Mark 9, 17–29 (Sav 78–79)

1члкъ единъ приде къ ісвї· 2кланꙗѧ сѧ ему и глѧ· 3учителю приведохъ снъ моі· 4къ тебѣ· 5имѫщъ дхъ нѣмъ и глухъ· 6иже аще колижьдо иметъ і· 7разбиетъ і· 8и пѣны тѣщитъ· 9скрьгьщѧ зѫбы своіми· 10и оцѣпѣнѣетъ· 11и

3 Cf. Večerka оплатъ 'Eucharist', borrowed from Latin *oblata*. This word is not part of the benchmark list in this grammar, absent from both PD and RD.

ръхъ очченикомъ тволлъ да ижденжтъ г 24 не възмогж: 33 онъ же рече имъ: 14 к роде невърьнъ! 15 доколъ бядж съ вали: 16 доколъ трьпля вън: 17 принесътте 1 къ мнгъ и принесоша къ немоу. 18 видъвъ дуъ сътрасе г. 3 и падъ на земн валъше са: 20пънът тъща: 21 въпроси іс оца его: 22колико лѣтъ отъщелиже: 23 се Бъютъ емь - 24 онъ же рече емог. 25 из отрочныхъ 20 м многаши и въ огнь въврьже: 27 в въ водж да с въ потопилъ. 28 нъ аще хоциеши помози мн. 29 милосоьдовавъ о мнъ: 30 % же рече емоу. 3 ацие можеши въровати. 32 вса възможъна въроуняциюмочъ 33 и авне възъпи оць: 34 отрочате съ слъзалин гла: 35 въроуск ги полюзи моемоу невърьствию: 36 видувъ же іс: 37гако сърищетъ са народъ 38 запръти дубу нечнстокмоу. 3 гла емог немъ гл8 учь дше. 40 зъ ти веля изиди из него. 4 н Възъпивъ и много са пряжавъ: ""нзиде и бъстъ гако мрътвъ" 43и мнози глахж гако Умреттъ: 41 с же ї ниъ за ряки: 45 въздвиже и поставн: 40 въшъдъщию емог вь домъ: "Тоученици его въпрашахж и единого" "Чако луд не могохомъ нзгнати єго 4°п рече имъ 5°тъ родъ ничилъжє не можетъ изнти. 51 тъкмо МОЛИТВОЈЖ И ПОСТОМЪ

#### Matthew 14, 15-21 (Sav 39v)

5°поздъ же бъвъшко припадях оучеци его гажцие 33поусто е место. 5 и годъ юже миня: 550 тъп8сти народъ: 5 да шъдъще въ окрыстьная вси: 57 коупатъ севъ врашъна: 59г. же рече нлуъ: 5 даднте въ ниъ гасти. 60ни же гааша: 6 не нмаль съде 62 тъкмо е хлъвъ 63 двъ ривъ. 64 онъ же рече принесъте съмо 65 и повелъ народоу: 66 възлеши по тръвъ: 67 приилъ с хлъвъ: 68 объ ръвъ: 6°и възьръвъ на нео сти: 7°н преломъ дастъ оученикомъ хлъбъ: 7 а оученици народомъ: 72и фша вси и насъстиша са: 73 и възаша извърша извършуюмъ 7461 75кошъници плънъ: 76 и адъшихъ съ: 7 мяжъ е тъсаци: 7 развъ женъ и дѣти:

#### Matthew 14, 22-23, 25-34 (Sav 40v-41)

7°016° С ОУЧЕНИКЪІ СВОА ВЪЛЪСТИ ВЪ КОРАБЪ. °° И ВАРИТИ ЕГО НА ОНЪ ПОЛЪ МООВЪ' 8 доньдеже отъпУститъ народъ 8 и отъпоущь народъ възиде на горж единъ. 83 пололитъ са: 94 поздъ же бъвъшю въ дж же годиня нощи: 8 приде къ нилъ ic по мороу хода: 86н видѣвъшє і фученнци єго по мороу χодащи: 87 оукогаша са гляцие: 88гако призракъ естъ: 9°и отъ страха възъпнша: 9°н гла имъ к: 2+надъктте са гако азъ ссмь и не когте са. 220 гъвъщая же петръ рече емоу ги. 99 аще тъв €60 повели мн прити къ севъ: 94,6 же рече емоу придн: 95излъзъ же ис коравлъ петръ: 96 хождаше по водъ и приде къ 168. 97вида же вътръ лютъ оугога сах 90 нача потаплетн са: 9% възъпи гла ги спс ма: 10°г же простьръ рякж АТЪ ЕГО: 101 г. ГЛА ЕМОУ МАЛОВЪРЕ: 102 ПОЧТО СА СЖЛИЧЪ: 103 И ВЪЛЪЗЪШЮ ЕМОУ ВЪ коравь оулеже вътръ 104а сяще! въ коравн поклонишь са емоу гляще: 105тако тъ сен въ истинж вжи сбъ 106 и препловъвъше: 107 придж въ земья генисарефь

#### §213. Basic writing system

The basic writing system of Sava's Book is given in Table 213.


Table 213. Basic writing system: *Sava's Book*

«ш» stands for shibilants, C stands for simple consonants.

§214. Departures from the basic writing system

Rare departures are possible. Here are some examples.


§215. Allographs of the letter щ

Only щ occurs in Sav.

§216. Allographs of the letter и

Three allographs are used in Sav: и is the main one, but also і, and, less frequently, ї can occur. See, in the text samples above: приде къ ісвї (1), іс же ї имъ (44), etc.

§217. Allographs of the letter ы

For ы we always have ы, in Sav; the only exceptions are: въїѭ Lk 15, 20 (68, 16), недѫжънъѧ Mt 14, 14 (39v, 2–3); note also the spellout ѧзоікомъ Mt 12, 18 (142v, 12).

# §218. Allographs of the letter у

Two allographs are used in Sav: у occurs normally, less frequently; cf., in the text samples above: ем (23), гл хы (39), мрѣтъ (43), отъп сти (55), прѣпл въше (106), etc.

# §219. A special spellout for the letter ѧ

Spellouts with the letter ꙙ as an allograph of ѧ can occur. Cf. бышꙙ Jn 15, 24 (104, 14), овьцꙙ Mt 25, 32 (70, 16–17), клꙙти сѧ Mt 26, 74 (98v, 5), etc.

# §220. Mixing of the letters з and ѕ

The letter ѕ is missing in Sav, only з occurs; cf., in the text samples above: помози (28), помози (35), мнози (43). Cf. however the letter ѕ in the numeric value of '6': отъ ѕѧ же годины Mt 27, 45 (119v, 2).

### §221. Rendition of kamorated letters

Kamora is not used in Sav. Thus, л҄ꙗ and р҄ꙗ are rendered as лѣ and рѣ, while н҄ꙗ is usually rendered as нꙗ. Such are, in the text samples above: кланꙗѧ (2), валѣше (19), морѣ (80), кораблѣ (95), потаплѣти (98); likewise испльнꙗѧ Lk 2, 40 (1392, 10), but испльнѣѧ Lk 2, 40 (144v, 1). Cf. also: учителю (3), трьплѫ (16), къ нему (17), въ огнь (26), велѫ (40), из него (40), въ окрьстьнѧѧ (56), къ нимъ (85), лютъ (97). However, cf. по мору (85), по мору (86) with ру in place of the expected рю; cf. also црю Mt 18, 23 (43v, 2), Mk 15, 18 (117v, 18).

# §222. Substitutive softening

In a number of spellouts the expected *l-epentheticum* is missing. Such are, in the text samples above: на земи (19), прѣломь (70), корабь (79), корабь (103), кораби (104), земьѭ (107). Cf. also: NSg земьꙗ (4×) and землѣ (2×); ASg земьѭ (3×) and землѫ (7×); GSg земьѧ (2×) and землѧ (1×); LDSg only земи.

See § 117, *Alternative pairings in the substitutive softening alternation*, § 141, *Modifying aberrations*.

# §223. Omission of intervocalic *j*

Relatively unfrequent, cf.: покаание Mt 9, 13 (36, 4), Mk 2, 17 (80, 2), Mk 1, 4 (145v, 19), сѣаниꙗ Mk 2, 23 (73v, 8), etc. However: дѣꙗшѧ Mt 26, 67 (98, 8), дѣꙗхѫ Mt 26, 67 (112, 14), знаꙗше Mt 1, 25 (137, 17–18), грѣꙗхѫ сѧ Jn 18, 18 (110v, 7); съмѣꙗше Mt 22, 46 (47v, 9) and съмѣше Jn 21, 12 (164v, 6).

§224. Rendition of the *yers*

All aberrations are possible with the exception of strengthening.

Fall: къ мнѣ (17), многащи (26), о мнѣ (29), вса (32), много (41), изгнати (48), etc.

Confusion: имѫщъ (5), скрьгьщѧ (9) (for canonical скрьжьтати ‹837›), възможъна (32), слъзами (34), вьздвиже (45), въшъдъшю (46), вь (46), ничимъже (50), постомъ (51), съде (61), вьзьрѣвъ (69), плънѣ (75), etc.

Regression is not attested in the text samples above. In the manuscript as a whole there are sporadic examples, cf. на трьꙗ (121, 1) in the title, and also недѫжънъѧ Mt 14, 14 (39v, 2–3).

§225. On spellouts for normalized combinations -ии- and -ыи-

Such are, in the text samples above, the following spellouts: невѣрьны (14) with the NSgmPlen, terminal ъ*i̯*.ь, likewise нѣмы (39), гл хы (39); ꙗдъшихъ (76); but: дѣтиі (78), приімъ (67), etc. Cf. also Imv умы Mt 6, 17 (73, 11); LSg ѡ окамененъі Mk 3, 5 (74, 14–15); больи Jn 15, 2 (27, 16; 102v, 2–3), дарьи Mt 26, 68 (112, 16), cf. удариі Mt 26, 68 (98, 10); GPl дьньі Mt 24, 29 (88, 7), NSgm искрьньі Lk 10, 36 (56v, 5).

§226. Additional commentary on samples


# Codex Suprasliensis

### ട്ട് 227. Text samples

### 217, 25-219, 94

СЪЛОУЧИ СА КОУПЬЦОУ НЪКОТОРОУОУМОУ ГЪНАВЪШОУ ВЕЛЬБЕЖДЪІ СВОЛ И ПАСТИ А на мѣсть томь' пасжштємъ жє са вєльсядомъ тоу. 3 по прилоучаю ютъ АНУЪ ШЬДШИ: "БЛИЗЪ ВЬСИ ВЪЛЪЗЕ ВЪ ЙНВЖ ГАСТИ ХОТАШТИ: "ТЯ ЖЕ ВИДЪВЪ господінъ нивъютом: чтече изгънати их: "изгонила же въпаде са въ пъровъ" 8 н изломи ноги предьням: 3и которъ въвъши междоу господиномъ вельежда 10 господиномъ fings: "повръже ня въ горъ: "на мъстть идеже лежаше сватью: 13невн||димь никъмьже: 44 шьдь оуво господинь ел вь курєстннь градь 15 повъда кьназор о нивъ: 19и како оклоснишл емоу вельвядь: "Вельвалоу же оставеноу бывьшоу тоу до четырь дьнин: 18 лежаштоу на муссть и ПОКОУШАЖШТОУ СА ВЬСТАТИ' 12 И РАЗЛОМЕНЫА НОГЫ: 2° НЕ МОГЖШТОУ ОУТВОЬДИТИ' 2 стъпи: 22 въниде кемог нога въ доупи | ня пештеръмяня: 23 плеже бъдуж мошти сватъюзъ 24 авне оутврьди са нога вельеждоу и бъютъ съдрава: 25 пришъдъ же господинъ его. 20 притече къ немоу. 27 вельвядъ же видъвъв и. 2 нетръже ногж свою и тече и сървте и. 20012 же видѣвъ и и въ оужасти във. 30 увай вше н славьташе господа. 3 въпрашаахъ же его дроужина кемоу. 32 что ест, преславъное се: 33 онъ же рече нлуъ азъ оставихъ и на мъстъ уромъ сящитъ 3 они же ръша кмор гради и покажи намъ: 35 да видимъ извъстьнъю мъсто то: 36 н шъдъше на мъсто обрътоша доупиня велики: 37и авни пъкотории отъ нихъ реша кмоу 39 по истинъ сватааго дометна место се естъ 39 оуго сьде ОБРАШТЕЛЪ И СТРЪЛА МОШТИ ИГО. 40И ШЪДЪШЕ ПОНВЕДОША ПОПЪЛ ДА ТВОДАТЪ молитвъ! на льстъ: 4-они же мотъжъ! и рълла въземъше копашь: 4-н оворътъше МОШТИ ИЗНЕСОША СЪ ПЪСНЕМИ И ХВАЛАМИ:

### 557, 23-559, 4

43 въ тѣхъ же лѣстъхъ развонници нѣцни тальштє са: 4 тако имѣньга обнлиє Бъяше въ манастъюн: 45 подъкопав ше стъня ноштьях: 4 вълъзоша въ манастърь: и вса съкроушив'ше: 47и въсьде понскавъше и ничсоже овръттъше: 48 HA CA OEDATTHILDA HEHCTOBOR OYCTOS.NEREHHE . 4 TELKTIN OTT HHY'S . SYOTA съспоспъшникъ своихъ оугасити въщенью: 5 въсувативъ съ земья камъык. 5 връже на главж пръподовьнааго: 3 и толли и газви: 5 такоже изволениимъ и дъло съконьчати: 5 правьдънъю же съ стлъпа хотъяше сълъсти: 5 стъин же анина авние объедъвъ сало то отъ стааго джа: 5 призъва кед'ного объчувъ 58 слоуго и тааго капо льва: 59 предъложи имог гади многъг 60 ггла немоу обило гаждъ: 6 хоштеши во се на пжть отити длъгъ: 6 и просивъ дартиях написа сице въ нен. 63 w оче гавыено ми въ! "Тако нападение нъкою отъ розвонникъ пострадавъ 6 хоштеши съ ставпа сьлъсти. 6 длъжънъ същитенние

4 Sever' janov gives a note to the text fragment marked by || (13)-(22): "written by the unskilled hand of scribe no. 2, whose writing teems with corrections".

маловрѣменьно въсприѧти· 67нъ устави устръмьѥнье то· 68да не вьзмьздьꙗ труднааго· 69рекꙿше црьствиꙗ вѣчьнааго лишиши сꙙ· 70тъгда же навꙙзавъ на выѭ звѣри· 71посъла ѥго заповѣдавъ· 72ни ѥдꙿному мимоидѫщиихъ пѫтьмь пакости сътвори· 73звѣрꙿ же устрьмьꙗше сꙙ къ манастыру· 74да съконьчаѥтъ повелѣноѥ· 75сълучаѭщии же сꙙ на пѫти· 76видꙙще издалеча· 77съ великыимъ устрьмьѥниимъ теченьꙗ приходꙙщъ· 78мнꙙще на изѣдениѥ имъ прѣдълежати· 79звѣрьнууму нашьствию· 80онъ же окы отъ пастуха женомъ· 81тако идѣаше на повелѣньѥ стааго· 82и ни ѥдному пакости творꙙ· 83пришъдъшу же ѥму къ манастиру· 84и ногътьми своими двьри дерѫщу· 85и їсшъдъ нѣкыи отъ живѫщиїхъ ту видѣ· 86и текъ повѣда учителю своѥму· 87глꙙ· ꙗко звѣрь прѣвеликъ стоитъ прѣдъ двьрьми· 88онъ же рече сътворивъ молитвѫ брате отврьзи ѥму да вьлѣзетъ· 89ѫтръ же бывъ звѣрь· 90и ногама прѣднима опъръ сꙙ о стлъпъ· 91выѭ же къ прѣподобьнууму горѣ вьздѣвъ· 92подаꙗше книжицꙙ прѣподобьнууму· 93онъ же приимъ и почьтъ· подиви сꙙ вьзвѣщению бывъшууму о томъ къ стууму· 94и послушанию дивиихъ звѣрии· 95и прославивъ ба· 96стави сꙙ· 97отъ начꙙтиꙗ· 98ѥже бѣ отъмъщеньѥ помыслилъ·

### 565, 14–566, 14

99юньць приведенъ на даръ бви· 100и ту кръмимꙿ въшьствиимъ лѫкааго бѣса вьздивьꙗ· 101и толꙿма гнѣвомъ движе сꙙ· 102ꙗкоже ни пастуха знати ни їного никогоже миловати· 103нъ вьсꙙ сълучаѭщꙙѧ сꙙ губити· 104иже урвавъ сꙙ отъ привꙙзаньꙗ· 105и їзъ ограды излѣзъ· 106ровы и гнѣваѧ сꙙ бе-чину· 107нападаѧ на сьрѣтаѭщꙙѧ ѥго· 108и устрьми сꙙ на вьнѫтрьн҄иї дворъ и вьлѣзе въ ꙗтъхульницѫ· 109на братьѭ ту дѣлаѭщѫѭ· 110на утѣхѫ довьлѣѭщаꙗ странꙿныимъ· 111ꙗкоже принуЖеномъ имъ быти боꙗзньѭ юньца· 112самѣмъ сꙙ въврѣщи въ пещъ хлѣбьнѫѭ· 113и їзволити паче въ огн҄и съгорѣти· 114нежели отъ того събоденомъ быти· 115видѣахѫ бо ѥго бечіслъноѥ бѣшеньѥ· 116чловѣколюбивыи же бъ· 117не опалимы ѧ съблюде· 118въ пламени огньнѣмь· 119молитвоѭ правьдꙿнааго пламы огн҄ьныи погасивъ· 120прочии же вьси отъ ужасеньꙗ юньча на хлѣвины горѣ вьлѣзошꙙ· 121правъдныи же учувъ плищъ бѣгаѭщиихъ· 122пытааже кыи извѣтъ сьмꙙтеньꙗ того· 123ꙗко юньць гнѣва испльнь бывъ и їс хлѣва искочивъ· 124гонитꙿ вьсꙙ на толицѣ· 125ꙗкоже ѥму и въ ꙗтъхульницѫ вьлѣсти· 126стыи же слышавъ глаголы ты· 127и въздѣвъ рѫцѣ на небо· 128и молитвѫ сътворивъ· 129иде на сьрѣтеньѥ юньцу· 130на хлѣвинахъ же стоѧщии· 131видꙙще ѥго устръмьѥниѥ къ юньцу· 132милуѭще правьднааго· 133ставьꙗхѫ и гласы· 134уклонити сꙙ гнѣва и възбѣшеньꙗ юньча· 135правьдныи же ничсоже никомуже не отъвѣщавъ· 136нъ вѣроѭ ꙗже на ба утвръдивъ сꙙ· 137приближивъ сꙙ къ юньцу· 138и коснѫвъ сꙙ его рѫкама· 139и въстлапивъ гнѣвъ словесы кротъкыми· 140рѫкоѭ поимъ· 141вьведе въ свои хлѣвъ къ ꙗслемъ·

# §228. Basic writing system

The basic writing system of the *Codex Suprasliensis* is given in Table 228.


Table 228. Basic writing system: *Codex Suprasliensis*

«ш» stands for shibilants, C stands for simple consonants.

## §229. Departures from the basic writing system

Rare departures are possible. Here are some examples:


### §230. Allographs of the letter щ

Two allographs are attested in Supr: щ and, rarely, щ, cf. in the text samples above пасѫщемъ (2), хотꙙщи (4), etc., but хощетъ 446, 20–21.

§231. Allographs of the letter и

In Supr three allographs are attested, mainly и, less frequently ї and і; cf. in the text samples above вьнѫтрьн҄иї (108), бечіслъноѥ (115), etc.

§232. Allographs of the letter ы

Among the allographs ъи and ы the latter is usually used; cf., помъишл҄ении

249, 18. In the text samples above we see ьі in (13)—(22). See § 240 below.

§233. Allographs of the letter у

Two allographs are attested in Supr: normally у and less frequently ; cf. ѥм 410, 30.

§234. A special spellout for the letter ꙗ

In Supr we find some spellouts with the letter а in place of the expected ꙗ or ѣ. Such are: гон҄аше 41, 16, alongside гонꙗаше 214, 16; лазара 431, 27, alongside лазарꙗ 457, 17 and лазара 303, 11.

§235. Mixing of the letters з and ѕ

In Supr only з is attested, never ѕ. Cf. however in the numeric meaning '6': ѕ златицъ 120, 5.

# §236. Rendition of kamorated letters

The use of kamora is common but inconsistent. Kamora is missing fairly often, see in the text samples above: о нивѣ (15) and въ н҄ивѫ (4), учителю (86), чловѣколюбивыи (116), съблюде (117); cf. also манастыру (73) and манастиру (83), likewise манастырѣ GSg 44, 1 and манастырꙙ APl 292, 6. Cf. also нынꙗ 287, 11; нын҄ꙗ 356, 11, and нын҄ѣ 317, 1; мол҄ꙗше 287, 7 and молꙗаше 298, 7; мьн҄ѫ 48, 7.

Note that the stem огн- presents only three occurrences without the kamora in Supr: in the text sample above огньнѣмь (118), and also огнъ 526, 8 and огн (DSg) 263, 12. Cf., in the text sample above: огн҄и (113), огн҄ьныи (119).

# §237. Substitutive softening

In a number of occurrences the regularly expected *l-epentheticum* is missing. Cf.: оставену (17), разломеньіѧ (19). Supr specific spellouts with ь occur frequently; cf. in the text samples above: славьꙗше (30), устрьмьѥниѥ (48), устръмьѥнье (67), устрьмьꙗше (73), устрьмьѥниимъ (77), устръмьѥниѥ (131), ставьꙗхѫ (133); likewise земьѧ (51), despite земл҄ꙙ 106, 2 and 129, 9. See § 117, *Alternative pairings in the substitutive softening alternation*, § 141, *Modifying aberrations*.

§238. Omission of intervocalic *j*

Fairly frequent, cf.: покаати 134, 9, покаавъшу 483, 16, but also покаꙗвъ 362, 25–26, покаꙗвъше 386, 26, all from покаꙗти; likewise сѣати 266, 5, сѣалъ 369, 23. Cf. also свꙙтааго (38) and similar.

# §239. Rendition of the yers

All aberrations are possible.

Strengthening: вьземъше (41), and also вьземъши 80, 17, вьземъ 25, 20, etc.; cf. also: весь 36, 21; 95, 5; день 14, 8; 15, 18, etc.

Fall: шьдши (3), что (32), вса (46), ничсоже (47), съспоспѣшникъ (50) (for съпоспѣшьникъ), толми (53), многы (59), труднааго (68), мнꙙще (78), прѣдн҄има (90), кн҄ижицꙙ (92), урвавъ (104), правъдныи (121), правьднааго (132), правьдныи (135), ничсоже (135).

Cases of confusion of the *yers* aberration are fairly frequent. Here are some examples from the text samples above.


§240. On spellouts for normalized combinations -ии- and -ыи-

Canonical spellouts predominate, the aberrant spellouts in the text samples above are the following: никымьже (13), свꙙтыхъ (23), видѣвы и (27, 29), прѣдн҄има (90), кротъкыми (139). Cf. canonical spellouts живѫщиїхъ (85), великыимъ (77), and many others.

# §241. Additional commentary on samples

In the text samples above various paradigmatic aberrations are attested.

1. Terminal deformation, as in such spellouts as некоторууму (1), свꙙтааго (38), etc.; or, such forms as извол҄ениимъ (54) and устрьмьѥниимъ (77) in place of извол҄ѥниѥмь, устрьмл҄ѥниѥмь, etc. (replacement of the terminal емь by the terminal ьмь); or, forms of the contracted imperfect, namely: лежаше (12), хвал҄ѣше (30), славьꙗше (30), подаꙗше (92), etc.; or, new ш-Part: съкрушивꙿше (46), въсхвативъ (51), etc.


PART II

Paradigmatics

# CHAPTER 8 **Acquaintance**

The part of the book entitled Paradigmatics describes paradigmatic synthesis (wordform generation) in OCS in terms of classical paradigmatic construction. The goal of this auxiliary chapter is to fix the terminology and to mention some particular issues which would be confusing without explanation. These clarifications concern only the sense of the notions under discussion, while their content1 is treated in the rest of this Part, and in the dictionary.

# §242. Lexemes and wordforms

A *paradigmatic construction* is a certain kind of classification of *wordforms*. Each wordform belongs to a particular *lexeme*, and each lexeme contains a fixed set of wordforms, ranging from one to several dozen. For example, рогъ, рога, роѕѣ, роѕи are wordforms of one lexeme, while роѕѣ and рожьць are wordforms of different lexemes. Likewise, the wordforms сѧ, себе, себѣ, си, собоѭ are different wordforms of one lexeme, while себе and тебе, or нама and вама are wordforms of different lexemes.

Lexemes are units of the lexicon; wordforms are linear strings in a text. In each lexeme one of the wordforms is selected as the *starting* one. This starting form, first, is used as the name of the lexeme itself (and thus lexemes are represented in the dictionary by this wordform), and second, the starting form supplies the primary segmental material for constructing all other (so-called *oblique*) wordforms.

### §243. Paradigms and grammatical categories

Within lexemes, wordforms are located in special containers. Some wordforms may be in more than one slot of the same container. These containers are known as *free paradigms*; a container filled with the wordforms of some lexeme is called

1 Following Carnap, the sense is the *intension* and the contents is the *extension*.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

a *paradigm* of that lexeme. Unwieldy paradigms can be broken up into subparadigms.2 A single slot in the paradigm is called a *paradigmatic cell*.

Paradigms are formed by *grammatical categories*. Each grammatical category is a finite set of *category values*. Accordingly, each paradigmatic cell receives a name, which is its address in the free paradigm, i.e. its address in terms of the categories that form that free pardadigm.

Table 243 shows a free substantive paradigm, and paradigms of the lexeme страна.



There are two *grammatical categories* here: (1) number, with three category values, Sg (singular number), Du (dual number), and Pl (plural number), and (2) case, with six category values, N (nominative), A (accusative), G (genitive), L (locative), D (dative), I (instrumental).

The free paradigm has 18 cells; each cell has a unique name: NSg, ASg, …, LDu, etc. There are 11 distinct wordforms in the paradigm of the lexeme страна. Some wordforms are found in more than one cell, e.g. the wordform странѣ is found in four cells: LSg, DSg, NDu, and ADu.

# §244. Grammatical description of a wordform

The list of names of all cells that contain a given wordform in a given paradigm is called its *grammatical description*. For example, the wordform страна has the description ⟨NSg⟩; the wordform странѫ has the description ⟨ASg⟩; the wordform странѣ the description ⟨LDSg, NADu⟩.

# §245. The paradigmatic address and the paradigmatic call

An expression of the form *X*(*L*), where *X* is the grammatical description of the wordform *A*, and *L* is its lexeme, is called the *paradigmatic address* of the wordform *A*. The name of the lexeme fixes which lexeme we are dealing with, and its description fixes the member of that lexeme. For example, LDSgNADu(страна)

<sup>2</sup> The separation into subparadigms corresponds to differences in the morphological composition of the corresponding forms. In particular, the forms within each subparadigm have the same workstem.

is the paradigmatic address of the wordform странѣ; DIDu (страна) is the paradigmatic address of the wordform странама.

An expression of the form *K*(*L*), where *K* is the name of one paradigmatic cell, and *L* is some lexeme, is called the *paradigmatic call* of a wordform *A*, namely the wordform which occupies cell *K* of the paradigm of lexeme *L*. For example, LSg (страна) is the paradigmatic call of the wordform странѣ. The same wordform can be generated by other paradigmatic calls, such as DSg (страна), or NDu (страна), or ADu (страна). In other words, a paradigmatic call mentions one specific paradigmatic cell, where a given wordform is found, while the paradigmatic address mentions all paradigmatic cells containing that wordform.

### §246. The paradigmatic construction

The term *paradigmatics* can refer both to an object of study (a component of the grammar of a language), as well as the result of the study (a part of the written grammar). In the latter sense, paradigmatics is a *paradigmatic construction*—a set of various grammatical objects (paradigms, lexeme classes, terminal sets, etc.), which are connected with each other in some way. Fundamental to any paradigmatic construction is a system of paradigmatic names of all wordforms under investigation, which is given in terms of category values of the grammatical categories.

It is important that the system of paradigmatic names of wordforms be created in such a way that

the same system of names works both for construction (building of forms), and for selection (syntax) of the needed wordforms of a given lexeme.3

The system of paradigmatic names adopted in the present grammar is syntactically effective, i.e. the conditions stated above are met. No demonstration of this claim is undertaken here, because syntax is left outside of the scope of this work.

## §247. A note on paradigmatic calls and paradigmatic addresses

A paradigmatic *call* contains *disjoined* grammatical descriptions (i.e. it points to one paradigmatic cell), while a paradigmatic address contains *conjoined* ones (i.e. it points to all paradigmatic cells containing a given wordform). It is easy to see that, on the one hand, paradigmatic calls correspond to synthesis and paradigmatic addresses to analysis; on the other hand, paradigmatic addresses are morphological names of wordforms, while paradigmatic calls are syntactic names of occurrences of wordforms. In constructing a text we need a certain paradigmatic cell. It does not matter that the same form may be found in other cells. In reading a text, we need to determine the paradigmatic address of a wordform,

<sup>3</sup> Cf. Čistovič's thesis for segmental grammar, which states that the system of phonetic features that defines the phoneme inventories must be created in such a way that the features allow both an articulatory and acoustic interpretation (see Čistovič).

and be ready to consider all corresponding syntactic hypotheses.4

# §248. A note on the term *wordform*

A wordform is commonly understood as a pair comprising the signifier and the signified. Under this view, странѣ(LSg), странѣ(DSg), странѣ(NDu), and странѣ(ADu) are four different wordforms. This is the treatment found in Zaliznjak 1967, for example. In this book, a wordform is a one-sided unit (cf. the notion of *segment* in Zaliznjak 1967). Below we sometimes use the term *form* as an approximate synonym of the term wordform.

# §249. Grammatical classes of lexemes

Different lexemes can have the same free paradigms or can differ in their free paradigms. For example, градъ, село, жена have the same free paradigm as страна—the same grammatical categories are applicable to these lexemes. On the other hand, lexemes новъ and градъ have different free paradigms: the free paradigm of новъ is formed not only by the categories of case and number, but also by the category gender (its category values are m 'masculine', n 'neuter', and f 'feminine'). All lexemes are classified into grammatical classes such that two lexemes belong to the same grammatical class if they have the same free paradigms.

The following grammatical classes are distinguished in OCS: V (verbs), S (substantives), and A (adjectives). There is also a trivial class containing lexemes whose free paradigms have just one cell. Lexemes and wordforms of the first three classes (V, S, and A) are called *paradigmatic*, while the forms of the trivial paradigmatic class are called *extraparadigmatic*.5

Paradigmatics studies only paradigmatic lexemes and wordforms.

# §250. Morphological composition of wordforms

Every wordform is representable as a string of formatives, which can be seen in its morphophonological representation, cf. пришьдъшу — при.шьд.ъш.у, ѥму — j.ему, странѫ — стран.ѫ, устрьми — у.стрьм.и, etc. The morphological composition of any paradigmatic wordform assumes two morpho-


logical components: the *stem* and the *terminal*.6 This bipartite morphological composition corresponds to the bipartite composition of the paradigmatic call and paradigmatic address: the stem is responsible for assigning a wordform to some lexeme, while the terminal places the wordform in paradigmatic cells. In fact, two different wordforms of the same lexeme in the general case have identical or similar stems and different terminals: стран.а, стран.ѫ, стран.ѣ, and рѫк.а, рѫк.ѫ, рѫц.ѣ.

# §251. Morphological skeleton and inflectional spellout

The morphological composition of a wordform is explicated by two morphological representations, of which the deeper one is called the *morphological skeleton*, and the one closer to the surface the *inflectional spellout*. The morphological skeleton is modified into the inflectional spellout by *boundary adjustment rules*, or simply *boundary rules*. Thus, the morphological skeleton рѫк+ѣ is modified into the inflectional spellout рѫц=ѣ; the morphological skeleton рек+ѣахъ into the spellout реч=ѣахъ, etc. An inflectional spellout is modified into the phonological or graphic one by the mph⇒ph/norm rules: рѫц=ѣ into рѫцѣ, реч=ѣахъ into речаахъ, etc.

# §252. Examples of morphological spellouts

As an example, consider a short passage from a text (Mt 8, 28–30), where each paradigmatic wordform is represented by its morphological skeleton (top row) and its inflectional spellout (bottom row). Extraparadigmatic forms and loans outside of the benchmark list of wordforms are parenthesized. Morphologically anomalous forms, for which no morphological composition is established, are placed in square brackets.

```
(и) пришьд.ъш+у j+ему (на) он+ъ пол+ъ· (въ) стран+ѫ
(и) пришьдъш=у j=ему (на) он=ъ пол=ъ· (въ) стран=ѫ
(герьгесиньскѫ)· сърѣт+осте j+ь дъв+а бѣсьн+а· (отъ)
(герьгесиньскѫ)· сърѣт=осте j=ь дъв=а бѣсьн=а· (отъ)
гробищ+ь j+ихъ· исход.ѧщ+а л҄ют+а ѕѣл+о· (ꙗко) (не)
гробищ=ь j=ихъ· исходѧщ=а л҄ют=а ѕѣл=о· (ꙗко) (не)
мог+ѣаше ни.к+ъ.то.же· минѫ.т+и пѫт+ьмь т+ѣмь· (и) с+е
мож=ѣаше ни.к=ъ.то.же· минѫт=и пѫт=ьмь т=ѣмь· (и) с=е
```
<sup>6</sup> The terminal can be null, cf. 2–3SgAor моли=0, and in exceptional cases can be altogether absent, cf. щ-Part молѧ|.

```
възъпи+сте глагол҄.ѫщ+а· ч+ь.то [ѥстъ] [нама] (и) [тебѣ]
възъпи=сте глагол҄ѭщ=а· ч=ь.то [ѥстъ] [нама] (и) [тебѣ]
(исусе)· сын+е божиј+ь· пришьд.л+ъ (ли) [ѥси] (сѣмо)· (прѣЖе)
(исусе)· сын=е божиј=ь· пришьл=ъ (ли) [ѥси] (сѣмо)· (прѣЖе)
врѣм.ен+е мѫчи.т+ъ [насъ]· бѣ+0 (же) (далече) (отъ) н҄+ею·
врѣмен=е мѫчит=ъ [насъ]· бѣ=0 (же) (далече) (отъ) н҄=ею·
стад+о свиниј+ь· мъног+о пас.ом+о·
стад=о свиниј=ь· мъног=о пасом=о·
```
As the examples show, the terminals in the morphological skeleton and inflectional spellout are always identical, while the stems may differ. The latter fact reflects the two-step nature of paradigmatic synthesis: on the first step, a morphological skeleton is constructed, which consists of a workstem and a terminal; on the second step, the workstem undergoes certain segmental rewrite rules, which are regulated by boundary adjustment rules. These rules, which leave terminals untouched, generate the inflectional spellout. For example, from the morphological skeleton мог+ѣаше we have мож=ѣаше (imperfect); from the morphological skeleton мог+ѣте we have моѕ=ѣте (imperative); from the morphological skeleton мог+ѫ we have мог=ѫ (1SgPrae); from the morphological skeleton мог+еши we have мож=еши (2SgPrae).

Workstems of the nominal forms of verbs (infinitive, supine, and participles) show separately the suffix of these nominal forms, cf. пришьд.ъш +у, исход.ѧщ+а, минѫ.т+и, глагол҄.ѫщ+а, пришьд.л+ъ, мѫчи.т+ъ, пас.ом+о.

### §253. Examples of paradigmatic addresses and paradigmatic indices

A paradigmatic address is assigned to a wordform. A lexeme containing that wordform receives a paradigmatic index that shows which grammatical class the lexeme belongs to (cf. below the superscript index before the lexeme's name), and, within the grammatical class, the lexeme's paradigmatic class (cf. below the symbols after the lexeme's name). Paradigmatic indices are shown in the PD. Table 253 (p. 129) shows examples for the wordforms of the three verses from Mt 8, 28–30 (§ 252).

Notes to Table 253

1°. The paradigmatic address of participles contains, first, the grammatical description of the participle as an ordinary A-lexeme (cf. DSgmnGLDumnfBrev in DSgmnGLDumnfBrev [ш-Part (прити)]); second, it shows the type of participle (cf. ш-Part in DSgmnGLDumnfBrev [ш-Part (прити)]), and finally, it shows the starting form of the parent verb (cf. прити in DSgmnGLDumnfBrev [ш-Part (прити)]). Cf. also исходѧща, глагол҄ѭща, пришьлъ, пасомо.


Table 253. Examples of paradigmatic addresses

Notes to Table 253 (continued)


### §254. Paradigmatic synthesis

Paradigmatic synthesis is the algorithm that matches inflectional spellouts to paradigmatic calls. Here are some examples.


As these examples show, different paradigmatic calls can correspond to the same wordform. However, in the general case, a paradigmatic call cannot correspond to more than one wordform.

Paradigmatic synthesis takes place in three stages: 1) building the workstem, 2) selecting the terminal, and 3) applying boundary adjustment rules. The first two steps result in the morphological skeleton; the third step gives the inflectional spellout.

All three stages can require information on the paradigmatic class of the lexeme, which is contained in the *paradigmatic dictionary* (PD). The first stage requires information on the morphophonological representation of the starting form, which is also contained in the paradigmatic dictionary. The second stage uses the terminal catalogs.7 Table 254 (p. 131) shows intermediate results in the synthesis of several wordforms.

<sup>7</sup> It is useful to note that, on the one hand, the segmental material for the needed workstem is found in the starting wordform of the lexeme that is retrieved from the paradigmatic call (roughly speaking, the needed workstem is the result of *segmental transformations* of the segmental component of the paradigmatic call), and on the other hand, the segmental material for the terminal is retrieved from the given catalogs according to the grammatical description of the paradigmatic call (i.e. the needed terminal is the result of *selecting* from a list).


Table 254. Examples of paradigmatic calls and inflectional spellouts with intermediate forms

### §255. Paradigmatic classes

Each paradigmatic lexeme belongs to a particular paradigmatic class. (Normally, the pardigmatic classes of the verb are called *verb classes* or *conjugations*, while the paradigmatic classes of the nominal are called *declension types* or simply *declensions*). With some simplification, we can say that two lexemes belong to the same paradigmatic class if, in constructing their wordforms, they follow the same alternatives among those that are supplied by the rules of paradigmatic synthesis. In other words, they follow the same *paradigmatic standard*. For example, the V-lexemes глаголати and исходити belong to different verb classes: they use different suffixes of the щ-Part, cf. исход.ѧщ+а and глагол҄.ѫщ+а.

The so-called *unique lexemes* form a peculiar class that stands by itself. Their paradigms are so irregular that it makes no sense to build oblique forms according to any rules. Their paradigms are given directly. Accordingly, formally speaking, their oblique forms are not assigned any morphological representation.8

It is important to stress that a lexeme's association with a particular paradigmatic class is its lexical feature, which in the general case cannot be derived from any other feature of the lexeme, i.e. not from its morphophonological, syntactic, or semantic features. This general observation is not undermined by the fact that in some special cases a lexeme's paradigmatic class correlates with some other property (e.g. all verbs with a root infinitive are either class 4 verbs, or unique).

<sup>8</sup> However, certain forms among unique lexemes may be sufficiently regular and their morphological spellouts may be sufficiently transparent. Such forms are supplied with morphological representations, but those representations are not built by rules but are given directly in paradigms. Cf. in the examples above such forms of unique lexeme as сърѣт=осте, възъпи=сте, врѣмен=е, etc.

# §256. Profiles

The profile of a lexeme is an abbreviated paradigm that contains only some *key forms*, i.e. a kind of questionnaire that covers those key forms. The profile includes forms that show the application of the nontrivial component of the paradigmatic synthesis rules. Thus, the profile is a convenient format for showing all paradigmatic particularities of a given lexeme. It visually shows the paradigmatic standard of a given paradigmatic class. Here is an example profile for the verb плакати.9


Profiles can be used in place of general rules of paradigmatic synthesis to build called forms of any lexeme.

§257. The paradigmatic dictionary and the benchmark list of lexemes and wordforms

All lexemes in the *benchmark list of lexemes* are represented in the *paradigmatic dictionary*. Each lexeme has a graphic and a morphophonological representation of its starting form, and a *paradigmatic index*, which indicates the paradigmatic class to which the lexeme belongs. Because paradigmatic indices of different grammatical classes do not intersect, the index can also easily indicate grammatical class. Paradigmatic indices of verbal lexemes begin with an Arabic numeral (from 0 to 7); paradigmatic indices of nominal lexemes begin with expressions of the form n/x, where n is an Arabic numeral (from 0 to 2). For example, we have пробости 4c, възбьрати 3°\*, ублажити 1 for verbs, and богъ 2/m, божии 2/a, боꙗзнь 1/f for nominals.

Each wordform of each lexeme from the benchmark list may be built according to the rules of paradigmatic synthesis. Thus, the benchmark list of lexemes defines the *benchmark list of wordforms*. In the general case, the rules build all oblique wordforms not only in the graphic and phonological shape, but also create their morphophonological and inflectional spellouts, as well as the morphological skeleta.

# §258. Workstems

As indicated above, three stages enter into building a called wordform: creation of the workstem, selection of a terminal, and the application of boundary adjust-

<sup>9</sup> The symbol | indicates the absence of a terminal in the forms of a paradigmatic lexemes; 0 stands for a zero terminal.

ment rules. The primary segmental content for building a workstem is given by the starting wordform, more exactly, its workstem.

The workstem of the starting wordform is retrieved from its morphophonological representation that is given in the paradigmatic dictionary. So, for the A-lexemes новъ and съпасовъ we find the morphophonological representations нов.ъ and съ.пас.ов.ъ in the dictionary. The workstems of these forms are нов and съ.пас.ов, respectively. To arrive at e.g. the forms ⟨LPlmBrev⟩, we build morphological skeleta: нов+ѣхъ, съ.пас.ов+ѣхъ.

In the simplest cases all wordforms of a lexeme use one and the same workstem, but often a paradigm contains several workstems. For example, the lexeme граЖанинъ has two workstems: граЖ.ан.ин and граЖ.ан. The former, shown in the starting form, is used in the forms of the singular; the latter is used in the forms of the plural (cf. NPl граЖане). In OCS, the existence of several workstems within a lexeme is commonplace for verbs and exceptional for nominal paradigms. For example, the V-lexeme плакати has two workstems: плач (in the personal forms of the present, imperative, and щ- and м-participles), cf. плач+еши (Prae) and плака (the rest of the forms), cf. плака+хъ (Aor), плака+въ (ш-Part).

The set of workstems and their distribution among paradigmatic cells is determined by the paradigmatic class. Also, the rules that build the workstems themselves may be different in different paradigmatic classes.

### §259. Selection of the terminals

In terminal inventories, both nominal and verbal, terminals are not unordered, but assembled into so-called *sets*. A set of terminals is a table whose cells are labeled in the same way as the paradigmatic cells of free paradigms. Terminals occupy those cells. Table 259.1 shows two sets of verbal terminals in the personal present form subparadigm. Table 259.2 shows two sets of personal aorist forms.



Table 259.1. Prae terminal sets




The terminals are collected in such a way that the forms of one lexeme take the terminals of one set, while different lexemes can take different terminals, each from its own set. For example, the lexeme л҄юбити takes the terminals ѫ, иши, итъ, etc., while the lexeme плакати takes the terminals ѫ, еши, етъ, etc.

The paradigmatic call specifies the name of the paradigmatic cell that corresponds to the paradigmatic address of the terminal. However, in the general case, there may be several terminals with that address, e.g. for the address ⟨2SgPrae⟩ we find two terminals from different sets, viz. иши and еши. As can be seen from the examples, even a single cell in a set can contain more than one terminal.

Terminal selection rules are mixed, in that they contain both paradigmatic and morphophonological mechanisms. If a given subparadigm corresponds to several sets, the selection of the set is determined by the paradigmatic class of the lexeme. The corresponding identically named terminals are called *paradigmatic variants*. Such are, for example, the terminals ⟨2SgPrae: иши⟩ (in л҄юб+иши, л҄юбити), and ⟨2SgPrae: еши⟩ (in нес+еши, нести).

If a given cell contains more than one terminal, their selection is determined by the morphophonological rules, i.e. rules that take into account the morphophonological information from a given workstem. Such identically named terminals are called *morphophonological variants*. Such are, for example, ⟨3DuAor: осте/сте⟩ (осте in e.g. мог+осте, but сте in e.g. плака+сте). Here, the selection is by CVC agreement. Likewise for personal terminals, e.g. ⟨LPlmn: ѣхъ/ихъ⟩ (ѣхъ in e.g. нов+ѣхъ, but ихъ in e.g. нищ+ихъ). Here, the selection is by the twofold rule.

### §260. Boundary adjustment rules

Boundary adjustment rules, or boundary rules, are segmental rewrite rules that apply in the transition from the morphological skeleton to the inflectional spellout. Boundary rules effect certain segmental transformations of the work stem, which are replacements by some alternation. The choice of the rewrite rule can be determined by the grammatical properties of the called wordform and by the paradigmatic class of its lexeme. For example, from мог+ъ we have мог=ъ (1SgAor), but from мог+е we have мож=е (2SgAor); from мог+ѣаше we have мож=ѣаше (2SgImf), but from мог+ѣте we have моѕ+ѣте (2PlImv); from трьп+ѫ we have трьпл҄=ѫ (1SgPrae for трьпѣти), but for теп+ѫ we have теп=ѫ (1SgPrae for тети [теп.т.и]).10

Note that, within each subparadigm, wordforms of a given lexeme have a single workstem. However, in some paradigmatic cells the final consonant of the workstem may undergo certain segmental rewrite rules in the transition between the morphological skeleton and the inflectional spellout (in the majority of cases, these rules involve replacement by an alternation). Thus, the inflectional spellouts of different forms of a subparadigm may contain stems that differ in

<sup>10</sup> At boundaries between formatives within a workstem, ordinary rules mapping morphophonological into phonological representations apply, i.e. the mph⇒ph/norm rules.

the segmental shape of their final formative. Cf. 1SgPrae мог+ѫ / мог=ѫ, but 2SgPraeмог+еши / мож=еши; трьп+ѫ / трьпл҄=ѫ, but трьп+иши / трьп=иши.

### §261. Morphologically anomalous forms

Certain forms in a paradigm may be irregular to such an extent that it makes no sense to endow them with a morphological composition. In the majority of cases, such *morphologically anomalous* forms (marked by ∇) are found in unique lexemes. However, anomalous forms are possible among lexemes belonging to standard paradigmatic classes (cf. the form 2SgImv ∇виЖь in the verb видѣти). Such, for example, are вѣстъ 3SgPrae (вѣдѣти); даси 2SgPrae (дати); нама DIDu (азъ); насъ GLPl (азъ); тебѣLDSg (ты). Anomalous forms are given by lists. (See further details in Ch. 24, *Supplement*, § 877–885, *Excursus on the grammatical regularity of canonical wordforms*).

All morphologically anomalous forms are canonical.

### §262. Secondary forms

Some lexemes may contain *secondary forms*, i.e. which are present only in some lexemes of a grammatical class. Some secondary forms belong to the ordinary free paradigm of their class as morphological doublets of the corresponding primary forms, while others have no primary analogs and have a special grammatical description. The distribution of secondary forms is not precisely defined.

Secondary forms are given by a list. (See more details in Ch. 24, *Supplement*, § 877–885, *Excursus on the grammatical regularity of canonical wordforms*).

All secondary forms are canonical.

### §263. Paradigmatic effects

Comparing paradigmatic standards of different paradigmatic classes makes it possible to assign some operations to a special class of *paradigmatic effects*. Paradigmatic effects are, so to speak, supplementary operations. First, they make it possible to describe the formation of marginal subclasses of the basic paradigmatic classes (the so-called *irregular verbs* and *deformations* of declension types). Second, they make it possible to represent paradigmatic aberrations. The same paradigmatic effect in some cases forms a paradigmatic standard of the main paradigmatic class, and in others forms a marginal subclass, and yet in others forms an aberrant paradigmatic derivation that creates an aberrant form. For example, the substitutive softening paradigmatic effect creates the standard for class 3 verbs (type плакати), creates a subgroup of irregular verbs in class 4 (group брати), and creates the paradigmatically aberrant form мещетъ (for canonical мететъ) in class 3°.

In terms of their content, paradigmatic effects can be classified according to six basic types, shown in Table 263 on p. 136.


# Table 263. Paradigmatic effects and their types

# §264. Aberrant forms and paradigmatic aberrations

Paradigmatic aberrations and aberrant forms generated by them are studied in paradigmatics. Most of the material concerning paradigmatic aberrations is treated in the special chapters (Ch. 13, *Aberrant nominal forms in sources,* and Ch. 22, *Aberrant verbal forms in sources*). However, overviews of some groups of aberrant forms are included in the main body of this grammar: in particular, aberrant forms of unique lexemes are treated in Chs. 12, *Unique nominal lexemes* and 21, *Unique verbs*. Ch. 20, *An overview of verb classes* examines some aberrant forms of the imperfect.

# §265. Benchmark task of paradigmatics

In this grammar, the benchmark task11 of paradigmatics of OCS is formulated as follows: supply a *grammatical analysis* of any wordform from the *benchmark list of wordforms*.12 Namely:


<sup>11</sup> A benchmark task is the content that (1) must be provided by the relevant part of the grammar, and (2) where success can be evaluated relatively formally. A grammar may pursue other aims that are more weakly controlled and lie outside of the benchmark task.

<sup>12</sup> See § 3 on the benchmark list of wordforms.

In case a manuscript text (from the benchmark corpus of texts) is being analyzed, a wordform may lack a paradigmatic derivation. In that case, one must first find a canonical analog for that wordform. In order to do this, it is necessary to (1) make a hypothesis about the paradigmatic address of the wordform, and (2) construct a canonical wordform with the same paradigmatic address. Then the wordform under study can be compared with the derived canonical wordform, and their differences can be attributed to the application of some *aberration* (see Part II, Ch. 13, and Ch. 22 on paradigmatic aberrations, and Part I, Ch. 6 on segmental aberrations). If the wordform being inspected is not representable as a result of aberrations, it must be treated as corrupt. In that case, hypotheses about its prototype are not controlled by grammar (cf. брѣгру for брѣгу, Sav Mt 8, 32).

The Nominal

# CHAPTER 9 **Free nominal paradigms**

§266. Nominal lexemes and their free paradigms

Lexemes of two grammatical classes, *substantives* (S) and *adjectives* (A), are nominal. Here are the free paradigms of adjectives (A) and substantives (S):


First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

### §267. Grammatical categories

Free paradigms of nominal lexemes are formed by the categories of *case* (both for adjectives and substantives), *number* (both for adjectives and substantives), and *gender*, only for adjectives. The category of case has six values: *nominative* (N); *accusative* (A); *genitive* (G); *locative* (L); *dative* (D); *instrumental* (I).

The category of number has three values: *singular* (Sg); *dual* (Du); *plural* (Pl). The category of gender has three values: *masculine* (m); *neuter* (n); *feminine* (f).

For adjectival lexemes, gender is an inflectional category, i.e. wordforms of the same lexeme can have different gender. Cf. новъ (m) ~ ново (n) ~ нова (f); онъ (m) ~ оно (n) ~ она (f). For substantives, gender is a word-classifying category, i.e. all wordforms of a lexeme have the same gender, cf. the masculine nouns градъ, кон҄ь, отьць, the neuter nouns село, пол҄ѥ, слово, and feminine nouns рѫка, рабын҄и, кость.

It is necessary to distinguish syntactic gender, which determines agreement with adjectival lexemes and anaphora, and morphological gender, which determines the paradigmatic shape of a lexeme, especially the choice of feminine endings. In the majority of cases syntactic gender coincides with morphological gender, but this is not always the case. For example, the lexemes кръмьчии and юноша have masculine syntactic gender and feminine morphological gender.

### §268. Substantives and adjectives

For many nominal lexemes, there is no sharp boundary between adjectives and substantives in OCS. The reason is that not only substantive free paradigms can be represented as a narrowing of adjective free paradigms (by eliminating the gender opposition), but also the paradigmatic shape of substantive lexemes for the majority of nominals can be represented as a narrowed form of the paradigmatic shape of adjectival lexemes. For example, the noun mгнѣвъ, nчрѣво and fдѣва are inflected in the same way as the adjective лѣвъ 'left' in the masculine, neuter, and feminine genders, respectively. Cf. GSgm лѣва and GSg гнѣва, GSgn лѣва and GSg чрѣва, GSgf лѣвы and GSg дѣвы.

For nominals without suffixes that follow the main declension type, the distinction between adjectives and substantives found in dictionaries is based on the word's meaning, and is to some extent arbitrary. For example, in Večerka, we find independent dictionary entries for другъ 'other', an adjective, другъ '(male) friend', a masculine noun, and друга '(female) friend', a feminine noun. Likewise, there is an entry for the neuter noun сухо 'dry land', and another for the adjective сухъ 'dry'. However, for nominals such as благъ, добръ, зълъ, or слѣпъ, the same dictionary gives an adjective entry with an indication of their substantive use.1

<sup>1</sup> The paradigmatic dictionary in this book treats this question in a way that is maximally similar to Večerka's dictionary. However, in contrast to Večerka, there are no separate entries for participles. Some other departures are discussed in comments to specific lexemes.

On the other hand, suffixed nominals in the majority of cases can be sharply identified as adjectives or substantives. Such are, for example, the suffixed adjectives адовъ, гор҄ькъ, плодьнъ, плътьскъ, and suffixed substantives such as ловьць, дльгота, вѣтвиѥ, падениѥ. Likewise, adjectives and substantives are neatly opposed if the nominal follows a properly substantival (1–simplex) or properly adjectival (2–pron) declension type. This is the case, e.g., for mпѫть (GSg пѫти), fкость (GSg кости), and for pronominal adjectives тъ (GSgm того), нашь (GSgm нашего).

# Long and short adjectives

## §269. General

In OCS, most adjectival lexemes show two sets of forms: the so-called *short* forms (also called indefinite, shown as Brev), and *long* forms (also called definite, shown as Plen). The terms "short adjective" and "long adjective" refer to short and long adjectival forms, respectively. For example, новъ is Brev and новыи is Plen for NSgm; нова is Brev and новаѥго is Plen for GSgm, etc. The paradigmatic, lexicographic, and grammatical statuses of this opposition are not correlated.

# §270. Paradigmatics

From a paradigmatic point of view, short and long adjectives can be treated as separate lexemes of the same grammatical class, namely adjectival, each of which has a full set of forms given by the free paradigm of A-lexemes. The corresponding identically named wordforms differ only in their terminals: short adjectives take the terminals of one set (2-base), while long adjectives take those of another (2-combi). Thus, each paradigmatic cell corresponds to two forms, a short and a long one: for NSgm нов=ъ and нов=ыи; for GPlm нов=ъ and нов=ыихъ, etc. Short and long adjectives form correlated lexeme pairs, whose members are only distinguished by their paradigmatic classes. Workstems for such pairs are identical. E.g. we have нов=ъ (Brev) and нов=ыи (Plen); плътьск=ъ and плътьск=ыи; л҄ют=ъ and л҄ют=ыи; мысльн=ъ and мысльн=ыи. The starting form of the short adjective serves as the starting form for building any short adjectival forms as well as any long adjectival forms.

Thus, from a paradigmatic point of view, there is a regular set of doublets. Similar paradigmatic doublets are observed for single lexemes, for example, in the case of parallel verb classes, cf. сѣти 4v and сѣꙗти 3.

### §271. Lexicography

In the lexicographic practice, long and short adjectives are usually unified in a single dictionary entry, whose main representative is taken to be the short adjective.

# §272. Grammatical status

Some adjectives lack short forms (e.g. которыи), while some other adjectives lack long forms (e.g. инъ). Wherever a pair exists, the lexical meaning of the short and long form is considered to be identical. At the same time, the problem of the rule that regulates the choice of the long vs. short adjective form in OCS has no clear solution. Accordingly, there is no clarity whether the long ~ short opposition is a grammatical one.

Treating members of correlative long ~ short pairs as separate lexemes entails that the opposition is word-classifying. Generally speaking, word-classifying oppositions can be grammatical.2 Treating members of correlative long ~ short pairs as a single lexeme entails that the opposition is an inflectional one. The final word on this problem remains for syntactic analysis, and is thus outside of the goals of the present work.

# §273. Technical conventions

This book treats the difference between short and long forms of adjectives as one of *representations*. Different representations, although they have the same free paradigms, belong to the same lexeme. The starting form of that lexeme is stipulated to be NSgmBrev. The representation type is included in paradigmatic addresses and paradigmatic calls, although formally these indications are not part of grammatical properties, because they are not appropriately correlated with grammatical categories that form free adjectival paradigms. This opposition is only found among adjectival lexemes that follow the basic twofold declension type. As discussed in § 275, л-participles are systematically defective. All other lexemes are treated as having long forms.3

# Participles

### §274. General

Participles comprise a special group of adjectival lexemes. Participles are deverbal adjectives that are connected in some way with their parent verb. Along with participles, some verbs may have derived adjectives that are not participles. Cf. for видѣти, видѣнъ (н-Part) and видьнъ, an ordinary adjective. The paradigmatic, lexicographic, and grammatical status of participles calls for special clarification.


### §275. Morphology and paradigmatics

Each participle belongs to one and only one parent verb. Morphologically, participles are regular derivates of the verb, which are connected with it by transparent morphological rules. Each of the six participles contains its own set of suffixes that are attached to the verb's workstem. Cf., e.g., the м-Part (л҄юбити): workstem л҄юб.им=: л҄юбимъ; м-Part (нести): нес.ом=: несомъ; н-Part (творити): твор҄.ен=: твор҄ѥнъ; н-Part (нести): нес.ен=: несенъ, etc. Participle formation rules are tightly intertwined with the general system of paradigmatic synthesis of verb forms.

All participles follow the main twofold declension type. щ- and ш-participles depart from this standard (see an overview of all participles in Ch. 25, § 918).

From a paradigmatic point of view, a participle is an ordinary adjectival lexeme that fits into the adjectival free paradigm. Just like other adjectives, it has long ~ short pairs. An exception is found in л-participles: they only have forms of the direct cases and are not paired by the long ~ short feature (they only have short forms).

However, participles are part of the free paradigm of the verb, thus forming special subparadigms.

### §276. Lexicography

Apart from a few exceptions, every participle is identified with a unique parent verb, although not every verb has all potential participles. Absence of a participle may be caused by a verb's syntactic, lexical, or, less commonly, by its morphophonological properties. While some restrictions can be obvious (e.g. intransitive verbs do not form passive participles, i.e. м-, н-, and т-participles), in the general case the paradigmatic dictionary in this book does not mark presence or absence of participles.4

The lexicographic convention, generally speaking, is not to give separate entries for participles, but to include them under their parent verb. The lexical meaning of the verb and the participle are treated as identical, and information on presence or absence of a participle for a given verb is not explicitly stated.5

5 Existing dictionaries are often inconsistent in handling participles. First, some participles are given as separate lexemes, both when finite forms of the parent verbs are not attested (cf., for example, раслабл҄ѥнъ: there is no entry раслабити in Večerka), and when finite forms are attested (cf. блаженъ: Večerka contains both of the entries блаженъ and блажити). Second, some participles are fixed in their nominal or adjectival use, and the lexicographic shape is not consistent: cf. съврьшенъ, marked *part*.*adj*.; there is the entry съврьшити. бѣшенъ is marked *adj*.; there is the entry бѣсити. Differences in lexicographic treatment in any case do not reflect any paradigmatic properties of the lexemes. In the present grammar, all adjectival lexemes, representable as some participle of some verb, are treated the same, namely simply as participles, and not given in the dictionary separately from their parent verbs.

<sup>4</sup> Note that participial stems actively participate in further derivations, and the presence of a participle is not necessary. Thus, we have притѧжаниѥ, where there is no н-participle for притѧжати. There is слутиѥ without a т-participle for слути, etc.

# §277. Grammatical status

Participles occupy an intermediate position: on the one hand, they are forms of a verbal lexeme, and on the other, they are independent adjectival lexemes. As forms of some lexeme, they should be opposed to non-participles by some category, but there is no such category. As separate lexemes, they should not be tightly connected paradigmatically with their parent lexemes, but such a connection is apparent (for example, participles reflect the division of verbs into paradigmatic classes). In this sense OCS participles are nothing special, sharing the sad fate of intermediate status common in other languages.

# §278. Technical conventions

In this book, all participles are treated as representing the corresponding subparadigm of the parent verbs. Grammatical descriptions and inflectional spellouts of participles reflect their grammatical duality. The symbol of the subparadigm that indicates which of six potential participles a given form belongs to is part of the grammatical description of the participle (both disjoined and conjoined). In the inflectional spellout, the stem shows a suffix: e.g. for the form пасомо we write пас.ом=о NASgnBrev [м-Part (пасти)]. For пришьдъшѫѭ, we write пришьд.ъш=ѫѭ ASgfPlen [ш-Part (прити)]. On the inflectional spellouts of syncopated forms of щ- and ш-Part see more details in § 307 and below (the overview of classes 2/a\* and 2/a\*\*).

# Comparatives

### §279. General

The so-called *degrees of comparison*, or *comparatives* (Compar), form a special group of adjectival lexemes. Comparatives are derived deadjectival adjectives, related to their parent lexeme by transparent morphological connections. For comparatives there exists a certain set of suffixes, attached to the parent adjective's workstem. The resulting suffixed stem takes adjectival terminals, the same ones that ordinary, unsuffixed stems take. Cf., for example, Compar (новъ) нов.ѣј.ьш=: ASgfBrev нов.ѣј.ьш=ѫ, ASgfPlen нов.ѣј.ьш=ѫѭ, APlmBrev нов.ѣј.ьш=ѧ, APlmPlen нов.ѣј.ьш=ѧѩ, etc. (see comparative suffixes in § 920, and peculiarities in declension in § 307–313). From a paradigmatic point of view, the comparative is an ordinary adjectival lexeme, possessing all case, number, and gender forms, and correlated by the long ~ short feature.

### §280. Lexicography

Except for a few isolated lexemes (e.g. вѧщии, without an ordinary adjectival parent lexeme), for each comparative there is a unique parent lexeme, although

not all adjectives possess comparatives. There is a systematic absence of comparatives for adjectives that follow the pronominal declension type; for others, it is hardly possible to distinguish grammatical defectivity from accidental absence in texts.

Lexicographically, comparatives are not shown separately in dictionaries, but are given under the entry of the parent adjective. The carrier of the lexical meaning is the workstem; the comparative suffix marks the syntactic possibilities of the lexeme, and possibly a transparent semantic shift.

# §281. Technical conventions

According to tradition, comparatives are not given as separate entries in the dictionary. The only exceptions are *comparativum tantum*. Such are the following 14: бол҄ии, вѧщии, гор҄ии, лучии, мьн҄ии, наивѧщии, наискорѣи, наитрѣбл҄ии, прѣвышии, сулѣи, сул҄ии, тачаи, ун҄ии, утрѣи. Accordingly, this grammar gives the rules for forming comparatives (see Ch. 25, *Summary*, § 919–924). The grammatical descriptions and inflectional spellouts of comparatives reflect their grammatical duality. The symbol Compar is part of the grammatical description of the comparative (both disjoined and conjoined). E.g. we write ASgfBrev [Compar (славьнъ)] for the form славьнѣишѫ. In the inflectional spellout, the stem contains a suffix: e.g. we write славьн.ѣј.ьш=ѫ, бол҄.ьш=и.

# Secondary nominal forms

# §282. The list of secondary forms

Secondary forms are attested for a small number of lexemes. Such are: 1) vocative forms (Voc), 2) personal dative forms (D2) with the terminal ови/еви, and 3) ad-prepositional forms of the pronouns \*и and иже (see § 318). Voc forms are considered possible for all substantive morphologically masculine and feminine lexemes, and for adjectival lexemes in the twofold declension type in case they are used as masculines. Such are, for example, отьче for отьць, сыне for сынъ, дѣво for дѣва. On the formation of vocative forms, see § 355, and also Vaillant, § 81. Personal dative forms are only possible for masculine and neuter substantive lexemes in the twofold declension type. For their formation, see § 356, and Vaillant, § 59.

# **Formation of nominal forms. Terminals and paradigmatic classes**

# **General**

### §283. Morphological composition of nominal forms

The majority of nominal forms have a morphological skeleton of the following shape: [workstem + terminal]. For example, рѫк+а for рѫка, NSg (рѫка); рѫк+ѫ for рѫкѫ, ASg (рѫка), рѫк+ѣ for рѫцѣ, DLSg (рѫка); л҄юбл҄.ьш+е for л҄юбл҄ьше, NPlmBrev [ш-Part (л҄юбити)]. The inflectional spellout corresponds to the morphological skeleton component by component, and can only differ from it in some cases in the segmental shape of the last formative of the workstem. In the examples above we have рѫк+а — рѫк=а, рѫк+ѫ — рѫк=ѫ, рѫк+ѣ — рѫц=ѣ, etc.

For a small number of morphologically anomalous forms, the inflectional spellout and the morphological skeleton is not established. Such are the forms of unique lexemes, e.g. насъ, GLPl for азъ, and syncopated forms of щ-, ш-Part, cf. NSgmnBrev for щ-Part: л҄юб.ѧ|, нес.ы|, and for ш-Part: л҄юбл҄.ь|, нес.ъ|. (See more details on morphologically anomalous forms in Ch. 24, *Supplement*, § 881, *Excursus on the grammatical regularity of canonical wordforms*).

### §284. Nominal form construction procedure

Three steps are necessary in order to build a nominal form: (1) construct a workstem (see § 286 below), (2) find the terminal (see § 287 below), and (3) apply boundary adjustment rules (see § 288 below). These steps build the inflectional spellout of the called wordform; transition from that spellout to the phonological or graphic one is handled by the general mph⇒ph/norm rules (see § 63–77).

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

The following information is required for these steps. The primary information includes (1) the grammatical description of the called wordform; (2) the graphic shape of the starting wordform of the parent lexeme of the called wordform. These facts are part of the paradigmatic call. The secondary information includes (3) the morphophonological representation of the starting wordform, and (4) the paradigmatic index of the parent lexeme of the called wordform. These data are retrieved from the paradigmatic dictionary.

After the necessary data have been collected, one can proceed to the synthesis of the called wordform.

Forms of unique lexemes are not constructed by rules. Their paradigms are given wholesale, or by samples (see Ch. 12, *Unique nominal lexemes*). Likewise, secondary forms are not built using general rules (see special rules in § 355 for Voc, § 356 for D2, and § 318 for ad-prepositional forms of the lexemes \*и and иже). Morphologically anomalous nominal forms either belong to unique lexemes, or are built by rules; the latter is the case for anomalous forms of щ- and ш-participles.

### §285. Starting forms of nominal lexemes

In the majority of cases the starting form is given in the dictionary. For adjectives it is the NSgmBrev form. For substantives it is the NSg form. For defective lexemes, the dictionary shows the first form in the representational order.1 For example, врата (2/n *plurale tantum*), дъва (2/p *duale tantum*), прочии (2/a *plenum tantum*).2

The situation is different for adjectival lexemes of participles and comparatives. Their starting forms are not shown in the dictionary, but are built by special rules from parent lexemes (verbs and ordinary adjectives). On the rules deriving starting forms of participles see Ch. 25, *Summary*, § 918; on comparatives, see Ch. 25, *Summary*, § 919 and ff. Note that щ- and ш-Part and Compar have not one but two starting forms, expanded and syncopated, representing two different workstems of the corresponding paradigm.

<sup>1</sup> Note that the order of the cells in free paradigms is not arbitrary. For example, the order of the cases, N, A, G, L, D, I, reflects the order of the observed joints.

<sup>2</sup> Let us list all such lexemes. *Pluralia tantum* substantives: ꙗдра, ꙗсли, букъви, вои, врата, гѫсли, дроЖиѩ, дръва, дѣти, кън҄игы, кън҄ижицѧ, лѧдвиѩ, л҄юдиѥ, ножьницѧ, ноздри, носила, оими, осъпы, плуща, прьси, пѣготы, пѫта, уста, чари, чрѣсла. *Pluralia tantum* adjectives: триѥ, четыре. *Dualia tantum* adjectives: дъва, оба. *Plena tantum* adjectives: которыи, прокыи, прочии. The adjectival lexemes ѕѣло and рано are represented by a single form NASgnBrev, which is shown in the dictionary. Defectivity of unique lexemes is not shown outside of the starting forms in these lists. For the anaphoric 3rd person pronoun, the fictional spellout \*и is taken to be the starting form; for the defective lexeme of the reflexive pronoun, it is the form сѧ (ASg).

### §286. The first step: building the workstems

In the general case, all forms in nominal paradigms have the same workstem, and thus finding the workstem for any oblique form amounts to extracting it from the starting form. This is not true in the following special cases: (1) in unique lexemes (their paradigmatic index has the symbol 0 above the slash, e.g. слово 0/n);3 (2) in lexemes of class 2/a\* (щ- and ш-Part), 2/a\*\* (Compar); and (3) in lexemes of class 2/m\*\* (type граЖанинъ). Lexemes in these three groups have two workstems, an expanded and a syncopated one. Table 286.1 shows examples for all such special cases, and gives references to their more detailed treatment.


Table 286.1. Expanded and syncopated stems in declension

As these examples show, the expanded stem contains the syncopated one, and in all cases except class 2/a\*, the syncopated stem can be derived from the expanded one by dropping the last formative. In class 2/a\*, the syncopated stem contains an uninterpretable sequence, apparent scraps of some formative. Wordforms of щ- and ш-participles that contain syncopated stems are morphologically anomalous, and thus it is not necessary to interpret those fragments.

If a lexeme contains both an expanded and a syncopated stem, it is necessary not only to construct those stems, but also to distribute them in the paradigm correctly. Forms with expanded stems are called *expanded*, while forms with syncopated stems are called *syncopated*. In this inflectional class, and in the relevant group of unique lexemes, the distribution of expanded and syncopated stems in the paradigm is identical.

In all other, normal cases, any oblique form's stem coincides with the workstem of the starting form. Thus, workstems can be derived by dropping the terminal in the starting form. Table 286.2 (p. 152) shows some examples.

<sup>3</sup> Note that finding the workstem of unique lexemes is not necessary for the application of paradigmatic synthesis rules. Formally, workstems of unique lexemes fall outside of the aims of this book.


Table 286.2. Workstems of nominal forms

### §287. Second step: selection of the terminal

The catalog of nominal terminals contains many different sets. The set is chosen according to the inflectional class of the lexeme using Table 299 (p. 161) or Table 302 (p. 163). Lexemes of class 2/a require two sets: 2-base and 2-combi, the former for short forms and the latter for long forms.

Once the terminal set is chosen, for adjectival lexemes it is necessary to select the terminal that answers the paradigmatic call, and for substantive lexemes, the terminal must also agree with the morphological gender of the substantive that is marked in the paradigmatic index. For substantive lexemes, the terminal must agree with the morphological gender of the substantive that is marked in its paradigmatic index. For example, in the set 1-simplex we have two terminals whose addresses are distinguished only by gender: ⟨ISgm⟩ ьмь (cf. пѫтьмь for пѫть for 1/m) and ⟨ISgf⟩ ьѭ (cf. мысльѭ for мысль 1/f). After the address is selected, if the cell contains only one terminal, the selection is made. However, in twofold sets (2-base, 2-pron, 2-combi), a single cell can contain two terminals, written with a slash. In that case the terminal is selected by the twofold rule (see § 86): the underslash terminal is chosen if the workstem ends in a morphophonologically soft consonant or epenthetic *i̯*, otherwise the overslash terminal is chosen. For example, in the set 2-base the address ⟨ISgf⟩ contains оѭ/еѭ. Accordingly, for stem нов the correct terminal is оѭ (новоѭ), while for нищ it is еѭ (нищеѭ).

For inflectional classes whose indices contain stars, we must take into account paradigmatic effects of alien terminals in the paradigm, and syncopated stems in the paradigm (see more details in § 301, *Deformations*).

On terminals of secondary forms see § 355–356.

### §288. Third step: boundary adjustment rules

Nontrivial boundary adjustment rules apply only in paradigmatic classes that take twofold terminals (sets 2-base, 2-pron, 2-combi), and only when the workstem ends in a velar (к, г, х). In these cases, the velar palatalization rule к→ц applies

before terminals beginning with и or ѣ. For example, for DLSg (рѫка) we have рѫк+ѣ⇒рѫц=ѣ; for NPl (врагъ) we have враг+и⇒враѕ=и, etc. On the other hand, boundary rules are vacuous in such cases as NPl (рѫка): рѫк+ы⇒рѫк=ы; DLsg (земл҄ꙗ): земл҄+и⇒земл҄=и; GSg (рѫка) рѫк+ы⇒рѫк=ы; Gsg (земл҄ꙗ): земл҄+ѧ⇒земл҄=ѧ. Boundary rules take morphological skeleta and produce inflectional spellouts—i.e. the output of paradigmatic synthesis.

In order to derive the final shape of the synthesized form, it is necessary to remove the sign = and apply standard mph⇒ph/norm rules. Thus, рѫц=ѣ ⇒ рѫцѣ; враѕ=и ⇒ враѕи; рѫк=ы ⇒ рѫкы; земл҄=и ⇒ земл҄и; земл҄=ѧ ⇒ земл҄ѩ; крај=ь ⇒ краи; крај=и ⇒ краи; крај=ѧ ⇒ краѩ, etc.

# **Nominal terminals**

§289. A catalog of terminals

The catalog contains seven terminal sets. Tables 289.1 and 289.2 (see pp. 154 and 155) show all sets of nominal terminals.

§290. Explanation of notation



Table 289.1. Twofold terminal sets


Table 289.2. Monovariate terminal sets

# §291. The paradigmatic meaning of terminal sets

Table 289.1. Twofold terminal sets

Standard sets

Basic twofold: 2-base

Number

Sg

L D

I NA

а

ѣ/и

а

ѣ/и

ою/ею

ој.у/еј.у

ѣма/има

Du

GL

DI N A

ы/ѧ

и

а

ы/ѧ

и

а

> ы/ѧ

ы/ѧ

Pl

L D

I

ы/и

ами

ѣми/ими

омъ/емъ

амъ

ѣмъ/имъ

ѣхъ/ихъ

ахъ

G

ъ/ь

ѣхъ/ихъ

ома/ема

ама

у

омь/емь

оѭ/еѭ

ој.ѫ/еј.ѫ

у

ѣ/и

ѣ/и

G

а

ы/ѧ

ого/его омь/емь

ои/еи

ој.и/еј.и

ому/ему

ѣмь/имь

оѭ/еѭ

ој.ѫ/еј.ѫ

Case

N

ъ/ь

> A

о/е

ѫ

ъ/ь

о/е

а

ыи/ии

оѥ/еѥ

о*i̯*.е/е*i̯*.е

ъ*i̯*.ь/ь*i̯*.ь

ѫ

оѩ/еѩ

аѥго

а*i̯*.его

ѣѥмь/иѥмь

ѣ*i̯*.емь/и*i̯*.емь

уѥму

у*i̯*.ему ыимь/иимь

ѫѭ

ѫ*i̯*.ѫ

ѣи/ии

ѣ*i̯*.и/и*i̯*.и

ую

у*i̯*.у

ыима/иима

ъ*i̯*.има/ь*i̯*.има

ии

и*i̯*.и ыѩ/ѧѩ

ы*i̯*.ѧ/ѧ*i̯*.ѧ

аꙗ

ыѩ/ѧѩ

ы*i̯*.ѧ/ѧ*i̯*.ѧ

а*i̯*.а

ыихъ/иихъ

овъ/евъ

ъхъ/ьхъ

ъмъ/ьмъ

ъми/ьми

ъ*i̯*.ихъ/ь*i̯*.ихъ

ыимъ/иимъ

ъ*i̯*.имъ/ь*i̯*.имъ

ыими/иими

ыими/иими

ъ*i̯*.ими/ь*i̯*.ими

ы*i̯*.ими/и*i̯*.ими

ъма/ьма

ове/еве

ъ*i̯*.имь/ь*i̯*.имь

аꙗ

а*i̯*.а

ѣи/ии

ѣ*i̯*.и/и*i̯*.и

ови/еви

ъмь/ьмь

ој.ѧ/еј.ѧ

а

m

n

f

m

n

f

m

n

f

аꙗ

а*i̯*.а

ѫѭ

ѫ*i̯*.ѫ

ыѩ/ѧѩ

ы*i̯*.ѧ/ѧ*i̯*.ѧ

Gender

Pronominal: 2-pron

Combined: 2-combi

nonstandard set

New twofold rule: 2-duplex

Each lexeme follows one set of terminals from one of the *standard sets*, i.e. all forms in the paradigm of the same lexeme take the terminals of the same set. Thus, standard terminal sets determine the paradigmatic standard of nominal lexeme classes. Accordingly, the opposition between standard sets is a paradigmatic one: the selection by a lexeme of a particular terminal set is determined by its paradigmatic class.

Nonstandard terminal sets do not create independent declension types. However, being part of particular nominal paradigms, they sometimes create *deformed paradigms* (paradigmatic classes 2/f\*, 2/a\*, 2/a\*\*, 2/m\*, 2/m\*\*), and in other cases they create *aberrant wordforms*. In the case of the personal dative (D2), they create *secondary wordforms* (terminals from the 2-duplex set). These terminals are also found in the paradigms of unique lexemes.

# §292. An overview of terminal sets

Terminal sets are classified, on the one hand, as standard vs. nonstandard, and on the other, as twofold vs. monovariate, as shown in Table 292. The table shows synonymous names of terminal sets that are found in the literature, and in this book.

Twofold terminal sets differ from monovariate ones in that some cells have two variants of terminals, cf. ъ/ь, ѣ/и, etc. Such terminals are called twofold. In these cases, the two versions of the terminals are each other's morphophonological variants distributed by the twofold rule.


Table 292. Overview of terminal sets

### §293. Overview of terminal types

All terminals are either twofold or monovariate, and bicomponential (the first component separated from the second component by a period in morphophonological representations) or monocomponential. Monocomponential terminals, in turn, can be monosyllabic or disyllabic.

In the combined set (2-combi) all terminals are bicomponential; in the basic set (2-base), only one terminal is bicomponential, viz. ISgf оѭ/еѭ. In the pronominal set (2-pron), there are four such terminals, viz. GSgf оѩ/еѩ, LDSgf ои/еи, ISgf оѭ/еѭ, and GLDumnf ою/ею. See Table 293.


Table 293. Overview of terminals in standard sets

# §294. Disyllabic monocomponential terminals in standard sets

All such terminals contain the consonant х or м, except for the peculiar terminal ого/его; cf. Table 294.


Table 294. Disyllabic monocomponential terminals in standard sets

# §295. Bicomponential terminals

*Bicomponential* are terminals that consist of two formatives. The first component of such terminals can be a suffixal formative or one of the inflectional terminals; the second component can only be inflectional. Table 295 (continued on p. 158) shows all bicomponential nominal terminals and their decomposition.

Table 295. Decomposition of bicomponential terminals



# Table 295 (continued). Decomposition of bicomponential terminals


### § 296. Simple and nontrivial compositions

Terminals in the 2-combi set may result from simple or nontrivial composition. In simple composition, two formatives are simply placed next to each other. In nontrivial composition, the result, i.e. the bicomponential terminal, cannot be represented as a sequence of formatives. The sequence is simplified: the suffixal formative oj/ej is deleted, and disyllabic monocomponential terminals that are the first member of the composition are replaced with the auxiliary formative '7/6. All nontrivial compositions are listed in Table 296.



Note that ISgf terminals use the arbitrary monocomponential terminal ऋ, which is not represented for that paradigmatic cell in any of the terminal sets, but is found in compositions. Such are:


§297. On the terminals លោភ/ពេក and ភាគ

The terminal oj.x/ej.x. as part of 2-base is a unique one: this set has no other bicomponential terminal. At the same time, 2-pron contains four bicomponential

terminals, and they all contain the formative ој/еј as their first component; these are: (ој/еј).ѧ, (ој/еј).и, (ој/еј).у, and (ој/еј).ѫ. Thus, it turns out that the terminal (ој/еј).ѫ in 2-base is secondary, in that it is a nonstandard terminal from a different set, namely from 2-pron. The primary member of 2-base is the monocomponential terminal ѫ. This interpretation fully agrees with the composition ѫ+ѫ⇒ѫѭ as part of 2-combi, and with the observed aberrant spellouts such as съ воѥводѫ Supr 72, 19; cf. also правъдѫѭ Ps Sin 30, 2. See § 403.

# §298. On the etymology of 2-combi terminals

As is well-known, long adjectival forms, i.e. forms that take 2-combi terminals, are reflexes of case-number forms of combinations containing a short-form adjective with the pronoun \*и, where those anaphoric pronominal forms function as a postpositive article. The synchronic description presented above does not contradict this accepted etymology, even though it is different in some details (see Van Wijk, p. 292–295; Lunt 2001, § 4.3–4.4).

# **Declension types**

§299. Paradigmatic classes of nominals: declension types and inflectional classes

The division of nominal lexemes into paradigmatic classes must line up with the similarities and differences in their paradigmatic behavior, in such a way that information on a lexeme's paradigmatic class ensures the correct application of the algorithm of paradigmatic synthesis. Roughly speaking, two lexemes belong to the same paradigmatic class if they are declined in the same way—in other words, if they follow the same paradigmatic standard. Membership of a lexeme in a paradigmatic class is shown in the paradigmatic dictionary.

In partitioning the set of lexemes into classes we distinguish larger classes, which are called *declension types* (there are three types, namely: basic twofold, pronominal twofold, and monovariate), and smaller classes, called *inflectional classes*, of which there are eight, each with its own paradigmatic index: 2/a, 2/p, and 1/a for adjectives and 2/m, 2/n, 2/f, 1/m, and 1/f for substantives. Declension types only take into account the terminal set, and reflect different morphological mechanisms (twofold vs. monovariate); inflectional classes further distinguish adjectives from substantives, and, in the latter case, morphological gender. The digit in the paradigmatic index shows whether the lexeme is monovariate (1) or twofold (2); the letter specifies the inflectional class: simple adjectives (a); pronominal adjectives (p); substantives that are masculine (m), neuter (n), and feminine (f). In the inflectional class 2/a, the set 2-base builds short forms, and the set 2-combi builds long forms. The partitioning is shown in Table 299.


Table 299. Standard inflectional classes and terminal sets

The eight inflectional classes shown above are basic, or standard, inflectional classes. Monovariate declension is also called *i-declension*. Twofold declension types (basic and pronominal) are sometimes unified under the term *twofold declension*. The pronominal twofold type can be simply called *pronominal*. Twofold types distinguish *soft* and *hard* subtypes (in the Table 299, the overslash form is an example of a lexeme of the hard subtype, and the underslash form is an example of a lexeme of the soft subtype). Whether a given lexeme belongs to the hard or soft subtype is determined unambiguously by the ending of its workstem: a lexeme belongs to the soft subtype if its workstem ends in a morphophonologically soft consonant or vowel, and otherwise belongs to the hard subtype.

The classification does not show marginal subclasses of the main classes (the so-called *deformations*; see § 301), and *unique* lexemes, which are outside of this classification (see details in Ch. 12, *Unique nominal lexemes*).

# §300. A note on the notion of declension type

The notion of paradigmatic class by itself does not take into account those differences in the paradigmatic behavior of lexemes that stand in complementary distribution in an obvious way according to some traits of the lexeme. In other words, differences in behavior are treated as properly paradigmatic only when they are a lexical feature of the lexeme, and are not derivable from any of its other features (including its other lexical features).

For this reason, substantive lexemes of the three genders are unified under a common declension type.4 Gender differences in terminals (cf. NSg in 2-base: masculine ъ/ь , neuter о/е, feminine а, e.g. mградъ, mкон҄ь; nсело, nпол҄ѥ; fжена,

<sup>4</sup> Trubeckoj was the first to introduce such a distinction in inflectional types for OCS, namely by unifying nouns of different genders into one declension type (Trubeckoj 1954); see also Lunt 1974, Lunt 2001. For modern Slavic languages this distinction can be found in Zaliznjak's *Russian nominal inflection* (Zaliznjak 1967).

Esenita) stand in complementary distribution by a word-classifying characteristic of substantive lexemes, namely by its morphological gender. The fact that substantives and adjectives are served by the same set of terminals (2-base) allows us to classify them in the same declension type. Likewise, the same declension type contains both the so-called hard and soft subtypes, cf. ™ΓΩΑΣ and 14 конь, and likewise "село and "полье, "жена and "зємла (see more details on the twofold rule in § 85-91).

The classification adopted here differs from the tradition found in Slavic studies (with etymological classification of lexemes by morphophonological stem types, into \*a-stems, \*ja-stems, etc.). Here, several traditional classes are unified into a single class. A system with so few classes is not always convenient in practice. The resulting working classification, i.e. the classification into inflectional classes, that is found in this book and fixed in the paradigmatic dictionary, is considered in more detail in § 302.


Table 301. An overview of deformations

### §301. Deformations

Deformations of standard inflectional classes are created by paradigmatic effects that destroy specific paradigmatic cells. Outside of the destroyed cells, the lexeme follows the paradigmatic standard that is defined for the superordinate standard class. Paradigmatic effects that create deformations amount to replacing a terminal with some other, "alien" terminal (the alien terminals in the par-

adigm effect), or in destroying the unity of workstem (the syncopated stems in the paradigmatic effect). An overview of deformations is shown in Table 301. See details in Ch. 11, An overview of nominal classes.

### ട്ട് 302. List of inflectional classes

Table 302 shows an exhaustive list of all inflectional classes, taking into account deformations and unique lexemes, and the terminal sets that correspond to them.


Table 302. List of inflectional classes

1) Here are all 22 unique lexemes with the index 0/1: агна, връма, жръса, има, ИСТО, КЛЮСА, КОЗЬЛА, КОЛО, НЕБО, ОВЬЧА, ОКО, ОСЬЛА, ОТРОЧА, ПИСМА, ПЛЕМА, СЛОВО, СЪМА, ТЕЛО, ОУХО, ЧИСМА, ЧОУДО.

2) Here are all 14 unique lexemes with the index 0/f: Εραдъι, Εογκъви, дъшти, жрьнъы, локъ, ЛЮБЪІ, МАТИ, НЕПЛОДЪІ, ПРЕЛЮБЪІ, СВЕКОВІ, СМОКЪІ, ХООЖЪІ, ЦЪЛЪ.

Clearly, apart from unique lexemes, every inflectional class corresponds to a particular terminal set. This is not the case only for adjectival lexemes 2/a, because each such lexeme has two paradigms: a paradigm for short forms (2-base), and a paradigm for long forms (2-combi). Conversely, the same terminal set can serve several inflectional classes distinguished by gender (in the case of substantival lexemes), or by the application of various paradigmatic effects.

# CHAPTER 11 **An overview of nominal classes**

# **Adjectives**

# Class 2/a: новъ/нищь

§303. Class membership

This class contains about 1000 lexemes.1 The distinction between hard and soft subtypes is made by the final sound of the workstem. Possible ends of the workstem are shown in Table 303.

Table 303. Possible ends of workstems in class 2/a


1 Večerka's dictionary sometimes lists wordforms with adverbial use as separate lexemes. Cf. the two entries тъчьно and тъчьнъ. The present grammar treats all such forms under the corresponding adjectival lexemes.

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

# §304. Profiles




### §305. A note on long/short forms

Some adjectival lexemes have no short forms. Such are прочии, прокыи, and которыи. For many lexemes, long forms are not attested; it is difficult to identify

2 On aberrant contracted forms of full adjectives see § 394–395.

among them the class of systematically defective ones. Some suffixes can create morphophonological problems for long forms, e.g. the suffix ьј (cf. божии). The dictionary does not mark such adjectives in any way (see details in Vaillant, §86).

# §306. Word-formation characteristics

This class can contain both root and suffixal adjectives. Table 306 shows examples.

Table 306. Word-formation adjectives subtypes


Many nominals belonging to the soft subtype are the result of *segmentless substitutive softening* (see § 864). Such are, for example, говѧЖь, лъжь, добл҄ь, and also such adjectives as осьл҄ь or вельбѫЖь, which are possessive for substantives (cf. осьлъ, вельбѫдъ).

# Class 2/a\*: щ-Part and ш-Part, and class 2/a\*\*: Compar

# §307. Class membership

These classes contain all щ- and ш-participles (2/a\*) and all comparatives (2/a\*\*). Recall that the starting forms of the adjectival lexemes under consideration are not listed in the dictionary.3 For participles, these lexemes are formed by the rules of paradigmatic synthesis of verb forms (see Ch. 25, *Summary*, § 918); for comparatives they are formed by the comparative formation rules (see Ch. 25, *Summary*, § 919 and ff).

# §308. Paradigmatic peculiarities

The paradigms of the lexemes under consideration show two differences from standard behavior of the class 2/a.

First, they distinguish *syncopated* and *expanded* wordforms and workstems (the syncopated stems in the paradigmatic effect), and, accordingly, have two starting forms: syncopated and expanded (see below, § 309–311).

Second, in some forms they take terminals from a nonstandard terminal set (the alien terminals effect). (See below in Table 312, p. 170).

<sup>3</sup> This is not the case for the 14 *comparativum tantum* lexemes, which are listed in § 281.

§309. Starting forms and workstems, syncopated and expanded

All three classes of adjectives under discussion have two starting forms: a syncopated one for forming syncopated forms, and an expanded one for expanded forms. The starting forms for all three adjective classes are the same:


Table 309 shows examples of syncopated and expanded starting forms and their workstems.

Table 309. Examples of syncopated and expanded щ- and ш-participles and comparatives (Compar)


Starting forms of syncopated forms of щ- and ш-participles are anomalous. They have no terminals, and the *syncopated workstem* is the whole wordform itself.

Starting forms of syncopated forms of comparatives have the standard 2-base soft subtype terminal. As usual, to extract the syncopated workstem from the starting form it is sufficient to remove the terminal.

*Expanded workstems* in both participles and comparatives are derived according to general rules, namely by removing the terminal и from the expanded starting form.

As the examples show, in class 2/a\*\* (comparatives), the syncopated stem differs from the expanded stem by the absence of the suffix ьш; in class 2/a\* (participles), the syncopated stem contains a poorly interpretable string, a kind of fragment of the suffix of the expanded stem. Syncopated wordforms of щ- and ш-participles are morphologically anomalous. However, unlike all other anomalous forms, they are formed according to rules.

In all three classes of adjectives, expanded forms contain shibilants in the suffix, while syncopated forms show no suffixal shibilants.

## §310. The distribution of syncopated forms

Tables 310.1 and 310.2 show the distribution of syncopated and expanded forms.

Cells that contain syncopated forms are marked by ⨂. While every participle has two distinct syncopated forms, NSgmnBrev and NSgmPlen, comparatives have three distinct syncopated forms, NAsgmBrev, NASgnBrev, and NASgmPlen.

Table 310.1. Distribution of syncopated forms in щ-Part and ш-Part


Table 310.2. Distribution of syncopated forms in Compar


# §311. Formation of oblique syncopated forms

The following holds for participles: the syncopated starting form occupies two paradigmatic cells, namely ⟨NSgmBrev⟩ and ⟨NSgnBrev⟩. In other words, ⟨NSgnBrev⟩ simply coincides with the syncopated starting form. The syncopated form ⟨NSgmPlen⟩ is formed by adding the terminal ь to the syncopated workstem (or, in other words, by adding the terminal ⟨NSgm⟩ from 2-pron to the ⟨NSgmPlen⟩ form, i.e. in the same way as it happens in the transition from ⟨NSgmBrev⟩ to ⟨NSgmPlen⟩ in ordinary adjectives). For example, мол.ѧ+ь⇒мол.ѧ*i̯*.ь⇒молѧи, or мол҄.ь+ь⇒мол҄.ь*i̯*.ь⇒мол҄ии.

The following holds for comparatives: the basic syncopated form occupies two paradigmatic cells, namely ⟨NSgmBrev⟩ and ⟨ASgmBrev⟩. In other words, syncopated forms (i.e. ⟨NASgnBrev⟩ and ⟨NASgmPlen⟩) are formed by adding to the workstem the corresponding 2-base terminals (for short adjectives) and 2-combi (for long adjectives) in their soft subtype, since syncopated workstems end in -*j* (the so-called new Compar), or a substitutively soft consonant (so-called old




### §312. Formation of oblique expanded forms

In expanded forms, expanded workstems (which are extracted from the second starting form, namely NSgfBrev) take 2-base terminals (for short forms) and 2-combi terminals (for long forms), except those forms where nonstandard terminals intrude. These forms and terminals are shown in Table 312.

See § 394-396 for aberrant forms.

Table 312. Nonstandard terminals in inflectional classes 2/a\* and 2/a\*\*


### §313. Profiles

Here are the profiles of the corresponding models (the symbol ® marks paradigmatic cells that contain syncopated forms).




\* Forms NASgmPlen (on the pattern of новѣии) in new comparative forms are not attested in sources.


\* In old comparatives, NASgmBrev (on the pattern of грѫбл҄ь) are not attested in sources. For this reason the form NSgmPlen usualy serves as the starting form in dictionaries. Likewise in the PD.

# Class 2/p: тъ/нашь

### §314. Class membership

This class contains 34 lexemes. These are: ꙗкъ, вашь, вьсакъ, вьсѣкъ, дъва, дъвакъ, дъвои, ѥдинакъ, ѥдьнакъ, ѥдинъ, ѥдьнъ, ѥликъ, \*и, иже, инакъ, инъ, какъ, коликъ, мои, нашь, оба, обоꙗкъ, обои, овъ, онъ, самъ, свои, селикъ, такъ, твои, толикъ, трои, тъ, чии. The workstem may end in a morphophonologically soft consonant (ш or j), or a plain consonant or к.

### §315. Notes on word formation

Some lexemes in this class are root lexemes, such as тъ [(т).ъ], овъ [(ов).ъ], дъва [(дъв).а], while others are suffixal. Note the specifically pronominal suffixes ак {13} and ој {10}: какъ [к.ак.ъ], обоꙗкъ [об.ој.ак.ъ], такъ [т.ак.ъ], etc.; обои [об.ој.ь], дъвои [дъв.ој.ь], трои [тр.ој.ь], etc. Pronominal lexemes with a suffix are attested in a few examples, which makes it difficult to determine their paradigmatic class (see Vaillant, § 99–100).

### §316. Combinations with clitics

The following complicates the question of the composition of wordforms for individual pronominal lexemes. In OCS, there are many combinations of the form *p*+*B*pron +*t*+*cl*, where *p* is the prefixal formative ни or нѣ, *B*pron is a lexical pronominal stem, *t* is a terminal and *cl* is a postpositive particle же, Же, or жьдо.4 These combinations are intermediate between a free sequences of wordforms and a single, unitary wordform. Accordingly, it is difficult to determine the set of items in the class under discussion (the number 34 in § 314 above shows the lexemes that are included in the paradigmatic dictionary). For example, this holds for the family какъ, никакъже, and нѣкакъ with the lexical stem как, and for the family къто, никъто, никътоже, нѣкъто. Syntactic behavior and the composition of these combinations is not altogether clear. The preposition is found between the prefixal formative (ни or нѣ) and root; cf. ни+о+к.омь+же не радиши, не бесѣдуѩ ни+къ+к.ому+же, etc.5

The paradigmatic behavior of these formations is determined entirely by the lexical stem, i.e. is identical for members of a single family. That is, e.g. никътоже is declined in the same way as нѣкъто and in the same way as къто.

In the present dictionary, such families are represented by a single item. Let us list all pronominal items that represent such families in the dictionary: ꙗкъ, ѥдинъ, ѥдьнъ, \*и (however, contrary to this convention, иже is given as a separate item), какъ, коликъ, такъ, тъ; къ.то (there is no isolated къ; this is a unique lexeme), and чь.то (there is no isolated чь; this is a unique lexeme); кыи and которыи. Thus, ꙗкъ represents ꙗкъ and ꙗкъже; къто represents къто, нѣкъто, никъто, никътоже, къжьдо and къЖе, etc.

<sup>4</sup> Some other particles are possible, cf. л҄юбо: въ н҄еже мѣсто аще въпадетъ какъ любо скотъ· то к тому живо не излѣзетъ Supr 193, 15–17; ꙗкоже аще и кꙿто начьнетъ отъ заградьныихъ· и отъ виноградъ· не тꙿчиѭ отъ плода нъ и отъ листвиꙗ· и отъ коѥго любо сѣмене кѫѭ любо винѫ да украдетъ немощьно ѥму ѥстъ Supr 42, 16–20.

<sup>5</sup> Syntactically anomalous combinations should be mentioned here, where the first element is in the absolute (NSgm) form such as онъ+сица (Supr), таковъ+сь (Cloz), the oblique forms are онъсицѧ (attested as азъ отьче· оньсицꙙ града епискупъ поставьѥнъ ѥсмъ Supr 286, 17–19) and таковъсѧѩ (attested as таковьсѧѩ бещініцѧ Cloz 2a, 24).

### §317. Profiles


Here are the profiles of some type representatives.

§318. The paradigms of the lexemes \*и and ижє

These two lexemes distinguish absolute (primary) and ad-prepositional (secondary) forms. The former begin with j-; the latter with i-. Cf. рече кмоу, but рече къ нимог, имъже глаголья, but къ нимъжє придє. The root has two alloforms, j and fi, cf. (j)=EMOY [IEMOY] and (h)=EMOY [HIEMOY].

The lexeme \* n (a 3rd person anaphoric demonstrative pronoun) is defective: it lacks nominative forms. The corresponding syntactic contexts contain TT ('that one, this one'), or own ('the aforementioned one'). The relative pronoun ижє is not defective in the same way.

Below is the complete paradigm of the lexeme \* n.


Illustrations

посъла дъва отъ огченикъ своцъ: и гла има: 1д въта въ вьсь ъже естъ пръмо ВАМА" 1 АБЬЄ ВЪХОДАШТА ВЬ ЯНЖ: ОБОЛШТЕТА ЖОВБЕЦЬ ПОИВАЗАНЪ: НА НЕМЬЖЕ НЕСТЪ НЕ ОУ НИКТОЖЕ ОТЪ ЧКЪ ВЪСЕЛЪ: ОТРЪШЪША I ПОНВЕДЕТА: L АШТЕ ВАМА къто речетъ чьто се дъета рьцета тко гъ тръведетъ: и с авье пакъи посълетъ I chno Mk 11, 1-3 ZOGR.

и пръшедъ отъ тъдоу од зьръ иъкова зеведеова: и ноана вратра его и та ВЪ ЛАДНИ ЗАВАЗАНЖШТА МОВЖА: 1 АБЬЕ ВЪЗВВА Ъ: 1 ОСТАВЛЬША ОТ"ВЦА СВОЕГО зеведеа въ ладни съ наемъникъ: по немь идете Mk 1, 19-20 мак.

И РЕЧЕ ОУЧЕНИКОМЪ СВОИМЪ' ДА ЕСТЪ ПОИ НЕМЪ ЛАДИНЦА НАРОДА РАДИ' ДА НЕ СЪТЖЖАНЖТЪ ЕМОУ' МНОГЪІ БО ИСЦЕЛИ' ΈΚО НАПАДАХЖ ЕМЬ ХОТАШТЕ ПОНКОСНЖТИ CA EME ENHKO HMEAX R PAHTI MK 3, 9-10 MAR.

ї съмњ рабъ твоїхъ оудръжітъ і ї любьяштє-і-мы твоє въсельнтъ сьо во N'T PS SIN 68, 37.

предо нь припаджтъ єттоптьні ї враді єго пръсть поліжять PS SIN 71, 9.

аште заповъди мото съедюдете: пръвждете въ любъви мост вкоже азъ заповъди оца моего съелюдохъ: и превъвания въ него лювъве Jn 15, 10 zogr.

# Class 2/p\*: вьсь

### §319. Class membership

The class contains only two lexemes, вьсь and сиць.

### § 320. Paradigmatic peculiarities

Terminals of the hard and soft subtypes from the 2-pron set are mixed in this class. For the pronoun BbCL, the twofold rule should produce hard subtype terminals, while for cnys, it should produce soft ones. The actual distribution is shown in the paradigms below. It is useful to note that both paradigms use all hard subtype terminals that begin with t. There are no dual forms; for BLCL, cf. the duale tantum lexeme OEA.


The wordforms BLCa and BLC's are inflectional doublets. SAV and SUPR only show BLCA; Glagolitic sources predominantly have BLCt. See the distribution in sources in Vaillant, § 99. In PS SIN and EUCH, only BLCH occurs (for canonical BLCK); cf. BbCIX HOWTL PS SIN 77, 14.6

<sup>6</sup> Vaillant mistakenly cites this example as PS SIN 87,14; see Vaillant, § 99.

# Class 1/a: трик

### §321. Class membership

The class contains only one lexeme, трик. This plurale tantum lexeme has the following forms.


### § 322. Illustrations

сь рече могж разорити црквь бжиж: и трьми дьньми созъдати их Mt 26, 61 MAR.

ЕКО БО БЪ СОНА ВЪ ЧОВЕЕ КИТОВЪ: ТОИ ДНИ ТОИ ЖЕ НОШТИ: ТАКО БЯДЕТЪ СНЪ чскъ въ срдци земла: три дни с три ношти Mt 12, 40 zogr.

чесо ради муро се не продано въстъ на тръхъ сътъхъ пеназь и дано HHILIMIN'S Jn 12, 5 SAV.

# A note on adjectival forms of the טייעסדs type

### §323. Group membership

Some grammars treat adjectival forms like 01/2056 as uninflectable adjectives, i.e. adjectives that belong to a separate inflectional class. However, the same grammars treat the same forms, or forms that are identical in relevant characteristics, as adverbs. Such are, e.g., испльнь, намель, особь, оттъврьнь, различь, свободь, coyroysb, orgoto (see Vaillant, § 138-139 and § 85). The present grammar considers such forms in all their uses as extraparadigmatic.

### Illustrations

КНАЗИ СОДОМЬСТИЇ" И ЛЮДИЕ ГОМООЬСТИН: ИМЖЕ ОЖЦЪ КОЪВЕ ИСПЛЬНЬ SUPR 136, 3-4; дьнесь отъ мрътвънуъ въстал лазарь: многоу и различъ гнъвоу раздовшение намъ даютъ SUPR 303, 3-5; соугоувъ дьнесь пришьствию господьяе СОУГОУВЬ СЪМОТРЕНИЮ СОУГОУВЪ ЧЛОВЪКОЛЮБИЮЛЮВИЮ. СОУГОУВЪ СЪНИТИК: ВЪКОУПЪ ЖЕ СЬМЪРЕНИЕ СОУГОУВО КЪ ЧЛОВЕКОМЪ ПОСЪШТЕНИЮ SUPR 449, 1-5, сf. соутоуво Дьнесь пришестье гне соугоуво съмотренье соугово чавколюбьствие соцтоуво съмптьее въ коупъ же і съмъренье: соцтово: къ чкомъ посъштенья CLOZ 12b, 37-13a, 1.

# Commentary on individual adjectival lexemes in the twofold declension

### §324. List of commented lexemes (class 2/a and its deformations)


горьн҄ь. See preceding word.


дрѣвѣнъ. See preceding word.


дьньшьн҄ь. See preceding word.


земл҄ьскъ. Cf. земьскъ; for distribution in sources see Večerka.


лучии. *Comparativum tantum*.


covers the hapax gloss in Euch нынѣщьнее раздрушение (65a, 7), aberrant form of ca- nonical нын҄ꙗшьн҄ь; the other occurrences

are in Supr. Variations of the type нꙗ//н҄ѣ are standard for the basic writing system of Supr, cf. нын҄ѣшьн҄ꙗ⟨NSgf⟩Supr 127, 29–30 and нынꙗшьн҄еѥ Supr 488, 23–24.


прѣдьн҄ь [(прѣ).д.ьн҄.ь] 'front'. Cf. прѣЖьн҄ь [(прѣ).Ж.ьн҄.ь] 'previous' with a substitu- tively softened basic component.

прѣЖьн҄ь. See preceding word.


сыръ1. Cf. the substantive сыръ2 2/m.


дивьноѥ и грьдое Supr 506, 16–17; Meyer treats it as a Voc (таибьна).

тачаи. *Comparativum tantum*. Cf. такъ 2/p.


ун҄ии. *Comparativum tantum*.

утрьн҄ь. Aberrant spellout ютр- in Euch 49a, 6.

утрѣи. *Comparativum tantum*.


#### §325. List of commented lexemes (class 2/p and its deformations)

вьсакъ. Contra Večerka, who normalizes the headword as вьсꙗкъ. This spellout, frequently found in the teaching manuals, is not attested in the OCS sources and contradicts its orthography. In the PD вьсакъ and вьсѣкъ are treated as doublet lexemes. On the suffix ак (ѣк) {13} see § 850. Aberrant forms following the adjectival declension are attested, see § 402.

вьсѣкъ. See preceding word.

дъва. *Duale tantum*. Aberrant form дъвама: никотеры же рабъ· неможетъ дьвама гнома работати Lk 16, 13 Zogr (cf. Mar дъвѣма, also двѣма Zogr Mt 6, 24).


ѥдьнакъ. See ѥдинакъ.

ѥдьнъ. See ѥдинъ.

ѥликъ. See толикъ.


NSgm form is иже. Secondary (ad-prepositional) forms show the prothetic н which gives rise to the nonstandard alloform (н҄), e.g.: ꙇ приближішѧ сѧ въ вьсь· въ н҄ѭже ꙇдѣашете Lk 24, 28 Zogr. Aberrant form ижь for иже in Zogr, e.g.: да бѫдете снове оца вашего· ꙇжь естъ на небесехъ Mt 5, 45. (See Vaillant, § 96).

инъ. Shows aberrant Plen forms with 2-combi terminals; cf. иниі же отъ стоѧщихъ т слышавъше· глахѫ (for canonical ини) Mt 27, 47 Sav. See § 402.

коликъ. See толикъ.

оба. *Duale tantum*.

самъ. Shows aberrant Plen forms with 2-combi terminals; cf. сьдеже и самоѥуныниѥ отꙙто быстъ Supr 493, 1. See § 402.

селикъ. See толикъ.


# **Morphologically masculine substantives**

# Class 2/m: градъ/конь

§326. Class membership

This class contains more than 700 lexemes. The hard and soft subtypes are distinguished by the right edges of the workstem. Possible workstem right edges are shown in Table 326 (p. 181).


Table 326. Possible workstem right edges in class 2/m

\* Open stems are found outside of the benchmark list, namely among compounds.

# §327. Profiles

# Here are the profiles of some type representatives:



# §328. Notes on word formation

Both root and suffixal substantives are possible in this class. Table 328 gives examples.

Table 328. Word-formation subtypes of masculine substantives


Many root nominals in the soft subtype result from *segmentless substitutive softening* (§ 864). Such are, for example, стражь, воЖь, ножь, въпл҄ь, сѫпьр҄ь, плачь, кличь.

# Class 2/m\*: дѣлател҄ь and class 2/m\*\*: граЖанинъ

# §329. Class membership

This class contains the following lexemes.


1) Plural forms of the lexeme жител҄инъ are the same as the plural forms of the lexeme жител҄ь; there are four Pl glosses in total: жителъ (Lk 15, 15 Zogr, Mar) and житель (Lk 15, 15 Sav, As) for GPl.

Note that the subclass граЖанинъ contains all the substantive ending in -инъ except the following: господинъ, чловѣчинъ, воинъ, съвоинъ, окринъ, чинъ. The lexemes жѧтел҄ь and жѧтел҄ꙗнинъ are derivational doublets. The граЖанинъ subclass contains also a large number of anthroponyms derived from place names, such as, самар҄ꙗнинъ, plural самар҄ꙗне; галилѣꙗнинъ, plural галилѣꙗне.

# §330. Profiles

Here are the profiles of the corresponding sample lexemes.


Other forms on the pattern of кон҄ь, i.e. by the standard twofold declension type in its soft subtype.


Other forms in Sg and Du with expanded stems (such as граЖан.ин), while Pl forms with syncopated stems (such as граЖан) on the pattern of градъ, i.e. by the standard twofold declension type in its hard subtype.

### §331. Paradigmatic peculiarities

The subclass дѣлател҄ь 2/m\* and граЖанинъ 2/m\*\* take the NPl terminal е from the nonstandard set c-simplex, instead of the expected и from the set 2-base. All other forms have the expected terminals from the 2-base set. The subclass гражданинъ 2/m\*\* distinguishes syncopated and expanded stems. Syncopated stems occur in Pl forms; expanded stems in Sg and Du forms. Syncopated stems can be derived from expanded stems, which are represented in the starting form, by removing the final suffixal formative ин. Thus, for the expanded stem граждан.нн, the syncopated stem is граждан; for the expanded stem жатейан.им, the expanded stem is жателан. See § 397 for aberrant form.

§ 332. Illustrations

а врагъ естъ въставъ плъвелъ диъволъ а жатва естъ коньчание въком" a жателе (for жатель) сятъ акли Mt 13, 39 ZOGR.

ВЕЧЕРОУ ЖЕ БЪ (ВЪ)шю. ГЛА ГНЪ ВИНОГРАДА. КЪ Приставъником своемочъ Призови дълателя: в даждь имъ мъздж: наченъ отъ последьникуъ до пръвъинуъ Mt 20, 8 мая.

W ДЖЕЄ БЛАЖЕНЪВ: ЕГОЖЕ ЦЪСАДЕ ЧЪТЖТЪ: ВЛАСТЕЛЕ И ВЕЛЬМЖЖА ЛЮБАТЪ SUPR 354, 5-6.

ЖАТЕЛЕНЕ (for ЖАТЕЛАНИНЪ) ЖЕ ОВИ КЛИЦААХЖ ОВИ ТЕЧААХЖ ВЬ СЛЕДЪ ВЛЬКА SUPR 43, 24-26.

СИ СЛЪЩАВЪ ЦЪСАРЬ ОУАЛЪ: ПОНЗЪВА ДВА БОЛГАРИНА: НИА ЕДНОМОУ САТООНИКЪТ a goveroyoumor of the rop supr 195, 17-19.

и сътворя та идного отъ водеръ своихъ SUPR 61, 3-4.

### Class 1/m: пять

### §333. Class membership

This class contains 29 lexemes. Workstems can only end in simple consonants. Here are all the lexemes in this class: воль, вєпрь, голявь, гость, гость, гость, грътамь, ДЬНЬ, ІЄЛЕНЬ, ЗВЪОЬ, ЗАТЬ, КАМЕНЬ, КОРЕНЬ, ЛАКЪТЪ, ЛЮДИЕ, НОГЪТЪ, ПАЗНЕГЪТЪ, ПЕЧАТЪ, ПЛАМЕНЬ, ПОКОНЬ, ПОЬСТЕНЬ, ПЖТЪ, ОЕМЕНЬ, СТЕПЕНЬ, ТАТЪ, ТОВХЪТЪ, ТЪСТЬ, оүшидь, чрьвь, жгль. See § 401 for aberrant spellouts that show alien terminals.

#### § 334. Notes on word formation

Substantives in this class may be root lexemes; cf., for example, гость, вяпоь, жгль, тьсть. They may also be suffixal, e.g. ногъть [(ног).ът.ь], прьстьнь [(прьст).єн.ь].

#### 8335. Profiles

Here are the profiles of some type representatives:



# Commentary on individual morphologically masculine substаntive lexemes

§336. List of commented lexemes (class 2/m and its deformations)


въторьникъ. See preceding word.


дрѫгъ. See другъ 'friend' ‹248›.

жидъ. Contra Večerka, who treats all the glosses under the headword жидовинъ, even including the following: жидомъ DPl (e.g., Supr 303, 7), жиды APl (e.g., Supr 315, 19), жидыIPl (e.g., Supr 398, 6). In the PD жидъ and жидовинъ are different headwords. Ab- errant forms are possible (cf. LPl жидохъ Supr 496, 7). See § 399.

жидовинъ. See preceding word.

жител҄ь. Contra Večerka, who gives two headwords, жител҄инъ and жител҄ь in one dictionary entry. These are all attested forms: 1) NSg жителинъ (Supr) for the headword жител҄инъ; and 2) GPl жителъ (2×) and житель (2×); the latter four glosses can be treated under both headwords. In the PD жител҄инъ and жител҄ь are different headwords.

жител҄инъ. See preceding word.

жупелъ 'sulfur'. Only three glosses: жюпелъ Lk 17, 29 Mar and the aberrant жюпьлъ Lk 17, 29 Zogr and зюпелъ Ps Sin 10, 6.

жѧтел҄ꙗнинъ. Cf. жѧтел҄ь; for distribution in sources see Večerka.

жѧтел҄ь. See preceding word.


исполинъ. Cf. сполинъ.


младенищь. Cf. младѣнищь. On the compe- tition between suffixes ен, ьн, ѣн see § 896.

младеньць. See preceding word.


попелъ. See пепелъ.


прѣдѣдъ. Cf. прадѣдъ.


размышл҄ꙗи. Contra Večerka, who gives the headword розмышлꙗи; hapax gloss with the aberrant spellout роз- in Supr 543, 14–15.

родъ. Contra Večerka, who gives two homon- ymous headwords, родъ1 'kinship' and родъ<sup>2</sup> 'Gehenna, hell'. The PD follows Sadnik, who treats both meanings under the root ‹‹761››.

скѫдълъ 'sherd'. Contra Večerka, who gives four headwords, скѫдьлъ, скѫдълъ, скѫдъль, скѫдьль, in the same dictionary entry. Five glosses altogether, for distribution in sources see Večerka. Cf. скѫдъльникъ. (from Latin *scandula*, *scandella*, see Vasmer, *скудель*).

скѫдъльникъ. Contra Večerka, who gives three headwords, скѫдьльникъ1 'potter', скѫдьльникъ2 'pitcher', and скѫдъльникъ 'pitcher'; the last two are in the same dictionary entry. Five glosses altogether.

сполинъ ‹338›. Contra Večerka, who gives the headword сполинъ as reference to исполинъ. Aberrant GPl form споловъ in Supr 485, 10. Cf. исполинъ, hapax gloss in Ps Sin 18, 6.

стѣн҄ь 'shadow' ‹909›. Contra Večerka, who gives the headword стѣнь 1/m. Cf. kamorated spellouts in Supr: сльньцувъсходꙙщу съкрыѥтъ сꙙ стѣн҄ь Supr 417, 28–29. 14 occurrences, all are NASg forms, with masculine agreement. Cf. сѣнь 1/f ‹929›.

съвѫзъ 'bond, union'. Aberrant spellout съѫз-Supr 400, 14. Cf. ѫза.

сынъ1 'son' ‹923› and сынъ2 'tower' ‹922› are homonyms.

сыръ2. Cf. the adjective сыръ1 2/a.

сѣчь 'battle'. Contra Večerka, who gives the headword сѣчь (m. or f.). Hapax gloss и прѣста сѣчъ Ps Sin 105, 30.

сѣчьць 'executioner, headsman'. Cf. сѣчьца; for distribution in sources see Večerka.

сѫдъ ‹936›. Contra Večerka, who gives two headwords сѫдъ1 'court hearing' and сѫдъ2 'vessel'; the two are etymologically cognate; cf. съсѫдъ 'vessel' ‹936› and сѫдии 'judge' ‹936›.

трудъ. Do not confuse: трудъ 'effort, labor' ‹971› and трѫдъ 'disease' ‹986›.

трѫдъ. See preceding word.

тысѧщьникъ. Cf. тысѫщьникъ; for distribution in sources see Večerka.

тысѫщьникъ. See preceding word.

херовимъ. Violation of the law of the velars, see § 869.

хитонъ. Violation of the law of the velars, see § 869.

хрьбьтъ. Aberrant spellouts хриб- and хръб-.

чари. *Plurale tantum*.


ѫзьникъ. See preceding word.

ѫродъ. Contra Večerka, where ѫродъ is an adjective. Only in Supr: Voc ѫроде 4×, NSg ѫродъ 1×.

§337. List of commented lexemes (class 1/m and its deformations)

вепрь. Contra Večerka, who gives the headword as вепр҄ь. Hapax gloss ѡзоба і вепрь отълѫга· їнокъ дивьеї поѣлъ естъ Ps Sin 79, 14.

дьнь. Aberrant spellouts with terminals from the c-simplex set. Cf. e.g.: дьне Mt 6, 11 Mar.

камень. Cf. камыкъ 2/m and камениѥ 2/n; aberrant form ⟨NASg⟩ камы in Supr.

л҄юдиѥ. *Plurale tantum*.

ушидь 'fugitive'. 1× in Supr 93, 11.

# **Morphologically neuter substantives**

# Class 2/n: село/полѥ

# §338. Class membership

This class contains slightly more than 1000 lexemes. Hard and soft subtypes are distinguished by the right edge of the workstems; possible workstem right edges are shown in Table 338.

# Table 338. Possible workstem right edges in class 2/n


# §339. Profiles

Here are the profiles of some type representatives:


# §340. Notes on word formation

This class contains both root and suffixal substantives. Table 340 shows examples.

Root nominals of the soft subtype are the result of *segmentless substitutive softening* (see § 864). Such are, for example, ложе, пол҄ѥ, мор҄ѥ, вѣще. A significant part of the nominals carries the suffixal determinant ьј=е, often with н- or т-Part as the basic component, cf. биѥниѥ, възнесениѥ, исхоЖениѥ, устроѥниѥ, убитиѥ, прослутиѥ, пѣтиѥ.


#### Table 340. Word-formation subtypes of neuter substantives

# Commentary on individual substаntive morphologically neuter lexemes

### §341. List of commented lexemes


ꙗдро. See preceding word.


довьльство [до.(вьл).ьств.о]. Contra Večerka, who gives the headword довъльство. Only two glosses, morphologically strange spellouts довл҄ьств- Supr 494, 20; 494, 22. Cf. довольнъ, довьлѣти.

дръва. *Plurale tantum*.


желѣниѥ. See preceding word.


лиць. Aberrant forms личесе SUPR 335, 14, личеса SUPR 397, 18, see § 400.


плоушта 'lungs'. Plurale tantum. Contra Vecerka, who normalizes the headword as плюшта. Hapax gloss supr 165, 10 плоушта.


n true. See preceding word.

пята 'fetters'. Plurale tantum.


# **Morphologically feminine substantives**

# Class 2/f: жена/землꙗ

# §342. Class membership

This class contains more than 450 lexemes. Soft and hard subtypes are distinguished by the right edge of the workstem. Possible workstem right edges are shown in Table 342

# Table 342. Possible workstem right edges in class 2/f


# §343. Profiles

Here are the profiles of some type representatives:







# §344. Notes on word formation

This class contains both root and suffixal substantives. Table 344 shows examples.


Table 344. Derivational subtypes of feminine substantives

In the soft subtype many root nominals are the result of *segmentless substitutive softening* (see § 864). Such are, for example, земл҄ꙗ, тѧжа, суша, притъча, одеЖа, купл҄ꙗ, зор҄ꙗ.

# Class 2/f\*: рабыни

# §345. Class membership

This subtype contains the following 30 substantives: балии, благостын҄и, благын҄и, богын҄и, вѣтии, господын҄и, гръдын҄и, корабьчии, крабии, кръмьчии, крьстиꙗнын҄и, кън҄игъчии, ладии, льгын҄и, милостын҄и, мльнии, поганын҄и, правын҄и, простын҄и, пустын҄и, рабын҄и, самъчии, свинии, свѧтын҄и, сокачии, сѫдии, сѫсѣдын҄и, тысѧщи//, тысѫщи//, шаръчии. The starting forms (NSg) end in и (рабын҄=и). The workstem ends in a morphophonologically soft consonant, viz. j, н҄, or щ (see examples in § 348 below).

# §346. Paradigmatic peculiarities

Class 2/f\* has the NSg и terminal from the nonstandard set c-simplex instead of the expected а from the set 2-base. All other forms have terminals from the set 2-base.

# §347. Profiles


Here are the profiles of some type representatives.


### § 348. Notes on word formation

### All substantives here are suffixal.


Note that while for nominals formed with the suffixes ፔוቡ and ጽሠጥ/ሐሠד, class 2/f\* is the only paradigmatic possibility, for substantives formed with the suffix bj this is not the case: such substantives contain lexemes that follow the main twofold declension. Cf., for example, 3мма [3м.ь], вратига [врат.ь].а], and some others.

# Class 1/f: кость

### §349. Class membership

This class contains slightly fewer than 200 lexemes. Possible workstem right edges are shown in Table 349.

Table 349. Possible workstem right edges in class 1/f


### ട്ട് 350. Notes on word formation

Substantives in this class can be both root and suffixal. Examples:


### §351. Profiles

### Here are the profiles of some type representatives.


# Commentary on individual substаntive morphologically feminine lexemes

### §352. List of commented lexemes (class 2/f and its deformations)

бездъна [без.дън.а] ‹264›. Frequently aber- rant spellout with ь alongside the canonical one. Aberrant spellouts безденъ for GPl in Ps Sin: въсходѩтъ до небсъ ї низъходѩтъ до безденъ (106, 26). Cf. бездънъ, бездъниѥ.

беспосагаꙗ. Contra Večerka, who gives the adjectival headword беспосагаи. Hapax gloss твоѥѧ беспосагаѧ матере и дѣвицꙙ Supr 391, 26–27.

братиꙗ. Cf. братриꙗ; for distribution in sources see Večerka.

братриꙗ. See preceding word.

брън҄ꙗ ‹54› 'armor'. Contra Večerka and following Sadnik. Večerka gives the starting form брънѩ (*Plurale tantum*). Do not confuse with брьна 'dirt' ‹56›.

брьна 'dirt'. See preceding word.

вльна1 'wave' ‹77› and вльна2 'fleece' ‹108› are homonyms.

вьдова and вьдовица. The spellout of the root vowel follows Večerka, although the overwhelming majority of the attested forms show the root as въд. See § 681.

геона. Violation of the law of the velars, see § 869.

господа 'inn'. Cf. госпоЖа, господын҄и 'lady'.

гостиница. Cf. гостиньница; for distribution in sources see Večerka.

гостиньница. See preceding word.

двьрьца. Contra Večerka and following Sadnik. Večerka gives only the headword двьрьцѧ (pl.f.).

дивиꙗ. Contra Večerka and following Sadnik and Meyer. Večerka gives only the adjective дивии. Hapax gloss (дивии DSg Supr 350, 17).

дира. Aberrant spellout дирѣ (NSg) in Zogr Mk 2, 21.

дроЖиѩ. *Plurale tantum*.

дрѧзга. Hapax gloss ꙗко левъ въ дрꙙздѣ Supr 12, 8.

дѫбрава. Cf. дѫброва.

дѫброва. Contra Večerka, who gives no headword дѫброва, only дѫбрава. Aberrant spellout with добров- 1× in Ps Sin 73, 5.

зар҄ꙗ. Cf. зор҄ꙗ.

змиꙗ. Cf. змии [зм.ьј.ь].

зор҄ꙗ. Contra Večerka, who gives two head- words зор҄ѩ (pl.f.) and зар҄ꙗ.

изун҄ьшина [из.(ун҄).ьш.ин.а] 'improvement'. Contra Večerka, who gives the headword without kamora. Hapax gloss Supr 314, 15–16 without kamora. Cf. ун҄ьшиина.

кл҄ѥвета. Contra Večerka and following Sadnik. Večerka gives the headword клевета and its derivatives without kamora.

кръма1 'stern' ‹448› and кръма2 'food, nourishment' ‹447› are homonyms. Cf. кръмл҄ꙗ ‹447› 'food, nourishment'.

кръмл҄ꙗ. See preceding word.

кръмьчии. Masculine agreement. Two glosses altogether, NSg душевьныи крьмьчи (Supr 474, 17) and ASg кръмьчиѭ убишꙙ (Supr 472, 13–14).

кън҄игы. *Plurale tantum*.

кън҄ижицѧ. *Plurale tantum*.

ладии ‹483›. Contra Večerka, who gives two headwords алъдии and ладии in the same dictionary entry. Aberrant spellouts with алъд- and алд-; for distribution in sources see Večerka.

лѧдвиѩ. *Plurale tantum*.

матица 'river of fire, hell'. Hapax gloss матицѫ огн҄ьнѫѭ Supr 525, 30.

мѫка1 'flour' ‹595› and мѫка2 'torture, suffering' ‹595› are homonyms. Aberrant spellouts of мѫка2 with мук- (2×) and мѵк- (1×) are attested in Cloz.

ножьницѧ. *Plurale tantum*.

нѫЖа. Aberrant spellouts with нуЖ- in Supr; for distribution in sources see Večerka.

осъпы. *Plurale tantum*.

паѫчина. Three glosses altogether: паѫ- (Ps Sin) and the aberrant паѭ- (Supr) and пау- (Ps Sin).

пиꙗница. Aberrant spellout пьѣн- 1× Zogr Mt 24, 49. Cf. пиꙗньство.

прапрѫда. Cf. прѣпрѫда, прѣпрѫдъ, прапрѫдъ; for distribution in sources see Večerka.

прѣгын҄ꙗ [⸨прѣгын҄⸩.а] 'mountain wilderness'. Contra Večerka, who gives two headwords прѣгын҄и and прѣгынꙗ in the same dictionary entry. Hapax gloss ѥстъ же въ прѣгын҄ѣхъмѣстото· и вь непрѣходъныихъ горахъ Supr 26, 19–20. See Durnovo 2000, p. 690–693.

прѣпрѫда. See above, прапрѫда.


слуга. Masculine agreement.

смокъвьница. Aberrant spellouts with смоков-.


тысѫщи. See preceding word.


юноша. Masculine agreement.

ѫза 'chains, fetters'. Aberrant spellout вѫз- 1× Supr 442, 27. Cf. съвѫзъ 'connection' and ѫже 'rope'.

### §353. List of commented lexemes (class 1/f)

#### ꙗсли. *Plurale tantum*.


гѫсли. *Plurale tantum*.

дрьколь. Contra Večerka, who gives the headword дрьколь (m.). Cf. ⟨NPl⟩ дръколи Cloz 12b, 13.

дѣти. *Plurale tantum*.

зъль 'evil deed'. Cf. зъло 2/n.

пѣготы. *Plurale tantum*.

къзнь. Cf. кызнь, for distribution in sources see Večerka.

кызнь. See preceding word.

мощь [(мог).т.ь]. Contra Večerka, who gives two headwords мощи 'relics' *Plurale tantum* and мощь 'power' 1/f (the latter covers four glosses). Only one lexeme is registered in the PD. Cf. also немощь 1/f.

мъдьлость [(мъд).ьл.ост.ь] 'indolence' ‹572›. Contra Večerka, and following Sadnik. Večerka gives the headword as мьдлость. Only two glosses, both in Cloz: мъдлост-2b, 30 and 5b, 9. The spellout in Večerka shows the prohibited cluster дл.

ноздри. *Plurale tantum*.

пропадь. Cf. пропасть [про.пад.т.ь], for distribution in sources see Večerka.

пропасть. See preceding word.

прьси. *Plurale tantum*.

сѣнь. Contra Večerka, who gives two homon- ymous headwords, сѣнь1 'shadow, umbrage' and сѣнь2 'tent, tabernacle'. Only one lex- eme is registered in the PD. Cf. стѣнь 2/m.

тѧгость. Cf. тѧжесть, for distribution in sources see Večerka.

тѧжесть. See preceding word.

# **Secondary forms**

§354. General

Nominal declension contains two kinds of secondary forms:7 the vocative form (Voc), and the personal dative (D2). Table 354 shows how they are distributed among inflectional classes.

Table 354. Distribution of nominal secondary forms


§355. Formation of vocative forms

These forms take special markers and follow special boundary adjustment rules. Formation of vocatives is shown in Table 355. The workstem is the same as in primary forms.

Table 355. Vocative formation


7 Ad-prepositional forms of the pronouns \*и and иже are also secondary; see § 318.

Vocative forms can be classified as paradigmatic only somewhat arbitrarily. Their special status makes them similar to adverbs and other extraparadigmatic forms. Accordingly, the markers у, е, о/е, и are neither ordinary terminals nor ordinary suffixes. Cf. the marker at end of adverbial extraparadigmatic forms, such as камо, къгда, and some others.

Boundary adjustment rules are morphophonologically anomalous. In part, instead of the standard replacement by the velar palatalization alternation к→ч, vocatives use the anomalous replacements ц→ч and ѕ→ж (see § 108).

Examples: (1) 2/m (substantives): мѫжь — мѫжу, отьць — отьче, кънѧѕь — кънѧже, сынъ — сыне, влькъ — вльче, богъ — боже, духъ душе; (adjectives): слѣпъ — слѣпе. (2) 2/f (substantives): душа — душе, дѣвица — дѣвице, жена — жено, владыка — владыко; (there are no such adjectives). (3) 1/f: милость — милости; for 1/m only господи (господь 0/m).

All other inflectional classes lack vocative forms; the corresponding syntactic positions contain NSg.

Some nominal 2/m lexemes of the hard subtype show aberrant vocative forms in у, cf. сыну for сынъ. E.g. помилуи мꙙ господи сыну давыдовъ Supr 392, 18–19.

### Illustration

фарисею слѣпе· очисти прѣЖе вьнѫтрьнее стекльници· ꙇ поропсидѣ да бѫдетъ и вьнѣщьнее има чисто Mt 23, 26 Mar; чудо велико объдръжитъ мꙙ кнꙙже Supr 157, 4–5 (cf. the aberrant form in не льсти сꙙ кнꙙзу Supr 156, 7–8); исъ рече еи· жено кьде сѫтъ иже на тѧ ваЖахѫ· никыи же ли тебе не осѫди Jn 8, 10 Mar; ѡ зависти· убоиству родителю· диꙗволꙗ обрѣтѣл҄ь· сьмрьти старѣишино Supr 388, 22–24; свꙙтыи артемонъ глагола· омрачене умомъ· змиѥ обьветьшавьшіꙗ· рабе дьбри огн҄ьныѧ· ѥще ли сꙙ отъмещеши не хотꙙ слышати Supr 231, 11–14; бе вьседръжителю· ѥдинъи блаже и милосрьде· послушаѧи вьсѣхъ въ истинѫ работаѭщиїхъ тебѣ· покажи и о мьнѣ недостоинѣ рабѣ своѥмъ· своѭ благын҄ѫ Supr 21, 26–22, 1.

ꙇ пристѫпл҄ь· примы д талантъ· принесе другѫѭ д таланꙿтъ· глѧ ги· д таланꙿтъ ми еси прѣдалъ· се другѫѭ д таланꙿтъ· приобрѣтъхъ ꙇми· рече ему гь его· добры рабе благы· ꙇ вѣръны· о малѣ бѣ вѣрьнъ· надъ многы тѧ поставл҄ѫ· вьниди въ радость ги своего· пристѫпль же· примы б таланꙿта· рече ги· б таланꙿта ми еси прѣдалъ· се другаѣ б таланꙿта приобрѣтохъ ꙇма· рече ему гь его· добры рабе благы вѣрьне· о малѣ бѣ вѣрьнъ надъ многы тѧ поставл҄ѫ· вьниди· въ радость га своего Mt 25, 20–23 Zogr.

### §356. Formation of the personal dative D2

The workstem is the same as in primary forms. The terminal ови/еви is from the 2-duplex set. For example, богъ — богови, мѫжь — мѫжеви, отьць — отьцеви, мор҄ѥ — мор҄ѥви.

#### Illustrations

ОНЪ ЖЕ ВЪДЪЧШЕ ПОМЪЩЛЕНИЪ ИХЪ. Г РЕЧЕ ЧЛВЪКОВИ ИМЖШТЖМОУ СОЦУЖ ржкя: въстани и стани по средъ: онъ же въста Lk 6, 8 м А в; достоитъ ли намъ кесаревн дань дати" или ни Lk 20, 22 м А R; гла емог нюда не искариотъскъ: гги н ЧЪТО БЪКТЪ ГАКО НАМЪ ХОЦЕШИ ГАВИТИ СА: А НЕ МИОВИ Jn 14, 22 SAV; Пръдајешн рьци мн вьса съдръжАШТААГО БЪСЪ! ВЛАДЖШТААГО\* МОРЕВИ ЗАПОЋШТАНЖШТААГО SUPR 412, 5-7.

# CHAPTER 12 **Unique nominal lexemes**

§357. A list of unique nominal lexemes

There are 49 unique nominal lexemes. All lexemes in this class are listed in Table 357.


### Table 357. Unique nominal lexemes

1 Only Sg forms. 2 Only Pl forms. 3 Indeterminate morphological gender.

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

Paradigms of model lexemes from each group are shown below. Doublet forms (//) can be present in the paradigms, i.e. forms that are both primary, both canonical, and are treated as equivalent (their distribution is not determined). The source of these doublets is the paradigmatic effect of alien terminals in the paradigm.

Also, the attribution of the terminals to their set is shown in the paradigms.

# Group имѧ 0/n

# §358. Membership and model paradigm

This group includes брѣмѧ, врѣмѧ, имѧ, писмѧ, племѧ, сѣмѧ, чисмѧ. Here is a model paradigm.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

# §359. Notes


# Group отрочѧ 0/n

§360. Membership and model paradigm

This group includes агнѧ, жрѣбѧ, кл҄юсѧ, козьлѧ, овьчѧ, осьлѧ, отрочѧ. Here is a model paradigm:


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

### క్రి 361. Note

The workstem is the same in all forms; the final suffix of the stem is the finally ambivalent ѧ/ѧㅠ, cf. отроч.ѧ/отроч.ѧт, овьч.ѧ/овьч.ѧт. See Vaillant, § 72 on the distribution of forms.

# Group СЛОВО О/п

### ട്ട് 362. Membership and model paradigm

This group includes исто, коло, нево, слово, тъло, чоудо. Here is a model paradigm.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

### ട്ട് 363. Notes


On the distribution of forms, see Vaillant, \$ 72-73. On aberrant forms with stems expanded by the suffix ec, in the lexemes диво, довео, дъло, лице, люто, see § 400.

# Group око 0/n

#### §364. Membership and model paradigm


This group contains око, оухо. Here are all the attested forms.

The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

### ട്ട 365. Note

Expanded and syncopated workstems are distinguished here; the expanded stem contains the suffix ec. The root formative stands in the u grade by the velar palatalization alternation before the suffix €c, as well as before the 1-simplex terminals (and in the anomalous forms Voyma, Voywama), but in the y grade before the corresponding 2-base terminals.

### ട്ട് 366. Illustrations

LSg - чьто же видиши сячьць ижь єсть въ оць (ZOGR, MAR, but очес AS, очесе SAV) вратра твоего а врьвна єже єсть въ оць (ZOGR, MAR, AS, but очесе SAV) TBOEML HE 4KOELIH Mt 7, 3 ZOGR.

GSg - 1.3%MH Пръвъе връвъю 1.3 очесе (ZOGR, MAR, but ока AS) тъсего и тъгда 0\360HILIH 13ATH 1 CRYLLE 13 OHECE (ZOGR, MAR, AS) EDATA TROETO Mt 7, 5 ZOGR. DSg - BB CANYX by Of You OYCARILLAWE MIA PS SIN 17, 45.

ISg - доврее ти есть съ едничемь окомы: вынити вь црсие вжие: нежели OE'S OYH LMALITY' ITTH B' TEOHS OTHERSER Mk 9, 47 ZOGR.

NAPI - источьницъ водъючъюхъ прикладаюсма съухъ очеса: и акъи съревро СЛЪЗЪІ ВЬОДШТА КАПААХЖ ПО ЛАНИТАМА НА ЗЕМЫЖ SUPR 397, 23-26; НЪІНІА ЖЕ очеса срьдьць съмъживъше SUPR 401, 12.

GPI - съуранен аплър свою отъ кляда [ ... с отъ възьръниъ на на очесъ любод виць EUCH 37а, 8-13.

IPI - стоудънося воленя: опреснакъ в дать ні обръзани срьдьці ні оушесъ! CBOIMI CLOZ 7a, 18-20.

For further details on the distribution of forms, see Vaillant, § 73.

### Group црькъї O/f

#### §367. Membership and model paradigm

This group includes врадъ, воукъвн, жрьнъ, локъ, люкъ, неплодъ, пръдюсть, свекоты, смокъю, хорягъ, цоькъ, целъ. Here is a model paradigm.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

#### ട്ട് 368. Notes


On the distribution of the forms and agreement peculiarities, see Vaillant, § 74.

# Group мати O/f

### ട്ട് 369. Membership and model paradigm

This group includes мати and дъшти. Here is a model paradigm.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

### క్తి 370. Note

Expanded and syncopated workstems are distinguished here; expanded stems contain the suffix ep. On the distribution of the forms, see Vaillant, § 74.

# Lexeme господь 0/m

### §371. Paradigm

Here are all the attested forms of the lexeme господь.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

### §372. Notes


On the distribution of forms, see Vaillant, § 66.

# Lexeme десѧть 0/m

### §373. Paradigm

Here are all the attested forms of the lexeme десѧть.


The source set of terminals of the given forms is shown to the right; cells are shaded if the corresponding forms are not attested for any lexemes in the group.

### §374. Notes


On the use of forms and agreement, see Vaillant, § 102.

# Pronouns หัวเท 0/p and сь 0/p

## §375. Pronoun къю

Here is the paradigm of the pronoun към.


### క్తి 376. Notes


#### \$377. Pronoun сь


Here is the paradigm of the pronoun cb.

### క్ర 378. Notes


# Pronouns чьто 0/p and къто 0/р

# ട്ട് 379. General

These lexemes have no opposition by number or gender. The forms are only opposed by case, as shown below. These lexemes belong to the agreementless numberless agreement class.



# §380. Notes


# Pronouns a3% O/s, TTI O/s, and CA 0/s

#### క్రి 381. Paradigms

Here are the paradigms of all three lexemes.



### ട്ട് 382. Notes


See Vaillant, § 97 on the distribution of forms.

# Lexeme четыре 0/a

# §383. Paradigm

Here is the paradigm of the lexeme четыре.


# §384. Notes


# CHAPTER 13 **Aberrant nominal forms in sources**

# §385. General

This chapter examines aberrant nominal forms that show paradigmatic aberrations.1 Paradigmatic aberrations of nominal forms are generated by the corresponding paradigmatic effects, most of which are of the type "alien terminals in the paradigm" (cf. дару for canonical дара GSg, or дьне for canonical дьниѥ NPl). Aberrant forms of unique lexemes are treated in Ch. 12, *Unique lexemes*.

### §386. Alien terminals

Aberrant nominal forms that are generated by the effect of the alien terminals in the paradigm contain terminals that differ from that which is assigned to the declension type of the lexeme. For example, the lexeme даръ 2/m should take the GSg terminal а, but the aberrant form дару has the terminal у. Every terminal belongs to a certain terminal set; accordingly, the aberrant form дару contains a terminal from a different (alien) set (u-simplex).

The alien terminal effect accounts for a whole series of paradigmatic aberrations, depending on the set to which the alien terminal belongs, and the set which is assigned to the corresponding lexeme canonically.

Terminals from nonstandard sets intrude in the majority of cases, viz. the sets u-simplex, c-simplex, or 2-duplex. However, in some cases, terminals from standard

1 Aberrant forms that show segmental aberrations are treated in Part I, *Segmental grammar*.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

sets function as alien. However, it is possible to have aberrations where a nonstandard terminal that is assigned by the canonical paradigmatic synthesis rules is replaced in the aberrant form by a nonstandard one. Such are, for example, the forms of ш-Part like бывъшии for canonical бывъшеи (see § 396 below).

# §387. Attestations of alien terminals in aberrant forms

Because terminal sets intersect, it is not always possible to determine the source set of an alien terminal in a given aberrant form. However, the assortment of alien terminals available to aberrant forms of lexemes in a given declension type is in part limited. An overview of possibilities is given in Table 387.


# Table 387. Intrusion of alien terminals

# §388. Set intersections

Terminal sets intersect, including standard and nonstandard terminal sets. For example, the ISg terminal емь belongs to the nonstandard set c-simplex and the standard set 2-base (soft subtype). Besides, in certain cases, the observed aberrant terminal can be treated not only as resulting from its intrusion from a different set, but as a certain deformation of the canonical terminal. For example, the terminal емь in the wordform печатемь (ISg) can be interpreted as the intrusion of the terminal емь from the c-simplex set, and as strengthening in the canonical form печатьмь.

Let us take for comparison all the sets from the masculine morphological gender (Table 388).

Some terminals are specific for a given set, in that they are not found in any other set. For example, the NPl е terminal is specific to the c-simplex set. Note that, if we exclude the 2-duplex set, all remaining nonstandard sets (c-simplex and u-simplex) would contain only terminals specific to those sets, except емь and емъ. The latter show their morphological independence in forms like словесемь (unique), печатемъ (aberrant), where the terminal емь (or емъ) from the 2-base set is inappropriate given the twofold rule. However, these forms may be seen as a result of the strengthening of the terminals ьмь and ьмъ (represented in the 1-simplex set).

Here, the spellouts in SAV are important, because SAV has no cases of strengthening. Indeed, in sav we find именемь Lk 10, 38, людемь Lk 3, 18, словесемъ Jn 17, 20 and some others.

The terminal set 2-duplex, on the contrary, mostly contains nonspecific terminals, combined from u-simplex terminals (quasi hard subtypes), and 1-simplex (quasi soft subtypes).


Table 388. Terminal sets in the masculine morphological gender

1 The terminal zx's is not attested; in its place we find ox's; see Vaillant, § 58-59.

### § 389. A note on 7-initial terminals

It is worth noting that terminals with initial yers with yer aberrations can be homophonous with eponymous terminals from other sets. However, the general problem of attributing each aberrant form to the type of aberration that it has undergone-and, accordingly, showing a terminal from one or another set-is neither posed nor resolved here. For example, the aberrant form πεναπελλω (for canonical пєчатьмь) allows to interpretations of equal status: intrusion of c-simplex and strengthening of 1-simplex. Note that the terminal was is not attested in OCS texts (in its place we find the terminal ox's, see more details in Vaillant, § 58-59).

### § 390. A note on new twofold rule

Among 2-duplex terminals there are those that are specific to this set. Such are DSg євн, NPI євъ, GPI євъ, and DPI ъмъ. All these terminals are attested by a small number of forms. These are, for example, мжжєви Lk 1, 27 мАR, ZOGR; морєвн

Supr 412, 6–7; змиѥве Supr 77, 5 (and also 227, 29 and 229, 24); зноѥве Supr 172, 29; врачевъ Mk 5, 26 Zogr, Sav; змиевъ Euch 4b, 19–20 and Supr 467, 3.

Note that the forms with DSg ови/еви in personal name lexemes are treated in this grammar as secondary forms of the personal datives (D2).

§391. Notes on the diachronic interpretation of the alien terminals in the paradigm aberration

1. Historical grammar establishes the existence of special declension types: *u*-declension2 and C-declension,3 and distributes all lexemes into declension types of equal status (*o*-, *a*-, *i*-, *u*-, and C-declensions). The distribution is paradigmatic, i.e. determined by the dictionary. Accordingly, all observable aberrations except the aberrant forms that show specific 2-duplex terminals (such as, e.g., зноѥве), are treated as resulting from contamination of declension types. Of course, the sets of aberrant forms from the diachronic and synchronic points of views do not match. Indeed, such forms as, for example, LSg дому are treated in a diachronically oriented grammar as canonical (they have terminals from the u-simplex set due to their membership in the *u*-declension), while in a synchronically-oriented grammar, they are treated as aberrant.

In historical grammar, forms resulting from the contamination of declension types have different status. Namely, the consonantal declension is considered moribund, and, accordingly, all consonant-declension forms are archaic, while forms of *o*-, *a*-, *i*-declensions among C-declension lexemes are innovations. Aberrant forms with C-declension terminals among lexemes that do not belong to this class are not expected. Likewise, *u*-declension is also considered moribund, but it is found among lexemes of the twofold declension type. Accordingly, from a diachronic point of view, the introduction of *u*-declension terminals into lexemes that do not belong to that class are treated as a contamination aberration

2 The u-simplex set has terminals that are specific to this set. Such are GLSg у, NADu ы and GLDu ову. Lexemes that show aberrant forms with these specific terminals, from a diachronic point of view, are treated as belonging to an older declension type, the *u*-declension. Such are волъ, даръ, домъ, полъ, родъ, санъ, сынъ1, сынъ2, чинъ (9 lexemes altogether). These lexemes show diagnostic aberrant forms in different numbers. The lexemes ꙗдъ, врьхъ, гласъ, длъгъ, дѫбъ, миръ, рѧдъ, станъ (8 lexemes altogether) show diagnostic aberrant wordforms only in Sg. Let us also list all the lexemes that show aberrant Pl forms with the terminals ове or овъ. These are бѣсъ, гадъ, градъ1, гроздъ, грознъ, грѣхъ, духъ, жидъ, плодъ, попъ, потъ, садъ, сѫдъ, трудъ, удъ, цвѣтъ (16 lexemes altogether). Some of these lexemes show other aberrant forms with terminals that belong to both the *u*-declension set and the 2-duplex set. See more details in Vaillant, § 58–59.

3 The c-simplex set has terminals that are specific to this set. Such are NSgmnfASgmn 0, GLSgmnfNPlm е, NSgfNAPln и, LPlmn ехъ. Lexemes that show aberrant forms with these terminals, from a diachronic point of view, are treated as belonging to the old consonant declension type (C-declension). Among masculine 1/m substantives, the following twelve are assigned to the old C-declension by historical grammar: дьнь, ѥлень, камень, корень, лакъть, ногъть, пазнегъть, печать, пламень, прьстень, ремень, степень. Not all of these lexemes show specific C-declension forms in OCS. See Vaillant, § 68–70.

(cf. the form ꙗду, Vaillant, § 59, p. 112), and the corresponding forms are expected innovations.

Synchronic grammar lacks an instrument for distinguishing between these types of aberrations and assigning a different status to them as archaic or innovative forms.

2. Historical grammar offers a complete reconstruction of the *u*- and C-declension sets. Indeed, every lexeme of the corresponding type must in principle contain a full complement of forms. Accordingly, the following reasoning can be appropriately used in reconstructing terminal sets. Some lexeme *A* shows its membership in the C-declension (through forms with terminals specific to the c-simplex set, e.g. дьне NPl); the same lexeme has a form, e.g. IPl with the terminal ы (дьны). Thus, the canonical IPl terminal in the c-simplex set is ы. The fact that this terminal coincides with one that is established for the 2-base set should not put in doubt the claim to its membership in c-simplex, because terminal sets often intersect. Thus, for the canonical c-simplex set, historical grammar sets up the following terminals, among others: APlmn ы, GPl ъ, IPlm ы. Accordingly, forms like GPl дѣлателъ, IPl ногъты are treated diachronically as containing C-declension terminals. On the other hand, in the present synchronic grammar, such forms are treated as containing 2-base terminals.

### §392. Deformed terminals

A special type of aberration involves the replacement of a canonical terminal with a *deformed* one. In this case, the aberrant form does not depart from the canonically assigned terminal set: that set corresponds to the canonical requirement, but the canonical terminal is replaced with a deformed one. The majority of such aberrant forms are found among adjectives. Cf. ѣѣмь, ѣмь for canonical ѣѥмь ⟨LSgmnPlen⟩, ууму, уму for canonical уѥму ⟨DSgmnPlen⟩, etc.

### §393. The order of examination of aberrant forms

Below, aberrant forms are discussed by declension type and inflectional classes, ordered by the number of forms and lexemes found in sources, from most to least numerous. The order is shown in Table 393 (p. 216). In each of the groups 1–7, we consider not only aberrations of the types shown in the table, but also aberrations of other types that are observed in the corresponding lexeme class that are found in a small number of aberrant forms.

On aberrations in unique lexemes, see Ch. 12, *Unique nominal lexemes*.


Table 393. Order of examination of aberrant nominal forms

# Aberrations in long adjectives: deformation of 2-combi set inflections

### ട്ട 394. General

The observed deformations of terminals in the 2-combi set have segmental underpinnings: elimination of intervocalic j, assimilation of vowels, and contraction. Some aberrant forms show each of these segmental operations in isolation. However, not all of the observed aberrant forms can be represented as a simple composition of certain segmental aberrations. The present grammar treats all such forms as derived by a paradigmatic aberration, namely terminal deformation.

Tables 394.1 and 394.2 show the most important models of aberrant forms. It does not include forms that are derived from canonical ones simply by removing the intervocalic j, and also forms that are homophonous with canonical forms of short adjectives. Thus, for example, the aberrant spellouts новла and нова are not included in the table.


Table 394.1. Possible deformations in long adjectives (hard subtype)

Table 394.2. Possible deformations in long adjectives (soft subtype)


\* Canonical forms of the нищиѥмь type are not attested.

### §395. Illustrations

Below we use the same fragment (Mt 12, 35) from different sources as an example. Zogr — добры чкъ отъ добрааго съкровища· ꙇзноситъ добраа· ꙇ зълы чкъ отъ зълааго съкровища ꙇзноситъ зълаѣ.

Mar — добры члвкъ отъ добраго съкровища износитъ добраѣ ꙇ зълы члвкъ отъ зълааго съкровища износитъ зълаа.

AS - довръм члкъ W добрааго съкровішта износттъ дократ: а зьлъни чакъ WT'S (3)ЛААГО СЪКООВІША ИЗНОСІТЪ ЗЬЛАА.

SAV - ENALBI YAK'S' OT'S ENALA C'BCALA H3HOCHT'S ENAFAR' H JAKKAB'S YAK'T ОТЪ АЖКАВА СЪСЖДА ИЗНОСИТЪ АЖКАВАГА.

The hard subtype LSgmnPlen forms, aside from canonical ones in -- ъкмь and contracted ones in - tmb, also show assimilated forms in - trkMb/- blank (as процеско SUPR 371, 18) only in SUPR and in - камь (ая въчьювам AS Jn б, 27) only in SUPR and AS.

тоу данинать въ пренсподниимъ гробъ адовъ: тоу неремна гако въ гамъ ТИНЬНЪИ ВЪ АДЬСТЪАЛЬ ГООБЪ: И СЪМОЬТЪНЪ ЕМЪ ИСТЪЛЪНИИ SUPR 460, 29-461, 2; H O Chrittin In Hallero Toosbarbians: no Conachish MALTS SUPR 447, 30-448, 1.

НИКОТОЖЕ БО РАБЪ НЕ МОЖЕТЪ ДЪВЪМА ГДІНОМА РАБОТАТІ' ЛЮБО ЕДІНОГО ВЪЗНЕНАВІДІТ' А ДОУГААГО ВЪЗЛЮБІТЪ' ЛИ ЕДІНОГО ДОВЖІТ СА' А О ДРОУБЪВАНЪ НЕ врещи вычьнетъ Mt 6, 24 as.

# Aberrations in 2/a\* adjectives (шт- and ш-Part) and 2/a\*\* (Compar)

§ 396. Aberrant form types in classes 2/a\* and 2/a\*\*

Table 396 shows the grouping of the aberrations observed here.


Table 396. Aberrations in classes 2/a\* and 2/a\*\*

Illustrations

1) Expanded in place of syncopated: вьсткя розгля о мынь не твориящиям ПЛОДА ИЗЪМЕТЪ И ВЪСЪКЖ ТВООАШИХАЯ ПЛОДЪ ОТРЪБІТЪ НА ДА ПЛОДЪ БОЛЬШІ СЬТВОРІТ Jn 15, 2 AS; cf. БОЛИН MAR, БОЛЫ SAV. Likewise: ЧЛКЪ ЄТЕРЪ имъ дъва снъг и рече ментьши снъ оцю Lk 15, 11-12 as, cf. мьнии снъ ею ZOGR, SAV; юнѣн єю MAR. Cf. also: 3 вло лють въсьштъ (NSgmBrev) са ЮНОША ПОНВЕДЕНЪ БЪСТЪ КЪ БЛАЖЕНОУОУМОУ АНТОНИЮ SUPR 171, 16-18; ВЬ ЗЕМЫЖ ЖЕ ВЪДОЖЖЕНОЕ КОПИЕ ВЪКОДЕНИ СА ОБЛИЧЕНЫЕ СЪР СЪГОВШЕНИЮ | ... ВЕЛИЧИЕ ЧОУДЕСЪ: АЖЕ СТЪИ ТВОРГАШЕ ВЪНИЯ НАЗНАМЕНОУЊЕШТЕ (NSgnBrev) до въка supr 560, 23-27.


Likewise in participles: н вси люднє радовах са ємог: о всьхъ славьнуъ E'SIBAIR LIE OT'S HETO Lk 13, 17 SAV, Cf. E'SIBARRIETHIN'S MAR, AS and БЪЛВАНЖШТИХЪ ZOGR; ДА ВЪСХВАЛЬАТО-1 НЕБЕСА Г ЗЕМЛЪ: МОРЕ ГВЪСВ ДВІЖНЫШТЪВ (NPInPlen) съд вь немь PS SIN 68, 35; помъщшления пръезыватиить (NPInPlen) НАМЪ ВЪ ДОЪЗОСТИ EUCH 24а, 11-12.

Likewise in comparatives: 1 бъвањетъ последьнеа чког томоч горьши пръвъюхъ Lk 11, 26 zoGR, ct. горьша MAR; и знамения же огню гаже видъ" И ЇНА БОЛЬШИ СНУВ СЪТВОЯЖ SUPR 23, 11-13; І ГЛАВЖ НА ВЪЩЪННИМЬ МЪСТЪ положи и на неи множеншить чювьствия оствовжаю: не застяпањящи ДРОУГЪ ДООУТА" ВЛАСЪ! ЖЕ ГЛАВЖ ПОКОЪЛЪ ЕСИ" ДА НЕ ВОЕДИТЪ СА ИЗМЕНЕННЕМЬ вътръ висн 7b, 6-12.

5) New comparative in place of the expected old comparative: пакъи жє помъщитьше към ннъ мякъ лютъмша изъоковсти: да въ тъкъ кго погоувитъ SUPR 50, 21-24; сf. егда же нечистър духъ изидетъ отъ чка […] ТЪГДА ДЕЧЕТЪ: ВЪЗВРАШТЪ СА ВЪ ХОАМЪ МОН: ОТЪНОУДОУЖЕ 13ИДЪ: […] ТЪГДА ГДЕТЪ І ПОИМЕТЪ СЪ СОБОЕЯ "Ж" ИЧЪХЪ ДХъ Люштьшь севе в ВЪШЕДЪШЕ живять Mt 12, 43-45 zogr; w великодоушыным фуме; тако на сядишти ПАКОСТЪНИКА НЕ ПОДВИЖЕ СА ТВОИ СЪМЪСЛЪ' НЪ ПАЧЕ КРЪПЪН БЪСТЪ SUPR 63, 29-64, 2, cf. грьдъ же по мы креплен мене естъ Mt 3, 11 As.

# Aberrations in substantives of classes 2/m\* (дълатель) and 2/m\* (гражданинъ)

### §397. General

Aberrant forms only in the plural. The variety of possibilities are shown in Table 397.


Table 397. Aberration types in classes 2/m\* and 2/m\*\*

### § 398. Illustrations


СЪЗЪДА СТАЪПЪ И ВЪДАСТЪ И ТАЖАТЕЛЕМЪ С ОТИДЕ И ПОСЪЛА КЪ ТАЖАТЕЛЕМЪ рабъ въ врема: да отъ тажателъ приметъ: отъ плодъ винограда Mk 12, 1-2 ZOGR, MAR.


ЧАКЪ ЕДИНЪ БЪ БОГАТЪ: ИЖЕ НАСАДИ ВИНОГРАДЪ И ИСКОПА ВЪ НЕМЬ ТОЧИЛО: Н ПОВДАСТ"В I ДЕЛАТЕЛЬМЪ И ОТИДЕ И ЕГДА ЖЕ ПОИДЕ ВОВЕМА ЕМАТИ Е ПОСЪЛА рабъя своя: къ делательмъ имать вина своего Mt 21, 33-34 SAV; архиерен же и старьци съеоръ всь: искахъ лъжа съвъдътелъ на іса: гако да оускнятъ I' и не овречтоша: и многомъ съвъдътельмъ пристяпьшемъ не оврътъ Mt 26, 59-60 SAV 97v.

# Nonstandard terminals in lexemes of the main twofold declension

§ 399. Intrusion of u-simplex or 2-duplex terminals

Possible aberrant forms are shown in Table 399 (p. 222).

<sup>4</sup> However, forms жидовъхъ (LPI) and жидовомъ (DPI) can be treated as aberrant forms with a restructured stem for the lexeme жизъ 2/m. The lexeme жидъ is attested only in Pl, and in NPI (жидовъ, жидови), GPI (жидовъ), and LPI (жидоуъ) only by aberrant forms. Cf. API and IPI жиды, and DPI жидомъ.



Illustrations


ЄН БЛАЖЕНЪІН ЊОМА ИСПЛЬНАИ СЛОВО ДЪЛЪМЪ: ПОВЪСЛЪДОУН ГЛАСА СВОЕГО SUPR 512, 23-24; воле оуво слъзъю оупоуштающи тако надъ морътвъмъ SUPR 457, 15-16; тогда събъютъ са реченое неремнемь преомъ гляшьмъ г(ла)съ оу рама слъншанъ бъютъ: плачъ и ръщание и въпль многъ Mt 2, 17-18 sav.

5) NADu, GLD, DIDu — тъгда пристяпи къ немоу мати сбовос зеведеовоу съ СНЬМА СВОИМА КЛАНТЕЖШТИ СА' І ПООСАШТИ НЕЧЕСО ОТТЪ НЕГО' ОПЪ ЖЕ ОСЧЕ ЕН ЧЪТО ХОШТЕШИ: ГЛА ЕМОУ ОБЦИ ДА САДЕТЕ СЪВ СНЪИ МОВ: ЕДНИВ О ДЕСНЯНЯ ТЕБЕ И ЕДИНЪ О ШЮНЖ ТЕБЕ: ВЪ ЦЕОСТВИИ ТВОЕМЪ Mt 20, 20-21 МАR; БЛАЖЖ їшсифа и никодима: въстє во […] преждє шестокрильнъють когоу слоузъ' не крильма нъ плаштаницетк бога покръввша SUPR 458, 4-8.6

<sup>5</sup> On secondary forms D2 see § 356.

<sup>6</sup> The forms NADu and GLDu are attested only for the lexemes съмъ" and полъ.


On the aberrant forms of the vocative of the type съмму see § 355.

<sup>7</sup> Vaillant for the same verse cites the gloss врачъ from Undolsky's fragments (Vaillant, § 59).

### §400. Expanaded stems with the suffix ¢ as part of substantive 2-base lexemes

Such are hapax glosses лютесе SUPR 341, 1 (ничсоже зъла ни лютєсе) from дюто 2/п, дівеса ps SIN 104, 5 (помъятете чоудеса его ъже сътвори: дивеса его и сядъ [ Бъ] съ съ съ съ съ бъ диво 2/п, личесе SUPR 335, 14, личеса SUPR 397, 18 from лице 2/n. Likewise, the lexemes дѣло and друве show isolated uses with expanded stems in ec in SUPR and EUCH; cf. доввесе SUPR 402, 9; дълесе SUPR 344, 11, дълесъ EUCH 89a, 17. Cf. оучааше и въсен мждрости; и дълесъв н CASE.

# Nonstandard and alien terminals in class 1/m substantives

§401. Intrusion of c-simplex and 2-base terminals

Possible aberrant forms are shown in Table 401.


Table 401. c-simplex and 2-base terminals in aberrant forms

Illustrations


4) страхъ твои затвори въ м'нъ EUCH 78b, 17-21.

Here also note the morphologically strange forms |Sg in SUPR: ዜДИНЪ ዜДНОН ПОНПОДГАЪ ЕСТЪ БОГЪ. А НЕ ЕДНЯ МНОЗЪХЪ' НИ ЕДНОМЬ МНОГЪНА: МЛЬЧИ ОУБО ГОСПОЖДЕ ВЕДЪ ТА КТО ТА ПЛАМЪМЪ РАЖДИЗАЕ НА МА SUPR 366, 19-23; ИЗВЕСТИТИ ЖЕ ЕМОУ ХОТА МАЛЪІ ОУБОГАВЪША СА: СТРАЖА ВИДОМА ПОУСТИ ЛЬВА ВЕЛИКА ЗЪЛО: И СТРАШЬНА ХранАШТА И: НОШТИЖ И ДЪННУК ОТЪ ОУБИВАНИШТИНУЪ ПОГАНЪНУЪ СЪВЕТА SUPR 292, 20-25; ВЛЪКЪ НОШТИНЖ присъдваше од тъла на съелюденые км8. Звъздъ свътьлъ снаажшти на МЕСТЪ ТОМЪ ИДЕЖЕ ЛЕЖААШЕ: ДЬНЫХ ЖЕ ООБЛОУ ТО ЖДЕ СЪКОНЬЧАВАЊЖШТОУ ALAO SUPR 537, 27-538, 1.


# Alien terminals in class 2/a and 2/p adjectives

### §402. Contamination of 2-base, 2-pron, and 2-combi terminal sets

Some adjectival lexemes of the main twofold declension show aberrant forms with 2-pron terminals, while some pronominal declension lexemes show aberrant forms with 2-base terminals. The variety of possible aberrations of this type are shown in Table 402 (p. 226).

For several lexemes with terminals from different sets, the choice of their declension type (2/a or 2/p) is arbitrary; see details in Vaillant, § 99-100.


#### Table 402. Terminal contamination in classes 2/a and 2/p

Cf. for example: члкъ єдннъ съхожда(ш)є отъ нєрслиа въ єрнух: и въ развотникъ! въпаде: иже съвлькъше ! и пъзвъ! Възложъще оставльше и кук живого сяща Lk 10, 30 SAV, cf. оставьше и ель живъ сящтъ ZOGR and octaranьше MEN'S WARA AS.

Among class 2/a lexemes, aberrant forms are most frequent in the lexeme тоуждь, мъногъ, дроугъ. Here are some examples.

TOYЖДЬ - КАКО СЪПОЕМЪ ПЕСНЬ ГНЕХ НА ЗЕМЛІ ТОУЖДЕЇ (for тоужди ог тогждии) PS SIN 136, 4. Also, the initial consonant varies: штоужд- and стоужд-. Cf. tako h a3's lecn's ballero племене' штоуждъ же васъ дъломъ SUPR 28, 2-3; КАКО ВЪСПОЕМЪ ПЪСНЬ ГОСПОДЪЩЖ НА ЗЕМИ СТОУЖДИИ: ЧТО ГЛАГОЛЕШИ ПЪСНИ ЛИ ГОСПОДЬНА НЕ ПОЕШИ: НА ЗЕМИ ШТОУЖДЕН SUPR 418, 29-419, 2. Also: ПО ТОУЖДЕМЬ ЖЕ НЕ ГДЖТЪ" НЪ БЪЖАТЪ ОТЪ НЕГО" "ЕКО НЕ ЗНАНЖТ" ШТЮЖДЕГО ГЛАСА Jn 10, S ZOGR (cf. the same passage in AS: TOY KAARO FAACA). In KIEV 1× with 2 in place of жд: и не отъдазь нашего тоузымъ 4b, 10-11.

МЪНОГЪ - СЪ МНОЗЪМЪ ПЪАНИИМЪ НА НЕЕО ВЪЗИДЕШИ SUPR 384, 3-4 (сf. обн ЖЕ СЛЪЩАВЪШЕ ИДОША СЪ МНОГОМЪ СТРАХОМЪ СЪТВОМТИ ПОВЕЛЪНОЮ SUPR 37, 24-26 and co co плачемъ и многътимъ ръданнемъ глаголааше: како възърж к ТЕБЪ БЖЕ SUPR 528, 1-2); НЕ ОУБОГТЕ СА ОУБО МНОЗЪХЪ ПТИЦЬ ЛОУЧЬШЕ ЄСТЕ ВЪІ Mt 10, 31 ZOGR (cf. i не сътвори тоу силъ многъ: за невърьствие цуъ Мt 13, 58 ZOGR and i 3a 0yWBNOWEHHE EESAKOHITE: iCAKHET'S ?ISBE'SI MBHOUSIN'S Mt 24, 12 MAR); CHNZ ЖЕ И ЇНЪМЪ МНОЗЪМЪ РЕЧЕНОМЪ БЪВЪШЕМЪ ПИОНИЕМЪ: ПОЛЕМОНЪ 16-19 (cf. MBNORONS HWE ПО ИСТИНЪ ВЪДОВАВЪШИЇМЪ ВЪ НЬ' СЬПОЖИВЪ СЬ ЯНИИ SUPR 11, 22-23); с тацели притъчами мъновъми глеше имъ слово: ъкоже можадж CABILLATH Mk 4, 33 MAR (cf. l "Tallemin прит"зчами многами" глааше имъ слово" ЕКОЖЕ МОЖААУЖ СЛЪЩИТИ MK 4, 33 ZOGR); ЕЛАГЪ! РАБЕ L ВЪЗМЕ: ВЪ МАЛЪ БЪКТЪ Въренъ надъ мнозъм та поставля CLOZ За, 19-20 (сf. довръ рабе класы и върьне о маль въ въренъ надъ мъногъ та поставля Mt 25, 23 мая, likewise in ZOGR, AS and SAV); глетъ во аплъч мнозъми скръбьми; подоваетъ намъ вынити въ цоство нескоє EUCH 69b, 11-13.

ДРОУГЪ - ПОИДЖ ЖЕ ВОНИ: И Прьвоумог же пръвнишь гольни: и дрУгомоу pacnaroyms co himb Jn 19, 32 SAV (cf. gpoyromor in the same verse in AS, f. 118b, 8, but дроугоумог in ZOGR, MAR and AS, ff. 106a, 19-20 and 110b, 29). On the distribution of forms see Vaillant, § 100.

Among class 2/p lexemes, aberrant forms by 2-base are found for the lexemes нединъ, инъ, самъ, такъ, вьсъкъ//вьсакъ, толикъ, коликъ, and селикъ. The last lexeme is attested in oblique cases only by aberrant forms, see Vaillant, § 100. Here are some examples.

IEДHN'S - shows aberrant forms Plen with 2-combi terminals. Cf.: како във МОЖЕТЕ ВЪРОВАТИ СЛАВЖ ДООУГЪ ОТЪ ДООУГА ПОНЕМЛЬЖШТЕ С СЛАВЪР ЕЖЕ ОТЪ EAMAATO EA NE HUFTE Jn S, 44 MAR; cf. also in the numeral sense: последь же ВЬЗЛЕЖАШТЕМЪ ИЛЪ ЕДИНОГЕМОУ НА ДЕСАТЕ ВВИ СА МК 16, 14 МАR (but ЕДИНОМОУ на десате in the same verse in AS); въ единяния на десате годиня Mt 20, 9 мая, SAV (but ENINE HA JECATE in the same verse in AS). One separate case is the aberrant form with the 2-base terminal: aштє οΥΕΟ ΤΈΛΟ ΤΈΛΟ ΤΈΛΕΤΈΣΤΙΣΥΣ NE HMBI YACTH EANN'S TOWBH' EX 2ET'S CRET'S CRETT'SAS BECE Lk 11, 36 MAR. See Vaillant, § 102-105.

инъ - shows aberrant forms Plen with 2-combi terminals. Cf.: ини жє отъ столцикъ ту слъпшавъше глахя (in place of ини) Mt 27, 47 SAV.

самъ - shows aberrant forms Plen with 2-combi terminals. Cf.: сьдє жє н CAMORE OVH'SINHE OTTATO E'SICT'T SUPR 492, 30-493, 1.

Takb - shows aberrant forms Plen with 2-combi terminals. Cf.: Ha Takble EQ Hanpacho rapHT's ca (APIm) supr 29, 18-19.

вьсъкъ //вьсакъ - shows aberrant forms Plen with 2-combi terminals. С .: пако же дастъ емоу власть всаць плъти (LDSgf) In 17, 2 sav 25v; in the same verse on folio 107 in SAV there is всаком; ст. въстком in the same passage in ZOGR, МАК.

толикъ, селикъ, коликъ - show aberrant forms Plen with 2-base and 2-combi terminals, of the type толика (GSgmnBrev) and толикаюто (GSgmnPlen). С.: алип" ГЛЖ ВАМЪ: НИ ВЪ ИЗЛИ ТОЛИКЪИ ВЪОТ НЕ ОЕОТЕТОХЪ Lk 7, 9 SAV (сf. ТОЛИКОВА ВЪЗБИ in the same passage in ZOGR and MAR); NH EA HA 04WB BL36MB Bb YACT TT+ HH ПОГОУБЫЕНЬГА ТОЛИКААГО ВЪЗДОВЖАНЬГА И ТРОУДА: ПОЛУБСЛИВЪ ИЛИ ВЪМЪНИВЪ SUPR 525, 27-29; દર્શાલય also belongs here. Only direct cases wordforms are attested, see details in Vaillant, § 100.

# Other aberrations in substantives

### ട്ട 403. An overview of other cases

Isolated aberrant forms, attested in individual lexemes, are discussed below in the order shown in Table 403.


1. A small number of aberrant forms breaks the twofold rule. Here are some glosses: не оставых вога сътворьшааго нево и земью: и поклония са коумиремъ SUPR 7, 8-9 (but коумиролъ SUPR 28, 11); и тръжьникомъ расъща пънаяви и ДЪСКЪГ ОПОВВОВЖЕ Jn 2, 15 MAR (but ПЪНАЗА Jn 2, 15 ZOGR); ТАКОЖДЕ ЖЕ И архнерен ржганжштє са: съ кънижьникъ и старьцъ Mt 27, 41 маr (but старци Mt 27, 41 zoGR). On lexemes ending in - TEAL and - aps (subtype 2/m\*) see § 397.

The twofold rule is violated with respect to o/e-initial terminals in the declension of borrowed substantives with stems ending in vowels or j. Such are, for example, in MAR: anbaotoon In 12, 22; apxrepeows Mt 20, 18; Mt 26, 14; Mt 27, 3; Mt 28, 11, etc.; apxffepeoBh Mt 26, 57; Mk 1, 44. Likewise we find фарисъемъ Lk 14, 3 ZOGR, фариссомъ in sav and AS, but фарисъемъ Lk 14, 3 in MAR; мосъевн EUCH 44b, 7 and MockoBh In 9, 29 MAR. Likewise MOCEOBH Mt 17, 4 MAR, etc., мосьомь Jn 1, 17 zogr, but мосеемъ Mk 9, 4 маг. Cf. in supr нюдешиъ 448, 27 and 469, 5, but василеемъ 538, 7-8.

2. Aberrant forms of the lexeme 38506 1/m. Here are some glosses: n ВЬЗЕМЪ ДЪТНШТА ОТЪ ЗВЪРЪ ДАСТЪ МАТЕОН' И МОЛИТВЯ СЪТВООНВЪ ОТЪПОУСТИ ВЬСЪХЪ СЪ ВЛЪКОМЪ SUPR 44, 28-45, 1 (likewise in SUPR 49, 22 and 509, 28); NPI "ко мог сятъ въс звърг джеравъни скот въ горахъ волови PS SIN 49, 10. С.f. the form with 1-simplex terminal: предлятъ выс звърье лжжыни PS SIN 103, 20.

Aberrant forms of 2/m lexemes (intrusion of 1-simplex and 2-duplex terminals). Here are some aberrant glosses:

ПРІ — разгитьваша же са тъмничьнии стражию: и разгитвавъще са ВЬВЕДОША ИХЪ ПАЧЕ ВЫНЖТОВ: ГАКОЖЕ НЕ ИМЪТИ ИМЪ НИКОКАЖЕ МИЛОСТИ SUPR 134, 5; и шъдъше стражию тъмничьнии полша стага и ведоша къ кназоу SUPR 184, 26.

GPI - H MHORO ПОСТРАДАВЪШІ ОТЪ МНОГЪ ВРАЧЕН MK 5, 26 AS; и много пострадавъши: отъ многъ врачевъ Mk 5, 26 zogr.

IPI - H OYTBOLAHIUA ПЕЧАТЛЪВЪШЕ ГРОБЪ СЪ СТРАЖЬМИ SUPR 440, 25-26.

The lexeme orish 2/m shows many aberrant forms of this type, often violating the expected distribution of the 2-base form with kamora vs. 1-simplex form without kamora. Cf., for example: feoneckaaro orfin supr 453, 3; слыньцоу и огну и вод supr 263, 12-13.

3. Aberrant forms of the lexeme кръвь 1/f. Here are some aberrant glosses: и жена едина сжции въ точении крьве: лъттъ ві· […] слъщавъшин іса: пришъдъши съзади прикосня са ризъ его глаше во тако аще прикоснях са ризъ его спена EXAR IN AEHE HCAKHZ HCTO4LHIK'S KOLBH EA Mk S, 25-29 SAV; in ZOGR: L ЖENA сжити въ точении кръви: отъ дъвою на десяте летор. [ ... си пристяпьши съ слъда: косня са въскрилии ризъе его" с авъе ста: теченье кръве ема Lk 8, 43-44 ZOGR; cf. also G or LSg: приобьщилъ єси плъти и кръвє єстъства твоєго EUCH 78b, 7-9.

4. ISg(f) terminal deformation. Here are some glosses: NE chem's com upb МНОГЖ СИЛЖ' И СПОЛИНЪ НЕ СПЕТЪ СЊА МНОЖЬСТВОМЬ КОВПОСТИ СВОЕЊА PS SIN 32, 16; сжитемъ стъюмъ въ темьници: приде доузъ отъ кесария: и пришедъ въ севастии ї въ и день съдъ съ воєводж повель привести стривести страла SUPR 72, 16-20; ПРЕДАНИ ЖЕ БЯДЕТЕ ООДИТЕЛЪ И БОАТРИНЖ: И РОДОМЪ И ДОУГЪВ LK 21, 16 МАR.

Cf. also forms with contaminated terminals: правъджия твоєм изваві мы PS SIN 30, 2; рякки же пльть дръжаахъ а доуших ба поразочићахъ SUPR 511, 20-21.

<sup>9</sup> Note that from the historical point of view, the lexeme orfis shifted its declesion type, changing from the monovariate to the twofold declesion. Subsequently, other 1/m lexemes also left that monovariate declension.

The Verb

# CHAPTER 14 **The free paradigm of the verb**

§404. The free paradigm

Here is the free paradigm of the OCS verb.








First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

# §405. Subparadigms: representations and systems

The verbal paradigm consists of 12 *subparadigms*, each of which belongs either to the *finite* or the *nominal representation*. Independently of the representation, all subparadigms are divided into three *systems*: the PRAE system, the IMF system, and the INF-AOR system. As a result, we have the following classification of subparadigms (Table 405).


Table 405. Subparadigms classes

The partitioning of subparadigms into two representations has both a syntactic and a morphological basis: different representations are formed by different sets of grammatical categories and have different morphological composition (see Ch. 15 below). The partitioning into systems also has a morphological basis. Namely, for every verbal lexemes, the wordforms of one system use the same workstem, while wordforms of different systems may differ in their workstems.1

# **Finite and nominal representations**

# §406. Finite representations

Forms in finite subparadigms (also called *finite*, or *personal* verb forms) show an opposition by *person* (with three category values) and *number* (with three category values). They are represented by four *bundles*: Prae, Imv, Imf, and Aor. The Imv bundle is defective, in that it lacks 1sg and 3pl forms. Other bundles are opposed to each other in tense/aspect: the Prae bundle carries the non-past meaning, and the Imf and Aor bundles carry the past meaning. The grammatical descriptions of finite wordforms contain three components: person, number, and an indicator of the bundle, cf. 3SgPrae, 2–3SgAor, etc. In all bundles except Prae, the cells ⟨2Sg⟩ and ⟨3Sg⟩ are represented by the same wordform; ⟨3Du⟩ and ⟨2Pl⟩ are represented by the same wordform in all four bundles.

# §407. Nominal representations

These are six *participles* and two *absolutives*, Inf and Sup. Participles are deverbal adjectives; absolutives are extraparadigmatic substantive forms, Inf (infinitive)

<sup>1</sup> Occasional departures from this principle are described as the alien stem expansion effect (notated ⤸ or ⤹ ); see more details in § 440.

with the substantive inflection и (from the 1-simplex set), and Sup (supine) with the substantive inflection ъ/ь (from the 2-base set).

The six participles in this book are named by their suffix: м-Part (м-participles), щ-Part (щ-participles), etc.2 Participle subparadigms have the free paradigm of the ordinary adjective. Their paradigmatic names (calls and addresses) contain (1) an adjectival grammatical description, (2) an indicator of one of the six participle, and (3) the name of the parent verbal lexeme to which they belong. For example, we have ASgfBrev [ш-Part (трьпѣти)] for the form трьпѣвъшѫ, ASgfBrev [ш-Part (л҄юбити)] for л҄юбл҄ьшѫ, ASgfBrev [м-Part (трьпѣти)] for the form трьпимѫ, etc.

Grammatical descriptions of absolutives contain only an indicator of one of the two absolutives (Inf or Sup), e.g. Sup(рещи) for the form рещь.

# **Secondary verb forms**

### §408. An overview of secondary forms

Some verbal lexemes contain so-called *secondary* forms. These are the following: (1) 1SgPrae (вѣдѣти): ∇вѣдѣ; see § 528; primary form вѣмь; (2) conditional (Cond), which are peculiar personal forms of the verb быти (e.g. 1Sg бимь); see § 546, and some other forms of this unique verb; see § 544; (3) nonstandard aorist forms (e.g. 1Sg идъ); see § 476–482.

<sup>2</sup> щ- and м-Part are the so-called active and passive present participles; ш- and н-Part are active and passive past participles; л-Part is the perfect participle, and т-Part is the infinitival participle, syntactically equivalent to н-Part.

# CHAPTER 15 **Formation of verb forms**

# §409. Morphological composition of verb forms

Finite and nominal verb forms have different morphological composition (see Table 409). The morphological skeleton of finite forms contains two components: [verbal workstem + finite inflection]. The morphological skeleton of nominal forms has three components: [[verbal workstem + subparadigm suffix] + nominal inflection]. The first two comprise the workstem of the nominal form, which, in turn, contains two components: the workstem of the parent verb is the *basic component* in relation to the suffixes of nominal subparadigms, i.e. the participial suffixes, and the infinitival and supine suffixes.


Table 409. Morphological composition of verb forms

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

### §410. Nominal forms in the paradigmatic synthesis of verbal paradigms

From the point of view of paradigmatic synthesis rules, the full morphological skeleton of finite forms—expressions of the form [verb workstem + finite inflection]—are equivalent to the workstem of nominal forms—expressions of the form [verb workstem + subparadigm suffix]. Within verbal synthesis rules, nominal forms are only built up to their workstems. Accordingly, nominal form suffixes are in a way equivalent to terminals of finite forms; selection and attachment of nominal terminals are outside of the responsibility of verbal synthesis: the construction of a nominal paradigm from a starting form and a nominal paradigmatic index is accomplished by nominal synthesis rules. Verbal synthesis is only responsible for building workstems.1

Such a distribution of responsibility between verbal and nominal paradigmatic synthesis is dictated by the morphology of the corresponding forms. The reason is that verbal workstems are the same for finite and nominal forms within a system, and are distinguished only by the system to which the synthesized form belongs and by its lexeme. The rules of building workstems are the same, as are boundary adjustment rules. The boundary that is adjusted by these general rules is between the verbal workstem and the terminal for finite forms, and between the basic component and suffix of the nominal subparadigm for nominal forms.

In other words, the synthesis of nominal subparadigms in the verbal paradigmatic synthesis algorithm stops midway, halting with the workstem synthesis for the starting form of the corresponding nominal paradigm.2

### §411. Verb form construction procedure

To build a called verb form, one must perform three steps: (1) build the workstem; see § 413 below; (2) find the terminal (for finite forms) or the suffix (for nominal forms); see § 414; and (3) apply boundary adjustment rules; see § 415.

Each of these steps is regulated by its own block of rules. As a result of the first two steps, we have a morphological skeleton of the called wordform, i.e. an expression of the form [workstem V + terminal] for finite forms, and the morphological skeleton of the nominal stem of the called wordform, i.e. an expression of the form [basic component + suffix] for nominal forms. In the latter case, the basic component is the workstem of the parent verb of the relevant system.


As a result of the last step we get an inflectional spellout of the called wordform for finite forms, and a morphophonological representation of the stem of the called wordform for nominal forms.

To derive the final shape of the called wordform, the standard mph⇒ph/norm rules must be applied. Note that boundaries between the workstem and terminal in the case of finite forms are adjusted twice: first by the boundary adjustment rules, then by mph⇒ph/norm rules. Thus, e.g., from рек+ѣахъ we have реч.ѣахъ by the boundary adjustment rules, and реч.ѣахъ⇒речаахъ by mph⇒ph/norm rules.

For unique verbs, forms are not built by rules. Their paradigms are given either wholesale or by individual profiles (see Ch. 21, *Unique verbs*). Morphologically anomalous forms are also not built by rules (see a list of morphologically anomalous forms in Ch. 19, *An overview of verb forms by system*).

The following input information is needed to perform the steps listed above. The primary information is (1) the grammatical description of the called wordform; (2) information on the membership of the synthesized wordform in a subparadigm in one of three systems, PRAE, IMF, INF-AOR (see above § 405); (3) the graphic spellout of the starting wordform of the parent lexeme of the called wordform. These data are contained in the paradigmatic call itself. Secondary information includes: (4) the morphophonological representation of the starting wordform; (5) a paradigmatic index of the parent lexeme of the called wordform. These data are extracted from the PD.

After gathering all of the necessary information, one can synthesize the called wordform (if it is finite), and the nominal stem of the called wordform (if it is nominal).3

### §412. Starting forms of verbal lexemes

The starting form of a verbal lexeme is the infinitive. Infinitives are shown in the paradigmatic dictionary, in both their graphic (normalized) spellout, as well as in the morphophonological spellout. So, for the verbs възл҄юбити, познати, въЖелѣти we find въз.л҄уб.и.т.и 1, по.зна.т.и 4v, въз.жел.ѣ.т.и 7, etc. The only defective lexeme that is not represented by the infinitive in the dictionary is the lexeme of the unique verb ѥсмь, where the 1SgPrae form—the first form in the paradigm—is chosen as the starting form that is shown in the dictionary.

### §413. The first step: building verbstems

In the general case, a verbal lexeme corresponds to several workstems, in such a way that different subparadigms use different workstems. Still, forms of a single subparadigm, and in most cases forms of a single system (PRAE, IMF, or INF-AOR) have the same workstem. Workstems are built in two steps: on the first step so-called *basic stems* are determined, and on the second step, workstems

3 To build the called wordform of participles one must go to the nominal synthesis rules.

are constructed from these basic stems. Thus, every workstem has as its source and antecedent some basic stem. Each verbal lexeme has either two or one basic stem. If there are two, they are distinguished as *expanded* and *truncated*. The assortment of basic stems of a given lexeme (one or two), and the shape of a basic stem, is determined by the segmental content of the inifinitive and the verb's paradigmatic class.

When the assortment of basic stems is determined, one must select the basic stem that serves as the source of the workstem of a given verb form. This depends, on the one hand, from the paradigmatic class of the verb, and on the other, on the system to which the given verb form belongs (see below, rules of assigment of basic stems, § 427). The transition from a basic stem to the sought workstem is accomplished using rules that can change the segmental content of the basic stem of the source. These rules also take into account, on the one hand, the verb's paradigmatic class, and on the other, the system of the given wordform. Thus, a PRAE workstem, a IMF workstem, and an INF-AOR workstem is defined for every verbal lexeme, although for most verbs the IMF workstem coincides with either the PRAE or the INF-AOR workstem.

### §414. The second step: the selection of the appropriate terminal or suffix

In the general catalog of suffixes and terminals (see § 455), all sets are distinguished as standard or nonstandard. The latter are used only among unique verbs, secondary forms, and aberrant forms. The sets are distributed among subparadigms. In the PRAE system, every subparadigm corresponds to two sets, where the choice is determined by the verb's paradigmatic class.

When the terminal set is determined, one must select the terminal from that set whose address corresponds to the paradigmatic call. The appropriate terminal is found if the cell contains only one terminal. However, some cells can contain several terminals or suffixes, where the choice is determined by the morphophonological features of the workstem. Specific rules governing that choice are shown for each such terminal set in the notes to the terminal catalog (§456–460).

### §415. The third step: boundary adjustment

After the morphological skeleton is constructed, i.e. the workstem is built and the appropriate terminal of suffix is selected, one must apply the boundary adjustment rules (see § 461–462). These rules use both morphophonological information (cf. рек +еши⇒реч=еши, but нес +еши⇒нес=еши), and properly grammatical information, such as information on the grammatical description of a form (cf. мог+ѣте⇒моѕ=ѣте (Imv), but мог+ѣаше⇒мож=ѣаше (Imf), and information on its paradigmatic class; cf. л҄юб+ѫ⇒л҄юбл҄=ѫ (class 1), but греб+ѫ⇒греб=ѫ (class 4).

# CHAPTER 16 **Paradigmatic classes of the verb**

# **Main paradigmatic classes**

# §416. Paradigmatic classes of the verb: acquaintance

The partition of the set of verbs into paradigmatic classes must be related to the similarities and differences in paradigmatic behavior in such a way that information on the membership of a verb in a paradigmatic class should ensure the correct application of the paradigmatic synthesis algorithm. Roughly speaking, two lexemes belong to the same paradigmatic class if they are conjugated in the same way, or, in other words, if they follow the same paradigmatic standard. Membership in a paradigmatic class is shown in the paradigmatic dictionary. This chapter shows the partition into classes used here and discusses its morphological underpinnings: the relationship between morphological composition of the infinitive (the left subtable) and the basic stems of the verb. The distribution of basic stems by system (the right subtable) determines the most important traits of the paradigmatic standard in a given class. The main paradigmatic classes and their type representatives are shown in Table 416 on p. 242.

### §417. unique verbs

The classification that determines the paradigmatic classes does not extend to unique verbs. They have individual paradigms that are neither similar to each other nor representable as a modification of some paradigmatic standard. These 19 verbs comprise a degenerate paradigmatic class: their paradigms are not built

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

# by rules, but are given wholesale (see Ch. 21, Unique verbs). The verbs датн, пасти, ксмь, въдъти, and имети are called athematic.


# Table 416. Main paradigmatic classes of verbs

### §418. Main paradigmatic classes

The seven classes shown in Table 416 are called main paradigmatic classes, or simply verb classes. The partition into these classes does not take into account particularities that have to do with the presence of secondary or anomalous forms in the paradigm, as well as the presence of marginal subclasses, or of so-called irregular verbs. Membership of a verb in one of the seven main classes is shown in the paradigmatic index with an Arabic numeral at the beginning of the index.

### §419. Splintering of the classification and irregular verbs

The partition is further somewhat splintered. First, the main class 4 is divided into three classes by the end of the truncated stem: C-final, V-final, or finally ambivalent. Accordingly, we have subclasses with the following type representatives: нести 4c, знати 4v, and клѧти 4h.

Second, within the main classes 3 and 4, there are subclasses of so-called *irregular* verbs, whose paradigms can be represented as certain deformations of the main paradigmatic standard of the corresponding class, which are caused by the application of paradigmatic effects (see a list of irregular verbs in § 434). *Regular* verbs are all verbs of classes 1, 2, 5, 6, and 7, and also (1) all class 3 verbs except irregulars, (2) all 4v verbs, and (3) all 4c verbs except irregulars. All 4h verbs are irregular. Roughly speaking, to build the paradigm of a regular verbs it is sufficient to know the segmental content of the infinitive and the main class to which the verb belongs, while to build the paradigm of irregular verbs, this information is insufficient.

All verbs in subclass клѧти 4h contain a labile root. Classifying all the verbs in the клѧти 4h class as irregular is necessary because the C-final version of the root cannot be predicted from the V-final version contained in the infinitive.

### §420. Paradigmatic indices

In the general case, the paradigmatic index contains the number of the main class (Arabic numeral between 0 and 7), and an extension. Irregular verbs contain indices with extensions that deform the main paradigmatic standard. For example, we have the following regular verbs in class 3: плакати 3, привѧзати 3, пострадати 3, and the following irregular verbs: бьрати 3°\*, полиꙗти 3\*, смиꙗти 3\*, сътъкати 3°. In class 4, we have regular verbs нести 4c, знати 4v, възвести 4c, and irregular verbs врѣщи 4c\*⤹, вълити 4h\*⤸, жрьти 4h•\*, начѧти 4h.

The paradigmatic index may contain the warning symbol ∇, which indicates the presence of morphologically anomalous forms. Cf. видѣти 2∇ (the form 2SgImv ∇виЖь), пещи 4c∇ (imperative ∇пьцѣте and others forms). However, this symbol is not used for unique verbs and for verbs where morphologically anomalous forms are part of the paradigmatic standard of their class (such is the class двигнѫти 5 with morphologically anomalous н-participles like двигновенъ).

### §421. The principle of paradigmatic equivalence of members of a family

Verbs concatenated with a prefix or the particle сѧ (less commonly си, cf. мьнѣти сѧ, пожалити си), and corresponding simple verbs, form a single *family*. Members of this family have equivalent paradigmatic behavior, and always belong to the same paradigmatic class. Accordingly, in paradigmatics the term "verb" often refers to an entire family. Quantitative data should be interpreted in the same way. For example, мьнѣти, мьнѣти сѧ, помьнѣти, помьнѣти сѧ, сѫмьнѣти сѧ, усѫмьнѣти сѧ, and съсѫмьнѣти сѧ, are counted as a single item. Both in the dic-

tionary and in the text of this book verbal lexemes with ca are shown without ca. Thus, members of pairs like мынути, мынути са are not distinguished. On the other hand, prefixed and prefixless verbs are shown in the dictionary separately.

# Classification of the infinitives

### § 422. Morphological composition of the infinitive

The infinitive has the following morphological composition:

$$
\begin{bmatrix}
\text{lexical} & \text{suffix + } \text{suffix + } \text{false} \\
\text{component} & \text{verb class}
\end{bmatrix} + \begin{bmatrix}
\text{infinityve} & \text{inflection} \\
\text{suffix + } \text{ufflex } \text{r}
\end{bmatrix}
$$

Here, the bracketed components are the basic component of the infinitive and its marker, respectively. The basic component, in turn, contains three morphological constituents, where the lexical component is present in every infinitive, while the suffixes of the verb class and themes may be absent. The lexical component may contain only the root, cf. âюE in âювити, a prefix with a root, cf. B23.(AHB) in BB30kSHTH, or can contain a nominal suffix with a root, cf. (E'S').bh in E'EchnoraTH [(EEC).bH.0B.a.m.h], or both prefixes and suffixes, cf. ИЗ.(БЪ).Т.ЪЧ.ЬСТВ İN ИЗБЪПТЪЧЬСТВОВАТИ [ИЗ.(БЪЛ).Т.ЪЧ.БСТВ.ОВ.А.Т.И].

### § 423. Morphological types of the infinitive

All possible combinations of verb class suffix with themes and the corresponding paradigmatic classes are shown in Table 423. Approximate count is shown for each class.


Table 423. Infinitive types

\* Strictly speaking, the n in въпити, and the t and a in the expanded stem in class 7 verbs are not themes, but verb classes suffixes. Accordingly, only verbs (and infinitives) in classes 1, 2, 3, 5, and 6 are thematic.

Clearly, the inifinitive shape by itself does not give full information on the paradigmatic class, although in classes 4, 5, and 6, the infinitive uniquely determines the class.

### §424. A note on class 4

In this class, the lexical component does not contain suffixes—there are neither verbal suffixes nor a theme. In other words, the infinitive marker т.и is attached directly to the root. The only formal exception is the verb (звѧ)щи [звѧ.г.т.и] ‹310› 'report, announce' for the hapax gloss звꙙгома (Supr 475, 11). The verb звѧщи, звѧжеши is found in Sreznevskij; for later forms, звѧг should be seen as a unitary root formative.

### §425. A note on class 7

Although in this class no subparadigm uses it, the truncated stem is present within the corresponding expanded stems. The expanders а and ѣ have a double function: these are ordinary themes from the point of view of the infinitive and expanded stems, and they are verbal suffixes from a paradigmatic point of view, since they are never truncated.

# **Basic verb stems**

### §426. Basic verb stems

Every verbal lexeme has a fixed *assortment* of basic stems, a full one (two basic stems), or a partial one (only one basic stem). There are two types of basic stems, *expanded* and *truncated*. A full assortment of basic stems contains an expanded and a truncated stem; a partial assortment contains either an expanded or a truncated one. Differences in the basic stem assortment and in their distribution across the paradigm determine the crucial equivalences and distinctions in the paradigmatic behavior of verbal lexemes, and, accordingly, the oppositions between paradigmatic classes, as shown in Table 416, *Main paradigmatic classes of verbs*.

The expanded basic stem differs from a truncated stem by the formatives it contains: an expanded stem can be represented as [truncated stem + theme]. Because the starting form of the verb (the infinitive) uses the expanded stem in all verbs that have it, the morphological composition of the infinitive correlates with the partition into paradigmatic classes, as shown in Table 416. The basic stem of a specific verb is defined to be the segmental string that instantiates the basic stem in the infinitive. So, e.g. бьр.а is the expanded basic stem of the verb бьрати; съ.пьс.а is the expanded basic stem of the verb съпьсати; бьр is the truncated basic stem of бьрати (cf. Prae берѫ, береши); въз.лег is the truncated basic stem of the verb възлещи (cf. Prae лѧгѫ, лѧжеши).

The segmental composition of basic stems of all regular and most irregular verbs is uniquely determined by the infinitive, as long as its morphophonological representation is given. In the graphic representation, the segmental composition of the truncated stem can be obscured in the subclass нести 4c (cf. пасти1 [пад.т.и] and пасти2 [пас.т.и]). In the subclass клѧти 4h the basic stem is truncated and finally ambivalent (there is no expanded stem); the infinitive contains its V-final version, while its C-final versions are given by lists (see § 434).

# §427. Distribution of basic stems by systems

The distribution of the basic stems by subparadigms is given by Table 416, *Main paradigmatic classes of verbs*. Clearly, the distribution is organized in such a way that, in each class, subparadigms of a system are served by a single basic stem, either truncated or expanded. The distribution of basic stems by systems will be referred to as the *basic stem allotment rules*. These rules are shown in Table 427.


Table 427. Basic stem allotment rules

The PRAE system uses the truncated basic stem whenever it is available (i.e. in all classes except 7); the INF-AOR stem uses the expanded basic stem whenever it is available (i.e. in all classes except 4). Meanwhile the IMF system uses the truncated stem in some classes and the expanded stem in others.1 The general basic stem allotment rule subsumes the rule selecting C-final or V-final versions of finally ambivalent truncated stems in the class клѧти 4h (although these verbs are considered irregular).

This selectivity of systems with respect to basic stems has morphophonological underpinnings. The issue is that systems are opposed by their typical terminals and suffixes: in the PRAE system, they are all V-initial, while in the INF-AOR system, they are mostly C-initial, and in IMF system they are mostly initially ambivalent. Meanwhile, expanded stems are V-final, while truncated stems are mostly C-final. (See more details in Ch.24, *Supplement*, §875, *On Jakobson's law*).

<sup>1</sup> Traditional Slavic linguistics uses the opposition between *present ~ infinitive stem*, which does not fully coincide with the expanded ~ truncated opposition which is adopted in this book.

### §428. Basic stems and workstems and subparadigm systems

As has already been indicated, in constructing paradigms, basic stems become workstems, and every workstem corresponds to a particular basic stem which is its source. Expanded basic stems generate *expanded workstems*; truncated basic stems generate *truncated workstems*. Expanded basic stems undergo no segmental changes, i.e. the expanded basic stems and its corresponding workstem are segmentally identical; a truncated basic stem can generate truncated workstems that differ from their source by its segmental composition. However, workstems maintain the crucial opposition: truncated workstems are in the vast majority of cases C-final; expanded workstems are always V-final.

The following proposition holds of the organization of verbal paradigms: whatever the verbal lexemes, all the forms of a subparadigm have the same workstems; for regular verbs, all forms of a system have the same workstems.

# **Profiles of the type representatives for the main paradigmatic classes**

### §429. The paradigmatic standard and the profile

For every paradigmatic class, there is a *paradigmatic standard*: all the verbs of a single paradigmatic class follow the same paradigmatic standard, while verbs of different classes follow different standards. Verbs that follow the same standard select the same alternative in constructing their wordforms in all the forks in the paradigmatic path which are given by the paradigmatic synthesis rules. Similarities and differences of paradigmatic standards are easily observable in *profiles* of the corresponding type representatives.

Table 429 (p. 248) shows a list of profile forms and profiles of type representatives of the main paradigmatic classes. Profiles show the forms in the inflectional spellout.2

For every subparadigm, except the supine and л- and т-Part,3 there is one or two forms in the profile. These key forms are chosen so that they can be used in a maximally simple way to build the rest of the oblique forms.

Participles are represented in the profiles by their starting forms; for м-, н-, and т-Part, the workstem of the starting form (NSgmBrev) is listed in its place (because some verbs may lack the participle itself). The period separates the suffix of the participle. The NSgmBrev forms of щ- and ш-Part are morphologically anomalous, in that they have no distinguishable terminal (symbolized by |).

<sup>2</sup> In other cases verb forms in profiles can be given in the graphic spellout.

<sup>3</sup> These forms are trivially reconstructible from the infinitive. On supine forms see also § 88.

Table 429. Profiles of type representatives of the main verb classes


# §430. From the profile of a verb to its full paradigm

Because inside the given subparadigm workstems are the same, in order to obtain the oblique forms given a certain key form it is sufficient to replace the terminals. Since every lexeme takes the terminals from the same set— and in case a set contains morphophonological variants of terminals, their selection is determined by the workstem—it is not difficult to select the needed terminals.4

<sup>4</sup> It is somewhat harder to build the profile of a verb given the profile of a type representative. So, practically speaking, profiles work as short summaries of paradigmatic synthesis rules, free of the grammatical apparatus. While useful as information carriers for beginners, they cannot reliably replace the full set of rules.

# CHAPTER 17 **Workstems of the verb**

# **Workstems of regular verbs**

### §431. The segmental content of basic stems

To build a workstem of a called verbal wordform, one must perform two steps: first, determine the *allotted basic stem* and its segmental content; second, transition from that allotted basic stem to the *sought workstem*.

The first step is the selection of the allotted basic stem. The search is performed according to the basic stem allotment rules (see Table 427), which indicate which basic stem serves the called wordform—expanded or truncated, and the case of the 4h class, whether the C-final or V-final version of the truncated stem is used in the given system. Once the selection of the truncated or expanded stem is made, one must find its segmental content.

The segmental content of the allotted basic stem for a given verb is determined by its infinitive. The expanded stem can be derived by removing the infinitive marker: въз.л҄юб.и.т.и\т.и: въз.л҄юб.и; двиг.н.ѫ.т.и\т.и: двиг.н.ѫ; мил.ов.а.т.и\т.и: мил.ов.а, etc. To arrive at the truncated basic stem in all classes except class 4, one must remove the theme from the expanded basic stem (i.e. the last vowel): въз.л҄юб.и\и: въз.л҄юб; двиг.н.ѫ\ѫ; двиг.н, etc. To arrive at the truncated basic stem of class 4 verbs, one must remove the marker of the infinitive ти, but in the subclass 4h, if the C-final version of the truncated stem is allotted, one must turn to the lists that indicate its segmental content (see the lists in § 434). Thus, for 4h: клѧ.т.и has a C-final truncated stem кльн; мрѣ.т.и

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

has the C-final truncated stem мьр; кла.т.и has кол. For 4v we have: зна.т.и has зна; по.чи.т.и has по.чи; на.сѣ.т.и has на.сѣ.1

The second step is the transition to the workstem. If the allotted basic stem is expanded, then the corresponding workstem is simply identical with it. This is the case for both regular and irregular verbs. If the truncated stem is allotted, then the transition to the workstem is accomplished through special rules (see § 432), in some cases changing the segmental content of the allotted basic stem.

These rules are separated into two blocks which apply sequentially: the first block ensures the creation of workstems for all regular verbs (§ 432–433); the second block introduces additional adjustments that are necessary for irregular verbs (§ 434–454). In the construction of irregular verb workstems, one first builds a half-baked product—the workstem of the verb as it would be if the verb were regular (by § 432), and then this half-baked product is completed by the rules of § 434 and ff.

### §432. Workstem construction rules for regular verbs

Table 432 on p. 253 shows which paradigmatic classes and which workstems undergo paradigmatic effects that ensure the transition from basic stems to workstems. In each class, the first half-row shows allotted basic stems; the second half-row shows the workstem of a type representative that has undergone the paradigmatic effect determining the segmental content of the corresponding workstem; the third sub-row shows the paradigmatic effects in action.

<sup>1</sup> It is useful to emphasize that, while in other cases for technical simplicity the graphic representation of the infinitive is sufficient, in class 4c the morphophonological representation is absolutely necessary, since the graphics in these verbs conceals the root-final consonant; cf. for пасти1 [пад.т.и] the stem is пад, while for пасти2 [пас.т.и] it is пас.


### Table 432. Workstems and paradigmatic effects

All subclass клати 4h verbs are considered irregular; see § 434 below.

§433. Paradigmatic effects and construction of truncated workstems of the main paradigmatic classes

1. The substitutive softening paradigmatic effect (indicated by ®). This effect calls for the replacement of the final consonant of the truncated basic stem by the pairings of the substitutive softening alternation. It applies in the workstem of the IMF system in the âювити 1 class, and in the workstem of the PRAE system in the плакати 3 class.

2. The *alien stem expansion* paradigmatic effect (indicated by ⤸ or ⤹ ). This effect expands leftward ( ⤸ ) or rightward ( ⤹ ) the scope of the workstems of neighboring subparadigms. The participants of this effect differ from case to case. In class 5, the INF-AOR workstem expands into the ш-Part subparadigm ( ⤸ ) (cf. двигнѫ.въш=и). This effect introduces adjustments into the opposition of workstems that is set up by the basic stem allotment rules (see Table 427), but it applies only after the workstems themselves have been built.2

3. The *labileness in the paradigm* paradigmatic effect (indicated by h). This effect only applies in the verbs whose truncated stem is represented by a labile root, as in the клѧти 4h class, or, as in миловати 6 class, by a finally ambivalent suffix. This effect requires the selection of one of two alternative stems: a C-final one (cf. кльн for клѧти 4h), or a V-final one (cf. клѧ for клѧти 4h). The choice is not only dictated by the CVC agreement rules, but additionally by the paradigmatic effect itself: in the IMF system, the stem is C-final, although some imperfect suffixes are themselves initially ambivalent. Thus, for н-Part, we have a stem that consists of a C-final basic stem (кльн for клѧти), and a V-initial participial suffix (ен); we have кльн.ен= (and likewise for ш-Part кльн.ъш=). This is the case despite the fact that the н-Part suffix is initially ambivalent, н/ен; the ш-Part suffix is also initially ambivalent, (ъш/ьш)/въш. Certain verbs, viz. in all the verbs of миловати 6 class and in some irregular verbs, in addition to the h effect, undergo the ⩨, see below.

4. The *C-final stem arrest* paradigmatic effect (indicated by ⩨). This effect only applies in those verbs whose truncated stem is represented by a labile root, or, as in the миловати 6 class, by a finally ambivalent suffix. This effect requires, contrary to the CVC agreement rule, a V-final version of the truncated stem before the V-initial terminals and PRAE system suffixes. (The hiatus is eliminated by the epenthetic *i̯*). Thus, in class 6, out of the two versions of the truncated stem like мил.ов and мил.у, this effect dictates the choice of the V-final one, мил.у for the PRAE workstem. Cf. also пл҄ьвати 3h\*⩨: the PRAE workstem is пл҄у, not пл҄ьв. (See below, § 465, *A note on the so-called* j*-present*).

### A note on indices

As was noted above, paradigmatic indices of irregular verbs show all the paradigmatic effects that apply in them. However, in the indices of regular verbs, paradigmatic effects that are part of the standard of the corresponding class are not shown.

<sup>2</sup> This effect applies widely among irregular verbs, see Table 440 (pp. 258–259). See that table also for the linear order of subparadigms within a system. Alien stem expansion can only take place into immediately neighboring subparadigms.

# **Workstems of irregular verbs**

# §434. The list of irregular verbs

The full list of irregular verbs is given in the Table 434. The family is represented by a prefixed verb only if the simple verb is absent in the family (otherwise only for ѩти and имати). The index is shown only for the type representative of the group. Along with the starting form of the infinitive, key forms are also shown: 1SgPrae, and in some groups also key forms of other subparadigms. For some forms, their morphophonological spellout is shown.


# Table 434. The list of irregular verbs

### §435. Classification of irregular verbs

Irregular verbs are sorted into groups in such a way that a group contains verbs of a single paradigmatic class that show identical paradigmatic effects in the formation of their workstems, and the same distribution of workstems among subparadigms.

The set of applicable effects is reflected by the index.3

### §436. Assortment and distribution of workstems of irregular verbs

Abnormalities in irregular verbs are limited to truncated workstems. On the one hand, the segmental content of the truncated stems can vary; on the other, different versions of truncated stems can be variously distributed among subparadigms.

### §437. Segmental content of irregular verb workstems

Expanded workstems segmentally coincide with expanded basic stems, just as for regular verbs. Truncated workstems can be created by various paradigmatic effects not envisaged by the paradigmatic standard of the main classes that the irregular verbs belong to. In addition, irregular verbs may undergo special effects that are not found in regular verbs. Those are the following effects.

The *unstable root vocalism* paradigmatic effect (indicated by \*). This effect requires the replacement of the root alloform shown by the infinitive (and, thus, in the basic truncated stem), by an alloform with different vocalism. For example, the PRAE workstem of the verb пьсати 3\* is not пьш, but пиш; the PRAE workstem of чисти 4c\*⤹ is not чит, but чьт.

Particular realizations of the unstable root vocalism effect for every irregular verb are shown in Table 434.

The *absence of expected substitutive softening* paradigmatic effect (indicated by °). This effect cancels the application of substitutive softening that is expected in the paradigmatic standard of class 3. For example, the PRAE workstem of the verb метати 3° is not мещ, but мет.

For irregular class 4h verbs, where all roots are labile, the PRAE workstem requires discovery of an alternative C-final alloform of the labile root whose segmental appearance in the general case cannot be deduced from the shape of the V-final alloform represented in the infinitive. Thus, for клѧти we have клѧ/кльн; for ѩти, ѧ/ьм. The labileness in the paradigmatic effect (h) also applies here.

The effects of unstable root vocalism and labileness in the paradigm can be seen as rules of replacement by fundamental alternations that require replace-

<sup>3</sup> Not all groups have different indices, however. Thus, the groups чисти 4c\*⤹ and влѣщи 4c\*⤹ have the same index, but are distinguished by the distribution of workstems. Likewise the verb пѣти 4h⤸ has the same index as the verbs in the плути 4h⤸ group, and has a different distribution of workstems. These groups have different realizations of the expansion of alien workstem in the paradigmatic effect.

ment of an alloform found in the starting form by an adjacent one in constructing the PRAE workstem. However, because the alternation is not free, it is impossible to predict whether a given formative shows that alternation (i.e. whether the root of the verb is stable or unstable); thus, the relevant verbs must be given by a list. And in this case it is easier to show the adjacent alloform directly instead of computing it by fairly arcane rules of replacement (see Ch. 24, *Supplement*, § 874, *On verbs with unstable root vocalism*).

The application of all these effects (except for the labileness in the paradigmatic effect that is part of the class 4h standard) is limited to the PRAE workstem, whose unity is not broken in any irregular verb.

The PRAE workstem is represented by the 1SgPrae form that is given in the lists of irregular verbs. Below, the workstem of the 1SgPrae form is indicated as «Prae», and the workstem of the Inf form as «Inf».

### §438. Redistribution of workstems

For irregular class 3, the workstem in the PRAE system is «Prae»; the workstem in the IMF and INF-AOR systems is «Inf».

The alien stem expansion effect is widely represented in class 4 irregular verbs. In many verbs, this effect shifts the standard distribution prescribed by the allotment rules (see § 427, 432), which opposes two workstems: the PRAE and IMF workstems to the INF-AOR.

Particular realizations of the alien stem expansion effects for every group of irregular verbs are shown in Table 440 (pp. 258–259).

In all irregular verbs, «Prae» and «Inf» stems are opposed. In labile verbs «Prae» is C-final and «Inf» is V-final. In verbs that undergo •, \*, h, and ⩨ effects, they shape the «Prae» stem, distinguishing it from the «Inf» stem.4 However, in some groups of irregular verbs, the combination of all these effects creates three distinct truncated stems in the paradigm (such are the groups брати 4h•, крыти 4h•\*⩨⤸ and мрѣти 4h\*⤹ ).

### §439. How to select the necessary workstem of irregular verbs

The needed stem can be found using type representatives of the group, whose workstems are shown in Table 440. If the stem is constructed by rules, they must apply in crucial order. First, PRAE and INF-AOR workstems must be built as if the verb had been regular; for class 4 verbs with a labile root, the C-final version is

4 Thus, the alien stem expansion effect either extends the «Prae» stem forward, outside of the PRAE system (up to Sup and Inf, as in the чисти 4c\*⤹ group verbs), or extends the «Inf» workstem backward, outside of the INF-AOR system (up to the Imf subparadigm, as in the case of пѣти 4h⤸). Because workstem expansion can only take place into immediately adjacent subparadigms, comparison of all observed realizations of the alien workstem expansion effect uniquely determines the order of subparadigms in the IMF system, and partly in the INF-AOR system.

determined only by lists. Then the PRAE workstem must be restructured with the adjustments required by segmental paradigmatic effects (all effects except alien stem expansion). Then, the resulting stems must be distributed by regular verb allotment rules (§ 427, 432), and then the adjustments required by alien stem expansion effect must be introduced according to Table 440.

§ 440. Distribution of irregular verb workstems by subparadigms

Table 440 shows the distribution of workstems for every irregular verb group.


Table 440. Irregular verb workstems

5 These lists are not explicitly given. Practically, the C-final version is determined by the «Prae» workstem that is shown in lists. This is the case for all groups except врати 4h ", кръгти 4h\*#2, EHTH 4h\* 2, and Motern 4h\* 5. In these groups, the effects ", \*, and # must be undone in the «Prae» stem.


### Table 440 (continued). Irregular verb workstems

The & effect indicates the absence of the expected \* effect.

# Commentary of particular groups of irregular verbs

### \$ 441. метати 3° group

Departs from the paradigmatic standard only in the formation of the PRAE workstem: contrary to class 3 standard, there is no substitutive softening. See details and illustration in § 494-498.

### \$442. пьсати 3\* group

Departs from the paradigmatic standard only in the formation of the PRAE workstem: there is substitutive softening, according to the class 3 paradigmatic standard, and alternation in the root vocalism. See details in Ch. 24, *Supplement*, § 874, *On verbs with unstable root vocalism*.

# §443. бьрати 3°\* group

Departs from the paradigmatic standard only in the formation of the PRAE workstem: contrary to class 3 standard, there is no substitutive softening; there is alternation in the vocalism. See details in Ch. 24, *Supplement*, § 874, *On verbs with unstable root vocalism*.

# §444. пл҄ьвати 3h\*⩨ group

Departs from the paradigmatic standard only in the formation of the PRAE workstem. The truncated stem is finally ambivalent; the vocalic realization of the root differs from the realization found in the infinitive (ь‖е); second, contrary to the CVC agreement rule, PRAE forms show V-final versions of the stem before V-initial terminals (see details in § 465, *A note on the so-called* j*-present*, and Ch. 24, *Supplement*, § 874, *On verbs with unstable root vocalism*).

# §445. клѧти 4h group

The choice of the C- or V-final truncated stem клѧ/кльн is made according to the paradigmatic standard of the 4h class (Table 445).


Table 445. Workstems of the клѧти 4h group verbs

# §446. брати 4h• group

The choice of the C- or V-final version of the truncated stem бор/бра is made according to the class 4h paradigmatic standard (Table 446). Substitutive softening that is not required by the paradigmatic standard applies in the PRAE system (see more details in § 465, *A note on the so-called* j*-present*).


# Table 446. Workstems of the брати 4h• group

# §447. плути 4h⤸ group

The following is worth noting concerning the distribution of workstems in the плути 4h⤸ group (Table 447). Initially, the choice of the C- or V-final version of the truncated stem плу/плов is made according to class 4h paradigmatic standard. However, in the ш-Part subparadigm, the V-final stem is used (expansion of the INF-AOR stem into the IMF system). CVC agreement is ensured by the choice of the C-final version of the initially ambivalent ш-Part suffix.

# Table 447. Workstems of the плути 4h⤸ group


# §448. бити 4h\*⤸ group

The following is worth noting concerning the distribution of workstems in the бити 4h\*⤸ group (Table 448). Initially, the choice of the C- or V-final version of the truncated stem би/[беј] is made according to class 4h paradigmatic standard (the shape беј is not found in the final version of the paradigm). In the PRAE system there is alternation in the root vocalism (the е‖ь grade).6 The same stem is found in the Imf and н-Part subparadigms, by the basic stem allotment rules. However, in the ш-Part subparadigm, the V-final version of the stem is used (expansion of the INF-AOR stem into the IMF system, the ⤸ effect). CVC agreement is ensured by the choice of the C-final version of the initially ambivalent ш-Part suffix. See details in Ch. 24, *Supplement*, § 874, *On verbs with unstable root vocalism*.

6 The ь vocalism in PRAE is supported by the systematic parallelism between the groups бити 4h\*⤸ and крыти 4h\*⩨⤸, and also by the development of these paradigms in the descendants (cf. Russian *бить*, *бью*, *крыть*, *крою*). The diachronic interpretation of these forms does not support the solution adopted in this grammar.


# Table 448. Workstems of the бити 4h\*⤸ group

# §449. крыти 4h\*⩨⤸ group

The following is worth noting concerning the distribution of workstems in the крыти 4h\*⩨⤸ group (Table 449). Initially, the choice of the C- or V-final version of the truncated stem кры/крыв is made according to class 4h paradigmatic standard (the shape крыв is not found in the final version of the paradigm). In the PRAE system, there is alternation in the root vocalism (the ы‖ъ grade). By the ⩨ effect, the C-final version is arrested, and thus the stem is кръ (and not кръв). The same stem is found in the Imf and н-Part subparadigms, by the basic stem allotment rule. However, the ⩨ effect does not apply in the н-Part subparadigm, and the stem is кръв, by CVC agreement. In the ш-Part subparadigm, the V-final version of the stem is used (expansion of the INF-AOR stem into the IMF system). CVC agreement is ensured by the choice of the C-final version of the initially ambivalent suffix of ш-Part.

Thus, the н-Part stem, which shows the C-final version of the stem (кръв.ен, мъв.ен) that is required by the paradigmatic standard, stands separately from the rest. See more details in § 465, *A note on the so-called* j*-present*, and also Ch. 24, *Supplement*, § 874, *On verbs with unstable root vocalism*.


Table 449. Workstems of the крыти 4h\*⩨⤸ group

# §450. пѣти 4h⤸ group

The following is worth noting concerning the distribution of workstems in the пѣти 4h⤸ group (Table 450). This group contains only the verb пѣти. Initially, the choice of the C- or V-final version of the truncated stem пѣ/пој is made according to class 4h paradigmatic standard. However, in the ш- and н-Part subparadigms, the V-final version of the stem is used (expansion of the INF-AOR stem

into the IMF system). CVC agreement is ensured by the choice of the C-final versions of the initially ambivalent ш- and н-Part suffixes.


Table 450. Workstems of the пъти 4h2 group

### §451. мрѣти 4h\*( group

The following is worth noting concerning the distribution of workstems in the мрѣти 4h\* group (Table 451). Initially, the choice of the C- or V-final version of the truncated stem мръ/мер is made according to class 4h paradigmatic standard (the shape Mep is not found in the final version of the paradigm).

In the PRAE system there is alternation in the root vocalism (the e|| s grade). There is expansion of the PRAE stem into the IMF subparadigms, and also into the 1- and T-Part subparadigms. In the latter, the V-final version of the PRAE stem мьр/мрь is used by the CVC agreement rule. See more details in Ch. 24, Supplement, § 874, On verbs with unstable root vocalism.


### \$ 452. влешти 4с\* group

The following is worth noting concerning the distribution of workstems in the BATHETH 4c\*s group (Table 452, p. 264). This group contains only two verbs, BATHETH and Ephurn. The paradigm contains two workstems opposed by vocalic realizations (the e|ls grade); the realization of ь is found in the н-Part, ш-Part, and A- and T-Part subparadigms. This distribution of realizations by subparadigm is unique and violates the principle of opposition of vocalic realizations in the «Prae» and «Inf» workstems. If the observed vocalic realization (the e grade, stems влѣк, връг) in the PRAE system (and also in Imf) is treated as secondary

(leveling of realizations by the Inf and Aor forms, the absence of expected vocalism change-a paradigmatic countereffect, indicated by ø), then the distribution of primary vocalic realizations (i.e. BALK, EPAr for the PRAE stem) corresponds to the ordinary alien stem expansion effect (here, of the PRAE stem into the IMF and INF-AOR systems): the distribution is identical to that found in the можти 4h\* / group. If the vocalic realization in the PRAE system and in Imf is not treated as secondary, then these two verbs must be considered unique.

Note that SUPR shows a generalization of the Inf vocalic realization to the participles in the IMF system (cf. съвлъкъше SUPR 103, 2); also Imf влъчаахъ SUPR 39, 5. Generally speaking, this makes it possible to consider aberrant all PRAE and IMF forms with the & grade vocalic realization (the Batk stem).


Table 452. Workstems of the влъшти 4c\* 4 group

The & effect indicates the absence of the expected \* effect.

### \$ 453. чисти 4c\*s group

The following is worth noting concerning the distribution of workstems in the чисти 4c\* group (Table 453). Two workstems are opposed by vocalic realizations: «Prae» чьт and «ከነ» чит. The vocalism of «Prae», though the alien stem expansion effect, is found in all forms except Inf and Sup.

Table 453. Workstems of the чисти 4c\* group


§ 454. Profiles of type representatives of irregular verb groups

Every group of irregular verbs yields its own special paradigmatic standard, representable as a deformation of the paradigmatic standard of the superordinate main paradigmatic class, class 3 or classes 4c and 4h. Similarities and differences

of these standards can be observed in the profiles of representatives which are shown in Table 454. The list of profile forms is also found there.


Table 454. Profiles of irregular verbs

# CHAPTER 18 **Verbal terminals and suffixes**

# **Sets of terminals and suffixes**

§455. a catalog of terminals and suffixes

Tables 455.1–3 (pp. 267–268) show the full catalog of verbal terminals and suffixes of nominal forms of the verb.








First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9


# Table 455.2. Terminals and suffixes in the IMF system


# Table 455.3. Terminals and suffixes in the INF-AOR system


тъ, стъ

(ѣ)ꙗше (ѣ)ꙗшета (ѣ)ꙗшете

(ѣ)ꙗшете (ѣ)ꙗхѫ

### Nominal subparadigm suffixes


# §456. On notation

The slash separates versions of twofold terminals and suffixes (the hard subtype is above the slash, or *overslash*, the soft subtype below the slash, or *underslash*). Standard initial CVC ambivalence of suffixes is indicated with a tilde (e.g. ен~н). Versions of terminals that differ in the presence/absence of the initial vowel are shown with parentheses. Notation like (о)хъ in the aorist should be understood as охъ or хъ. In the imperfect, notations like (ѣ)ахъ should be understood as ѣ.ахъ or ахъ; (ѣ)хъ as ѣ.хъ or хъ (see selection rules below). The sign | in suffix tables shows that the corresponding wordform lacks a terminal. Parentheses enclose morphophonological variants of suffixes.

# §457. Standard and nonstandard sets in a subparadigm

*Nonstandard sets* serve secondary, anomalous, or aberrant forms. All main class verbs, both regular and irregular, take terminals and suffixes from *standard sets*.

The Prae *athematic set* is shown, for reference, as it is normally found in grammars. This set is found in the forms of unique athematic verbs that are treated as anomalous and not subject to an analysis of the type [stem + terminal].

# §458. Terminals and suffixes in the present system

The choice between *i*-conjugation and *e*-conjugation sets is determined by the verb class. *i*-conjugation sets are used in classes л҄юбити 1 and трьпѣти 2, and *e*-conjugation sets in all others, except unique verbs. In the *e*-conjugation, terminals in the Imv subparadigm and suffixes of participles show twofold variance: the underslash forms are used with stems ending in a morphophonologically soft consonant or a vowel, and overslash forms are used elsewhere.

Unique verbs may combine *i*- and *e*-conjugations. The division into conjugation does not extend to anomalous forms with athematic presents (дати, ꙗсти, ѥсмь, вѣдѣти, имѣти) and athematic imperatives: ∇даЖь (for дати 0), ∇ꙗЖь (for ꙗсти 0), ∇вѣЖь (for вѣдѣти 0), ∇виЖь (for видѣти 2).

# §459. Terminals and suffixes in the imperfect system

The choice between terminals of the type ахъ and ѣ.ахъ in personal forms of the imperfect is made according to the so-called *morphological balance rule* (Table 459): short stems are combined with long terminals and long stems with short terminals.



In the workstems, the period in ѣ.ахъ-type terminals is often omitted. On the hiatus in Imf see § 69 and ff.

Nonstandard terminal sets in personal forms of the imperfect are only found among aberrant forms; the rule of morphological balance applies in nonstandard imperfects as well (see § 467–472).

The ш- and н-Part suffixes are initially ambivalent and distributed by the CVC agreement rule. The marker ов.ен is only found in the двигнѫти 5 class, in the model двигн.ов.ен.ъ.

§460. Terminals and suffixes in the infinitive-aorist system

All standard set terminals are initially ambivalent, e.g. охъ~хъ etc., and distributed by the CVC agreement rule.

Nonstandard sets are only found in secondary forms (see § 476–482).

Suffixes of nominal forms are C-initial. In case the workstem is C-final, mph ⇒ ph/norm rules apply that eliminate some clusters (§ 74), cf.: влад.л=ъ⇒влалъ.

# **Boundary adjustment rules**

### §461. General

These rules replace the alloform of the last formative of the workstem in the morphological skeleton by the alloform that shows up in the synthesized wordform in its inflectional spellout. For example, for the form 2SgPrae of the verb мощи, in the morphological skeleton we have the alloform мог (cf. мог+еши); in the inflectional spellout we have the alloform мож (cf. мож=еши). The needed alloform can be obtained from the alloform found in the morphological skeleton using rules of segmental replacement by an alternation. Boundary adjustment rules indicate, first, which forms must undergo adjustment (this depends on both the paradigmatic class of the lexeme and on the grammatical address of the synthesized wordform); second, they indicate which replacement rule and by which alternation must effect the needed segmental change. In the example above, this alternation is velar palatalization, and the rule is к→ч.

Note that boundary adjustment may generate segmentally diverse stems in the inflectional spellout where the workstems are identical, cf.:





# §462. Boundary adjustment rules

The boundary adjustment rules require replacement of workstem-final consonants as shown in Table 462.1.

Table 462.1. Boundary adjustment rules: consonants (verbs)


\* Not to be confused with substitutive softening that applies in the derivation of PRAE system workstems, in class 3 and irregular verbs of the брати 4h• group, and IMF system workstems, class 1.

\*\* Cf. anomalous forms of the athematic Imv such as ∇виЖь.

In constructing secondary aorists, viz. the old sigmatic 1 (type нѣсъ) and old sigmatic 2 (type рѣхъ), boundary adjustment rules require compensatory vowel lengthening1 as shown in Table 462.2.


Table 462.2. Boundary adjustment rules: vowels (verbs)

Not to be confused with the change in vocalic realization that takes place in irregular verbs and morphologically anomalous imperatives such as ∇рьцѣте.

1 This is a replacement by fundamental vowel alternations: vertical pairings for pure vocalism and horizontal ones for sonant vocalism; see § 125.

# CHAPTER 19 **An overview of verb forms by system**

# **The present system**

§463. A general overview of forms

Table 463.1 shows all possible forms in the PRAE system.



\* Only aberrant forms that form a separate paradigmatic standard of some subparadigm are shown here.

The distribution of anomalous, secondary, and aberrant forms are shown in Table 463.2 (p. 274).

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9


#### Table 463.2. Distribution of nonstandard forms in the PRAE system

### § 464. Anomalous imperative forms

All imperative forms of the verbs ρεшти 4cv, πεшти 4cv, από жεшти 4cV are anomalous; there are no other anomalies in these verbs. The morphological anomaly is the replacement of the root allotorm with € vocalism by an alloform with b vocalism. Thus, for example, we have pek + the > pbq=''Te, рєк + и > рьц=и. Note that all these verbs have truncated stems ending in єк, єг.

The verb видети 2 has a morphologically anomalous ImvSg form Vвиждь, analogous to athematic verbs, cf. даждь, въждь etc. (prefixed verbs with видукти have standard imperatives).

### Illustrations

и рече емог исъ виждь никомоуже не повъждь: нъ шедъ покажи са архнереови: н неси даръ иже повелъ моси въ съвъдъмие имъ Mt 8, 4 мак; слъщи дъщи н ВИЖДЪ И Приклони оухо ТВОЕ И ЗАБЯДІ ЛЮДІ ТВОЊУ І ДОМЪ ОЦА ТВОЄГО PS SIN 4, 11; but заповъди въси не прелюсъ дъи не оукради: не лъжа СЪВЪДЪТЕЛЬСТВОУИ" НЕ ОБИДИ" ЧТИ ОТЦА ТВОЕГО И МТОЬ Mk 10, 19 MAR; НЕ ОЄВЪНОУПТЕ ЛЖКАВЬНОУНИИИЪ: НИ ЗАВИДИ ТВОДНАШИЙМЪ БЕЗАКОНЕНИЕ PS SIN 436, 1.

НЕ ПЬЦЕТЕ СА ОУБО ГЛЕЖЦЕ: ЧТО ЋАУЪ ЛИ ЧТО ПИЕМЪ: ЛИ ЧИМЪ ОДЕЖДЕМЪ СА Mt 6, 31 MAR.

И НАЧАША ЕТЕОНІ ПЛЪВАТИ НА ЯВ І ПОНКОВІВАТИ ЛИЦЕ ЕГО І МЖЧИТИ Г І ГЛАТН EMOY" прорьци намъ χε къто єсть огдаривъ! та Mk 14, 65 zogr.

### § 465. A note on the so-called j-present

Forms in the PRAE system are mostly derived from truncated stems (otherwise only in the genam 7 class, and in some unique verbs). In some cases substitutive softening applies, while in others it does not. Substitutive softening is the defining characteristic of class 3. However, special paradigmatic effects apply in several cases that destroy the paradigmatic standard of the class. First, in the groups мєтити 3° and вьрати 3°\*, contrary to expectation, substitutive softening does not apply (METR instead of the expected MEUTZ); second, in some class 4h

verbs substitutive softening does apply (the клати 4h•); cf. 1SgPrae кол҄ѭ, instead of the expected колѫ, and the imperative кол҄ите instead of the expected колѣте. Diachronically, substitutive softening can be interpreted as the transformation of the C+*j* combination. Accordingly, Prae forms that show substitutive softening are called *j*-present, which is understood as a special morphological means of present formation. Prae forms which in the present grammar are treated as subject to the C-final stem arrest effect (noted by ⩨) could also be considered as *j*-presents. Such are, first, all standard PRAE forms in class 6 (милуѭ, милуѥши; милуѩ, etc.), and, second, some verbs which are classified here as irregular (such are the groups пл҄ьвати 3h\*⩨, крыти 4h\*⩨⤸). Introducing *j*-present in the description would have obviated violations of the CVC agreement rules: V-final variants would be selected before the C-initial formative *j*, which initiates PRAE terminals and suffixes in the *j*-present. However, setting up *j*-presents would be in contradiction to the cluster constraints that prohibit all C+*j* combinations.

# **The imperfect system**

§466. A general overview of forms

Table 466 shows all possible forms in the IMF system.


Table 466. Forms in the imperfect (IMF) system

\* Only those aberrant forms that form a separate paradigmatic standard of some subparadigm are shown here.

For nonstandard imperfects see below (§ 467 and ff.). On other aberrant IMF forms see § 612 (terminals), § 609 (paradigmatic classes 5 and 4). For new ш-Part see § 473.

For н-Part of the verb обути 4v see § 509. For н-Part of the unique verb быти see § 548.

# Nonstandard imperfects (personal forms)

# \$467. Contracted imperfect

Forms of the contracted imperfect use the nonstandard "Contracted imperfect" terminal set with terminals of the type th.yb ~ xb. The workstem is the same as in the standard imperfect. The choice between terminals of the xx vs. type in personal forms of the imperfect is made according to the morphological balance rule: truncated stems are combined with long terminals like typ, while expanded stems are combined with short ones like yo, i.e. with C-initial ones. Cf. standard âюблаауъ, contracted люблауъ; standard трьптахъ, contracted Tpurty's standard Heckay's, contracted Hecky's etc. In all classes except 4v, contracted imperfect terminals are distributed by the CVC agreement rule. See more details in § 914-915, Excursus on the morphology of personal forms of the imperfect.

# § 468. Illustrations and distribution of contracted imperfects

Contracted imperfect forms (see Table 468) are found in all sources. SAV is special in that it only shows contracted imperfects.4


Table 468. Contracted imperfect: illustrations

амиб амир гля тевъ: егда юнъ въ: погасаше са: и хождаше: гамо же хотъше: егда же състаръеши са въздеждеши ожцѣ (с)воич и инъ та погашетъ: и ведетъ rano we ne youremin In 21, 18 SAV.

НОШТЪЕ СРЪДЬЦЕМЬ ГЛОУМАЕХЪ СМА 1 КЛЪЦАШЕ ДХЪ МОГРУ SIN 76, 7.

скемонъ же рече имъ: чьто зъло сътвори: они же сзлиха въпитехъ гажште: да распатъ бядетъ Mt 27, 23 ZOGR.

# § 469. Present imperfect

Present imperfect forms use the PRAE workstem and imperfect terminals, both standard and contracted. Because only truncated stems are used, all terminals

<sup>1</sup> Note the gloss искаахъ SAV Mt 26, 59, which is the only apparent non-contracted imperfect example, but the second letter a is subscribed; cf. nckayx SAV Lk 5, 18; SAV Jn 11, 8.

are ѣ-initial (ѣ.ахъ/ѣхъ). For example, for зъвати 3°\* we have зов+ѣаше: зовѣаше, given the canonical form зъва+аше: зъвааше. Forms of the present imperfect are the result of the alien stem expansion effect. Note that this effect can distinguish present imperfects from canonical ones only in classes л҄юбити 1, плакати 3, and миловати 6 (which canonically distinguish PRAE and IMF workstems), and also in some unique verbs. In class 3, present imperfect is only found with irregular verbs.

As for the л҄юбити 1 class, where canonical imperfects follow the model л҄юбл҄ꙗаше and present imperfects follow л҄юбѣаше, present imperfect forms have a double analysis in case the truncated stem ends in a labial or л, н, р. Forms such as л҄юбѣаше (see below млъвѣаше As and мльвѣше Sav) can be interpreted as resulting from a segmental aberration (alternative pairings by the substitutive softening alternation, see § 117). Forms such as творѣаше an be interpreted as the graphic spellout not only of phonemic /rě/ (as present imperfect), but also of phonemic /r̕a/ (as canonical imperfect).

# §470. Illustrations and distribution of present imperfects

Isolated forms of the present imperfect are found in all sources (see Table 470). Supr is special in that present imperfect forms are common. Note that the unique verbs дати 0, ꙗсти 0, стати 0, об.рѣсти 0, and гънати 0 show present imperfect, but the corresponding forms of these verbs are treated in this grammar as canonical; all other present imperfect forms are aberrant.


Table 470. Present imperfect: illustrations

Notes to Table 470

1) Cf. contracted forms in Sav: и се еі бѣ сестра именемь мариꙗ· ꙗже и сѣдъши при ногу ісу слышаше слово его· марта же мльвѣше о мнозѣ служъбѣ· ставъши рече· ги не брѣжеши ли ꙗко сестра моꙗ единѫ мѧ остави служити тебѣ· рьци убо еі да ми поможетъ Lk 10, 39–40 Sav. Cf. люблѣше бо іс марьтѫ· и сестрѫ еѧ· и лазора Jn 11, 5 Sav. Note also those spellouts in Sav that are ambiguous because of graphics (see § 221): чюдеса ꙗже творѣше Mt 21, 15; ови же за ланитѫ і ударѣхѫ Mt 26, 67 (f. 112); and also молѣше Mt 18, 29; Lk 8, 38; Lk 8, 41; Lk 15, 28; Lk 18, 11; Lk 22, 44, молѣхѫ Mt 8, 31; Mt 15, 23; Lk 7, 4; Lk 8, 31, χογλέιμε Lk 23, 39, χογιτέχε Mt 27, 39; Mk 15, 29.


### §471. lotated imperfect

Iotated imperfect forms use the IMF workstem, which takes terminals from a nonstandard set, viz. the "lotated imperfect" set. These are terminals of the type th.ax's/ax's. They are distributed by the morphological balance rule, just as in the standard imperfect. Cf. standard трыптахъ, iotated трыпълуъ etc.

Note that iotated imperfect terminals can also attach to PRAE stems, forming the present iotated imperfect (regular or contracted).

§ 472. Illustrations and distribution of the iotated imperfect

Fewer than 25 aberrant iotated imperfect forms are attested (see Table 472, and also Diels, § 113.5; Vaillant, § 158).


Table 472. Iotated imperfect: illustrations

Notes to Table 472


phonetic treatment of these spellouts as indicating a long vowel, see Vaillant, § 158).2

# New ш-participle

# §473. New ш-Part of the любивъ type

New ш-participle is an aberrant ш-Part form in the л҄юбити 1 class. It uses the expanded, not the truncated stem. Cf. л҄юби+въ (л҄юбивъ) for standard л҄юбл҄+ь (л҄юбл҄ь). These forms result from the alien stem expansion paradigmatic effect, namely, the expansion of the INF-AOR workstem into the IMF system.

# §474. Illustration and distribution of the aberrant forms

Some sources show no examples of the new participle. These are Ps Sin and As, as well as Cloz. In Zogr, Mar, and Sav, the new participle is represented by a few examples (see Table 474), while in Supr the new participles are widely distributed (see details in Vaillant, § 162).


Table 474. New ш-participle: illustrations

In Supr, as already said, new participles predominate. See, e.g., in Vita of St. Paul and St. Juliana: отълѫчивъшу бо сꙙ адаму на оно мѣсто породы· и еугѫ ѥдинѫ оставивꙿшу· пристѫпи диꙗволъ· и уподобивъ сꙙ змиї прѣльсти ѭ· и та жена прѣльсти адама Supr 9, 12–16; however, old participles are also attested, cf. in the same lectionary: и рече зови паула и улиꙗниѭ· и постави ꙗ прѣдъ сѫдиищемъ· и прѣтꙙ има рече· пристѫпьша пожьрѣта богомъ· а не надѣита сꙙ избѣжати рѫку моѥю· стъи же паулъ осклабивъ сꙙ (new ш-Part) рече· не оставьѭ бога сътворьшааго небо и земьѭ· и поклон҄ѭ сꙙ кумиремъ Supr 7, 2–9.3

2 Supr shows a great variety of morphological and graphic variants of the imperfect. For example, for творити we find: творꙗаше 391, 18–19; творааше 200, 23; творѣаше 18, 27; творꙗ ꙗше 473, 20–21; творѣꙗше 438, 19; творѣѣше 198, 5; творꙗше 274, 27; творѣше 19, 29.

3 Cf. the homily of St. Epiphanius of Cyprus (Supr 447, 28–471, 12): […] не начинаѥтъ великыими рѣчьми нѣкыими къ пилату· да не вь сего раздражъ отъпадетъ прошениꙗ· ни глагол҄етꙿ к н҄ему даЖъ ми тѣло їсосово· омрачьшааго сльнце

# The infinitive-aorist system

### \$ 475. A general overview of forms

Table 475 shows all possible forms in the INF-AOR system.



For secondary aorists see § 476-482 below. For aberrant forms by paradigmatic class see § 609-610; for other aberrations in the INF-AOR system, see \$612-614.

# Nonstandard aorists

### & 476. General

The forms of the aorist bundle include a series of secondary forms. All these forms are declared nonstandard aorists (or old aorists). In the description of these non-

ПОТЕЖДЕ МАЛА ВРЕЛЕНЕ: И КАЛЕНИЕ РАЗ ДОВБЫША: И ЗЕМЫЖ ПОТРАСТЬША: И ГООБЪ! ОТВРЪЗЗЪША SUPR 454, 6-12;

[ ... ] воле оуво їщенфе испрошии и принлиз: въси пристяпії къ кристоу " и сьнемъ исоуса. въси ли оуво кого понесе supr 456, 9-12;

[...] обаче клажи ряцъ твон и нисифе послоужъшни: [...] клажи оуста твою: ве-съптости НАСЪЩЕТЪША СА И ПРИЛОЖЬША СА КЪ ИСОУСОВОМЪ ОУСТОМЪ' И ДОУЖА СТА ОТЪТЛАДУ ИСПЛАВЪША СА ЕЛАЖЖ ОЧУ ТВОИ ПОНЛОЖШИИ СА КЪ ЇСОУСОВАМА ОЧИМА' И СВЕТЪ ИСТИНЪНЪМ ОТЪТЪДОУ приимъши supr 457, 20-30;

[...] онъ истъпи естъ съкроушивън въ їщрда "Бокъ" главъ злиевъ злиевъ вашихъ съ истъ IECTT'S OEMYHB'SH KOLETON'S II ПОЗООУ СЕТВОРИВЪ' И ОСЛАЕЛЬ ЖИЛЪ! ВАША' Т'Ъ ИСТЪИН ЕСТЪ' СЪВАЗАВЪИ И ОМОАЧЪ' И ВЪ БЕЗДЬНЖ ПОСЪЛАВЪ ВЪ СУРРА 467, 2-6.

standard aorists, the bundle is divided into two subbundles: the first contains all forms except 2–3Sg (called the *main subbundle* below), while the 2–3Sg form belongs in the second subbundle.

The main subbundle has three types of nonstandard aorists: *root* (идъ type), *old sigmatic 1* (нѣсъ type), and *old sigmatic 2* (рѣхъ type).

For 2–3Sg there are two kinds of secondary forms: with the nonstandard terminal стъ (быстъ type; primary form бы=0), and with the nonstandard terminal тъ (клѧтъ type, primary form клѧ=0), both of which replace the standard zero terminal.

The secondary aorist is only possible among verbs with an athematic infinitive (class 4 and some unique verbs), and also among нѫ-dropping verbs of class 5. See more details on the distribution of secondary forms in § 481 below.

### §477. Root aorist (идъ type)

The workstem of the root aorist is the truncated C-final stem (the root). For unique verbs the stems are: ид for ити, ꙗд for ꙗти ‹1114›, сѣд for сѣсти, лег for лещи, рѣт for °рѣсти (other unique verbs have no root aorist forms). Irregular verbs with unstable vocalism have the same vocalic realization as in «Prae».

Terminals are from the nonstandard "Root" set. All terminals of the root aorist are V-initial, while all stems are C-final.

Boundary adjustment follows the general rules (see § 462), namely: velars undergo replacement by the к→ч rule before е-initial terminals. Thus, for мощи: мог+ъ gives мог=ъ, мог+овѣ gives моговѣ, мог+ета gives можета etc.

Below are paradigms of personal forms of standard and root aorists of the verb ити 0.


### §478. Old sigmatic 1 (нѣсъ type)

The workstem of the old sigmatic aorist is the truncated stem. It can be C-final (cf. нес: нес+съ), or finally ambivalent, in which case the V-final version is used (cf. клѧ/кльн: клѧ+съ). The only unique verb to use the old sigmatic aorist is ꙗсти: the workstem is ꙗд ‹1113›. Irregular verbs with unstable vocalism have the same vocalic realization as in «Prae».

Terminals are from the nonstandard "Old sigmatic 1" set. All terminals in this set are C-initial.

Boundary adjustment follows the general rules (see § 462). Compensatory vowel lengthening applies (by the following pairings: ь‖и, е‖ѣ, о‖а, рь‖рѣ, ль‖лѣ). Thus, for нести we have нес + съ: нѣс.съ; нес + совѣ: нѣс.совѣ, etc. Clusters are eliminated by the standard mph⇒ph/norm rules: нѣс.съ⇒нѣсъ, нѣс.совѣ⇒нѣсовѣ (see § 74).


Below are paradigms of the personal aorist forms of the verb нести 4c.

# §479. Old sigmatic 2 (рѣхъ type)

The workstem of the old sigmatic aorist is the C-final truncated stem. Irregular verbs with unstable vocalism use the same vocalic realization as in «Prae».

Terminals are C-final from the standard set.

Boundary adjustment follows the general rules (see § 462). Compensatory vowel lengthening applies (by the following pairings: ь‖и, е‖ѣ, о‖а, рь‖рѣ, ль‖лѣ). Thus, for рещи [рек.т.и] we have рек+хъ: рѣк.хъ; рек+ховѣ: рѣк.ховѣ, etc. Clusters are eliminated by the standard mph⇒ph/norm rules: рѣк.хъ⇒рѣхъ, рѣк.ховѣ⇒рѣховѣ (see § 74).

Below are paradigms of the personal aorist forms of the verb рещи 4c.


# §480. Secondary 2–3Sg forms

Secondary forms take the terminals -стъ or -тъ, which attach to the infinitive workstem, быти 0: бы+стъ; клѧти 4h: клѧ+тъ; мрѣти 4h\*⤹: мрѣтъ.

# §481. Distribution of nonstandard aorist forms

As noted, the secondary forms described above are possible for a limited set of verbs. Such are, first, some verbs with an athematic infinitive from class 4c or unique (see lists below); second, 4h verbs in -ѧти; third, нѫ-dropping class 5 verbs. Note that the latter also show nonstandard aorists by class 4, which are treated here as aberrant. Cf. the aorists of ужаснѫти: ужаснѫшѧ (by class 5, the canonical, primary form); ужасошѧ (by class 4, an aberrant form), and ужасѫ (root aorist, a canonical, secondary form).

Distribution limits of secondary forms are shown in Tables 481.1 and 481.2. See Vaillant, § 153–156; on the distribution in sources see Vaillant, § 157.


Table 481.1. Distribution of nonstandard aorist forms: the first subbundle

Table 481.2. Distribution of nonstandard aorist forms: 2–3SgAor


The observed constraints on the distribution of nonstandard aorists have a morphophonological, namely accentual, basis, as shown in Dybo's work (Dybo 2000, p. 304–309, 366–370).

### §482. Nonstandard aorists: illustrations

Tables 482.1 (p. 284) and 482.2 (p. 285) show a small selection of examples illustrating parallel uses of secondary (nonstandard) and primary aorist forms of the same verbs.


### Table 482.1. Nonstandard aorist (main subbundle): illustrations

Notes to Table 482.1



Table 482.2. Nonstandard aorists (2-3Sg forms): illustrations

Notes to Table 482.2


2–3SgAor agrees with the preceding aorist разгорѣ. Cf. Greek ὅτι ἐξεκαύθη (3SgAor) ἡ καρδία μου καὶ οἱ νεφροί μου ἠλλοιώθησαν (3PlAor); οἱ νεφροί 'entrails'. Večerka treats the form измѣтъ as corrupt for измѣнитъ. The verb измѣти is absent in Večerka. Cf., however съмѣти, измѣниѥ with the same root ‹586›. Večerka's treatment is based on taking the translation from the Greek too literally (the Greek verb ἀλλοιόω can mean not only 'change', but also 'be corrupted'). In addition, Večerka's solution leads to a syntactic contradiction (the same phrase contains an aorist and a present, which is not found in the Greek).

# т-participles

# §483. General

т-participles are equivalent to н-participles syntactically and semantically, but differ from them radically in their morphology. Stems of н*-* and т-participles figure in nominal derivations, the most common of which is the deverbal noun in ьј=е (усъпениѥ, бытиѥ, etc.). Cf. also: клѧтва, бритва, умрьтвиѥ [у.мрь.т.в.ьј.е]. Sometimes the participle itself is not attested in the sources, but its existence can be inferred through derivatives (see § 865, *A note on nominal stems and verbal platforms*).

For most verbs these participles are in complementary distribution: a verb that shows н-Part has no т-Part forms, and vice versa. Otherwise for the verb сѣти 4v, cf. вьсѣкъ иже слышитъ словеса цсарествиѣ· ꙇ не разумѣваатъ· приходитъ же неприѣзнь и въсхытаатъ сѣное въ срдци его· се естъ сѣное при пѫти Mt 13, 19 Mar, and несѣта н҄ива Supr 243, 8; and the verb вѣдѣти 0 (see § 527–531).4

# §484. The morphology of the forms

The suffix т is attached to the INF-AOR system workstem. If the workstem is finally ambivalent, then т-Part uses the V-final version of the stem (cf. клѧтъ for клѧти 4h). In case the truncated stem is C-final, clusters are transformed by general rules. In verbs of the мрѣти 4h\*⤹ group and чисти 4c\*⤹ group, the workstem has the same vocalic realization as in «Prae».

# §485. Illustrations and distribution of т-participles

т-participle stems are only possible for verbs with athematic infinitives. It is attested for most verbs of class 4h, and also for the verbs сѣти 4v, увѧсти 4c,

<sup>4</sup> Dybo established the accentual distribution of т- and н-participles, and, accordingly, their original complementarity (see Dybo 2000, p. 309–312). However, substantives derived from participles can show н- and т-Part stems for the same verb. Cf. пѣниѥ and пѣтиѥ, убиѥниѥ and убитиѥ.

отврести 4с\* 1, извъдувти 0, and съмити 0.5 Although the absence of attested forms does not always indicate the morphological impossibility of the corresponding stem, let us list the 4h verbs without T-Part stems: гнити, дляти, клати, and EpaTH, and also the group Kparrh 4h\*#2.

For some verbs, the v-Part stem is attested but the participle itself is not. Such are, e.g. Завъвенъ, but also завъютикама завъють (въса си въ забъють въложн сьревролюбьствия тъгда SUPR 412, 1).

#### Illustrations

гла низ пилатъ: что створя its нарицаемочив уди: глаша вси да распатъ БЖДЕТЪ Mt 27, 22 SAV; НАЧАТАМИ ТАННАМИ EUCH 162, 22-23; ВКО СТРЪЛЪ ВЪ ржці сильнааго тако снве отнать рус ро SIN 126, 4, обраштете младьныць повитъ: лежашть въ тъслехъ Lk2, 12 zogr; пръдъ очима нашима: месть кро[вь] ВИ РАБЪ ТВОІХЪ ПРОЛИТЪВ РЅ SIN 78, 10; ПЪТА БЪДЖ МЬНЪ ОПРАВЪДАНИЪ ТВОЪ PS SIN 118, 54; ДІВЪ ТВООГТЪ ОУМО' ЗЕМЛЪ ВІСАШТІ ПОВЕЛЪНЬЕМЬ НА ВОДАХЪ' А ТАЖЪКА ВЕШТЪ СЖШТ!' ЧЪТО ЖЕ КЪТО ЛЕЧЕТЪ' МООЕ ПООСТРЪТО ВІДА' І ПЪСЪКОМЬ съвазано CLOZ 10а, 3-6.

СВЕТЬЛИ СВЕТЪЛО ОУВАСТИ: ВЪ ЦОЬКВЫНЪВМЪ ЛИКОИНЖТЪ ЧОЬТОЗЪ SUPR 335, 7-9 (for oyBactin 4c); гробъ отвръстъ грътани ихъ: њазънкъи своими ЛЬШААХЖ РУ SIN 5, 10 (for ОТВРЕСТИ 4c \* ); ИЗВЕСТЪ СА СЪТВОМ SUPR 363, 3-4 (for извъдувти 0); cf. also in nominal stems: шьстию for ити 0), as well as сънит (съпитик for сънити 0).

<sup>5</sup> The forms carry, carry 'says, say', which can be interpreted as m-Part for the verb carry (without any other forms; see Vaillant, § 216), in the present grammar are treated as extraparadigmatic and are thus not examined here.

<sup>6</sup> Vaillant (§216) assigns the form отматыхъ to the verb тати, тьмя 4h, while Večerka assigns it to the verb overn, oThem 4h. The verb TATH, This not found in this grammar; accordingly, the interpretation follows Večerka.

# CHAPTER 20 **An overview of verb classes**

# любити 1 class

§486. Class membership

This class contains about 800 lexemes. There are slightly over 300 distinct verbs. Possible right edges of the lexical component are shown in Table 486.

Table 486. Possible right edges of the lexical component in л҄юбити 1


### §487. Notes on word formation

This class can contain both root and suffixal lexical components. Examples:


This class contains many denominal verbs (cf. дъЖити, съгрѣшити, дарити), causatives, and iteratives (cf. избавити for избыти, носити for нести, водити for вести, уморити for умрѣти, поити for пити, etc.).

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

# §488. Connections with other classes

There are systematic correlations 1≈7, cf. побѣдити≈побѣЖати. There may be competition between classes 1 and 7. Note that exclusion from the benchmark dictionary of lexemes of class 7 verbs that are represented only by non-diagnostic forms (Imf) is normally not discussed, unlike in Večerka. Cf. глумити and глумл҄ꙗти in Večerka, but only глумити. in the PD. See more details, including the meaning of the term *non-diagnostic*, in Ch. 24, *Supplement*, § 910– 913, *Excursus on the contamination and competition between paradigmatic classes*.

# трьпѣти 2 class

# §489. Class membership

This class contains just under 150 lexemes; there are about 40 distinct verbs. Possible right edges of the lexical component are shown in Table 489.

Table 489. Possible right edges of the lexical component in the трьпѣти 2 class


# §490. Notes on word formation

In this class, only the verb слышати [(слы).ш.ѣ.т.и] ‹844› contains a suffixal lexical component.

§491. Connections with other classes

Correlations 2≈7 are possible, cf. възьрѣти ≈ възирати, съгорѣти ≈ съгарати, бльщати ≈ блистати.

# Commentary on individual verbal lexemes

## §492. List of commented lexemes of л҄юбити 1 class


възъꙗрити. Nontrivial spellout of the starting form, see § 640–641.


and обрѣмен҄ꙗти. The spellout обрѣмьꙗѧ in Supr is treated as morphologically strange form for обрѣменити.


нждити, принждити; likewise понждати, принжждати.


свътити. Cf. свыткти 2.


оумьнити. See мынити.


#### § 493. List of commented lexemes of трыгъти 2 class

видети. Anomalous form Imv2Sq: Vвиждь in place of the expected види, only for the lexeme видети; for the others: завиди PS SIN 36, 1; OEHAH Mk 10, 19 ZOGR and MAR; възненавіді Mt 5, 43 AS. On such aberrant forms as вижди see § 620.

octorn [06.(croi).t.r.h] 'surround'. Only three glosses, all with aberrant spellouts of the prefix: 95%CT- (MAR and SUPR), 066CT-(ZOGR). Cf. or'scrolarn.


свъткти. Cf. свътити 1.

сътрьпчти. Večerka gives the headword HECTPLITEN' covering just two glosses: HECTTOLITHA MAKA (SUPR 375, 3) and несътръпътна мжка (supr 377,30-378, 1). The former is treated, in the PD, as a regular H-Part from the verb сътрыпъти; the latter is considered an erroneous spellout for

несътрьпѣнаꙗ or a morphologically strange form (a contamination of н- and т-Part).

# плакати 3 class

# §494. Class membership

This class contains just under 300 lexemes; there are just over 80 distinct verbs. Possible right edges of the lexical component are shown in Table 494.

Table 494. Possible right edges of the lexical component in плакати 3 class


## §495. Notes on word formation

Normally the lexical component ends in a root formative; otherwise only for the following items: клокотати, кльчьтати, ристати, ръпътати, скрьжьтати, исходатаꙗти, шьпътати. Cf. also: глаголати, кл҄ѥветати, колѣбати, лобъзати, трепетати.

# §496. Connections with other classes

The following correlations are possible: 3≈0 (даꙗти ≈ дати), 3≈4 (сѣꙗти ≈ сѣти), 3≈5 (въстрьѕати ≈ въстръгнѫти), 3≈7 (възьмати ≈ възимати). On the competition between classes 3 and 5, as well as between 3 and 7, see details in Ch. 24, *Supplement*, § 910-913, *Excursus on the contamination and competition of paradigmatic classes*.

# §497. Irregular verbs

This class contains the following groups of irregular verbs: метати 3°, бьрати 3°\*, пьсати 3\*, пл҄ьвати 3h\*⩨.

# §498. A note on the метати 3° group

Many lexemes that belong here are poorly attested; some entirely lack the PRAE system. Table 498 (p. 294) shows the source data by family. Data from Sreznevskij's dictionary of Church Slavic are shown for comparison.


Table 498. Attestation of PRAE forms in the метати 3° group

"Yes" in the Prae columns means that the attested forms in the PRAE system lack substitutive softening; "no" means that such forms are not attested. Conversely, "yes" in the Prae" columns means that the attested forms in the PRAE system show substitutive softening; "no" means that such forms are not attested. The symbol ° after the prefix, as usual, indicates the verb with the corresponding prefix from the relevant family. Likewise, the symbol 0 shows the presence of a prefixless verb in the family.

Commentary and illustrations


ковати with root ‹398›. While not found in OCS, its sonant vocalism shows up in Church Slavic forms such as куѭ.

уръвати ‹773› is represented by a single gloss, юньць […] иже урвавъ сꙙ отъ привꙙзаньꙗ· и їзъ ограды излѣзъ Supr 565, 18–19. Večerka shows the present уръвѫ, уръвеши (likewise Sadnik).

ковати ‹398›: PRAE forms are represented by the hapax gloss ковома (Supr 161, 29).

основати ‹863›: PRAE forms are not attested.

In this grammar, PRAE with the ° effect is selected as canonical for these three verbs.

# нести 4 class

§499. Class membership

This class contains just under 350 lexemes. There are just under 100 distinct verbs. Possible right edges of the lexical component are shown in Table 499.

Table 499. Possible right edges of the lexical component in нести 4 class


Let us list all 4v class verbs: отъ.вѣти, °грѣти, на.дути, °знати, съ.мѣти 'dare' and из.мѣти, у.ныти, °спѣти, °сѣти, об.ути, по.чити and ис.по.чити, °чути.

§500. Notes on word formation

The lexical component ends in a root formative; otherwise only for звѧщи [звѧ.г.т.и]; see § 424. There is no expanded stem.

# §501. Connections with other classes

4≈7 correlations are possible (възмощи ≈ възмагати). On the competition between class 4 and 5 verbs, see Ch. 24, *Supplement*, § 910-913, *Excursus on the contamination and competition between paradigmatic classes*.

# §502. Irregular verbs

This class contains the following groups of irregular verbs: чисти 4c\*⤹, влѣщи 4c\*⤹, мрѣти 4h\*⤹, бити 4h\*⤸, крыти 4h\*⩨⤸, клѧти 4h, брати 4h•, плути 4h⤸, пѣти 4h⤸.

Verbs in the влѣщи 4c\*⤹ group (such are влѣщи ‹105› and брѣщи ‹58›) are not distinct from verbs in the чисти 4c\*⤹ group in their paradigmatic index. The difference between the paradigms is in the distribution of workstems; see details in Table 440, *Workstems of irregular verbs*. Note the substitution of the vocalic realization of the root in н-Part, ш-Part, and л-Part: вльченъ and брьженъ.

The verb пѣти 4h⤸ is not distinct from плути 4h⤸ verbs in its paradigmatic index. The difference between the paradigms is in the distribution of workstems; see details in Table 440, *Workstems of irregular verbs*.

# двигнѫти 5 class

### §503. Class membership

This class contains just under 150 lexemes. There are about 60 distinct verbs. Possible right edges of the lexical component are shown in Table 503.

Table 503. Possible right edges of the lexical component in двигнѫти 5 class


Verbs are called нѫ*-dropping* if they show root stem forms without the suffix н. For example, for the verb двигнѫти, the secondary form of the root aorist is двигъ. Such are all the verbs with a closed root. All other verbs are нѫ*keeping*. Such are обинѫти [об.(ви).н.ѫ.т.и] and по.винѫти, °дунѫти, зинѫти, на.кынѫти, по.манѫти, °минѫти, по.мѧнѫти and въз.по.мѧнѫти, въспланѫти [въз.(пла).н.ѫ.т.и], пл҄инѫти, °пл҄юнѫти, °ринѫти, исунѫти [из.(су).н.ѫ.т.и].

This class has a nonstandard marker of н-Part: ов.ен. Cf. дрьзновенаꙗ Supr 560, 9.

When a root-final labial combines with a suffix-initial н, the result of cluster simplification is not straightforward (see details in § 871, *On the combinations*  бн*,* пн *and* мн).

### §504. Notes on word formation

The marker нѫ is immediately preceded by the root, including in the opaque verbs гонезнѫти ‹185› and посагнѫти ‹698›.

# §505. Connections with other classes

Correlations 5≈7 are possible (ичезнѫти ≈ ичазати). On the competition between class 5 and 3 verbs, as well as between classes 5 and 4, see Ch. 24, *Supplement*, § 910-913, *Excursus on the contamination and competition between paradigmatic classes*. On aberrant forms by class 3 (cf. гыбл҄ѭ for гыбнѫти) and by class 4 (cf. усъпе for усънѫти) see § 609.

# миловати 6 class

# §506. Class membership

This class contains about 130 lexemes. There are slightly over 100 distinct verbs. The class marker is the twofold finally ambivalent suffix (ов/ев)/у. Possible right edges of the lexical component are shown in Table 506.

Table 506. Possible right edges of the lexical component in миловати 6 class


### §507. Notes on word formation

There are many denominals among the verbs of this class, cf. вѣровати, даровати, дрѧселовати, трѣбовати. Many denominal verbs contain nominal suffixes as part of the lexical component. Such are, for example, шьствовати [шьд.ств.ов.а.т.и], пришьльствовати [при.шьд.л.ьств.ов.а.т.и], ликъствовати [лик.ъств.ов.а.т.и], лѫкавьновати [лѫк.ав.ьн.ов.а.т.и].

# Commentary on individual verbal lexemes

### §508. List of commented lexemes of плакати 3 class

алъкати. On the aberrant spellouts of the root see § 807. Cf. ги· къгда тѧ видѣхомъ· лачѫща ꙇ натрухомъ Mt 25, 37: лачѫща in Zogr, but алчѫща in Mar, As, алъчѫща in Sav.

въЖѧдати . Aberrant spellout in Mar: вьжѧЖетъ Jn 4, 13.

въсльпати 3\* 'stream'. Contra Večerka, who gives the headword въслѣпати 3; two glosses only: щ-Part въслѣплѭщѧѩ Jn 4, 14 As and вьслѣплѫщѫѭ Jn 4, 14 Zogr

(cf. Church Slavic въслъпахомъ; see Vail- lant, § 191).

дъхати 3\*. Contra Večerka, who has no headword дъхати, but has духати 3 and дыхати 3/7. The PD gives the headwords дъхати 3\* and дыхати 7. The glosses of PRAE with base душ are assigned to дъхати in the PD, cf. дхъ ꙇдеже хощетъ душетъ Jn 3, 8 Zogr, Mar. The spellout дышетъ (same verse in As) is treated as aberrant spellout for душетъ in the PD (see § 613), while Večer- ka refers this gloss to the verb дыхати 3/7.

исходатаꙗти. Contra Večerka, who gives the headword исходатаꙗти according to class 7. Hapax gloss послѣдь же плодъ миренъ ꙇсходатаетъ Euch 88b, 15–16. The PD as- signs it to class 3 following Lunt (Lunt 1974, § 15.43).


Church Slavic texts, to the verb плакати<sup>2</sup> (analogous to Večerka's плакати2). Only one headwords in the PD, following Sadnik. Note that Vasmer suggests here two differ- ent etymologies (see Vasmer, *плакать* and *полоскать*).

поковати 3°. See § 498.


уръвати 3°. See § 498.

#### §509. List of commented lexemes of нести 4 class


сѩ. Večerka treats this gloss as erronous for измѣнитъ сѧ. See § 482.


ковати 3°. See § 498.

be treated as correlated to вѣꙗти 3, on a par with the correlation сѣти 4v/сѣꙗти 3. Nevertheless, in this grammar the roots ‹139› and ‹137› are kept separate, follow- ing Sadnik.

отърыти ‹773›. Contra Večerka, who has no headword отърыти. Hapax gloss отърьвена: тетъка боитъ сꙙ да не отърьвена бѫдетъ Supr 133, 11–12. Večerka assignes this gloss to the lexeme отъръвати (отъръвѫ, отъръвеши), and is treated as an erroneous spellout for отъръвана. This opinion seems shared also by Sadnik's dictionary attesting the infinitive отъръвати but not the infinitive отърыти. The postulation of the lexeme отъръвати is favored by the hapax gloss of the deverbal noun отъръваниѥ (дръжꙙщи же ѥго савиниза ризы· отъръваниꙗради отъ народаSupr 133, 9–10). However, in Church Slavic the deverbal noun отръвениѥ is attest- ed (помѧнѫ іерслимь дни […] ѿръвеніа своего *Book of lamentations* 1, 7, *Upyr' Lix- oj* quoted in Sreznevskij, 766; the mean- ing is 'exile'). Note that Vaillant (§ 204) inexplicably assigns the gloss отърьвена to the verb отъринѫти, while, at the same time, suggesting that the above-mentioned gloss отъръваниꙗ «sûrement il faut corri- ger […] en отъръвениꙗ». Instead, the form отъръвениꙗ certainly requires the infinitive

отърыти, as well as the participle отърьвена (Supr 133, 12).


Do not confuse the following pairs of lexemes:

дѫти, дъмѫ 4h ‹268› 'blow' ~ °дути, дуѭ 4v ‹259›: надути 'become pround' жѧти , жьн҄ѭ 4h• ‹302› 'reap' ~ °жѧти , жьмѫ 4h ‹290›: съжѧти 'compress' жрьти, жьр҄ѭ 4h• ‹292› 'sacrifice' ~ °жрѣти 4h\*⤹: пожрѣти, пожьрѫ ‹295› 'devour' грети [греб .т.и], гребѫ 4c ‹190› 'row' ~ грѣти , грѣѭ 4v ‹154› 'warm' съпѣти [съ.(пѣ).т.и], съпоѭ 4h⤸ ‹738› 'sing' ~ спѣти [(спѣ).т.и] 4v ‹872› 'succeed' съмѣти [съ.(мѣ).т.и], съмѣѭ 4v ‹586› 'dare' ~ смиꙗти, смѣѭ 3\* ‹858› 'laugh' °лити, лиѭ 4h\*⤸ ‹491›: пролити 'pour out' ~ лиꙗти, лѣѭ 3\* ‹491› 'pour' брати [бра.т.и], бор҄ѭ 4h• ‹42› 'make war' ~ бьрати [бьр.а.т.и], берѫ 3°\* ‹24› 'gather' клати [кла.т.и], кол҄ѭ 4h• ‹364› 'split' ~ класти [клад.т.и], кладѫ 4c ‹375› 'put' °чрѣти [чрѣп.т.и], чрьпѫ 4c\*⤹ ‹1081›: почрѣти 'scoop up' ~ чрьпати [чрьп.а.т.и], чрѣпл҄ѭ 3\* ‹1081› 'scoop'

обити2 [об.ви.т.и], обиѭ 4h\*⤸ ‹95› 'entwine' ~ обити1 [об.и.т.и], обидѫ 0 ‹329› 'surround'

#### §510. List of commented lexemes of двигнѫти 5 class

възньзнѫти ‹609›. Contra Večerka, who gives the following doubled headwords: възньзнѫти , възнисти; въньзнѫти , вънисти; проньзнѫти, пронисти; уньзнѫти, унисти. From among class 4 verbs, the PD

has only вънисти for the gloss Imv въньзи. With other prefixes there are only class 5 verbs. However, the PD treats glosses of the root aroist (cf. уньзѫ), as well as AOR and IMF forms by class 4 (cf. възньзъ, пронъзъшогочмоу) as secondary or aberrant for class 5 lexemes. Cf .: исплънь гжья ФЦЬТА ВЬЗНЬЗЪ НА ТОЬСТЬ НАПАВАШЕ I МК 15, 36 ZOGR, MAR, etc. See § 910-913, Excursus on the contamination and competition between paradigmatic classes. Cf. BBHL3HXTH, проньзняти, фуньзнжти.

въньзняти <609>. See above, възньзняти.

BBCN0MAHATH <550>. Aberrant spellouts with BBCHOMEH-; for distribution in sources see Večerka.

oTTBMINATH <770>. Spellouts with the root in the shape pour are not attested, contra Sadnik. Večerka mentions a spellout with ρъιг in Enina Apostle.

плинжти. Cf. плюняти; for distribution in sources see Večerka.

плоняти. See preceding word.


проньзняти <609>. See above, възньзняти.

съгъняти [съ.(гъб).н.ж.т.и] <210>. Večerka gives also the headwords съгъвати 7 and HETBEÂL for the hapax gloss from EUCH: простеръ на распонъ вся оудъ свот де сже НАШЪ' СЪГЪБАЛЪ ЕСИ ВСА НЕГЪЕЛА 35b, 1-3. In the PD both съгъвалъ and негъвля are treated as erroneous or morphologically strange; contra Večerka, the PD does not register the headwords съгъвати and негъвль.

ФУНЬЗНЯТИ <609>. See above, възньзняти.

§ 511. List of commented lexemes of MHARBATH 6 class

даровати. Aberrant даровая in SUPR 481, 22, cf. canonical gapora SUPR 535, 16 and elsewhere.

постовати wail' < 931>. Cf. сътовати lament, sorrow' < 932>. The distribution of roots follows Sadnik.

сътовати. See preceding word.

цѣловати. Contra Večerka, who gives two headwords: цѣловати 6 and цѣлъївати 7. The PD does not register the headword цълъвати; the forms цълънвати Mk 15, 18 AS and целъванятъ SUPR 542, 22 are treated as morphologically strange spellouts for ЦЕЛОВАТИ 6.

# дълати 7 class

### §512. Class membership

This class contains more than 750 lexemes. There are more than 300 distinct verbs. Possible right edges of the lexical component are shown in Table 512.

Table 512. Possible right edges of the lexical component in дълати 7


\* Such are the following 10 verbs: на.въщати, клацати, по.мрыцати; °двизати, \* ЖИБАТИ, ПО.МИЅАТИ, ПО.СТОИЅАТИ, ОСАՏАТИ, "ТАЅАТИ.

After morphophonologically soft consonants and vowels, the phonological opposition /a/ ~ / t/ is neutralized. In the morphophonological representation,

OYTObHEBATH 'get up early'. Aberrant spellout ютрън- 1× in PS SIN 126, 2.

in case of neutralization the spellout with а is adopted by convention;1 in all other cases, а or ѣ is used; cf. насыщати [на.сыщ.а.т.и], съмѣшати [съ.мѣш.а.т.и], but дѣлати and ицѣлѣти, cf. also питати and питѣти.2

### §513. Notes on word formation

A significant number of verbs show an open root closed by the suffixal prothetic consonant в; cf. подавати, помавати, познавати, убивати (cf. убиꙗти), поливати, почивати, одѣвати, припѣвати, успѣвати, съгрѣвати, насѣвати, унывати. A small number of verbs shows the prothetic consonant в after the suffixal formative ѣ (originally a verbal theme): оцѣпѣнѣвати, съдолѣвати, прѣдолѣвати, одолѣвати, измѫдрѣвати, разумѣвати, проразумѣвати, повелѣвати (see § 865, *A note on nominal stems and verbal platforms*). In other cases the root vowel is closed by the nominal suffix, cf. вѣньчати, продльжати, вльн҄ꙗти, дѣлати, слушати.

### §514. Connections with other classes

The following correlations are possible: 7≈1, 7≈2, 7≈3, 7≈4, 7≈5. On the competition of classes 7 and 1, as well as 7 and 3, see Ch. 24, *Supplement*, § 910–913, *Excursus on the contamination and competition between paradigmatic classes*.

Many verbs ending in -а.т.и form correlative pairs of the form "starting verb ≈ secondary imperfective verb", where the secondary imperfective verb belongs to class 7 and its correlate starting verb belongs to a lesser-numbered class. Such are, for example, the following. Class 1 verbs: ꙗзвити≈ꙗзвл҄ꙗти, побѣдити≈побѣЖати. Class 2 verbs: съгорѣти≈съгарати, назьрѣти≈назирати. Class 3 verbs: избьрати≈избирати, напьсати≈написати, възьмати≈възимати. Class 4 verbs: възмощи≈възмагати, пропѧти≈пропинати. Class 5 verbs: избѣгнѫти≈избѣгати, ичезнѫти≈ичазати, растръгнѫти≈растръгати. As can be seen from these examples, class 7 verbs often show the lengthened grade of the vocalism compared to the correlate of the lesser-numbered class, and, in some cases, the substitutively softened variant of the truncated stem.

# Commentary on individual verbal lexemes

### §515. List of commented lexemes of дѣлати 7 class

авл҄ꙗти. Cf. the doublet lexeme ꙗвл҄ꙗти; for distribution in sources see Večerka. ꙗвл҄ꙗти. See preceding word.


напитѣти. See preceding word.


съмотр҄ꙗти. See preceding word.

# CHAPTER 21 **Unique verbs**

§516. General table of unique verb profiles

Tables 516.1–3 (pp. 304–305) show the profiles of all unique verbs and the paragraphs where the corresponding verb is examined in more detail. For reference, the inventory of profile forms is shown below:


Table 516.1, along with the main profiles, shows the forms of the Prae bundle, because these forms cannot be built using only profile forms.

Table 516.2, along with the main profiles, shows the forms of the Prae bundle for 'to be', and the additional bundles (Cond and Imf-Aor) for the verb быти, because these forms cannot be built using only profile forms.

An m-dash (—) shows that there are no forms that permit the reconstruction of the corresponding subparadigm.

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9


Table 516.1. Profiles and Prae of the unique verbs дати, ꙗсти, вѣдѣти, имѣти

Table 516.2. Profiles of the unique verbs ѥсмь 'to be' and быти


Table 516.3. Profiles of other unique verbs



### Table 516.3 (continued). Profiles of other unique verbs

# The verb дати

### §517. The main profile

Here is the profile of the main forms:


### § 518. Addendum to the main profile

Here are the paradigms of the personal forms of Prae and Imv:


### § 519. Morphology of the forms 1

1. The root <216> shows up in two shapes in the paradigm of the verb дати: C-final дад and V-final да. Only the truncated stem occurs in the paradigm. The distribution of C-final and V-final stems is nonstandard. Forms of the PRAE system show a mix of i- and e-conjugations. Anomalous forms of the PRAE system represent the so-called athematic present.

2. Morphologically, imperfect дадъахъ is a present imperfect (cf. дадать), while дагаахъ is infinitival. The latter coincides with the Imf of the verb длати 3.

3. In addition to the standard 2-3Sg aorist of the да type, old aorists like дастъ also occur (see § 480).

### §520. Family membership

There are 9 lexemes altogether. These are дати, въздати, въздати, издати, ОТЪДАТИ, ПОДАТИ, ПОИДАТИ, ПРОДАТИ, ПОЕДАТИ.

Cf. also дагати 3, °давати 7.

### §521. Illustrations

PRAE - рече црь дъвици емоуже аште хоштеши дамь ти Mk 6, 22 zogr ; ХОШТЖ ДА МИ ДАСН ОУСЕЧЕНЖ НА БЛЮДЪ ГЛАВЖ: СОАНА КОЬСТИТЕЙ МК 6, 25 ZOGR; что мн дасте и азъ вамъ пръдалъ і Mt 26, 15 sav; и авне принесоша ємор. ославена на одръ леждиа: видѣвъ же іс въря ихъ: рече ославеньмом надъ са ЧАДО ОТЪДАДАТЪ ТИ СА ГОЕСИ Mt 9, 2 SAV; ВЪЗДАЖДЬ МИ РАДОСТЪ СПИЪ ТВОЕГО" ї думъ владъючьнемъ очтвръді мы PS SIN 50, 14; не дадите стаго псомъ: нн помътаите висьов вашихъ предъ свиньеми Mt 7, 6 мая; и гако аште кто ВИДИТЪ МАТЕОЕ ЧАДОЛЮВЖ МАТЕКА НЕ ИМЖШТИ: И СВОЕ ЧАДО ВИДАШТИ ПЛАЧЖШТЕ: И ПОДАДЖШТЪ СЪСЪ' ДА НЕ НАДОН ДЪТНШТА" НЪ ДА ОТЪ ПЛАЧА ОУСТАВИТЪ' ТАКО же и насъ видите supr 384, 18-23.

IMF - в не дадеаше никомоуже мило нести" съсждъ сквозъ цръковь Mk 11, 16 ZOGR; NE JANEALIE IN'S MATH Lk 4, 41 ZOGR, but HE JATEWE in MAR; cf. also: такожде и ъкоже бъстъ вь дьни лотовъ: ъдъях и питухъ: коуповаахж и продавахъ: саждаахж зъдаауж Lk 17, 28 мая; можаше во се миро продано БЪГТИ НА МЬНОЗЪ" ↓ ДАНО БЪГТИ НИЦИИМЪ Mt 26, 9 МАR; БЪ ЖЕ КАНАФА" ДАВЪНИ съвътъ июдеомъ вко очнъе сстъ единомог чакоу очиьръти за люди Jn 18, 14 AS; ГАКО СЕ МАТИ ЧАДОЛЮБА" ПОДАВЪШИ СЪСЪ МЛАДЕНЬЦОУ" ВЕСЕЛИТЪ СА ДЪТНШТУ ПОНВЛАЧАШТОУ МАКЪКЖ ПИШТЯ МАТЕКА SUPR 312, 3-6.

INF-AOR - BEATEAWE EO EKO 3ABHCTH DAZH HOTEAAWA H Mt 27, 18 ZOGR ; ДЪВА ДЛЪЖЬНИКА БЪСТЕ ЗАНМОДАВЪЦОУ ЕТЕРОУ · ЕДИНЪ БЪ ДЛЪЖЕНЪ ПАТНИЖ СОТЪ ДИНАОЬ' А ДООУТЪП ПАТЪК ДЕСАТЪ' НЕ ИМЖЦИЕМА ЖЕ ИМА ВЪЗДАТИ ОБЪЗДАТИ ОБЪЗДАТИ ИМА отъда Lk 7, 41-42 мак (likewise ZOGR, SAV, cf. объма отъдастъ Lk 7, 42 AS); Възалъкахъ во са и дасте ми псти: въждадауъ са и напосте ма Mt 25, 35

On the distribution of the forms, see Vaillant, § 222.

Sav; юже двьри затворены сѫтъ· ꙇ дѣти моѩ съ мъноѭ на ложи сѫтъ· не могѫ въстати датъ тебѣ Lk 11, 7 Zogr; аще не бы былъ злодѣі не быхомъ его прѣдали тебѣ Jn 18, 30 Sav.

# The verb ꙗсти

§522. The main profile

Here is the profile of the main forms:


# §523. Addendum to the main profile

Here are the paradigms of the personal Prae and Imv forms:


§524. The morphology of the forms2

1. The root ‹1113› has two shapes in the paradigm of the verb ꙗсти: C-final ꙗд and V-final ꙗ. The paradigm of the verb only shows the truncated stem. The distribution of the C-final and V-final stems is *nonstandard*. The forms of the PRAE system are a mix of *i*- and *e*-conjugation. Anomalous PRAE system forms represent the so-called athematic present.

2. Along with the standard 2–3Sg aorist of the ꙗ type, there is the old aorist of the ꙗстъ type (see § 480). In other forms, alongside the standard aorist of the ꙗхъ type, there is the sigmatic aorist of the ꙗсъ type (see § 478).

3. The canonical м-Part is ꙗдомъ (cf. Supr 396, 5). However, there is a compound звѣроꙗдимъ, once in Supr: воинъ ушидь храбърыи· плѣньникъ овьчꙙ хво· звѣроꙗдимо Supr 93, 11–12.

4. On the aberrant forms of the щ-Part, see § 622.

2 On the distribution of the forms, see Vaillant, § 223.

5. Alongside the w-Part a gb, there is the adjectival formation 14.118. (cf. the compounds плътогадивъ, кръвогадивъ: сне сотонинъ: объщтъниче нечистъютъ Бъсовъ ЕЕСЪНЪК ПЪСЕ: КЪВВОПИВЪИ ЗМИЮ. ПЛЪТОВДИВЪМ МЕЧОУ. СВЕРЪПЪН ЗВЕДНИ: НЕ СТРЫДНШИ ЛИ СА ЖЪРА КАМЕННЮ ПОВДЪ НАРОДОМЪ SUPR 115, 26-30; АЕНІЄ ЖЕ АКЪІ ЛЬВИ КОВВОГАДИВНИ' ОУСТОВМИША СА НА ЙВ СКРЪЖЬШТЪЖШТЕ ЗЖЕЋІ SUPR 216, 22-24).

#### §525. Family membership

There are 5 lexemes in total. These are actry, orterry, orterry, and сънести.3

### § 526. Illustrations

PRAE - къде єстъ обить деже пасуж съ одченнкъ моглин сънъмь Мк 14, 14 ZOGR; къде уоштеши шедъше оуготовимъ да вси пасуж Mk 14, 12 мАR; НЕ ПЦЕТЕ СА ОУБО ГЛЕЖШТЕ: ЧЬТО ЕМЪ ЛИ ЧЬТО ПИЕМЪ: ЛИ ЧИМЬ ОДЕЖДЕМЪ САТ всъуъ во снуъ мзъщи штятъ Mt 6, 31-32 zogr; и глаголаша кмоу паждъ се во третиї день юсть понеже ничьсоже въкомси SUPR 19, 9-11; вдящемъ же ИМЪ: ПОНИМЬ ИС ХЛЕБЪ И БАГСШТЪ: ПРЕЛОМИ И ДАВШЕ ОУЧЕНИКОМЪ СВОНМЪ: 1 ОЄЧЕ ПОНИМЪТЕ ЕДИТЕ СЕ ЕСТЪ ТЪЛО МОЕ Mt 26, 26 МАR; ВДЪ ПЛЪТЪ МОНА И ПЊАИ кръв моня: въ мнѣ пръстываетъ и азь въ немъ Jn 6, 56 AS; ъкоже во въехух: ВЪ ДЫН ПОВЖДЕ ПОТОПА: ЪДЖШТЕ І ПИНЖШТЕ: ЖЕНАШТЕ СА І ПОСАГАНЖШТЕ: ДО Негоже дьне вынде ное въ ковиегъ Mt 24, 38 ZOGR.

IMF - WKAE(BE) ТАВАНЖШТАГО ТАІ ИСКОВНЪГО СВОЕГО" СЕГО ВЪИГЪНАХЪ" ГОЪДОМЬ окомь і несъітьномь срдцемь съ симь не вдвахъ PS SIN 100, S; чловъкъ нъкто БЪ БОГАТЪ: ИМЪА СКОТЪГ И ЧОЕДЪ! МНОГЪ!' ЧЛОВЪКЪ ЖЕ ПАКЪІ БЪ ОУБОГЪ' ИМЪ ТЕЛИЦИ ЕДНИХ: ТАЖЕ НА ТРЕПЕЗЪ СЬ ЯНИЪ ГАДЪАШЕ: И ЧАШЖ ЈЕДИНЯ СЪ нимь пигааше SUPR 359, 27-360, 1; егда же снъ твом сь изъдъ твоє нлувне съ любодьнцами приде и закулалъ емор есі телець пітомъ Lk 15, 30 as; н МОБТВЪНИИ ВРАТЪР ИЗВАЪК'ШЕ ТЪЛЕСА СТВИНУЪ: ПОВОЪГОША НА ВЪСТОЧЬНЪИНУЪ странахъ гакоже изъденомъ бъюти имъ отъ пьсъ SUPR 538, 24-26.

INF-AOR - тогда начьнете глати тахомъ поведъ товом и пихолуъ и на распяттихъ нашихъ огчилъ еси; и речетъ гла валиъ; не въдъ васъ отъкждоу ECTE OTTECTMINTE OTT MENE BCH ДЕЛАТЕЛЕ НЕПОАВЬДЬНИI Lk 13, 26-27 SAV; WIJH НАШИ ЪША МАННА ВЪ ПОУСТВИНИ: ЕКОЖЕ ЕСТЪ ПСАНО: ХЛЪБЪ СЪ НЕССЕ ДАСТЪ IN'S ECTH Jn 6, 31 ZOGR; ПЕПЕЛЪ АКЪІ ХЛЪВЪ ГАХЪ SUPR 297, 25 (cf. ПОПЕЛЪ ЪКО УЛЪБЪ ЪСЪ РЅ SIN 101, 10); ТОИ СМОКВИ ООДИ' АЖЕ ВЪЗЪМЪ СТАРЦЪ И СЪ СЛЪЗАМИ ОБЛОБЪ!ЗАВЪ: ИЗЪ SUPR 300, 24-25; W30EA I ВЕПОЬ ОТЪ ЛЖГА: "ПОКЪ ДИВЬЕ" ПОВЛЪ ЕСТЪ PS SIN 79, 14; L БЪКТЪ ЕГДА ВЪНИДЕ ИСЪ ВЪ ДОМЪ ЕДИНОГО КЪНАВА фарисънска въ совотъ хлъва встъ: и ти въвахъ назираняште и Lk 14, 1 мак.

<sup>3</sup> Večerka also lists overve (once in Hilandar folios).

# The verb вѣдѣти

### §527. The main profile

Here is the profile of the main forms:


### § 528. Addendum to the main profile

Here are the paradigms of the personal Prae and Imv forms:


§529. The morphology of the forms4

1. The root <140> in the paradigm of the verb въду и shows only a C-final shape. The paradigm has two stems, a V-final one (the expanded stem BEA.B), and a C-final one (the truncated stem BBA). The expanded stem occurs in the IMF and INF-AOR systems. The PRAE system forms show a mix of i- and e-conjugations. Anomalous PRAE system forms represent the so-called athematic present.

2. The canonical M-Part is BEAMS (cf. for example, SUPR 504, 30), but cf. BEAMMB and compounds: 00/8/2/2/2015 господа: истинъною слово въдимов SUPR 212, 29-30; W CBTTE HEHCHOB'EДHMBIH SUPR 426, 18-19; H HE OTTOMEWITZ CA СВАТААГО ДОУХА: НЪ ТОЕПЕШТЪ ИГО КОВПОСТИ: И НЕИСПОВЪДИМЖЕЖ СИЛЖ: ПОНЯ И СЛАВЫЖ ВЪННЖ SUPR 115, 17-20; ДА ЗАПРЪТИТЪ ТИ ГЪ ДИЪВОЛЕ БЕЗНАЧАЛЬНЪГ и невидимъ сжщьствомъ: и недовъдимън силож EUCH S6b, 11-14.

3. On aberrant шт-Part forms see § 622.

### §530. Family membership

There are 12 lexemes in total. These are въдети, заповъдъти, извъдувти, ИСПОВЪДЪТИ, НАВЪДЪТИ, НЕДОВЪДЪТИ, ПОВЪДЪТИ, ПОВЪДЪТИ, ПОООУВЪДЪТИ, съвъдъти, съповъдъти, очвъдъти.

Cf. also °въдати 7, °въдовати 6.

4 On the distribution of the forms, see Vaillant, § 224.

### §531. Illustrations

PRAE — идѣмъ въ ближьнѧѩ въси и градъи· да и ту проповѣдѣ· на се бо прид Mk 1, 38 As; онъ же отъвръже сѧ глѧ· не умѣѭ ни съвѣмь что ты глши Mk 14, 68 Mar; тъгда увѣси въ часъ бѣды твоѥѧ· ꙗко естъ богъ на небесехъ иже тобоѭ хулимыи Supr 195, 14–17; вѣстъ бо отецъ вашъ ихъже трѣбуете· прѣЖе прошениѣ вашего Mt 6, 8 Mar; отъвѣщасте же имъ родителѣ его и рѣсте· вѣвѣ ѣко сь естъ снъ наю· ꙇ ѣко слѣпъ сѧ роди· како же нынѣ видитъ не вѣвѣ· ли кто ему отвръзе очи вѣ не вѣвѣ· самого въпросите· въздрастъ иматъ· самъ о себѣ да глетъ Jn 9, 20–21 Mar; да увѣсте ѣко власть ꙇматъ снъ чскы на зем҄и· отъпущати грѣхы Mt 9, 6 Zogr; си же естъ жизнь вѣчънаꙗ· да съвѣдѧтъ тебе истиньна ба· и егоже посъла іс ҃ха Jn 17, 3 Sav 26, 1–4; вышънее слово слыші· ꙇ въспоі· слыші и прославі· слыші и проповѣЖь· бжіѣ вельѣ чюдеса Cloz 13a, 34–37; възвеселите сѩ праведьні о гі· їсповѣдіте памѩть стынѩ его Ps Sin 96, 12; ꙇродъ бо боѣаше сѧ иоана· вѣды и мѫжа праведъна и ста Mk 6, 20 Mar; рече къ другомъ своимъ· кыѧ мѫкы нанесѫ на зълосъмрьтънааго сего не вѣдѣ· онѣмъ же недовѣдѫщемъ· и не отъвѣщаѭщемъ ничсоже Supr 111, 27–30; горе вамъ кънижьници ꙇ фарисѣі· ꙇ лицемѣри· ꙇЖе есте ако ꙇ гроби невѣдоми· ꙇ чци ходѧщеі връху не видѧтъ Lk 11, 44 Zogr; ты небесьныи и земьныи· ты не видимыи ї видомъ· ты не вѣдомы и вѣдомъ быстъ мене дѣл҄ꙗ Supr 504, 28–30.

IMF — не у бо вѣдѣахѫ кьнігь· ѣко подобаетъ ему от мрътвъиихъ въскрснѫті Jn 20, 9 As; прѣЖе бо вѣдѣаше ꙗко сему ѥстъ быти· и кон҄ꙙ оседъланы уготова Supr 51, 1–3; и отътоли увѣдѣнъ быстъ родъ їѡанновъ и епискупьство Supr 295, 5–6; сътворивъ же четыри десꙙти дьнии отиде чистъ къ господу· проувѣдѣвъ прѣЖе трии дьнии· съконьчаниѥ своѥ Supr 168, 15–18.

INF-AOR — ты заповѣдѣ заповѣди твоѩ· хранити ѩ ѕѣло Ps Sin 118, 4; бже ушіма наші(ма) услышахомъ· оці наші повѣдѣшѩ намъ Ps Sin 43, 2; посъла мѧ· ꙇсцѣлитъ съкрушеныѩ срдцемь· проповѣдѣтъ плѣньникомъ отъпущение· ꙇ слѣпымъ прозьрѣние· отъпуститъ съкрушеныѩ въ отърадѫ· проповѣдѣти лѣто гне приѩто Lk 4, 18–19 Mar; ослѣпі бо імъ умъ· хотѧ ѣвіті своѭ сілѫ· ꙇ хотѧ сътворіті· да бѫ прѣсталі· отъ зълобы своеѩ· ꙇ да бѫ увѣдѣлі· ѣко невъзможънаѣ начінаѭтъ Cloz 5a, 12–16; быстъ нѣчто сіце невѣръныимъ не извѣсто· вѣрьныимъ же и зѣло не извѣсто Supr 210, 21–23.

# The verb имѣти

§532. The main profile

Here is the profile of the main forms:


### § 533. Addendum to the main profile

Here is the paradigm of the personal Prae forms:


§ 534. The morphology of the forms>

1. The root <334> in the paradigm of the verb nutsTh shows only a C-final shape. The paradigm has two stems: a V-final one (the expanded stem HM.h), and a C-final one (the truncated stem им). The distribution of the C-final and V-final stems is nonstandard. Anomalous PRAE system forms represent the socalled athematic present.

2. Note that 3PIPrae, шт-Part, and the personal Imf forms coincide with the forms of the verb with 4h, cf. им=жтъ and ibM=ятъ [имять]. However, the Imf forms of nattru show the expanded stem (unt=ax etc.), while the forms of wTH show the truncated stem (HM= BAY'L etc.).

3. 3PI Prae and wr-Part forms have aberrant alternatives by class 7. These are: имъжтъ for имятъ: имъжтъ моусиня и пророкъ SUPR 351, 29 (cf. ИМЖТЪ МОСЕА И ПОКЪІ ДА ПОСЛОУШАЊУТЪ НУВ LK 16, 29 МАВ); ДА ИМЪНИЯТЪ ПАМАТЪ ЖИТИПА СВОКА ЗЕМЬА SUPR 418, 26-27 (cf. CLOZ 7a, 36-37 да имятъ ПАМАТЬ ДОМАШЪНЪГО ЖІТЬЪ); ИМЪНЖШТ- FOr ИЛЖШТ-: ПКОЖЕ АШТЄ К'ТО ВИДИТ' ВъДро ЗЛАТО ИМЪНЖШТЕ ЖТОВ КАМЕНИЕ ЧЫСТЪНОЮ. НОСАШТЕ НА СЕБЪ СИ ТАННА НАПИСАНА: ХОШТЕТЪ ВЪДО ВИДЕТИ: БЛЫШТАННА ДЕЛЬМА: СТАВЫЖЕТЪ СА ПАКЪГ О ЖТОБНИМЪ: КАМЕНИИ: ТЪШТИТЪ БО СА И КАМЕНЬНЯЯ ВИДЪТИ ДОБОФТЯ: И IEWE HA REME ПИСАНО ТАННОЕ ОУВЪДЕТИ SUPR 344, 3-10; HANTERS FOR HANDI: NE T'B БО ЕСТЪ ТОВПЕЛИВЪ ИЖЕ ТОВЕОВАНИИ НЕДОИМАТЪ: НЪ ИМЪЖИ ОБИЛИКЪ ТИ ВЪ страдании трьпа SUPR 92, 28-30; азъ члекъ есмъ подъ властелъв оччиненъ INTEM INDA CORON BOLA Lk 7, 8 MAR (in the same verse HNTSM also in ZOGR, AS and SAV), cf. a3' чавкъ єсмь подъ влком: чиві подъ совом вонны Mt 8, 9 м AR (in the same verse имъ also in ZOGR, AS and SAV).

On the attestation of the forms, see Vaillant, § 225.

### §535. Family membership

There are 2 lexemes. These are имѣти and недонмъти. Cf. also hmath 3\* [bm.a.T.h], tath 4h, °HMATH 7 [HM.a.T.H].

### § 536. A note

The verb หมายาท plus infinitive creates a construction with deontic modality or futurity. Cf. cha so vekbl imat" no barn car bo price vout Lk 9, 44 zogR, ВЪРОУЊИ ВЪ МА: НЕ ИМАТЪ ВЪЖДАДАТИ СА НИКОГДАЖЕ JΩ 6, 35 ZOGR; К ТО ИМАТЪ БА ВИННА СЪТВОРИТИ SUPR 432, 19.

### § 537. Illustrations

PRAE - LMAML EO ПАТЪ БРАТЪНЬ: 'ЕКО ДА ЗАСЪВ'ЕД'ЕТЕЛЬСТВОУ ОУТЪ ИМЪ' ДА не и ти приджтъ на мъсто се мячъное гла же емоу авраамъ имятъ моста н пркъі да послоушаютъ ихъ Lk 16, 28-29 мая; лисі "Езвінъ имять" и пьтица НЕСЬНЪНЬА ГНЪЗДА: А СНЪ ЧАЧЪ НЕ ИМАТЪ КЪДЕ ГЛАВЪН ПОДЪКЛОНІТ! Mt 8, 20 АS; АЦИЕ БО ВЪЗЛЮБИТЕ ЛЮБАЦИУЪ ВАСЪ' КЖИХ МЪЗДЖ ИМАТЕ' И МЪГТАОН ТАКОЖДЕ творатъ Mt 5, 46 Sav; оуста їмять ї не възглежтъ: очи їмять ї не видьють ወሃши їмятъ ї не слъщнять ноздри имять и не овомъжтъ: рящъ їмя тъ ї пе оснажить ноять їмять ї не поїджт ps sin 113, 13-15; имѣить върла вжиж Mk 11, 22 MAR; LNARDTRANDY EO BLOGE ANNO EXAFT'S I H3ERAJETT'S A OTT'S HE ИМЖШТААГО' І ЕЖЕ АШТЕ МНИТЪ СА ИЛЪГ ВЪЗАТО БЯДЕТЪ ОТЪ НЕГО Mt 25, 29 ZOGR.

IMF - вьси во имеах иоана тько прка Mk 11, 32 мАR; видъшл БЪСЪНОВАВЪШААГО СА: СЪДАЩА ОБЛЬЧЕНА: 1 СЪМЪКЛАШТА ИМЪВЪШААГО ЛЕЋЕОНЪ' L OYEO BUA CA Mk S, 15 MAR.

INF-AOR - все елико имаши продаждъ: и даждъ нициииъ: и имъти имаши съкровнице на несув Lk 18, 22 SAV; вко помняк слово стое свое еже имъ къ авраамог равоу своемоу PS SIN 104, 42; пать во мжжь имъла єси' і нувін'є єгожє IMAWH HECT'S TH MAND Jn 4, 18 ZOGR.

The verb ഭ്രോപ

### ട്ട 538. The profile

Here are the personal forms of Prae and шт-Part:


#### § 539. A note on defectivity

In most grammars, the verb ​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ verb въти <16>, with a distinction in the present system between imperfective 'to be' forms ( комь, иск) and perfective гъгти forms ( кладяши). The present grammar separates these verbs into two independent lexemes. The lexeme ​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​

#### § 540. The morphology of the forms

1. The root <276> is anomalous, with two shapes: c and ie, both occurring in the Prae paradigm. 3PPrae and шт-Part are formed from the C-final stem following the class 4c standard. Anomalous Prae forms represent the so-called athematic present.

2. On the aberrant wr-Part forms, see § 622.

#### §541. Family membership

There are no prefixed forms.

Concatenation with the particle ne in Prae forms that begin with is-, gives the + ecmb > HECMb etc. 3Pl usually shows ne carry; a single aberrant spellout N'E CRT'S In AS: HNI ЖЕ ГЛААХХ СІН ГЛІ Н'В СЖТ"В Б'ЕСЪНОУНЯШААГО СА: ЕДА Б'ЄСЪ слепънимъ очи отврести можетъ Jn 10, 21.

#### § 542. Illustrations

Prae - въ глете ма оучитель и га и довов глете сомь во Jn 13, 13 SAV; ГИ ВЪДЪКЪ ТА ЕКО ЖЕСТОКЪ ЕСИ ЧАКЪ: ЖЬНА ИДЕЖЕ НЕСИ СЪЛЪ Mt 25, 24 МАR; друже на неже єси пришълъ' то створи Mt 26, 50 SAV; нли нѣстъ ми лъть' СЪТВООГТИ ЕЖЕ ХОШТЖ ВЪ СВОНУЪ МИ' АЩЕ ОКО ТВОЕ ЛЖКАВО ЕСТЪ' НЪ АЗЬ ЕЛАГЪ ЕСМЬ Mt 20, 15 AS; ВАМЪ ЕСТЪ ВЪДЪТИ ТАІНЪІ ЦОСТВИГА НЕСКАГО' А ПРОЧИИЪ ВЪ притъчахъ Lk 8, 10 SAV; г рече одченикомъ своимъ: да естъ при немъ ладница Mk 3, 9 MAR; CBATAA WE DEKOCTA ΓΛΑΓΟΛΑχΟΒ' ΤΗ ΜΗΟΓΑΙΙΔΗ: ΓΑΚΟ ΚρΣΟΥΠΟΝΑ IECB'S II B'S XOHCTOCA HMAB'S HAZEWAR E'SCOM'S WE HEYHCT'SIMN'S HE ПОКЛОНИВЪ СА: НИ ЕОГОМЪ ТВОНМЪ СЛОУЖИВЪ SUPR 181, 11-16; ОНЪ ЖЕ ГЛАШЕ ИМЪ НЪСТЕ ЛИ ЧЬЛИ НИКОЛИЖЕ ЧТО СТВОРИ ДДЪ ЕГДА ТЪВЕОВА И ВЪЗАЛЪКА САМЪ И ЇЖЕ ЕВУЖ съ нимь Mk 2, 25 Sav; не дьвъ ли на десате годинъ єстє въ дьни: аштє кто ХОДИТЪ ВЪ ДЬНЕ НЕ ПОТЪКНЕТЪ СА: "ЕКО СВЪТЪ МИРА СЕГО ВИДИТЪ Jn 11, 9 МАR; ЧЬТО СЖТЪ СЛОВЕСА СИ' О НИУЪЖЕ СЪТАЗАЕТА СА КЪ СЕБЪ LДЖШТА: L ЕСТА ДАЖЕЛА Lk 24, 17 ZOGR; tko moi czTT'B BBCI 3B'B01 AMBOABBHII PS SIN 49, 10; ENAKEHH ECTE ЕГДА ПОНОСАТЪ ВАМЪ Mt 5, 11 ZOGR; ЧАДО ТЪГ ВЪСЕГДА СЪ МЬНОЕЖ ЕСИ" и ВЪСЪ MOT TBOT CATT'S Lk 15, 31 MAR.

шт-Part - свътелъ ссточьникъ. сжуми. присно съ съ оцемь висти 63а, 17-19; ТЪ БО ИСПОЬВА НЪР СЪТВОРИЛЪ' ОТЪ НЕ СЖШТААГО ВЪ СЖШТЕЕ ПОНВЕДЪ SUPR 479, 27-29; същ на селъ да не възвратитъ са въспать MK 13, 16 мак.

# The verb гъти

### § 543. The main profile

Here is the profile of the main forms:


§ 544. Addendum to the main profile: aorist and imperfect

The verb Eurn has three bundles that can be identified with the subparadigms of the personal forms of the aorist (Aor) and of the personal forms of the imperfect (Imf). The A and B bundles are the primary imperfect and aorist forms; the C bundle contains secondary forms.




§ 545. Morphology of the imperfect and aorist forms

1. The A bundle has an anomalous stem shape. The forms \* E' ax b, \* E' ax or , \*Бъахомъ, \* Бъашета from the A bundle are not attested (cf. Vaillant, § 158, 221).

2. The forms EtyB, EEYOB'E, and ETEXON'S from the C bundle can be interpreted either as aorists from the aberrant stem ETS, or as contracted imperfects from the stem E. Likewise, E'ETTA and E'ECTE can be interpreted as aberrant replacements of \* E hamera and E bauere. Paradigmatically, A bundle forms are treated as primary Imf forms of the verb ETTH, while B bundle forms as secondary Aor forms of вътти and corresponding prefixal verbs, and C bundle forms as secondary with respect to bundle A or bundle B forms. There are aberrant contracted imperfect forms corresponding to bundle A forms (in some cases coinciding with bundle C forms), cf. ETEWE SAV Jn 12, 2, ETKX SAV Lk 8, 45.

3. Bundles A and C are not attested in prefixal verbs.

4. Note the aberrant 3Du form бѣаста in Supr for canonical бѣашете.

5. Alongside the standard 2–3SgAor form бы, there is the old Aor form быстъ (see § 480).

§546. Addendum to the main profile: conditional

Here are all the attested forms (Vaillant, § 172):


The forms ∇бихомъ and ∇бишѧ show aorist terminals (cf. the B bundle above) with the secondary stem би (alien terminals effect). Conditional forms enter into irrealis constructions with л-participles of the shape conditional + л-Part (see illustrations below in § 549).

### §547. Family membership

There are 6 lexemes altogether. These are быти, забыти, избыти, прибыти, прѣбыти, събыти.

### §548. Commentary on individual lexemes

1°. Most grammars unify the verb быти ‹16› into a single paradigm with the verb ѥсмь ‹276›, with a distinction in the present system between imperfective 'to be' forms (ѥсмь, ѥси) and perfective быти forms (бѫдѫ, бѫдеши). This is the case in Večerka and Sadnik. In the present grammar the corresponding forms are distributed between two independent lexemes.

2°. The prefixless verb быти is *sui generis* in that it is the only one with secondary forms; however, the secondary 2–3SgAor form with the terminal стъ is possible for prefixed verbs as well.

3°. The verb забыти shows an anomalous н-Part: ∇забъвенъ. Derivatives show both an н-Part stem, cf. забъвениѥ, and a т-Part stem, cf. забытиѥ, забыть. For other verbs, н-Part is not attested, while т-Part is attested in derivatives, cf. бытиѥ, избытиѥ, избытъкъ, побыть, прибытъкъ.

### §549. Illustrations

PRAE: аще бѫдетъ етеру члвку ·р· овецъ· ꙇ заблѫдитъ едина отъ нихъ· не оставитъ ли девѧти десѧтъ и девѧти на горахъ· и шедъ ищетъ заблѫЖьшѧѩ Mt 18, 12 Mar; да свѩты твоі въсѫдъ пріемлѭце достоіні бѫдемъ очішчениѣ твоегоKiev 3a, 11–13; молѧще же сѧ не лихо глѣте· ѣкоже 1 1633146481411111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 Mt 6, 7 ZOGR; БЖДИ ЖЕ СЛОВО ВАШЕ: ЕН: ЕН: 1 НИ НИ: ЛИХОЕ БО СЕЮ ОТЪ НЕПОНЪЗНИ ЄСТЪ Mt 5, 37 ZOGR; НЕ ЗАБЯДІ ОУБОГЪНУВ ТВОІХЪ ДО КОНЪЦА PS SIN 9, 33; ЕГДА ПОСТИТЕ СА: НЕ БЖДЕТЕ ЋКО УПОКОИТИ: СЕТОУНЖШТЕ: ПРОСМОАЖДАНЖШТЕ БО ЛИЦА СВОЕ: ДА БИША СА АВИЛИ ЧКОМЪ ПОСТАШТЕ Mt 6, 16 ZOGR; А НЖЕ ОЕЧЕТЪ НА ДХЪ стъпне отъпоуститъ са емор" ни въ сь въкъ ни въ къджции Mt 12, 32 мая; ЕГДА ЖЕ ОУЕО ОНИ ПОДВИЗААХХ СА: ОНЪ ЖЕ БИЮДЪЛШЕ: СЪВЖДЖШТЕЕ СА ВИДЪ ВИДЕНИЕ ДИВЬНО СИЛЪ НЕКЪГА СЪ НЕЕСЪ СЪ НЕЕСЪ СЪХОДАШТА SUPR 92, 30-93, 3.

Cond: подобааше ти офео въдати сърещо моє тръжъникомъ' и пришьдъ АЗЪ ВЪЗАЛЪ ОУБО БИМЬ СВОЕ СЪ ЛИУВОЈЖ Mt 25, 27 МАR; ВКО ВШТЕ НЕ ЗАКОНЪ ТВОЇ ПООУЧЕННЕ МОЕ ЄСТЪ' ТОГДА ОУЕО ПОГЪВЛЪ БИЛЬ ВО СЪМЪРЕНИЇ МОЄМЬ РЅ SIN 118, 92; ЪКО АШТЕ НЕ ГЪ БИ БЪЛЪ ВЪ НАСЪ' ВЪНЕГДА ВЪСТАТИ ЧАВКОМЪ НА НЪГ ОУВО ЖИВЪЇ ПОЖОВЛІ НЪЇ ВЪШЊА PS SIN 123, 2-3; АШТЕ НЕ ЕН СЬ БЪНЛЪ ЗЬЛОДЪН' не книз предали его текъ In 18, 30 as (cf. аште не ки бългъ: съ зълодъ: не ЕНХОМЪ ПРЕДАЛИ ЕГО ТЕБЪ In 18, 30 ZOGR); АЩЕ ЛИ ЕИСТЕ ВЪДЪЛИ" ЧЪТО ЕСТЪ МИЛОСТИ ХОШТЖ А НЕ ЖОВТВЪ: НИКОЛИЖЕ ОУБО БИСТЕ ОСЖДИЛИ НЕПОВИНЪНЪЦЪ MIT 12, 7 MAR; аштє въ содомъю съ силър бъюзшата въ тебъ: пръбъють ся до (дьнес)ьнъго дьне Mt 11, 23 мак; горе тебъ хоразних. горе тевъ видъсагда ЕКО АШТЕ ВЪ ТУОВ С СИДОНЪ СИЛЪ ЕНША ЕВИЛЪ! БЪЛВЪШАЊА ВЪ ВАЮ. ДОЄВЛЕ OYEO BB BOTTHIUTH I ПОПЕЛЪ: СЪДАШТЕ ПОКААЛИ СА ЕНША Lk 10, 13 ZOGR; OHH ЖЕ ИСПЛЪНИША СА БЕЗОУМЬЕ' І ГЛУХ ДООУГЪ КЪ ДРОУГОУ' ЧТО БИША СЪТВОДИЛИ HOBH Lk 6, 11 MAR.

Imf (bundles A and C): и се лице кго образът изменташе: овогда оуко видъти и възше съда: овогда же отрока SUPR 121, 24-26; и не въ ниа чада: понеже бъ ЕЛИСАВЕТЬ НЕПЛОДЪ!' І ОБА ЗАМАТТООБВЪША ВЬ ДЬНЕХЪ СВОИУЪ БЪЗШЕТЕ L.k 1, 7 МАR; ал ченъ во въръ и дасте ми псте (for canonical исти) · жадъмъ бъхъ и напонсте МА СТРАНЕНЪ БЪХЪ И НАВЕДОСТЕ МА БОЛЪХЪ И ПОИСЪТИСТЕ МЕНЕ: НАГЪ БЪХЪ И ОЕЛЪКОСТЕ МА SUPR 123, 20-23; ТАКО: И ВЪР ЕГДА СТВОРИТЕ ПОВЕЛЪНЬНАГА ВАМЪ" ГЛЕТЕ ГАКО РАБИ НЕДОСТОІНИ ЕСЛЬ' ЕЖЕ ДЛЪЖЬНИ БЪХОМЪ' СТВООИТИ' СТВОДИХОМЪ Lk 17, 10 SAV; пришъдъ пакъі обрътъ а съпаціа: въстъ во очи имъ такотъмъ Mt 26, 43 SAV; E'Bax We TOY WEN'SI M'SHOP'SI H3 AAREYE 360AUTA' MAXE HIJA по исъ отъ галилена слоужашта ємоу Mt 27, 55 мая; и съвъра са къ немоу НАООДЪ МЪНОГЪ: ЕКО САМЪ ВЪЛЪЗЪ ВЪ КОРАБЪ СЪДЪАШЕ ВЪ МООН: И ВЪСЪ НАООДЪ БЪША ПОИ МОРИ НА ЗЕМИ MK 4, 1 MAR; И БЪЛШЕ СЪ АДАМОМЪ ЕУА ВЪ ДАН" НЖЕ ВИДЪВЪ ДИГАВОЛЪ И ВЪЗДЕВЪНОВАВЪ: СЪТВОМ ПОГЪНЕТАЪ ИМА: ОТЪЛЖНИВЪЩОУ БО СА АДАМОУ НА ОНО МЪСТО ПОДОДЪГ И ЕУГЖ ЕДНИЖ ОСТАВИВШОУ" ПОНСТЪПИ ДНІАВОЛЪ' И ОУПОДОЕНВЪ СА ЗМИЇ ПРЕЛЬСТИ НА НЕНА ПРЕЛЬСТИ АДАМА' Н Бълста (for canonical въашете) ова въ пръслоушаниї вжин въвъща: и їзгънана БЪСТА ИС ПОРОДЪНЪМ ПИШТА" ДЕЛАТИ ЗЕЛАТИ ЗЕМЫЖ ОЖКАМА СВОИМА SUPR 9, 9-19.

ш-Part, н-Part: текоже пръдашта намъ бъвъшен искони самовидьци и слоутъв словесе Lk 1, 2 мак; не пать ли пътиць: вънитъ са пъназема двъма: 1 ни едина отъ ниуъ: нестъ завъвена предъ вли Lk 12, 6 ZOGR; завъвенъ Бъютъ ЕКО МОЪТВЪ ОТЪ СОЪДЪЦА PS SIN 30, 13.

INF-AOR: єгда же придж къ немоу самарене: молівахъ і да ви пръбъвалъ of finx t i noterait To Toy goba gbhin Jn 4, 40 ZOGR (likewise in MAR, AS); i

азъ очнічьженъ і не разоумъю" теко скотъ бъютъ об текс ру тебе ру зго 72, 22; рекж БОГ ЗАСТУПЬНІКЪ МОІ ЄСІ ПОЧЬТО МІЖ ЗАЕТЫ PS SIN 41, 10; КЪ ИНЪМ' ЖЕ ВСЪМЪ И дрочток преславно чоудо приези върьнъимъ усовомъ оугодьникомъ SUPR 565, 12-14; аще оуго вь неправьдью семь житиі не бъюте върьни: Въ истиньмъемь КЪТО ВАМЪ ВЪОХ ИМЕТЪ И АЦИЕ ВЪ ТОУЖДЕМЬ ВЪЗЬНИ НЕ БЪКТЕ ВАШЕ КТО ВАМЪ дастъ Lk 16, 11-12 SAV; и просветъ са лице его слънъце а ризъл его въщин Бълъ вко свътъ Mt 17, 2 м д ; въ распаленью въпасти и страхъ вжин забъетти SUPR 521, 12-13; aurre B's cogom'sy's EMILIA CHA'SI E'SIB'SWAHA B'S TEE'S' ПОВЕЧИЛЪ БЪЩИ: ДО ДЬНЕСЪЙ ВАГО ДНЕ Mt 11, 23 ZOGR.

# The verb χงтъти

§ 550. The main profile

Here is the profile of the main forms:


### §551. Addendum to the main profile

Here is the paradigm of the personal Prae forms:


§552. The morphology of the forms

1. The paradigm of the verb χοπ' έτη <1038> shows two stems, a V-final one (the expanded stem you.b), and a C-final one (the truncated stem you). The distribution of the C-final and V-final stem follows the трытути 2 standard. PRAE system forms show a mix of e- and i-conjugations.

2. SUPR contains aberrant spellouts with you vocalic realization. Here are all the glosses: и ини мнози отъ вратия о ниуъже уъшти глаголати SUPR 169, 1-3; ч'то же хъштж глаголати SUPR 534, 11-12; и ншедъ же сватъти антомии Въпраша и чьто хъште глагола паулъ чръноризецъ хошти въ " SUPR 169, 20-22; рыци ми что се хъште въти SUPR 153, 7-8; они же оумъюше кемог

НОЗЪ: И ХЛЪВЪ ПОВДЛОЖИВШЕ МОЛГАХХ И ВЪКОУСИТИ: ОНЪ ЖЕ НИ СЛЪЩАТИ ТОГО X'ETA' ГООКО ВЪЗДЪХНЯВЬ рече' ю горе мьнъ окапаноуримог SUPR 523, 22-25; ПОВЪСИВЪ ЖЕ ГА ПОВЕЛЪ ДЪРАТИ' ЖЕЛЕЗНЪ НОГЪТЪТ ДОНДЕЖЕ ЧОВВА НАЧЪНЖТЪ ХъТЪТИ ИЗВАЛИТИ СА НА ЗЕМЫЖ SUPR 113, 29-114, 2.

3. SUPR contains yourn (2Sglmv in appearance) instead of the expected 2SgPrae хоштеши, сf. рече къ немоу аште ми са хошти извъстити принесх ТИ ІЄГОЖЕ ВЪІ ГЛАГОЛЕТЕ КОВСТА" И ВИДАШТУ МЬНЪ ПОПЕРИ Ї И ОТВЕРЬЗИ СА ІЄГО' H CE TROOD EX. JEWIN MH ADONT'S NOHCH'SUPR 65, 20-24.

4. The H-Partstem is represented in derivatives, cf. your hnur and похотъ ник.

#### ട്ട് 553. A note

The verb your tru plus infinitive creates a construction with deontic modality or futurity, cf. чьто єстъ знаменьє єгда χотатъ си въюти Lk 21, 7 ZOGR; и полиъ ИС ПАКЪІ ОБА НА ДЕСЬЖТЕ НАЧАТ" ИМЪ' ГЛАТИ" ЕЖЕ ХОТЪАШЕ БЪЩИ ЕМОУ МК 10, 32 ZOGR; KANZI BENHK'S MA YOT'S ПОГНЕСТИ SUPR 276, 1-2.

#### §554. Family membership

There are 3 lexemes altogether. These are: χολήθηκετα, ποχοπτέπη, ποχοπτέπη.

#### \$555. Illustrations

PRAE - om a r b crongs ykoy ichoaran's cha i namb ero becyoujet" stano PS SIN 36, 23; и предъ нимь идете "паковъ и иша" сна зеведеова гляща ЕМОУ ОУЧИТЕЛЮ ХОЦИЕВЪ ДА ЕГОЖЕ АЦИЕ ПРОСИВЪ СТВОРИШИ НАМА" ІС ЖЕ РЕЧЕ НЛАА" что хоциета да творя вама Mk 10, 35-36 SAV; мъю не хоштемъ тевъ зъла НИКАКОГОЖЕ' АШТЕ НЕ ТЪІ САМЪ ВЪСХОШТЕШИ ЗЪЛО ПОНОЕОВСТИ СЕБЪ SUPR 259, 2-5; Съвиранятъ же Бръмена ТАЖЪКА И НЕ ФУДОБЬ НОСИМА" 1 ВЪЗЛАГАЊУТЪ НА плешта члеска: а пръстомь своимь не хотатъ двигняти ихъ Mt 23, 4 мая; поштади оудовъ своихъ варахнсние: и не въсхошти безгодьмъ осядити себе SUPR 269, 13-15; въдъяше ко исконинсъ: кто сятъ не въростящи: и кто естъ хотаи повдаті и In 6, 64 as; пасжштемъ же са велься домъ тоу" по прилогчаю IEANA OTT ANY'S WEALLY ENH3'S BLCH BLACK BE RE RETH YOTAWITH SUPR 217, 28-218, 1.

IMF - а граждане емоц: не χотъахъ его и посьлаша м°литвя въ слъдъ его глце не хоцемъ семоу" да цоствоуєт надъ намі Lk 19, 14 AS; влънъ і же вьливахъ СА ВЪ ЛАДИЯ: ЕКО ЮЖЕ ПОГОАЗНЯТИ ХОТЪАШЕ МК 4, 37 МАR; СВЪХОТЪВЪ НАМЪ ОУТВОЬДІ СОЬДЬЦЕ НАШЬ КІЕУ 2b, 14-16.

INF-AOR - Manosu пророци и цере въсхотъша видъти чеже въ видите и НЕ ВИДЪША Lk 10, 24 MAR; син хотъли бъще да большемъ чьстемъ достоини БЯДЖТЪ SUPR 73, 20-21; ЪКО АЩЕ БІ ВЬСХОТЪЛЪ ЖОВТВЪ ДАЛЪ БИМЪ ОУБО ВЬСЕСЪЖАГАЕМЪЇХЪ ЖЕ НЕ БЛАГОВОЛІШІ PS SIN 50, 18.

# The verb довычити

### ട്ട 556. The profile


### \$557. The morphology of the forms

The verb довьльти [40.(вы). в.т.н] <85> is poorly attested (see Table 559 опр. 320). The attested forms show class contamination.

### §558. Family membership

There are no prefixed verbs with the root <85>.

\$ 559. A note on the attestation of the forms of the verb довычъти

The sources have only 33 glosses (see Table 559 on p. 320).

\$560. Illustrations

PRAE - не п'ц'вте са оуво на оутрен оутрьни во днь совож печетъ сах довь?етъ дьни зълова своѣ Mt 6, 34 ZOGR; гла ємь филипъ ги покажи намъ ОЦА И ДОВЬЛЕТЪ НАМЪ In 14, 8 SAV; ЕДА КАКО НЕ ДОВЬЛЪКЕТ"В ВАМЪ И НАМЪ SUPR 369, 8-9; дьвъма сътома пъная улъбъ не довълятъ илиъ: да къжьдо ихъ МАЛО ЧЪТО ПОНИМЕТЪ Jn 6,7 MAR; ИСХОДА ЖЕ ИЗ ДОМОУ РОДИТЕЛЬ СВОНУВ: НИ ХАТБЕА ВЬЗАТИ ВЪСХОТЪ: ПОНЪ ВЬ ЕДИНЪ ДЬНЬ ДОВЪЛЪЖШТЪ НА ПИШТЯ SUPR 547, 6-8.

IMF - HE довьльахъ водъю развъ на мало днии SUPR 549, 19-20; тръвж ПРЪДЪ ДВЪЗЬМИ ГРОБИШТА СВОЈЕГО ДАСТЖШТЕМ ОБОВТАА: И ТЪ: ИСТИОАА ГАДЪАШЕ кликоже кмоу довьльнше supr 529, 1.

INF-AOR - длъженъ естъ въстькъ кръштенъ: самъ себе чиста дранити: твко црквє воу станк: в о своєї жєнь довъльти са CLOZ Za, 39-2b, 1; нн аштє н м АЗЪЩИ ДОВЬЛЕЛИ БЪЩА ТОЛІКЖ ДОЕОЖ ДЪТТЪЛЬ МЖЖЬ ВЪСПЕТИ SUPR 82, 25-27.


Table 559. Glosses of the verb довьлѣти

# The verb ити

### § 561. The profile


### \$562. The morphology of the forms

1. The root <329> in the paradigm of the verb urn shows two shapes of the truncated stem: C-final ид and V-final и. Their distribution follows the CVC agreement rule.

2. Alongside the standard aorist, there is the root aorist of the идъ type (see \$ 477).

3. л- and ш-Part are formed from the stem шьд (root <1099>); there is no n-Part; the T-Part stem is found in сънитик; cf. also шьстик [шьд.т.ь], and the prefixed verbs пришьстик, ошьстик, дошьстик.

### §563. Family membership

There are 19 lexemes altogether. These are: ити, възити, домти, заити, заити, ИЗИТИ, НАИТИ, НИЗЪНТИ, ОБИТИ, ОТИТИ, ПОДЪИТИ, ПОИТИ, ПОИТИ, ПОМТИ, превъзити, предъити, преити, разити, сънити.

### & 564. Illustrations

PRAE - БЪДИТЕ И МОЛИТЕ СА ДА НЕ ВЬНИДЕТЕ ВЪ НАПАСТЬ Mt 26, 41 МАВ; ГІ ОУслъщи молитея мося: и въпль молкъ тевъ да придетъ PS SIN 101, 2; положилъ ECI TIMA I E'SICT'S HOW'S B'S HENE ПОВІДЖТ'' BLCI 3B'EDSE ЛЖЖЬНИ PS SIN 103, 20; ДА ПОІДЕТЪ СЪМОВТЬ НА НЊА СЪНІДЖТЪ ВЪ АДЪ ЖИВИ PS SIN 54, 16; ВЪСКОВСНЕТЪ БЪ И РАЗИДЖТЪ СЊА ВРАЗИ ЕГО PS SIN 67, 2; ТТЫ ЖЕ ИЕГДА МОЛИШИ СА: ВЫНДИ ВЬ КЛЕТЪ свою: и затвори двъри свол Mt 6, 6 SAV; въстанъте погдъемъ: се привлижи сл предаван ма Mt 26, 46 zogr; то жде и тър сътвори аште видиши чловъка ДНІАВОЛА СЪТВОЛЬША СА: И ПОНДЖШТА К ТЕБЪ: ТАКО ЖДЕ И ТЪІ ОДОЛЪН SUPR 381, 14-17; W чловъколювьствия многа въсота: W трыпънна непрендомое множьство SUPR 394, 14-16.

6 This form contains a nonstandard prefix contraction. There are no forms with hiatus in OCS.

IMF - и проидъеше грады вса и вса очча въ сьньмишихъ Mt 9, 35 SAV; старьца наю заповъдь съконьчавањевъ: се помъкливъша идъаховъ: и се львъ ИДЪАШЕ НА СЪРАШТЪ НАМА: И ОУБОГАХОВЪ СА ЗЕЛО SUPR 296, 29-297, 2; С ИШЕДЪ вонъ петръ плака са горько Lk 22, 62 мая; множьствоу оуво ратьничьског рашъдъшог са по поусттини сен SUPR 292, 2-3; и къмижьници низъшьдъще отъ селма глаахж: теко вельзъютиъ сматъ: и вко о къмази въсъ: 13гонитъ EBC'EI Mk 3, 22 ZOGR.

AOR - не придоуъ призъвать правьдьникъ нъ гръшьникъ въ покаанье Lk 5, 32 ZOGR (cf. не придъ MAR); в изндосте одченика: и придосте въ градъ: 1 овретосте вкоже рече има: и оуготовасте пасуж Mk 14, 16 ZOGR (сf. изидете ... ПОНДЕТЕ | ... | ОБОВТЕТЕ МАR ); ВЬ ПЕНЪІ ВЛЬНЪ ИХЪ ДАЗИДОША СА SUPR 448, 19-20; друже на неже еси пришълъ: то створи Mt 26, 50 SAV; приде жена отъ самариња почреттъ водъг гла ен іс: даждь ми пити: оученнци во его ошьли Бъахи въ градъ In 4, 7-8 ZOGR; вълъзь же въ единъ о" коравицю" сже бъ симоновъ: моли и от земча възити мало Lk 5, 3 а.

# The verb гати


§ 565. The profile

### § 566. The morphology of the forms

1. The root <1114> in the paradigm of the verb ram shows two shapes, a V-final one a and a C-final one гад. Only the truncated stem is found in the paradigm. The distribution of C-final and V-final stems is nonstandard.

2. Along with the standard aorist, there is the root aorist of the ag& type (see \$ 477).

### §567. A note on the distribution of the forms

Generally, the verb is poorly attested; there are altogether 16 forms and 25 glosses, including aberrant ones, for the entire family; the Inf, Sup, a-Part, Prae, M-Part,

Here we list all 25 glosses: ади: възъди Lk 5, 4 мАR, въъді Lk 5, 4 ля; пуан: +хаг Lk 5, 4 ZOGR; ГАДЪМЪ: ПОВЪДЪМЪ Lk 8, 22 МАR; ГАДЖШТ-: ЕДЖШТА Mk 6, 33 ZOGR, ВДЖШТЕМЪ

and н-Part subparadigms are not found at all. Table 567 shows the distribution of the attested forms by subparadigm and the number of the corresponding glosses.

Glosses with the stem ꙗха are treated as aberrant (marked with \*); the same in Lunt 1974, § 16.4. In Imv the stem ꙗха occurs 1 time out of 4; in ш-Part, 2 times out of 7; for the whole verb, ꙗха is found 3 times out of 25 for ꙗ and ꙗд together. The choice of the type ꙗ/ꙗд is treated as variation in the root formative, and is found also in other paradigms (cf. ити, быти, дати, ꙗсти), while the ꙗ/ꙗха choice is the variation between a simple and suffixal stem, which should not be found within a lexeme outside of нѫ-verbs.


Table 567. Distribution of the forms of ꙗти by subparadigm

Unfortunately, the starting forms Inf and Prae are reconstructed. Given that the forms with the stem ꙗха are aberrant, the reconstruction can be based either on the V-final ꙗ or the C-final ꙗд. Of course, the simplest solution is to assume the V-final stem for Inf, and the C-final stem for Prae. This is the reconstruction adopted in the present grammar.

Other authors make different assumptions. Vaillant (§ 214) uses the spellout ꙗ(ха)ти, without a grammatical interpretation. Diels lists the infinitive ꙗхати in the index, and a full list of forms in § 134.26. Večerka forms all infinitives from the stem ꙗха. Sadnik does likewise, but in the root dictionary lists the reconstructed infinitive ꙗти (\*jati).

The present grammar's paradigmatic dictionary contains cross-references: ꙗхати see ꙗти, възѣхати see възѣти, etc.

### §568. Family membership

There are 5 lexemes altogether. These are ꙗти, възѣти, въꙗти, приꙗти,8 прѣꙗти.

Lk 8, 23 Zogr, Mar; ꙗде: прѣѣде Mt 9, 1 Zogr, Mar; прѣꙗде Mt 9, 1 Sav f. 37; ꙗдошѧ: прѣѣдошѧ Lk 8, 26 Zogr; ꙗдѫ: въѣдѫ Lk 8, 23 Mar, прѣѣдѫ Lk 8, 22 Zogr, Lk 8, 26 Mar; ꙗдѣахѫ: ѣдѣахѫ Jn 6, 21 Zogr, Mar; Jn 6, 17 Zogr, Mar; ꙗвъш-: прѣѣвъшуму Mk 5, 21 Zogr; прѣѣвъшю Mk 5, 21 Mar, прѣѣвъше Mk 6, 53 Mar, Mt 14, 34 Zogr, Mar; ꙗхавъш-: прѣѣхавъше Mt 14, 34 As, приѣхавъше Mk 6, 53 Zogr.

8 The PD lists the headword приꙗти2, contra Večerka and following Sadnik. Večerka assigns the form приѣхавъше Mk 6, 53 Zogr to the item прѣꙗхати.

# §569. Illustrations

PRAE — вьзѣди въ глѫбинѫ Lk 5, 4 Mar (cf. виЖь ѣхаі въ глѫбинѫ Zogr, въѣді въ глѫбінѫ As); ꙇ рече къ нимъ прѣѣдѣмъ на онъ полъ езера· ꙇ въѣдѫ (3PlAor) ѣдѫщемъ же имъ· усъпе Lk 8, 22–23 Mar.

IMF — ꙇ абие быс корабл҄ь на земл҄и· вь н҄ѭже ѣдѣахѫ Jn 6, 21 Zogr; ꙇ прѣѣвъше придѫ на землѭ ћенисаретъскѫ Mt 14, 34 Mar (cf. и прѣѣхавъше As).

AOR — ꙇ прѣѣдошѧ на земл҄ѭ ћенисаретьскѫ Lk 8, 26 Zogr (cf. ꙇ прѣѣдѫ Mar).

# The verb стати

### §570. The profile


### §571. The morphology of the forms

1. The paradigm of the verb стати ‹879› has two stems, a C-final one with the suffix н (ста.н), and a V-final one (ста); both are truncated. The C-final and V-final stems are distributed as follows: PRAE uses the C-final stem, while all other systems use the V-final one (but only the present Imf occurs). C-final stem forms follow the class 5 standard; V-final stem forms follow class 4v.

2. The н-Part stem is represented in derivatives, cf. станиѥ and станъ; пристанище, останъкъ, въстаниѥ and some others. However, the т-Part stem is also found, cf. остатъкъ and остатиѥ.

### §572. Family membership

There are 10 lexemes altogether. They are: стати, въстати, достати, настати, остати, пристати, прѣдъстати, прѣстати, състати, устати.

Cf. also стаꙗти 3, стоꙗти 2.

### §573. Illustrations

PRAE — еда како не достанетъ вамъ ꙇ намъ·ꙇдѣте же паче къ продаѭщиімъ ꙇ купите себѣ Mt 25, 9 Zogr; ꙇ въставъ запрѣти вѣтру· ꙇ рече морю мльчи и устани· и улеже вѣтръ и быстъ тишина велиѣ Mk 4, 39 Mar; ꙇс же рече останѣте еѩ· по чьто ѭ труЖаете· добро бо дѣло съдѣла о мнѣ Mk 14, 6 Zogr; блаженъи же чудивъ сꙙ вѣрѣ ѥго· на молитвѣ станевѣ рече ѡ друже Supr 276, 15–16; въстані славо моѣ· въстані пьсалътыръ и гѫслі· въстанѫ рано ї исповѣмъ сѩ т(е)бѣ въ людехъ гі Ps Sin 56, 9–10.

IMF — и сице не останѣахѫ сꙙ стыдости Supr 413, 18–19; многу же часу минѫвъшу· и нощи уже прѣполовꙙщи сꙙ· не прѣстанѣѣше тлъкѫщі Supr 515, 17–18; хлѣбъ нашъ наставъшааго дьне· даЖъ намъ дънесь Mt 6, 11 Mar; анꙿг҄елъ же гнь ꙗви сꙙ стуум съшъдъ сь небесе· въ образѣ мльниѧ· и приде к н҄ем ставъ на въздусѣ· окы трьми лакты движенъ отꙿ земьѧ· и прѣбы стоѧ на мѣстѣ томь многы часы Supr 567, 23–26; излѣзъ же ис корабьꙗ приде на кон҄и кꙿ н҄имъ· ставъ же рече слугамъ своимъ· приведѣте ми ѥдꙿного оть н҄ихъ Supr 60, 10–13.

INF-AOR — лобьзаниѣ ми ты не дастъ· сі же ѡтнелі вънидъ· не прѣста облобызаѭщі ноѕѣ мои Lk 7, 45 As; нога моѣ ста на правъдѣ· въ цркъвахъ благословесъстуѭ тѩ Ps Sin 25, 12; мы же въстахъмъ (for canonical въстахомъ) і прості быхомъ Ps Sin 19, 9; състашѧ сѧ архиереи и кънижъници съ старьци· ꙇ рѣшѧ глѭще къ нему Lk 20, 1–2 Mar; сь не бѣ присталъ съвѣтѣ· ꙇ дѣлѣ ꙇхъ Lk 23, 51 Zogr; хотѧ ѣвіті своѭ сілѫ· ꙇ хотѧ сътворіті· да бѫ прѣсталі· отъ зълобы своеѩ Cloz 5a, 12–14; двѣ на десꙙте бо бѣста кошници укрухъ осталѣ Supr 428, 29–30; ту падѫтъ въсі творѩщеі безаконеніе· въїрїновені бъїшѩ і не могѫтъ статі Ps Sin 35, 13.

# The verb съпати


§574. The profile

§575. The morphology of the forms

1. The paradigm of the verb съпати ‹918› shows two stems: a C-final one (the truncated stem съп), and a V-final one (the expanded stem съп.а). The C-final and V-final stems are distributed as follows: the PRAE system uses the C-final stem, while the INF-AOR and IMF systems use the V-final stems. The PRAE forms follow the i-conjugation.

2. The N-Part stem is found in the derivative съпанию.

### §576. Family membership

There are 2 lexemes altogether. These are: съпати, посъпатн.

### §577. Illustrations

PRAE - с приде с обръте ка съплашта: и гла петрови: симоне съпиши ли НЕ ВЪЗМОЖЕ ЕДИНОГО ЧАСА ПОБЪДЕТИ Mk 14, 37 MAR; НЕ ПЛАЧАТЕ СА: НЪСТЪ БО 8 мрьла дъвица нъ съпитъ Lk 8, 52 SAV; и рече имъ чьто съпите въставъще помолите са да не вьнидетъ въ напасть Lk 22, 46 ZOGR; ї въста чеко съпья гъ I \*KO CINEH'B ШЮМЕНЪ ОТЪ ВІНА PS SIN 77, 65; БЪ ВЪ МАЛЪ ОУСЪПЕ' I СЪПАШТАЊА отъ въка: отъ адама въскресс CLOZ 12b, 7-9; вще посъпите мюждю (for canonical междоу) пределъ: криль голжен посъревренть PS SIN 67, 14.

IMF - въздовмаша са вьса и съпаахж Mt 25, 5 мая; їаковъ съпавъ тако ЛЬВЪ: И ВЪСТА ГАКОЖЕ СКУМЬНЪ SUPR 478, 25-26; ЧЛОВЪКЪ ЖЕ ПАКЪІ ЕЪ ОУБОГЪ" ИМЪ ТЕЛИЦИ ЕДНИХ: ТАЖЕ НА ТРЕПЕЗЪ СЬ ЯНИЪ ГАДЪАШЕ: И ЧАШЖ ЈЕДИНЯ СЪ нимь пипаше и на лоне кмоу съпааше SUPR 359,28-360,2.

AOR - a3' x e of chaxx a i c'onax's: b'estray's the r's sactimin's min PS SIN 3, 6; 0T'B CO'EA'SI CKOYMEN'S ПОСЪПАХЪ СЪМЖШЕНЪ PS SIN 56, 5.

# The verb къпити


\$578. The profile

<sup>9</sup> Večerka also gives оусъпати for the hapax gloss оусъпи и повелъниемь твоимъ: на ложи єго вUCH 42b, 5–6), which in the present grammar is assigned to the lexeme оусъпити 1, as is the case in Sadnik.

#### § 579. The morphology of the forms

1. In the paradigm of the verb въпити [въп.н.т.н] <129> there is one stem: the V-final expanded stem въпн. The morphology of the forms follows class 7, but instead of the class 7 suffix (a or t) the verb shows a suffix from class 1, namely the theme n (see Vaillant, § 198).

2. The canonical imperfect is въпи=чахъ, въпи=ваше (contra the morphological balance rule), cf. the contracted imperfect выпиаста in Mt 20, 31 SAV.

#### § 580. Family membership

There are 2 lexemes altogether. These are: въпити [въп.и.т.н], възъпити [въз.ъп.и.т.н]. The root is morphophonologically anomalous, with two shapes, ъп/въп.

#### §581. Illustrations

PRAE - сего дъльма лазаре гради вънъ: тебе юдного нъмутноу възъпиня: да съ странты ходаштинмъ покажж силж SUPR 311, 7-10; осченнци его молъхж и глажие: отъпоусти их вко въпиятъ въ следъ насъ Mt 15, 23 мая; и се дуъ ЕМЛЕТЪ I' И ВЫНЕЗАПЖ ВЪЗЪПНЕТЪ И ПОЖЖЕТЪ СА СЪ ПЪНАМИ Lk 9, 39 SAV; Проповъдникъ спаса нашего въсковшения. Паче же чаше същасенние възыния SUPR 479, 13-14; 04 TOOYA(H)X'B CHA BONICA 13MA'SHE FO'BTAHL MOI PS SIN 68, 4; ГЛА ВЪПІНЖЦІАЛГО ВЪ ПОУСТЪННИ: ОУГОТОВАНТЕ ПЖТЪ ГНЪ: ПРАВЪН ТВООНТЕ СТЪЗА Ero Mt 3, 3 AS.

IMF - onin же излиха въпитьахъ глире да распатъ сждетъ Mt 27, 23 as (cf. BBIHTEXX in ZOGR and MAR); XL женамъ възъпьвше CLOZ 14a, 23-24; н ВЪЗВПИВЪ И МНОГО СА ПОЖЖАВЪ: ИЗИДЕ МК 9, 26 SAV; КНАЗИ ЖЕ ВЕЙМИ ВЪЗПИВЪШЕ BECMHIAWA CA SUPR 138, 1-2.

INF-AOR - възпихъ к тевъ ги ги оуслъши гласъ мон SUPR 461, 21-22; и ВИДЪВЪШЕ И ОУЧЕНИЦИ ХОДАШТЪ ПО МОРЮ СЪМАША СА ГЛЕЖШТЕ ЪКО ПОИЗОАКЪ ECT'S I OT'S CT ST CT ST LE MAR; I H3N 26 MAR; I H3N ST IC KODAEN'S NET B' ХОЖДАШЕ НА ВОДАХЪ" I ПОИДЕ КЪГ ИСВИ" ВИДА ЖЕ ВЪТОВИ КОВПЪКЪ! ОУБОЪ СА: 1 НАЧЫ В ОУТАПАТИ: ВЪЗЪПИ ГЛА ГИ СПИ МА Mt 14, 29-30 ZOGR; ГЛЕЖ ВАМЪ: ЕКО ацие и сни оумлъчатъ: камение въпити иматъ Lk 19, 40 мак.

# The verb състи

### § 582. The profile


\$583. The morphology of the forms

1. The paradigm of the verb chern <798> shows two truncated stems that only differ in their vocalism: cha and cag. The pairing between the vocalic realizations is nonstandard: (ג, ৳) is not a pairing by any alternation (cf. °ף৳ং۳۲ <791> and лешти <482>). The root is anomalous. Both stems are C-final; forms follow class 4c standard.

2. Some verbs in this family have correlates in the трыпъти 2 class; these are chern > chartern, and nockern > nockytry. Note that the personal lmf forms of these correlates coincide. Accordingly, there is no basis for assigning particular glosses to one or another lexeme. For example, in съдъяше съ слоугами BHATH KONSYHHA Mt 26, 58 MAR, the form chatame can be assigned either to the lexeme състи 0 or to the lexeme съдъти 2.

3. Along with the standard aorist, there is the root aorist charbetc. (see \$ 477).

4. The N-Part stem is found in the derivative съдению.

### §584. Family membership

There are 7 lexemes altogether. These are: chern, BBC CTH, nochern, просъсти, повдъсти, съсъсти, съсъсти.

Cf. also charm 2 and caguru 1.

### § 585. Illustrations

PRAE — възненавідъхъ цоковъ лжкавънъръ: и съ нечъстивыми не съдж PS SIN 25, 5; даждъ нама да єдниъ о десням тєєє а дроутъ о людутъ о лъвжих тєєє садевъ въ славъ твоєн Mk 10, 37 зау; югда са сьседе (for canonical съсадетъ) МЛЕКО И ОСЪЮЋЕ: ЗАГОАДИТЪ ПЖТЪ СЬСОУ: ТЪГДА И ДЕТИШТЪ ПЛАЧЕТЪ И МАТН EONHT'S SUPR 312, 6-9; NH BLAMBAINT'S BHHA HOBA" B'B N'EX'SI BET'SX'' ALLE AH же ни просаджтъ са мъси; с вино пролъетъ са: и мъси погъвлиятъ Mt 9, 17 MAR; N'B єгда з'ван'ї бждєши; ш'їд'ї слди на посл'єдьнинимь м'єст'''; да єгда

придетъ зъвавыи тѧ· речетъ ти друже посѧди выше Lk 14, 10 Mar (likewise Zogr, Sav, As); и гла ученікомъ· сѩдѣте ту· доньдеже шедъ помлѭ сѧ тамо Mt 26, 36 As.

IMF — въздрѣмашѩ сѩ въсѣдъшеі на конѩ Ps Sin 75, 7; вьзгнѣщъшемъ же огнъ по срѣдѣ двора· ꙇ въкупѣ сѣдъшемъ имъ· сѣдѣаше петръ по срѣдѣ ихъ Lk 22, 55 Mar; ръцѣте дъщери сионовѣ· се цсръ твои грѧдетъ тебѣ· кротокъ и вьсѣдъ на осьлѧ· ꙇ жрѣбѧ сна ѣрьмьнича Mt 21, 5 Mar.

INF-AOR — ѣко обідѫ мѩ пъсі мноѕі· и сънемъ зълобівыихъ осѣде мѩ Ps Sin 21, 17; на рѣцѣ вавулоньстѣ ту сѣдохомъ и плакахомъ сꙙ помꙙнѫвъше сиѡнъ Supr 418, 20–22 (cf. сѣдомъ Ps Sin 136, 1 and Cloz 7a, 30–31); и помоливъша сꙙ· сѣдоста оба ꙗсти хлѣбъSupr 524, 18–19; обрѧщета жрѣбьць привѧзанъ· на н҄емьже нѣстъ не у никтоже отъ чкъ вьсѣлъ Mk 11, 2 Zogr; на н҄емъже убо мѣстѣ господь съподобилъ мꙙ ѥстъ сѣсти· то на томъ подоба ми ѥстъ сѣдѣти мльчꙙще дожи и до коньца жизни моѥѧ Supr 205, 11–15.

# The verb лещи

§586. The profile


### §587. The morphology of the forms

1. The paradigm of the verb лещи ‹482› shows two truncated stems that only differ in their vocalism, лег and лѧг. The pairing between the vocalic realizations is nonstandard: (ѧ, е) is not a pairing by any alternation (also found in сѣсти ‹798› and °рѣсти ‹791›). The root is anomalous. Both stems are C-final; forms follow the 4c class standard.

2. Some verbs in this family have correlates in the трьпѣти 2 class; such are: лещи≈лежати, възлещи≈възлежати and облещи≈облежати. Note that the personal Imf forms of such correlates coincide. Accordingly, there is no basis for assigning particular glosses to one or another lexeme. For example, in мьздоꙇмьци пришедъше възлежаахѫ съ исомь Mt 9, 10 Zogr or съ мытари вьзлежааше· и сь любодѣицами живѣашеSupr 474, 1–2, the Imf forms can be assigned either to the lexeme възлещи 0 or to the lexeme възлежати 2.

3. Alongside the standard aorist, there is the root aorist легъ (see § 477).

### §588. Family membership

There are 4 lexemes altogether. These are: лешти, възлєшти, облєшти, оулєшти. Cf. also лежати 2 and лѣгати 3.

### \$589. Illustrations

PRAE - въсгъ слънъце и собърашњу см. и въ ложихъ своютъ людить люди в SIN 103, 22; мноен W въстокъ: и отъ западъ приджтъ: и възлагитъ съ авраамомъ: И ИСАКОМЪ ИЕКОВОМЪ: ВЪ ЦОСТВИИ НЕСЬНЕАЛЪ Mt 8, 11 AS; и няждаашете н ГЛІЖШТА: ОБЛАЗИ СЪ НАМА ЪКО ПОИ ВЕЧЕРЪ ЕСТЪ: И ПОВКЛОНИЛЪ СА ЕСТЪ ЮЖЕ ДЕНЬ" І ВЬНИДЕ СЪ НИМА ОБЛЕШТЬ (Sup) Lk 24, 29 МАR; НЕ ДЕН ЕМОУ ПАКОСТИ" НН ржкама ни ногама: ни всемоу тълеси: нъ въ едино мъстъв лази съвивъши CA EUCH 36b, 9-13.

IMF - кгда же очмъ нозъ нутъ и прилтъ ризър своя: и възлегъ пакът рече ИМЪ ВЪСТЕ ЛИ ЧТО СТВОРИХЪ ВАМЪ Jn 13, 12 SAV; ЕГ ДА ВЪЗЛЕГТЫИ ВЪ ГООБЪ ГАКОЖЕ ВЬ ЧОВЕВ: ХОТЪАШЕ ОТЪВАЛИТИ КАМЪКЪ ОТЪ ГООБА SUPR 472, 27-29.

INF-AOR - L BBCTAB'S Запретти В'ЕТРОУ' и рече морю: млльчи' и оустани: и ОУЛЕЖЕ ВЪТРЪ: 1 БЪКТЪ ТИШИНА ВЕЛЬЪ Mk 4, 39 ZOGR; 1 ПОВЕЛЪ 1.МЪ ПОСАДИТИ њ все: на сподът: на сподът на тръвъ зеленъ: възлегоша на лъута: на лъдът по сътоу и пати десатъ Mk 6, 39-40 ZOGR (сf. възлегж MAR); сьде петръ: сьде ПАВЕЛЪ' СЬДЕ СТОЕ ЕВЕЛИЕ' СЬДЕ АЗЪ ПОКЛАНЪВЪ СА: ЛЕЦИ ХОЩИЯ ЕUCH 37b, 1-3; и въниде съ нима облець: и бъг възлеже съ нима: приемъ хлъвъ клас: н преломъ давше има Lk 24, 29-30 AS.

# The verb OE.precTH


\$ 590. The profile

§591. The morphology of the forms

1. The paradigm of the verb °ptcTH <791> shows two truncated stems that differ in their vocalic realization and the final consonant: phy and pourr. The pairing between the vocalic realizations is nonstandard: ( ^ , t) is not a pairing by any alternation (cf. chern <798> and newnn <482>). The root is anomalous. Both stems are C-final; PRAE forms follow the class 3 standard; the present Imf and all other forms follow the class 4 standard.

2. Alongside the standard aorist, there is the root aorist of the type (06) рътъ (see \$ 477).

3. The H-Part stem, in addition to participles themselves (cf. coptivena ERAEBES SUPR 472, 12), is widely attested in derivatives: съръттєнние, приоеръттяние, ИЗОБОВТЕНИЮ, ОБОВТЕНИЮ.

#### § 592. Family membership

There are 4 lexemes altogether. These are: изоврести, приовръсти, съръсти.

#### § 593. Illustrations

PRAE - осклангания са страхъ! и отъвьсядъ съдовжя са како сьрашти се дъплине: към ли овраштя исходъ словоу SUPR 508, 15-17; шедъ въ море Въвръчи ждиция: в тяже имеши пръжде ръкъх възьми и отвръзъ фуста ен и ОБОЛШТЕШИ СТАТИОВ: ТЪ ВЪЗЕМЪ ДАЖДЪ ИМЪ ЗА СА МА И ЗА СА МЕ 17, 27 МАR; LATTA B'S TOANS' I C'SPALITET'S B'SI YK'S B'S CKAANSHILE BOAR HECA (for canonical несъ!)' по немь гдета Mk 14, 13 ZOGR; мядростия нашеж изовраштемъ ИСТИН(Х) SUPR 265, 30; сего раді обратьятъ сь людне могсъмо' и дьнь(Е) ИСПЛЪНЬ ОБОНШТЯТЪ СЊ ВЪ НУУЪ PS SIN 72, 10; ПОИОБРАШТИ ЖИЗНЬ СЕБЪ" ДОВЬЛЕКТЪ ТИ ГАЖЕ ПОИА ДОСАЖДЕНИИ SUPR 253, 2-3.

IMF - рякжя же пльть дръжаахъ' а доуших ба поразоммъахъ: и вычъыхъ ОБОДШТААХЪ ЧОУДЕСНО: А ЖТОВЖДЪ СТРАШНО SUPR 511, 20-22; ПОДОБЪНО ЕСТЪ Црствие невское съкровнштю съкръвеное на селъ: еже овретъ чкъ съкоръ. и отъ радости его идетъ: и все елико иматъ продаетъ Mt 13, 44 ZOGR; овръчтъв ДШЯ СВОЕК ПОГОУБІТЪ НЖ' А ІЖЕ ПОГОУВІТЪ ДША СВОЕЖ МЕНЕ РАДІ" ОБРАЩЕТЪ НЖ Mt 10, 39 AS; 6' Hamb привежиция и сила: помощъмикъ въ скръеехъ обовътъ НЪ БЪЛО PS SIN 45, 2; ОУКЛОНИВЪ СА ОТЪ СЕГО ПЖТИ" ДА НЕ СЪРЪТЕНА БЯДЕВЪ сверъпъюнин њзъкъ SUPR 472, 11-12.

INF-AOR - днаволъ: нно коумирослуженние арнанъскжен хоулж изоворъте SUPR 186, 13-15; 1 пришьдъшюмор на онъ полъ въ странж кер кесинсках сърътосте и дъва въсъна: отъ гребншть суъ судашта люта 5 го Mt 8, 28 ZOGR (cf. CBSTETETE MAR and the aberrant spellout CSSETTOCTA SAV); aEIE %E СЪХОДАЩИЮ ЕМОУ СЕ РАБИ ЕГО СЪОВТОША И ГЛУДЕ. ВКО СПЪ ТИ ЖИВЪ ЕСТЬ In 4, 51 AS (cf. cbptima in ZOGR and MAR); I'm mmm Tanant's MI ECH AAN'T' H CE ANYTHER пыть талантъ приостетохъ ими Mt 25, 20 As (сf. приоеретъхъ zogr and приововтъ in MAR and SAV); колико са бъ троудилъ да същитькако обрълъ понъ ІЕДНОГО ЗА ТА МОЛАШТА ГА SUPR 95, 11-13; ПРЕВЪЗВЪРЪВЪ ГАРОСТИЯ: СЪМАТРАШЕ КОК БИ ПОИМЪЩИЕНИЕ ИЗОЕРЪСТИ" ДА ЕН ДЛЪГЖ И ЛЮТЖ ИМ' СЪТВООНАЪ СЪМОБТЪ SUPR 88, 28-89, 1.

# The verb гънати

### § 594. The profile


### § 595. The morphology of the forms

1. The paradigm of the verb гънати <153> shows two stems that differ in their vocalic realization: the truncated stem with € vocalism, жєн, and an expanded stem with & vocalism, гъма. The root is anomalous. The stems are distributed as follows: the truncated stem in PRAE; all other systems use the expanded stem (in Imf the stem жєм is the present imperfect).

2. The H-Part stem, in addition to participles themselves (cf. изгънани Jn 12, 42 MAR), is attested in the derivative изгънання.

### §596. Family membership

There are 7 lexemes altogether. These are: гънати, въпъмати, изгънати (Prae иждєня, иждєнєши), отъгънати, погънати, прогънати, разгънати (Ргае ражденя, ражденеши).

Cf. also гонити 1.

### § 597. Illustrations

PRAE - граджштаго къ мьнъ: не ижденя вънъ Jn 6, 37 ZOGR; влажени ЕСТЕ ЕГДА ПОНОСАТЪ ВАМЪ' И ИЖДЕНЖТЪ ВЪГ И РЕКЖТЪ' ВЪСЪКЪ ЗЪЛЪ ГЛЪ НА ВЪР ЛЪЖЖШТЕ МЕНЕ РАДИ Mt 5, 11 ZOGR; ОУКЛОНИ СЊА ОТВ ЗЪЛА I ТВОРИ ДОБРО" ВЬЗИЦИ МИра I ПОЖЕНИ I PS SIN 33, 15; И рекжить вамъ се сьде се овъде убъ не ИЗИДЕТЕ НИ ПОЖЕНЕТЪ Lk 17, 23 Mar; ЖЕНЖШТИИ ЖЕ ПОГАНИ ВИДЪВЪШЕ ТАКО ВЪ ПЛЕВЬНИЦЯ ВЪСКОЧИ БЪЖА: ВЪЗЪМЪШЕ ОГНЪ ЗАПАЛИША ПЛУВВЫНЦА SUPR 196, 19-22; вьзьпи гласомъ въсъ идл идж идж ноуждею (for canonical НЖЖДЕНХ) И МЖКОН ЖЕНОМЪ ЕСМЪ: ОТЪХОЖДЖ ОТ В ЧЛОВЪКА: ОУЖЕ К ТОМОУ НЕ ПОНЕЛИЖЯ СА КЪ СЪМЪРЕНОУОУМОУ И ПОВЕПРОСТОУОУМОУ ПАУЛОУ" ЖЕНЕТЪ БО МА И HE BEAT KAMO HAX SUPR 173, 15-22.

IMF - и съвазавъше кго сътвориша кмог оноушти. и гвоздна остръв ВЪНОЗИША ВЪ ОНФУШТЪ: И ФЕФУША И: И ЕНЕЖШТЕ ИГО ЖЕНЕАХХ: ПАКОЖЕ КОВВИ ИГО ЗЕМЫЖ ПОЛИГАТИ SUPR 17, 18-22; ПОДАЖДИ МИ ВЛАГЖ ТВОНУВ ЦЕДООТЪ' 13горъвъшее гръхъ срце мое и отъпънавъ мракъ гръеховънъи. и въжьзи въ НЕМЬ: ССКОЖ ТВОЕГО МАДИЪ ЕUCH 78а, 3-8; БЛАЖЕНИ СЗГЪНАНИ ПРАВЪДЪЩ РАДИ" тько тъхъ естъ црство ньскоє Mt 5, 10 ZOGR; движжщен сьо да пресельную сь CHBE EFO H BBCXAMILART'S' H B'SIT'SHAHI EXAMIT'S 13 ДОМОВ'S CROIX'S PS SIN 108, 10.

INF-AOR - и дастъ намъ вгъ силя: да ово отъ них оченуомъ: овъи прог нахомъ SUPR 72, 28-29; и ютро проврезгог съсло въставъ изнде исъ и ИДЕ ВЪ ПОУСТО МЪСТО І ТОУ МОЛИТВЯ ДЕАШЕ І ГЪНАША И СИМОНЪ И ИЖЕ БЪЗДЖ C' C' C' C' C' C' C' C' T' T' L' C' C' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' аште мене изгънаша: и васъ сжденятъ Jn 15, 20 ZOGR; віноградъ 13 егоупьта пръчессъ въпъна назъкъ и насаді I PS SIN 79, 9; въровавъше же молахъ са " ГАКО ДА БЪР ШЕЛЪ И ОТЪГЪНАЛЪ ЛЖКАВАТО ТОГО БЪСА: ДДОУЗИИ ЖЕ ИСКОУШИНЖШТЕ гакоже не имать съдолъти SUPR 35, 28-36, 2; дипаволъ [...] вида сеге отъ МНОГЪ! БЛАГЪ!А ДЪТЪЛН МЯЖА И ПОАВЪДНААГО ЖИТИГА НА МНОЗЪ ПООГОНИМА" Въста на нь: хотаи прогнати отъ прежденареченааго мъста SUPR 514, 19-23.

# The verb плъти

### \$ 598. List of forms


This verb is represented in OCS by two forms (3 glosses):

### §599. The morphology of the forms

The root of the verb плъти <685> is sonant, with H(l) vocalism. In OCS it is only found with the vocalic realization ats. The noun not. B.a contains the nominal suffix в. The accepted reconstruction of the infinitive плетти is based on Church Slavic (and Russian полоть, полешь). The isolated PRAE forms observed in OCS could be assigned to the infinitive \*плъвати 3° (cf. °ръвати 3°, ковати 3°, искати 3°, and others; see details in § 498). The PD lists the infinitive плъти, following the tradition (Vaillant, Večerka, Sadnik).

### §600. Family membership

There are 2 lexemes altogether. These are: плъти, исплъти.

#### §601. Glosses

хоштеши ли оуво да шьдъше сплъвенъ м Mt 13, 28 ZOGR, MAR; да […] НАВЪЖНЯТЪ ГАКО ПЛЪВОМА ЕСТЬ СЬМОЬТЪ ДА СА ИСТОДСКТЪ ГРОБИ" А МОБТВИН Въперени БЖДЖТЪ НА ВЬСТАНИЕ SUPR 424, 29-425, 2.

# The verb дъти

#### § 602. The profile


#### \$603. The morphology of the forms

1. The paradigm of the verb Attu <225> shows two stems: a C-final one ( дежд) and a V-final one ( дъ). The stems differ in their vocalic realizations ( ( 1 ). The C-final stem is substitutively softened (PRAE follows class 3); the INF-AOR and IMF system forms follow class 4v.

2. Some verbs in this family have correlates in the плакати 3 class. Cf. BB3ATTH and BB3ATHATH, 30ATETH and 30251077. Note that the personal Imf forms for such correlates coincide. Accordingly, there is no basis for assigning particular glosses to one or another lexeme. For example, ютро продобзгог 5-кло: въставъ изнде ись и иде въ полсто мъсто и тог молитвя деаше Mk 1, 35 мая. 10

3. Alongside the participles themselves (cf. og thin supr 271, 11-12), the N-Part stem is found in the derivative og them.

#### §604. Family membership

There are 5 lexemes altogether. These are: въдъти, въздъти, задъти, одъти, придѣти.

Cf. also делати 3.

<sup>10</sup> Contra Sadnik and following Večerka, there is no headword дѣти 0 in the PD, but only дъюти 3, which means that the form дъаше belongs to дѣюፐи 3.

§605. Illustrations

PRAE: ї о илені твоємъ въздєждя ряць мої рѕ SIN 62, 5 (сf. въздъюж ряць MOI PS SIN 27, 2); NE ПЬЦЪТЕ СА' ГЛЯЦИЕ ЧТО "ВМЪ ЛИ ЧТО ПИЕМЪ" И ВЪ ЧТО ОДЕЖДЕМЪ са Mt 6, 31 SAV; васъ одеждятъ ризъл: мене же правда SUPR 19, 19-20; въдеждн пръстъю въ фуши єго висн 31b, 18-19; въ ноштєхъ вьздєждувтє (for canonical ВъЗДЕЖДИТЕ) рякъ вашы PS SIN 133, 2.

IMF: MATH ЖЕ ЕГО И НЕ ИСПОЧИВЪШИ ВЪЗДЪВЪШИ ОЖЦЪ И ТЕКЖШТИ ИДЕ плачжшти са до стааго мяжа: повъдати кмоу свотя въдж SUPR 43, 26-29; пръдъста црца о десноя тебе въ ризауъ позлациенахъ одъна пръкогушена PS SIN 44, 10; BLATE3E WE \TO2JERK'S' OML'S MATS DACHOOCTTONT'S H BL3ATH'E HA HEEO SUPR 555, 14-15.

INF-AOR: 3BBAX' K' TEE'S FI BLCb 26HL' B'B3A'S' B'B3A'S' DRI PS SIN 87, 10; СТРАНЬНЪ БЪХЪ И ВЪВЕДОСТЕ МА: НАГЪ И ОДЪСТЕ МА Mt 25, 35-36 SAV; КОГДА ЖЕ TA BHATEXON'S CTOANTHA H BEBECOM'S' AH HAFA H OATEXON'S Mt 25, 38 MAR; CENOY Задъша понести крстъ кго Mt 27, 32 ZOGR; приемъ глоудааго приведенааго КЪ ТЕЕЋ: ВЬДЪ ПОВСТЪК ВЪ ФУШИ ЕГО EUCH 316, 20-22; ТО СЛЪШАВШИ ВАСИЛИНА НА МНОЖАНШИ ЛЮБЬВЬ ПОНДЕ' ГАКОЖЕ ВИДЕТИ СВОИМА ОЧИМА СТАРЦА' И ПОМЪСЛИ ОДЕТИ СА ВЪ МЖЖЪСКЪМ ОБРАЗЪ' И ИТИ К НЕМОУ ВЪ ЛАВОЖ' И КЖЕ О СЕБЪ съповъдати томоу SUPR 299, 8-13.

# CHAPTER 22 **Aberrant verbal forms in sources**

# §606. General

The present chapter treats aberrant forms of the verbs that show paradigmatic aberrations. The source of the observed aberrations are various paradigmatic effects.1 Some paradigmatic effects apply both in derivation of canonical forms, where they shape the paradigmatic standard of some class, as well as in the generation of aberrant forms, deflecting their derivation from the assigned paradigmatic standard. For example, in building the PRAE workstems for the метати 3° group of irregular verbs, the absence of substitutive softening effect (° symbol) applies, and the same effect leads to aberrant forms such as колѣте (for canonical кол҄ите) in the брати 4h• groups of irregular verbs.

### §607. The order of examination of aberrant verbal forms

Some aberrant forms are treated above in Ch. 19, *An overview of verb forms by system* (aberrant imperfect forms and the new ш-Part of the л҄юбивъ type), and in Ch. 21, *Unique verbs*. All other aberrant forms are treated by paradigmatic class and system, as shown in Table 607 (p. 338).

1 Some paradigmatic effects, namely those that lead to deformations in terminals, have segmental underpinnings, but are treated together with properly paradigmatic aberrations, for compactness of the overview of all aberrant verbal forms. For segmental aberrations, see Part I, *Segmental grammar*.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)


### Table 607. Order of examination of aberrant verbal forms

# Aberrations by paradigmatic class

\$608. Contamination of classes 3 and 7

In the PRAE system, плакати 3 class verbs can show aberrant forms from the expanded stem by the class 7 paradigmatic standard. (On the paradigmatic attestation of the classes, see Ch. 24, Supplement, § 910–913, Excursus on the contamination and competition between paradigmatic classes). Table 608 shows some examples.


Notes to Table 608


§609. Contamination of classes 5, 3, and 4

This aberration applies to class 5 C-final мя-dropping verbs. In the PRAE system, aberrant forms follow the плакати 3 class paradigmatic standard, while in the INF-AOR system and in ш- and н-participles they follow the нести 4c class paradigmatic standard. (See more details in Ch. 24, Supplement, § 910-913, Excursus on the contamination and competition between paradigmatic classes).


The system of forms is as follows (aberrant forms are shown after the // sign).

Among Aor forms there are canonical standard ones by class 5 (type гъвняхъ), canonical secondary by class 5 (root aorists of the r'us's type), and aberrant by class 4 (type гъгохъ).

Aberrant Imf forms are not attested in OCS, although for Church Slavic Vaillant gives the present imperfect form by class 3 (гыблѣахѫ, see Vaillant, § 207).

Note that class 5 is a new class relative to both classes 3 and 4c. The abundance of aberrant forms speaks to a not yet fully stabilized system of classes and paradigms during the examined time period. Subparadigms of different paradigmatic standards coexisted within a single lexemes; or, to put it differently, subparadigms had considerable autonomy (see Van Wijk).

# §610. Illustrations and distribution of aberrant forms

The aberrant forms under discussion are found in all sources (Table 610, p. 340), and are possible for all нѫ-dropping verbs, i.e. all verbs that contain a C-final root. On the distribution of the forms, see Vaillant, § 205–207.


Table 610. Illustrations of the class 5 and 3, 5 and 4 contamination aberration

Notes to Table 610


# Aberrant personal terminals

# §611. Decomposition of verbal terminals

Most personal terminals are combinations, i.e. can be represented as sequences of several formatives. Decomposing combined terminals allows a compact view of their aberrations. In combined terminals, the bundle marker is initial, followed by person and number markers. Root aorist terminals can be considered simple (and this type of aorist is sometimes called simple aorist), as long as initial е and о are treated as auxiliary formatives (in this case prothetic vowels). How combined terminals are representable as sequences of several components is shown in Table 611 (for brevity, 1Sg and 2Sg terminals and nonstandard imperfect bundles are not shown).

Within wholesale combinations, the mechanisms of alloform selection for individual formatives, as well as boundary adjustment rules, are, so to speak, forgotten. The whole combinations play the role of terminals, i.e. are distributed among sets, etc., as ordinary terminals, as if forgetting their bicomponential nature. The majority of observed aberrations are those of the final components (see aberrations 1–6, see Table 612.1 on p. 343); there is redistribution of components in aberration 7 (see Table 612.1).2

<sup>2</sup> Some aberrations, namely aberrations 1–3 from Table 612.1, can be interpreted as having a segmental origin (confusion of final *yers*).



y is replaced with w before front vowels £ and ▲. ㅇ in ٥.૪ ០٢ ٥.٥ is used in consonant-final stems. €/0 behaves by CVC-agreement with the final segment of the preceding formative.

### §612. Aberrant personal terminals

Tables 612.1 and 612.2 (pp. 343-344) give an overview of possible aberrations of personal verbal terminals, and illustrations. (On the distribution of aberrant forms, see Vaillant, § 147, 158).


Table 612.1. Forms with aberrant personal terminals: an overview

### Table 612.2. Aberrant forms of personal verbal terminals: illustrations



Table 612.2 (continued). Aberrant forms of personal verbal terminals: illustrations

Notes to Table 612.2


са: онъ же пристяпьши: њсте са за нозъ его: и поклонисте са емог М128, 8-9 ZOGR.

4) маріа жє магдалъні н марна носіфова: зьрѣаста кьде н полагааχж Mk 15, 47 AS (360 BALLETE ZOGR, MAR).

# Aberrant root vocalisms

### §613. General

This aberration is observed only in those verbs that canonically allow different root vocalism (see Table 613). These are irregular verbs with the unstable root vocalism paradigmatic effect (indicated by \*), and 4c class verbs that show root vocalism alternation in Imv (the решти 4с, тешти 4с, жешти 4с group). 3


Table 613. Aberrant root vocalisms: illustrations

Notes to Table 613


See § 552 on spellouts of the forms of the verb yours with the root you.

3) I слъщавъ цер тъ разгнъва сл и посъла воњ своњ: погоуви оувица тъл и градъю ихъ зажьже Mt 22, 7 мак; сf. с слъщавъ цръ тъ разгичва са: и посълавъ вока свою погоруви оувинца тъи и градъ суъ зажеже EUCH 106а, 24-106b, 3.

Note here also the form дъщетъ for canonical доушетъ, representing a contamination of the verb stems of дъхати 3\* and дъхати 7: дхъ иждєжє хощеть дъшетъ Jn 3, 8 AS.

# Alien workstem expansion

# § 614. Aberrant application of the alien workstem expansion effect

This paradigmatic effect applies in building canonical forms; cf. the paradigmatic standard of двигняти 5 class (expansion of INF-AOR workstem into ш-Part subparadigm), and also irregular class 4 verbs. A small number of aberrant forms that show this paradigmatic effect is found among irregular and unique verbs that lack an expanded stem (see Table 614).

Table 614. Aberrant application of the alien workstem expansion effect: illustrations


Notes to Table 614

1) Cf. заколень пръди лежн (for canonical лежитъ) усть того дълма закланъ Бъютъ и почто: да съмиритъ невесьнага: да и аггеломъ та сътворитъ дроуга SUPR 421, 30-422, 3; cf. also въранние SUPR 86, 13 и въранита SUPR 86, 14 (with yer insertion), but воренни SUPR 486, 27-28. Cf. also VoEOVEEHH Mk 6, 9 ZOGR (VOEOYBEN'SI MAR; cf. OB'BEN'S in Church Slavic).

Note that for the verbs клати 4h°, врати 4h°, млѣти 4h° ш-Part are not attested at all. However, cf. Church Slavic EpaBz (ши), cf. Vaillant, § 201. Here note the aberrant spellout простръвъ SUPR 311, 6 (cf. canonical простьоъ SUPR 353, 10), where, contrary to allotment rules, the V-final version of the labile root is used in the IMF system, and, accordingly, the C-final version of the w-Part suffix.

2) С. С. С. С. С. Маловръменьнааго сего житью оттъвръгъ СА: живе нъ Въ коен пештеръ еї чуть SUPR 514, 8-10, and: и обрътъ пештерж великах вь EDES SE PENSHE EN'S KANTENA 'A' ATT'S KH TOY SUPR 519, 1-2.


Cf. also in the Gospels:

# Other aberrations in the PRAE system

§615. Aberrations in substitutive softening in PRAE system

Aberrant forms are only observed among irregular verbs of the метати 3° group (substitutive softening contra expectation) and брати 4h• group (absence of substitutive softening contra expectation); see Table 615. (On Imv forms, see § 620 below).

Table 615. Aberrations in substitutive softening in PRAE workstem: illustrations


§616. Contamination of *i*- and *e*-conjugation in PRAE system forms

In this aberration, forms take terminals or suffixes of an alien conjugation (see Table 616). Such forms can be personal (Prae and Imv), as well as participles (see also § 620–622 below).


Table 616. Contamination of *i*- and *e*-conjugation

Notes to Table 616 (p. 347)

1) ERA 40%CAA BAWA Преповсана' 1 светильници горяштє Lk 12, 35 ZOGR, likewise in MAR and AS, but: и свътильници ваши горация in SAV; cf. also: СТОВЛЪЇ СИЛЬНААГО ЇЗОШТОЄНЪЇ: СЪ ГООЖШТІМИ ЖГЛЬМИ ПОУСТВЫННЪЇЇМИ PS SIN 119, 4; и погаста благословашта ба въ ровъ огнынъ кориштиїмъ SUPR 5, 24-26, but: и повель принести свъщта горашта: и прижагати лице пауде SUPR 13, 1-3.

§617. Aberrations involving initial vowel assimilation of the personal terminal (forms of the дълаатъ type)

In e-conjugation verbs, in vowel-final PRAE stems (ending in a, t, or of), the terminal-initial vowel є assimilates to the final vowel of the stem: дѣла + €™ъ → ДЕЛАЙ.ЄТЪ > ДЕЛАЕТЪ > ДЕЛААТЪ.

Further contractions of the type делатъ for делаатъ, ицелътъ for нцѣлѣатъ. See illustrations in Table 618.

Necessary conditions for these modifications are only found among verbs of the classes 3HaTh 4v, MHAOBATH 6, and ATEAATH 7, and also class 3, subtype дляти 3.

§618. Illustrations

Table 618. Aberrations involving initial vowel assimilation of the personal terminal: illustrations


Notes to Table 618

<sup>1)</sup> Cf. in nearby verses: същи слово съатъ си же сятъ час пяти: цдеже СЪЕТЪ СА СЛОВО" С ЕГДА СЛЪЩИТЪ" АБИЕ ПОНДЕТЪ СОТОНА" С ОТВИМЕТЪ СЛОВО

СЕНОЕ ВЪ СОВДЪЦИХЪ ИХЪ. Г СИ ТАКОЖДЕ СЖТЪ ГЖЕ НА КАМЕНЪНЪЦЪ СЪЕМИ Mk 4, 14-16 in MAR; cf. the same verses in ZOGR: съвм слово съетъ: се же СЖТЪ ЪЖЕ НА ПЖТИ СОТОНА' L ОТ'ЪРМЕТ'Ъ СЛОВО' СЪАНОЕ ВЪ СОЪДЬЦИХЪ ГУЪ' L СИ ТАКОЖДЕ СЖТЪ' ИЖЕ НА КАМЕНИХЪ СЕЕМИ.


### \$619. Note on the forms of the verb дълати in Supr

Among 2SgPrae forms with the quotative meaning 'you say' in place of canonical gruun (1x supr 402, 3-4) we find gewn (1x supr 303, 23), and frequently gthun (e.g., supr 306, 29-30; 307, 8). In the same syntactic function we find canonical 2Sglmv дъи (дъї supr 438, 23). Сf.: что оуво· деши волин ПРЕДАВЪН ИЮДА ХОНСТОСА" ПОНЕЖЕ ОУМЫ ХОНСТОСЪ НОЗЪ ЕМОУ" НЕ БЯДИ SUPR 303, 23-25; югда сьръте нъкого износима въ врать въ мрьтва: не тъчыж ли ПОНКОСНЯ СА ОДОБ И ВЪСТАВИ МОБТВААГО ДЪВШИ ЛИ МОЛИТЕТИ ТОВЕОВА ТЪГДА НА ВЪСТАВЬЮЕНИЕ ОУМЬОВШААГО" И ПАКЪІ ИНЪГДА" НЕ ТЪЧЫЯ ЛИ СЛОВО ДЕЧЕ НАДЪ ДЪВИЦЯ: ТАЛИФА КОУМЪГ ДАСТЪ РА РОДИТЕЛЕМА СЪДРАВЖ: ДЪВЪШИ МОЛИТИ СА ТОББОВА ТЪГДА SUPR 307, 1-8.

### §620. Aberrant forms of the imperative

Here we note, on the one hand, aberrant forms of the athematic imperative (вижди for canonical виждь), specific to EUCH, and on the other, imperatives that show 's-initial terminals, contrary to expected n-initial ones (BAЖATE for canonical BANHTE); see Table 620 (p. 350).

In the first case, a nonstandard terminal is replaced by a standard one (alien terminals in the paradigmatic effect).

In the second case, the nature of the paradigmatic effect depends on the class. For example, in class 3, the twofold rule is broken; in the 4h ° class, substitutive softening is absent (Kontre for canonical κολητε). In other cases, the effects combine somewhat differently. Generally all the forms show confusion of tautonymous terminals with initial n (cf. 2Pllmv итє) and initial ቴ (cf. 2Pllmv ቴፕዌ).

<sup>4</sup> The form покитъ Mt 21, 41 sav should be treated as an error for погоукитъ, сf. зълъ зълъ погоувитъ на in MAR and AS (not in ZOGR), rather than as a contraction from покиєтъ, as suggested by Ščepkin because in this verb class neither the SAV text, nor any other source, shows such contractions.

In sav, class 3 imperatives are represented exclusively by aberrant forms of this type (cf. Vaillant, § 149); other sources show aberrant forms alongside canonical ones.


Table 620. Aberrant forms of the imperative: illustrations

Notes to Table 620

1) гг гг призьри съ неси вижди с посъти винограда своего висн 11b, 10-11 (сf. EЖE СІЛЪ ОБОАТ!" ї Призьрі съ невесі ї виждъ: ї постеті вінограда своєго PS SIN 79, 15), but: ВИЖДЬ СЪМЪРЕНИЕ МОЕ С ТРОУДЪ МОН' И ОТЪПОУСТИ ВСА Гръузи мом виждь врагът мом теко очизножника са на мл EUCH 75а, 21-25 (cf. BIЖДЪ СЪМЕРЕНЪЕ МОЕ И ТРОУДЪ МОГ И ОТЪПОУСТИ ВЪСЊА ГОВУЋИ MOHA BINA BOAT'SI MOLA [ ... ] PS SIN 24, 18-19).

Note also in MAR: BH3%b Jn 20, 27, a single corrupt form.



Cf. in other gospels:

Cf. also BB31111"BTE EA I WIBA ERAET" AWA BAWA PS SIN 68, 33.


### \$621. Aberrant forms of M-participles

A small number of adjectival forms with the suffixes on or им that do not agree with the paradigmatic class of their verb (see Table 621) can be seen as aberrant M-Part forms, as well as extraparadigmatic deverbal derivatives.


Table 621. Aberrant forms of M-participles: illustrations

Notes to Table 621


### §622. Aberrant forms of шт-participless

Some class 4c verbs, as well as some unique verbs, show aberrant shapes of suffixed syncopated št-participle forms; see Table 622. First, these are forms with the suffix x or a instead of the expected 'L (forms like нєся, нєсѧ). Second, these are forms with a unique shape. Namely, in Glagolitic sources we find forms where instead of the letter '11, a special letter & is used (cf. €, the normal correspondent of Cyrillic a). This special letter is only found in such participle forms; in academic publications it is transcribed as 'p (nasal 2; Lunt 1974, 1.25, 4.311) or A (Vaillant, § 16). In Cyrillic editions of Glagolitic sources it is rendered as A, in forms such as неса, несам (for canonical несъм). The шт-Part of the verb ropern belongs here as well.


Table 622. Aberrant forms of шт-participles: illustrations

Notes to Table 622

1) не сал съ мноск на ма естъ' и не съвирали съ мном растачаетъ Mt 12, 30 SAV (IЖЕ НЪСТЪ СЪ МНОИЖ НА МА ЕСТЪ' 1 ИЖЕ НЕ СЪБИРАЕТЪ СЪ МНОЖ pactayaers Mt 12, 30 ZOGR, likewise in MAR and AS); cf. also: 046 живы

On forms that violate the distribution of шт-Part suffixes by conjugation (плачашти for navammin, ropmum for ropaurrh), see § 616 above.

и саз вы мнѣ: тъ творитъ дѣла Jn 14, 10 sav f. 25 (оць же пръстывалі вь MHE " T'B TBOMT'S Atha Jn 14, 10 SAV f. 100, likewise in ZOGR, MAR and AS).


PART III

Addenda

# CHAPTER 23 **Formative inventories: prefixes, roots, suffixes**

### §623. General

The goal of this part of the book is to give by list and to overview formative inventories for the three positional classes: prefixes, roots, and suffixes (terminal inventories are given in Part II, *Paradigmatics*, § 289 for nominals and § 455 for verbs). For prefixes and suffixes, complete lists are given, and for roots, lists of nontrivial classes. A complete list of roots is given in the root dictionary (RD).

For every positional class of formatives, the inventory uses a particular format for representing data, which is described before it is used.

# **On morphophonological representations**

### §624. Types of data

Full information on the morphophonological representation of a wordform assumes that the following data is available: 1) the sequence of formatives making up the wordform, i.e. the location of the formative boundaries and the segmental content of each component of that string (hereafter *parse*); 2) the position class of each component of the string (hereafter the *pRs scheme*); 3) which element of the inventory underlies the given occurrence of the formative (hereafter *identification*); see § 632–634.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

# §625. Two aspects of the problems

For each type of data, there are two kinds of questions, external and internal. The external aspect concerns the reader and how they can discover the solution adopted in the grammar. The internal aspect concerns the author and the choice of one solution over another. Formally the grammar author is not responsible for explaning all the ins and outs of the grammar, but basic politeness compels the researcher to open up some of the backroom secrets. Below, short explanations are given for all of these data types, first from an external, and then from an internal point of view.

# §626. Extraction of data from morphophonological representations

As long as some concrete morphophonological representation is given, its parse can be readily obtained, because it is given explicitly by periods. Cf. дѣвица, дѣ.в.иц.а; съвъплъчениѥ, съ.въ.плъч.ен.ьј.е; невѣста, не.вѣд.т.а; сластьнъ, сла.д.т.ьн.ъ; рещи, рек.т.и etc.

To derive the pRs schema, it is sufficient to find the root formative, which is given in the dictionary in parentheses; all formatives preceding the root are prefixal, and all formatives following the roots are suffixal, except for the last one, which is the terminal.1 Usually a wordform contains exactly one root formative; otherwise only in the rare cases of reduplication, cf. въз.⸨гла.гол⸩.а.т.и for възглаголати.

The following can be said on the identification of formatives. Identification of prefixal and suffixal formatives is not difficult. Every occurrence of a prefix, e.g. the prefix из, represents a single prefix из, cf. из.ѣд.т.и, из.бод.т.и, из.бы.т.и, из.мѣн.а, из.чит.т.и. Likewise, every occurrence of a suffix, e.g. the suffix т, represents one and the same suffix т, cf. из.бы.т.и, ѣд.т.л.и, за.бы.т.ь, из.бы.т.ък.ъ, без.чит.т.м.ен.ьн.ъ, дѣ.т.ѣл.ь, дѣ.т.ищ.ь, жѧ.т.ел҄.ь (indeterminacies that arise in some isolated cases can be resolved by list; see § 843). Identification of roots is less straightforward, because there are many homonyms among root formatives. However, every root is assigned a number that is shown in the dictionary. For example, we have съ.жѧ.т.и ‹290›, съ.жим.а.т.и ‹290›, and жѧ.т.ел҄.ь ‹302›, по.жѧ.т.и ‹302›; бедр.а ‹21› and бедр.ьн.ъ ‹20›, раз.бој.ьн.ик.ъ ‹23› and бој.ѣ.зн.ь ‹38›, etc.

### §627. Etymology and synchrony

Morphophonological representations treated in this book reflect morphophonology as part of the synchronic grammar of OCS, and as such need not fol-

<sup>1</sup> In some special cases the terminal is represented by two formatives, which is explicitly stated in terminal catalogs. A special case are some unique nominal lexemes, whose starting form contains zero inflection; see Ch. 12, the groups имѧ 0/n (имѧ=0), отрочѧ 0/n (отрочѧ=0), црькы 0/f (црькы=0). Wordforms of the so-called syncopated forms of щ- and ш-Part have no inflections at all (the sign |, cf. мол.ѧ|).

low etymological facts.2 Clearly, however, other things being equal, a choice of morphophonological representations that does not contradict etymology is preferable, even if only on the grounds of elegance.

In morphophonological representations, both the parse (and the entailed pRs schema that follows in most cases), and the identification can contradict etymology. Note that classic etymology is usually concerned with the origins of the root, focusing on external connections. In the general case, etymology not only fails to clarify problems of the parse, but does not even answer the question of whether some two words share a root. For example, the etymology that establishes the existence of two independent roots каз ‹362› (казнь) and каз ‹363› (казати), can remain silent on the attribution of the root formative in particular words to one or the other root, cf. проказьлѣти, проказа, or казньць. Luckily, Sadnik's dictionary takes the responsibility for establishing the "share a root" relation. But this dictionary does not show the parse. Thus, using this dictionary, one can determine whether two words, e.g. воскъ and вощага share a root (they do not: воскъ ‹112›/‹‹1100››, вощага ‹113›/‹‹1101››), but the segmental boundaries of the root and the pRs schema for вощага cannot be determined using this dictionary. Thus, the parse finds no support in existing lexicographic resources.

### §628. Parsing problems and roots: opaque stems

A general principle that forms the basis for dealing with non-obvious cases amounts to economy of inventories of atomic elements—roots, suffixes, and prefixes. Accordingly, a parse with smaller divisions is always preferable, i.e. a block of suffixes is never treated as a single unit in the inventory. For example, we have т.ел҄ (about 100 occurrences for the block), and not тел҄; ьн.ик (about 200 occurrences for the block), and not ьник; а.н.ьј (more than 200 occurrences for the block), and not аньј, etc.

Of course, with respect to the parsing problem, hardest are the stems that are represented by isolates with unknown or dubious etymology, cf. агода, бесѣда, чел҄ꙗдь, чловѣкъ, огавиѥ, голѣнь, халѫга. A significant number of such words (in particular, all the ones listed above) are treated as having an *opaque parse*, in that any attempt to discern a non-atomic structure and register the root in the root inventory is considered inappropriate. Instead of a morphophonological representation, such words are assigned an arbitrary spellout of the following type: ⸨агод⸩.а, ⸨бесѣд⸩.а, ⸨чел҄ад⸩.ь, ⸨чловѣк⸩.ъ, ⸨огав⸩.ьј.е, ⸨голѣн⸩.ь, ⸨халѫг⸩.а, and likewise in derivatives: ⸨чловѣч⸩.ьск.ъ, ⸨бесѣд⸩.ов.а.т.и.3


A subsequence of the stem can be opacified even in some words where the root is identified without effort. From a parsing point of view, the preor post-radical remainder is treated as opaque. This is the case, for example, in овысити [⸨овыс⸩.и.т.и] ‹133› (the prefix is opaque: от+ высити should have given отвысити, об + высити should have given обысити; by general rules; see also commentary to the lexeme овысити, § 492), and пѣсньнъ [⸨пѣсн⸩.ьн.ъ] ‹738›.

Subsequences with an opaque parse can be called *opaque stems* or *opaque roots* in this book. In the root dictionary and in the morphophonological representations of the paradigmatic dictionary all such subsequences are explicitly indicated by double parentheses. In the text of the grammar in mph representations double parentheses can be dropped or substituted by single parentheses.

### §629. Parsing problems and suffixes

Because the solution adopted and specified in the suffix dictionary makes it unnecessary to examine each individual lexeme—suffix тел҄, ьник, ан (see previous section) are simply absent from the dictionary—the parsing problem is solved mechanically. However, in some cases this generalization is violated, and parsing cannot be accomplished mechanically. Sometimes two adjacent suffixes turn out to be segmentally identical to a single suffix, cf. а.т in плак.а.т.и and ат in бог.ат.ъ, ѣ.н in трьп.ѣ.н.ьј.е and ѣн in гѫс.ѣн.иц.а. In every such case, the parse is accomplished individually for each lexeme or a grammatical class of lexemes on the basis of the totality of considerations which may not be explicitly spelled out. While establishing the suffix sequences а.т and ѣ.н for плак.а.т.и and трьп.ѣ.н.ьј.е is grounded in verbal paradigmatics, any attempt to identify a suffix ат in бог.ат.ъ or ѣн in гѫс.ѣн.иц.а inevitably leads into the derivational jungle.4

### §630. Parsing problems and prefixes

Outside of the cases examined in § 628 (a prefix can be pulled into an opaque stem, e.g. овысити etc.), there are no nontrivial situations involving prefixes.

opaque and etymologically transparent items. Among the latter, on the one hand, there are those such as говор (говорити), попел (попелъ), and, on the other, obvious borrowings such as порода 'paradise' or къмотра 'godmother' (Latin *commater*). On the other hand, monosyllables, regardless of etymology, are not assigned to the opaque class, cf. бичь, (бич).ь ‹26›. On opaque stems containing pronominal roots, see § 785 and ff. below.

4 This jungle, aptly dubbed *бужениноведение* (which can be loosely translated as "cranberry studies") by Sergej Krylov, fortunately is outside of the responsibility of the present grammar. Individual solutions are presented as arbitrary, i.e. not requiring supporting arguments. It is important, however, that the parsing task is not placed on the reader's shoulders, but given explicitly in the dictionary.

# §631. Problems with establishing of the pRs schema

Outside of the cases examined in §628, there are no nontrivial pRs scheme issues.

# §632. Problems with the identification of prefixes

In setting up morphophonological representations it is often difficult to determine whether a given lexeme contains the yerless variant of the prefix от or the prefix об, namely, in cases of stems beginning with the consonants п, б, м, к, г, х, с, з, ч, ж, ш, щ, Ж, ц, and ѕ. For example, the representation [об.(зъл).об.и.т.и] is established for озълобити, while [от.(сыр).ѣ.т.и] for осырѣти. Likewise for the prefixes въ and въз, cf. въсыпати, [въ.(сып).а.т.и], but въсиꙗти, [въз.(сиј).а.т.и]. See details in § 641.9°.

# §633. Problems with the identification of suffixes

In setting up the morphophonological representations, suffix identification problems are reduced to the following three situations: 1) one suffix vs. a string, cf. а.т or ат, see § 629 above; 2) alloforms of one suffix or different suffixes, cf. ак and к, see § 842, *Alloformy of suffixes*; 3) intersection of assortments, cf. ощ‖ост and ощ‖от (specific cases are noted *passim*).

# §634. Identification of roots

The choices for root identification made in this dictionary are close, to the extent possible, to Sadnik. In fact, for the purposes of the segmental grammar of OCS, it is unimportant how the lexemes containing the root каз are sorted into the two etymological bins, каз ‹362› and каз ‹363›. For many solutions adopted by Sadnik there are equally valid alternatives, but the choice of one solution over another may not always be justifiable convincingly. In such situations, of course, it is prudent to follow choices established earlier.

# Splintering

The case is different for roots where Sadnik, following etymology, adopts nonstandard alloformy that violates certain systemic principles. For example, consonantal alternations are possible in OCS only in the final-C segmental position, but not in the initial-C position; thus, formatives like град and жрьд cannot be connected by regular segmental pairings (i.e. morphophonological ones). Accordingly, Sadnik's unified root град/жрьд ‹‹253›› in the present dictionary is separated into two roots, град ‹192› and жрьд ‹293›, contrary to their known etymology. Let us list all roots that are separated contra Sadnik:


\* All lexemes with х-initial roots (cf. хрьстъ) are outside of the benchmark list of wordforms.

### Merger

In some cases, on the contrary, Sadnik's roots were unified (usually Sadnik's dictionary itself indicates such a possibility). For example, the root ‹‹573›› is given for words невѣста, невѣстьникъ, невѣстьство, and, separately, ‹‹1065›› for вѣдѣти and 73 other words. In this dictionary, these roots are unified. Let us list the roots that are unified contra Sadnik, ordered by number.


For pronominal roots, a technical solution is adopted without any pretense at etymology. Sadnik's dictionary gives incomplete information on etymology. See details in § 805, *A note on personal pronoun roots*.

# §635. The problem of establishing root alloforms

In some cases, morphophonological spellouts that differ in the segmental content of the root can be assigned to the same normalization. For example, for вѣꙗти ‹139›, along with the established morphophonological representation вѣ*i̯*.а.т.и

(вѣ alloform), there can be an alternative representation вѣј.а.т.и (вѣј alloform); for виꙗлица ‹95›, both the alloform вьј (as in the dictionary, вьј.а.л.иц.а) and виј can be set up. The choice can be more or less motivated in various cases, but a dose of arbitrariness is inevitable here.

# §636. The problem of borrowings

Clearly, classifying a given formative on the native ~ foreign scale is based on a variety of principles and yields different results depending on the overarching goals—etymological or synchronic. Indeed, from a synchronic point of view, external, morphophonological signs of alienness are important, such as an amorphous stem: violation of the CVC norm (a greater than monosyllabic "root", cf. вилитисъ 'type of pernicious insect' < Greek φυλλίτης, 1× Euch 59a 17), violation of edge conditions (V-initial stems, cf. ароматъ below), and prohibited sequences of phonemes (cf. геона below), and some others. A large class of new Greek and Latin borrowings, richly represented in the OCS corpus and excluded from the benchmark list of lexemes, shows such morphophonological signs of alienness. Such are, for example, акротомъ, апостольствовати, ароматъ, архитриклинъ, афедронъ, вилитисъ, газофилакиꙗ, ганъгрена, гистерьна, дидрагъма, димосии. However, from an etymological point of view, morphophonological anomalies are not a necessary condition of foreignness. Possible combinations are shown in Table 636.


Table 636. Morphophonological anomalies among borrowings

Because excluding borrowings from the benchmark list of words serves a synchronic, not an etymological goal, there was no attempt to draw a precise and justified boundary between these classes; it has been drawn arbitrarily. The native ~ foreign division in the present dictionary almost coincides with the division fixed in Sadnik (in the root dictionary): there, a large number of obviously alien words is excluded. More precisely, classes 1 and 2 are represented in Sadnik completely; classes 3 and 4, with a small number of lexemes (e.g. мрѣкориꙗ, крьстиꙗнъ, куциꙗ, кумиръ, луна, хрьбьтъ, гобино, въсѫдъ, порода are present, but адъ, връхъ, and връхосъ are not).5

5 Some obvious borrowings not included in Sadnik are found in our dictionaries—PD and RD. These are the following 18 lexemes: адъ, адьскъ, адовъ, адовьнъ, адовьскъ, аеръ, аерьнъ, архиереи, архиереискъ, архиереовъ, геона, геоньскъ, декѧбръ, китъ, китовъ, херовимъ,

# **Prefixes**

§637. On the autonomous use of prefixal formatives

Most prefixal formatives have a double function, as a prefix that is part of a wordform, and as an autonomous form, a preposition or a conjunctive particle.

All prefixal formatives that have autonomous uses as prepositions are shown in Table 640 (p. 365) with an asterisk.

The prefixal formative не functions as a conjunctive particle.

The prefixal formatives ни and нѣ combine only with pronominal stems, cf. ни.къто, нѣ.къто, and ни.къто.же.

In combinations of the form prefixal formative (без, из) + S or A, where (без, из) function as autonomous forms, namely prepositions, the boundary between the prefixal formative and the following stems is treated by mph⇒ph/norm rules (just as in combinations of prefixes with stems). Cf. бездрѫку as бездръпътьнъ, ичрѣва as ичисти, etc. (See details in Ch. 24, *Supplement*, § 868, *On wordform boundaries…*). Aberrant spellouts are possible in combinations with ad-prepositional forms of the pronoun \*и: canonical без н҄ѥго and aberrant бежн҄его (see § 76, *Aberrant cluster rewrite rules*).

# §638. Prefixes and opaque stems

Often it is impossible to reliably distinguish a construction of the type p+R from an amorphous root; in such cases it is treated as a morphophonologically opaque one. Recall that morphophonologically opaque words can contain etymologically transparent ones, cf. e.g. порода ‹697› 'paradise' (from Gk. παράδεισος), or онуща ‹637›, as well as roots with unknown or unclear etymology, as, for example, овоще ‹629› or потьпѣга ‹701›.

### §639. Prefixes and roots

Some prefixal formatives act as roots in some nominal stems. Cf. e.g. об: (об).ьщ.ь, (об).ьщ.ьств.о ‹623›;6 also (от).ьц.ь ‹647›.

херовимьскъ, хитонъ. Note that completely excluding all alien items (broadly defined) from the benchmark lists would lead to the absence of all amorphous and etymologically opaque words (including such lexically integrated items as господь and чловѣкъ). The resulting sterile wordlist would give a misleading picture on the reality of OCS.

<sup>6</sup> обьдо 'treasure' [⸨обьд⸩.о] ‹626› can also be assigned here. Sadnik treats it differently: обьдо and обьщь are assigned to different roots; cf. Vasmer's etymology of *обдо*.

# §640. An inventory of prefixal formatives

There are 28 distinct prefixal formatives. Morphophonological variants are listed across slashes; in other cases alloforms are separated by commas. The character of the alloformy is explained in notes.


Table 640. Prefixal formatives

# §641. Commentary

1°. C-final prefixes that lack V-final alloforms often show aberrant spellouts with ъ, more rarely with ь. Cf. безъврѣменьноѭ Supr 86, 30, изъходꙙщеSupr 267, 5, обьстоꙇмъ Lk 21, 20 Zogr.

In exceptional cases aberrant spellouts are shown in the dictionary. These are: безъчувьствьнъ, изъчитати, объходити, объходьнъ.

2°. All з-final prefixes behave relatively indeterminately before C-initial stems, with multiple repairs of resulting clusters. For example, for the verb из.чистити in the sources we find spellouts like изчист- (cf. Mt 8, 3 As), like ищист- (cf. Mt 8, 2 Zogr, Mar), and like ичист- (cf. Mt 8, 2 As). For без.числьнъ we find spellouts like бечисльн- (cf. Supr 565, 26) and like бещисльн- (cf. Supr 448, 13).

The lexicographic tasks of the choice of the lexeme name for the dictionary is often not a simple one: grammatical consistency conflicts with avoidance of fictional spellouts. The practical compromise struck here is based on the following: all possible spellouts, including aberrant ones, are ordered by loyalty to the declared canonical form, which is ordered first. Least systematic are spellouts in case of boundaries of the type з+(ц, ч) (for the boundary of the type з+ж, the result is stable, cf. въз.жѧд⇒въЖѧд 18×). See details in Ch. 24, *Supplement*, § 872, *On combinations with* з*-final prefixes*.

3°. въ/вън and съ/сън are morphophonological variants distributed by CVC agreement. Here are all lexemes with н: 1) вън: вънити, вънѧти, вънимати, въньмати, вънушити, вънѫтрьн҄ь; 2) сън: сънити, сънитиѥ, сънискати, сънорити, сънузьнъ, сънимати, съньмъ, съньмище, съньмати, сънѧти, сънѧтиѥ, сънѣдати, сънѣдениѥ, сънѣдь, сънѣдьнъ, сънѣсти.

Note that these formatives can be C-final also when used as prepositions. Cf. въ+ꙗдра⇒вънѣдра: молитва моѣ вь нѣдра моѣ възвратитъ сѩ Ps Sin 34, 13, but also cf. въ истинѫ Lk 1, 3 As.

However, V-final forms before vowels are possible, cf. съоблѣщи 1× Euch, съобразьнъ 1× Supr, въобразити 2× Supr, въорѫжити 2× Euch, 1× Supr. Cf. also the aberrant spellout съѫзи Supr 400, 14 (cf. canonical съвѫзъ).7 Likewise the adjectives въиспрьн҄ь and въиньнъ, formed from adverbial combinations въ+испрь and въ+инѫ. These forms are represented by isolated glosses in Cloz, Euch and Supr, cf. вынꙿнаѣ Cloz 6b, 39; въинъно Euch 89b, 1; отъ выспрьниихъ Euch 32a, 2–3; выспрьн҄иихъ Supr 458, 26–27.

4°. In the combination въ+ѣти, the prefix appears in the V-final variant. By the mph⇒ph/norm rules, the outcome must be въ#ѣ→въ#*i̯*ѣ→въꙗ, which agrees with existing Glagolitic spellouts: въѣді Lk 5, 4 As and въѣдѫ Lk 8, 23

<sup>7</sup> This form is absent in Večerka but is found in Meyer and Sadnik.

Mar. In the PD, the infinitive is shown as въꙗти, which corresponds to the normalized version of Glagolitic sources, cf. прѣѣхавъше Mt 14, 34 As, прѣѣвъше Mt 14, 34 Zogr, Mar and others.

5°. In the combination въз+ѣти, we have the standard sequence of a C-final prefix and a V-initial root. Sources have spellouts like възѣ-; in the PD, the infinitive is written as възѣти.8

In a combination of a C-final prefix with a j-initial root, the threat of a prohibited combination is eliminated by a nonstandard alloform with a final ъ; cf. въз+јарити⇒възъ+јарити (Supr възъꙗривъ 112, 17, възьꙗривъ 164, 21), из+јаснити⇒изъ+јаснити (Euch 53a, 25 ꙇзъѣсньшиимь). In the PD, the starting forms of these lexemes are shown with the following spellouts: възъꙗрити [възъ.(јар).и.т.и] ‹346› and изъꙗснити [изъ.(јасн).и.т.и] ‹349›.9 These spellouts are anomalous, because they contain nonstandard alloforms of the prefixes, възъ and изъ (these alloforms in their other occurrences indicate an aberrant spellout, cf. изъходити for canonical исходити, изъобрѣсти for canonical изобрѣсти). However, the spellouts възъꙗрити and изъꙗснити cannot be treated as aberrant, because canonical analogs that do not violate syntagmatic rules cannot be constructed: these forms are canonical, albeit morphophonologically anomalous.10

6°. The prefix вы only occurs in Ps Sin (выврѣщи, выгонити, выгънати, and выринѫти), and Cloz (вынести).

7°. The formatives не and ни are sharply distinct from all other prefixal formatives in their syntactic and derivational behavior. The particle ни, when combined with another lexeme, creates a separate lexeme only when the base is pronominal: къто and никъто, as well as ни.къто.же etc. The particle не can sometimes form a new lexeme, cf. невѣста or недѫгъ, while in other cases it shows a free association with a full lexeme, cf. не пьцѣте сѧ убо глѭще· что ѣмъ ли что пиемъ· ли чимъ одеЖемъ сѧ Mt 6, 31 Mar etc.11

The formative нѣ is only found with pronominal bases, cf. нѣкакъ, нѣколикъ, нѣкыи (see § 316).


<sup>8</sup> In Večerka's dictionary, the entry for възѣхати refers to the infinitive възꙗхати — a form that is graphically and morphophonologically impossible. Evidently, the reconstruction is based on an analogy with въꙗхати.

8°. While the rewrite rules for the combination E+ B require a resulting E, with B eliminated (as in OE + BHATTH = OEHATTH), in a few cases a different outcome is possible for this cluster, namely the elimination of E and a preserved B. Cf. obsichth [(obbic).h.m.n] (supr 1x: obbillaame ca 511, 27), and obztophtry [((овътор).н.т.н] (SUPR 3х: овъторити 569, 13; 570, 2; ов' торити 569, 22).12 See also the commentary to the lexeme овъкити \$ 492.

9°. In the transition from the morphophonological representation to the graphic one, oppositions between some prefix pairs are neutralized. These are 06 ~ 0T, BB ~ BB3.

Distinguishing o + X from o'r + X is quite difficult in cases where the stem begins with п, к, м, к, г, х, с, з, ч, ж, ш, шт, жд, ц, and s. In a few cases the indeterminacy is resolved by aberrant spellouts that reveal the prefix. For example, we have spellouts with объ (обь) for одрьжати, озирати, осъщити: объдоъжитъ SUPR 224, 26, 0663HDAA SUPR 502, 25, 056CBHHTTB SUPR 10, 25; spellouts with отъ for оходити, окръгти: отъхода Mt 25, 14 zogr, отъкръвша Mk 2, 4 zogr. However, for a significant number of lexemes, a reconstruction of the morphophonological representation as or + X or or + X is arbitary.13 This is the case, for example, for oE.( жег).т.н for ожешти, ов. ( дежд ).а for одежда, ов. ( зъл).ов.и.т.и for 0336106HTH; VS. OT.( LEWA ).A.T.H for outh MATH, OT. Про. ( BOTET ).T.H for ongoBOBLUTH, от. (съю). в.т.н for осъювти. For several words, a spellout with овъ, which is found in the texts, is preserved in the lexical entry, even though it contradicts the canonical standard. This is the case in or broad ( org. ( yog).bn.bl along with оходыъ for [от.(ход).ьн.ъ], and объюдити [объ.(χολ).н.т.н] along with oyogumm for [ov.(xo2).h.T.h]. Several items remain homonymous: octranema [0E.(cTmn).H.T.H] 'surround', and ocTRUHTH2 [oT.(cTRRI).H.T.H] 'step away'.

Distinguishing BB + X from BB3 + X is difficult in those cases where the stem begins with c, 3, ч, ш, шт, жд, ц, and s. In such cases the choice made in the dictionary is arbitrary. Cf. въ.съпати, but въз.сверъпъчи.

10°. The prefix or/om has two variants (alloforms) whose distribution is lexical. The same lexeme or even wordform shows spellouts in sources with over as well as or. Cf. orBptcTh: usually without the yer, but in SUPR, MAR, SAV and PS SIN we find forms with yers: 0T'BB0L3ET'' Ca Lk 11, 9 SAV 123; 0T'BB0'L23; 0T'BB0'L23 Mt


12 овоштє for ог + воск сап регhaps be included here; cf. Belorussian обоч. 13 Let us list the material for the stems (ov, on's) + χολ and οπ + χολ:

25, 11 Mar; отъвръзошꙙ Supr 301, 1; отъвръзъшю Ps Sin 103, 28, ѡтъвръзе сѩ Ps Sin 105, 17. In such cases the spellout is treated as aberrant. Note that in yerless cases, cluster transformations create the threat of homonymy between от+*X* and об+*X* (see 9° above).

Cf. aberrant forms with отън- in the lexemes отьмати, отѧти: отънемьѭщъ Supr 297, 4; отънемьѧи грѣхъ миру Supr 331, 25; отънꙙ ти убожьство Supr 348, 5–6.

11°. The prefix па only occurs in nominal lexemes. These are пагуба, пагубьнъ, пагубьникъ, пажить, памѧть, памѧтиѥ, памѧтивъ, паѫчина (and aberrant паѭчина). Cf. пазуха, пазнегъть, папрьтъ with opaque stems.

12°. The prefix пра only occurs in nominal lexemes, and all have doublet forms with прѣ: прадѣдъ//прѣдѣдъ, etc. For the distribution of doublet forms among sources see Večerka.

13°. In при+ити, the resulting form is always contracted: прити (unlike the later Church Slavic приити). Cf. ꙇ бываатъ дрѣво· ѣко прити птицамь нбскымъ· ꙇ витати на вѣтвехъ его Mt 13, 32 Mar; помѣни мѧ ги· егда придеши въ цсрствии твоемь Lk 23, 42 Mar. The contracted form is treated as canonical, and is given as the name of the lexeme in the PD, even though приити does not violate any morphophonological constraints (cf. прииЖити).

14°. The variants су and сѫ are lexically distributed. The stem сугуб- always has су; both variants are possible with the root мьн ‹550›, and spellouts with су are treated as aberrant in the PD; in other cases, сѫ is found.

15°. The variant ѫ only occurs in nominal stems; these are ѫдоль and ѫдолиѥ, ѫтьлъ, ѫсобица, and ѫродъ, with derivatives. Thus, у and ѫ are lexically distributed alloforms.

16°. Words with the prefix раз can have aberrant forms with роз, cf. разбити and розбити, различьнъ and розличьнъ. Spellouts with роз are mainly found in Supr, but even there forms with раз predominate. For example, разбоиникъ occurs about 30 times, while розбоиникъ only twice.

17°. The formative наи is found in the benchmark corpus only in three adjectives: наитрѣбл҄ии, наивѧщии, наискорѣи, the latter two only in adverbial uses. Cf. also the extraparadigmatic form наипаче (see Vaillant, § 91).

18°. For прѣди, cf. прѣди текъ Lk 19, 4 Mar, Zogr and прѣдтекъ As, see Vaillant, § 242–243.

19°. In autonomous use:

\*EE3 - EE3 OTTBLA Mt 10, 29 MAR (cf. EE36 OLA AS, EE30 OLJa ZOGR); ETAA ПОСЪЛАХЪ ВЪ! БЕЗ ВЪЛАГАЛИШТА" 1 БЕС ПИРЪ!" 1 ЕЕСАПОГЪ' ЕДА ЧЕСО ЛИШЕНИ БЪЮТЕ Lk 22, 35 MAR; N'ECT'S ПООООКЪ БЕШТЪСТИ' ТЪКЪМО ВЪ СВОЕМЬ ОТЪЧЬСТВИI Mk 6, 4 ZOGR (cf. EEYECTH MAR); cantent Bpations EE3A0MKO' 0 CEE'S BLSATH CA BEAHT'T SUPR 465, 29-30;

\*ВЪ - ТЪ ЖЕ ЕГДА МОЛИШИ СА ВЪНИДИ ВЪ КЛЪТЪ ТВОИЯ И ЗАТВОИ ДВЪОН ТВОЊ: ПОМОЛИ СА ОТЦОУ ТВОЕМОУ ВЪ ТАННЪ: С ОТЦЪ ТВОИ ВИДАН ВЪ ТАННЪ" ВЪЗДАСТЪ ТЕБЪ АВЪ' МОЛАЦЕ ЖЕ СА НЕ ЛИУО ГЛТЕ' ЪКОЖЕ И ЊАЗЪРУВНИЦИ' МЫНАТЪ EO CA EKO BO MBHOS' FANH CROEML OYCATHIANH EXAMT'S Mt 6, 6-7 MAR;

\*BB3 - принахомъ: слагодъть въз благодъ тъ Jn 1, 16 ZOGR; въздаахж ми 36, 30500 A PS SIN 34, 12; BBC 47570 MLCTHTH XOWTETE EOSOBH SUPR 388, 15-16; въс кран же въаше извра вань: раждежена SUPR 76, 17-18;

\*до - прискръсъна єстъ діпа мов до съмръти Mt 26, 38 ZOGR; аштє имаши что до врага своюто: истръгни гневъ SUPR 421, 26-27; въсходнатъ до небсъ ї НИЗЪХОДНАТЪ ДО БЕЗДЕНЪ PS SIN 106, 26;

\*3a - вьен оусо оученици развъгошь са за страха нюденска SUPR 483, 11-12; слъщасте тько речено въг. око за око с зявъ за зявъ Mt 5, 38 ZOGR, MAR; H MHOSH ПАЧЕ ВЪРОВАША ЗА СЛОВО ЕГО: ЖЕНЪ ЖЕ ГЛААХЖ: ВКО ОУЖЕ НЕ ЗА ТВОНА БЕСЕДЖ ВЪРОУЄМЪ: САМИ ЕО СЛЪЩИХОМЪ Jn 4, 41-42 AS; ЕГДАЖЕ 13ГЪНАНЪ Бъ Народъ: Въшьдъ же мутъ тя за рякж: и въста дъвнца Mt 9, 25 zogr;

\*H3 - 0EAYE E'S H3EABIT'S ADAR MOHR H3 ORK'SI AAOB'S PS SIN 48, 16 (but: ЇЗВАВІ МЊА ИЗДОЖКЪІ ВРАГЪ МОНУЋ PS SIN 30, 16); СЖТЪ ВО КАЖЕНИЦИ ИЖЕ ИС ЧОВВА МГЕОВ БЪША ТАЦИ Mt 19, 12 SAV (but: ИЧОВВА MAR, AS), "ЧОВВА ПОВЖДЕ ДЕНЬНІЦІА РОДИУЪ ТИ PS SIN 109, 3 (but: ШТРЪВА PS SIN 21, 10); ЗАПОВТИ ЕМОУ ИСЪ ГЛА: ОУМЛЪЧИ ИЗИДИ ИЗ НЕГО: 1 СБТОЖСЪ 1 ДУЪ НЕЧИСТВИ: 1 ВЪЗЪПИВЪ гласомъ велиемь сзиде нж него Mk 1, 25-26 мая (сf. г запретти емоу нс глах оумльчи" и изиди ижнего" и сътрасъ и дуъ нечистър" и възъпнвъ гласъмь велнемь" I H3HZE I.WHETO ZOGR);

\*на - да вждетъ воль твов: вко на нен с на земли Mt 6, 10 zogr; вко слъньце СВОЕ СЪЪАТЪ НА ЗЪЛЪЩ И ЕЛАГЪР С ДЪЖДИТЪ НА ПРАВЕДЪНЪ! И НА НЕПОАВЕДЪНЪВ Mt 5, 45 MAR; и не хотъаше на длъя времени Lk 18, 4 AS;

\*надъ - самарьнинъ же єтєръ градуві придє надъ нь Lk 10, 33 ZOGR; кто ОУБО ЕСТЪ ВЪЗЪНЪ РАБЪ И МЖДОВЪ: ЕГОЖЕ ПОСТАВИ ГЪ НАДЪ ДОМОМЬ СВОИМЪ Mt 24, 45 MAR; гъ вь сюнъ велен и въсокъ есть: надо вьсъм людьмі PS SIN 98, 2;

\*06 - в пръедж на землях гадаринъски (ня). Еже єсть об онъ полъ галилън Lk 8, 26 MAR; L 075B'SWTAB'S CHMON'S DEYE EMOY: HACTABLHHYE' 06 HOWTL BLCZ Троуждыше са: не њудить ничьсоже по глоу же твоемоу въвръжемъ мръжат L CE COTECODE OEALIA MHONKETTES POLES MHOLO' ΠΡΟΤΏΣΖΑΛΥΣ ЖЕ СА МОЋЖА IX' Lk 5, 5-6 ZOGR;

\*OT'S - L CABILLAB'S HE OTHILE OTT TOYA'S B'S KODAEAH' B'S noyer's M'ECT'O единъ с слъщавъше народн: по немь идж пъщи отъ градъ Mt 14, 13 мак; и НЖЕ НА КРОВЪ ДА НЕ СЪЛАЗИТЪ ВЪ ДОМЪ: НИ ДА ВЪНИДЕТЪ ВЪЗАТЪ ЧЕСО ОТЪ ДОМОУ своєго Mk 13, 15 мАR;

\*по - wtrabtsштавъ же петръ рече емоу ги аште тър еси; повели ми прити къ тевъ по водамъ Mt 14, 28 мак; тамо осво: иже снидеаше по провъемъ: оуже

НЕ ИЦЪЛЪАШЕ: |... | НЪ СЬДЕ ПО ПОВЕЕКСМЪ ВЪТОРЪИ СЪЛАЗИТЪ: ПО ВЪТООЪЪВА третии и четврътъю SUPR 496, 15-20;

\*ПОДЪ - КОЛЬ КРАТЪН ВЪСХОТ" БУЪ СЪБЪРАТИ ЧАДА ТВОВ "ЕКОЖЕ КОКОШЪ ГИЪЗДО своє подъ крилъ: и не въсхотъсть Lk 13, 34 мая; гла ємоу натаналь: како ма ЗНАЕШИ" ОТЪВЪ IC И рече емоу" прежде даже не възгласи теве филипъ: съща подъ смоковных видехъ та In 1, 49 SAV; оушрилъ сстъ стопъл мом подо мъном PS SIN 17, 37; BBC E ПОКОРИЛЪ EU ПОДЪ НОЕ ЕГО: ОВЪЦИ I ВОЛЪІ ВВАТЫ BS SIN 8, 8;

\*ПОН - НИШЪ ЖЕ БЪ ЕДИНЪ ИМЕНЕМЪ ЛАЗАРЪ' СЖЕ ЛЕЖААШЕ ПОН ВРАТЪУЋ ЕГО гноннъ Lk 16, 20 мая; и нжждаастє и гліжціа: облади сь нама: "вко прі вєчєр'в ECT'B Lk 24, 29 AS;

\*ПОВДЪ - ПЕТОЬ ЖЕ СТОГАШЕ НА ДВОРЪ И ПОИСТЖПИ КЪ НЕМОУ ЕДИНА РАБА ГЛЖЦИ: И ТЪІ БЪ СЪ І́СМЪ ГАЛИЛЕІСКЪІМЪ: ОНЪ ЖЕ ОТВОЬЖЕ СА: ПОВДЪ ВСЪМИ ГЛА не въдъ что глешн. ишъдъшко же емор предъ врата: оузъръ и другата рава и гла ємоу тоу и сь въ чакъ съ смъ назаръннюмъ Mt 26, 69-71 sav 112; ОУГОТОВАЛЪ ЕСІ ПРЕДО МНОЖ ТРАПЕЗЖ ПРЕДЪ СЪТЖЖАЊЕТИМИ МНЕ PS SIN 22, 5;

\*съ - и огнь великъ нача жешти и пакоже съпадъщоу са конта оумирати SUPR 221, 27-29; не погоуві съ нечъстівънням дшых моєма и съ мжжі Крывы (for кръвии GPI) живота моего PS SIN 25, 9; ї скомьчаших свя въ соуєт'в дьні ихъ ї лѣта іхъ со тъштаниемь ps SIN 77, 33;

\*01 - МАРИЪ СТОВАШЕ ОУ ГООБА ВЪНЪ ПЛАЧИЯЩИ СА: ПОЛНИЕ ВЪ ГООБЪ: И ВИДЪ Дъва анкла: Въ Евлахъ ризахъ съдици: едного од главъ! и единого от ногоч" НДЕЖЕ БЪ ЛЕЖАЛО ТЪЛО ИСВО Jn 20, 11-12 AS.

# Roots

#### 8642. General

Roots are the largest and most diverse class of formatives. There are about 1000 distinct roots in the benchmark corpus, including opaque ones, of which there about two hundred. They are all presented in the root dictionary (RD), where for each root its alloforms are shown explicitly, and for each alloform, all lexemes that represent that alloform. Most roots are morphophonologically trivial and require no special discussion. At the same time, a relatively small group of root formatives (about 250) show various morphophonological peculiarities that require grammatical notes. The purpose of this section is to show and in some cases comment all such nontrivial roots.

Trivial are those roots where the vocalism is stable, the CVC schema is stable (i.e. the same in all alloforms), and where it agrees with the CVC norm. The diversity of assortment for trivial roots is limited by the application of two free alternations: velar palatalization for roots ending in K, r, or x, and substitutive softening for others. Here are examples of trivial roots: < 69> Ertr (ESTATH), <70> БЪД (БЪДА), < 215> ГЖС (ГЖСЕНИЦА), < 217> ДАВ (ДАВИТИ).

# §643. Root classes by main morphophonological features

Nontrivial roots are divided into several classes; below each class is treated separately. Even though certain morphophonological features of roots are at the basis of this classification, it is not a properly grammatical one; its purpose is to provide a clearer overview of a rather diverse mass of nontrivial root formatives. The membership of all classes is given by lists.

Two groups of *unique* roots stand separately: *anomalous* and *pronominal* ones. These groups include roots with peculiar assortments of alloforms, which do not fit with standard types of alloformy (see § 82 above). The division of these roots into anomalous (see list in § 806) and pronominal (see list in § 785) is relatively arbitrary.

All other nonstandard roots are divided into groups by two features crucial for this classification. The first feature concerns the vocalism: it can be stable or unstable, and in the latter case pure or sonant (cf. § 121 above).

The second feature concerns the CVC structure. For roots with sonant vocalism, the division by CVC structure type differs from the division that applies to roots with stable or unstable pure vocalism.

For sonant roots we have: 1) stable CHC structure (e.g. врѣг ‹121›); 2) CH structure (e.g. мрѣ ‹539›); 3) HC structure (this is only the case for the root льп ‹516›); 4) H structure (such are only ор/ра ‹639› and ьм/ѧ ‹334›). All H-final roots are called *labile* (see § 128 above). All other possibilities are found only among unique verbs.

For other roots we have: 1) stable CVC structure (e.g. бѣд ‹70› or рок ‹766›), the so-called *closed* roots; 2) stable CV structure (e.g. зна ‹319› or сто ‹879›), the so-called *open* roots; 3) stable VC structure (e.g. ок ‹634› or ор ‹12›). All other possibilities are found only among unique verbs.



### §644. Order of exposition

Table 644 (p. 372) shows the classes of roots and the order of exposition below. The paragraph number contains the membership list for the corresponding class. For sonant vocalism roots, the order corresponds to the following order of the sonants: H*(n*), H(*m*), H(*n*/*m*), H(*u*), H(*j*), H(*r*), H(*l*). Within each group, roots are treated in order by number. The counts are given without opaque roots.

### §645. Order of illustration for alloformy

The overview below presents alloformy data as follows. First, alloforms generated by consonantal alternations are not noted in the general case, because all consonantal alternations are free. For example, for root ‹766› the alloform роц is not explicitly shown, but only the alloform рок is (cf. пророкъ, NPl пророци). On the other hand, for root ‹23› (бити), the alloform бьј, represented only in oblique wordforms (forms of the PRAE system) is noted, because the alloforms би and бьј correspond by the fundamental vowel alternation, which is not free. Note that alloforms found only in oblique wordforms of some lexemes cannot be discovered from the root dictionary, which only lists starting forms of the corresponding lexemes.

### §646. A note on unattested alloforms

Not all roots show all possible alloforms by all licit alternations. A pedantic approach to attestation in sources of various alloforms does not serve a substantive purpose from the point of view of grammar. Let us list a few examples. The root клок ‹385› is only represented by the lexeme клокотати; сох ‹870› only by the lexeme посоха, which in turn is found in the hapax gloss посохами (Supr); the root лук ‹507› is found in the hapax gloss луцѣ (Euch).14 Likewise for substitutive softening: the root бѣд ‹70› shows the alloform бѣЖ (cf. побѣЖати), while the root блѣд ‹32› does not show a substitutively softened alloform.

In the overview of roots, unattested alloforms are not shown; e.g. the root ‹570› is represented by two lexemes, муха (мух.а) and мъшица (мъш.иц.а). According to the velar palatalization alternation, we get the following alloforms: мух, мус, муш and мъх, мъс and мъш. The alloform мус is expected in the oblique forms of the lexeme муха (e.g. LSg мусѣ), but this lexeme is only found in a few NAPl glosses (мухы). In fact, of the six expected alloforms, only two are attested. Thus, in Table 712, the ъ vocalism grade is illustrated by the alloform мъш (and not \*мъх).

On the contrary, if several alloforms corresponding to different consonantal alternation grades are attested, the table illustrates only the leading grade alloform. For example, in Table 705, the alloform губл҄ is not shown, even

<sup>14</sup> The starting form лукъ is reconstructible for this gloss, accordingly, the RD lists the root ‹507› by its alloform лук, which is, formally speaking, unattested.

though it is expected in the paradigm of e.g. the verb погубити, and is represented in the lexeme погубл҄ꙗти. For root ‹810›, Table 674 lists the alloforms свьт, свит, and свѣт, while the alloforms свьщ, свищ, and свѣщ are not listed. Likewise, for root ‹1050› in Table 676, only the alloforms цвьт, цвит, and цвѣт are listed. However, while for root ‹810› the benchmark corpus attests substitutively softened alloforms, this is not the case for root ‹1050› (only цвьт, цвит, and цвѣт are attested).

# §647. A note on reconstructions of fictional alloforms and sources of alloformy

A stipulation that consonantal alternations are free, in place of fixing all attested alloforms, leads to the risky business of segmental reconstructions. We should, first, distinguish a reconstruction from a leading to a following grade (from observed мух to unobserved мус) from the reverse direction (from observed выс to unobserved вых), and treat the former as better-grounded in synchronic grammar. Note that following grades of the substitutive softening alternation have intersecting series, and thus establishing the leading form of roots represented only by following grades in the majority of cases is impossible without etymological evidence (cf. дъщ ‹267›: дъщи, нищ ‹611›: нищь, плищ ‹678›: плищь, вещ ‹94›: вещь, вѣЖ ‹142›: вѣЖа, дъЖ ‹263›: дъЖь, мѫж ‹601›: мѫжь, and many others). Second, we must distinguish intraparadigmatic reconstruction (e.g. лукъ for the hapax gloss луцѣ Euch 54a, 10) from interlexemic reconstruction (cf. мъш in мъшица to мъх), treating the former as better-grounded in synchronic grammar.15

### §648. A note on variant spellouts of roots

1°. Some roots have unstable spellouts in the same lexemes, which show variability, not alloformy. Such variants are either (1) doublets, where both competing spellouts are canonical (cf. авити//ꙗвити, авл҄ѥниѥ//ꙗвл҄ѥниѥ), or (2) aberration, where one form is canonical and the other aberrant (cf. нѫЖа and нуЖа), or, finally, results from accidental mistakes by the scribe (cf. хлѫбаѩ Mk 10, 46 Mar for canonical хлѫпати or хлꙙбьныи Supr 135, 6 for canonical хлѣбьнъ).

2°. Some types of variation are grouped as segmentally analogous, where a few roots show the same variation, cf. рыб~риб, and кры~кри. Some authors treat these as segmental aberrations. Such are, for example, ры~ри, ра ~ро, рѣ~ра, ла~лѣ; see details in Vaillant, § 26–29 and ff., and also Diels, § 9–40. Such variants are listed in commentaries to particular lexemes that are found in the overview of nominal and verb classes.

<sup>15</sup> Note that the reconstruction of all paradigmatic forms is taken as an axiom in synchronic paradigmatic grammar both for dead and living languages (albeit usually an implicit one). Otherwise only in special cases, where the grammatical investigation focuses on a very limited corpus of texts or on unstable grammatical parameters, as e.g. accent.

# Roots with stable vocalism: standard

### §649. Examples

Here are some roots of this class:




Alloform чищ is also in the paradigm of чистити, cf. чищѫ, 1SgPrae (чистити). Alloforms цѣст and цѣщ are assigned to a different root ‹1059›, contra Sadnik.

### §650. A note on roots ending in к, г, х

Some roots show interlexemic alloformy in the final C position by the pairings к‖ц, г‖ѕ, and х‖с. For example:


The same effect is found in some suffixes, e.g. ик and иц (see § 862). The number of roots with this effect is strictly limited: outside intrapardigmatic alloformy, replacements by the к‖ц and г‖ѕ alternations lacks the freedom of the velar palatalization alternations. See also *A note on the pairings* ц*‖*ч *and* ѕ*‖*ж in § 108, 115. The effect noted here in historical grammar is attributed to the socalled *third palatalization*.

# Roots with stable vocalism: nonstandard

### §651. General

All alloforms of some root can be nonstandard, i.e. violate the CVC norm (e.g. ок and оч ‹634›); or, one alloform can be standard and the other nonstandard

(e.g. да and дад <216>). All roots that contain nonstandard alloforms along with standard ones are in the unique class. In the present section we treat all roots whose alloforms all have the same CVC schema that departs from the CVC norm. These are: first, V-final, or open roots (e.g. 3Na <319>; see Table 652), and, second, V-initial ones (e.g. orn <631>; see Table 654).

### §652. Open roots

Table 652 shows all stable V-final roots; there are 29 of them. Commentary to particular roots follows.



### §653. Commentary

1°. про <713>, прв <719>. Prefixal formatives function as roots (see об <623>, oT <647> below).

2°. Not to be confused: na <480>, naram+, natish 'seek, catch', and na <479>, лагати2 'bark (of dogs)'.

3°. чи <1072>. С. кој <401>. On distinguishing the roots see § 634.

4°. доу <259> надоути. Contra Sadnik, the lexeme дъмъ is excluded from the root goy <259>, and listed under root <268> in the present dictionary.

5°. wt <586>. Contra Sadnik, in addition to the verb contrant the lexemes нзмѣти and измѣннє are also treated under this root here. Večerka and Sadnik assign these lexemes to the headwords измънити <589> and измъявния < 589>. For H3METH Cf.: tko pa3ropt cha cprzdele noe mirpora mots 13MtT's cha, 2-3SgAor PS SIN 72, 21; for n3Mthile cf.: поносишь измение ха твоего ps SIN 88, 52.

### §654. Vowel-initial roots

Table 654 contains all V-initial roots, excluding unique ones; there are 41 such roots. Commentary on particular lexemes follows.




### §655. Commenary

1°. arn <3>, aB <1>. The vowel-initial anomaly can be repaired: cf. aberrant trubus (CLOZ); canon has doublet forms aBHTH and aBHTH (and others sharing this root); see details of distribution among sources in Večerka.

2°. EA0 <1115>. In combination with the preposition Bb in SUPR and PS SIN, editions show a space, e.g. BB HEADA. Cf. H MONHTBA MOTE BL HEADA MOTE ВЪЗВОАТИТЪ СНА PS SIN 34, 13.

3°. нг <330>, иг <331>, игр <332>, им <333>, иск <335>, искр <336>, нскр <337>, испр <339>, ист <340>, ист <341>. All these roots have a morphophonological и; an alternative treatment (jь or is) is only possible in roots that

show clear evidence of ь. These are the pronominal root ьн ‹1112› (инокъ) (cf. отьнѫдьнъ, see § 804), and ьм/ѧ ‹334›, a root with unstable sonant vocalism (cf. възьмати, see § 691).

4°. об ‹623›, от ‹647›. Prefixal formatives act as roots (see above про ‹713›, прѣ ‹719›).

5°. огн҄ ‹631›. On the vacillation between огн and огн҄ see § 403.

6°. ап ‹10›. Cf. the aberrant вънезаѣпѫ Mk 13, 36 Zogr; ѫч ‹1121›, cf. aberrant spellouts, паѭчины Supr 274, 14–15, and паучина Ps Sin 89, 10.

7°. уз ‹1012›. Only in сънузьнъ, hapax gloss пѣши и сънузни по н҄ему хоЖаахѫ Supr 90, 7–8. Sadnik notes the possibility of identifying this with the root вѫз/вѧз ‹148› (cf. съвѫзъ 'connection').

8°. утр ‹1019›. The aberrant spellout ютр is possible; for distribution among lexemes and sources see Večerka.

# Roots with unstable pure vocalism: standard and nonstandard

# §656. General

Roots with unstable pure16 vocalism differ in the number and composition of the vocalic realizations found in them. However, this property cannot form the basis of a useful classification, because the absence, in the benchmark list, of lexemes showing one or another vocalic realization of a root is entirely random. For example, the lexeme блѣскъ is not attested in OCS, but is found in Church Slavic, which is why in OCS this root ‹31› блиск has a different assortment of vocalic realizations from ‹1050› цвьт and ‹810› свьт.

Nonstandard roots with pure unstable vocalism that violate the CVC norm also belong here. They are ‹12› ор‖ар (раз.ор.и.т.и), ‹819› си‖сѣ (си.л.о), and ‹879› сто‖ста (сто.ѣ.т.и).

Table 656 (p. 379) lists the roots in order by number. There are 68 roots with unstable pure vocalism.

§ 657–676 provide commentary on individual roots and groups of roots.

<sup>16</sup> On the difficulties of matching a given root with a series of fundamental alternations, as well as on the opposition between pure and sonant roots, see § 677–680 below.


Table 656. A list of roots with unstable pure vocalism





# §658. Roots with unstable є|| ৳ vocalism


# §659. Roots with unstable ь||и vocalism



# §660. Roots with unstable ъ‖ы vocalism


number <sup>ъ</sup> <sup>ы</sup>

Cf. among morphophonologically opaque roots лобъз/лобыз ‹500›.


# §661. Roots with unstable е~о


The vocalic change е~о (the old ablaut) is found in combination with a wider variety of vocalic realizations, such as грети/гробъ ‹190›, °мести/мотыла ‹545›, плести/оплотъ ‹676›, рещи/рокъ ‹766›, тещи/токъ ‹941›, вести/воЖь ‹75›, нести/поносъ ‹607›. Note that the е~о pairing is not an adjacent one.

# §662. Roots with unstable ѣ~а vocalism


# §663. Roots with unstable и‖ѣ vocalism




\* Not to be confused with стрѣг ‹891›; see § 767.

# §664. Roots with unstable ъ‖о vocalism


\* Contra Večerka and Sadnik, where the headwords contain the spellout грьм even though sources only have гръм (Supr, Ps Sin). See details in § 681.

Cf. хът in aberrant forms from the root хот ‹1038›, e.g. хъщѫ Supr 534, 11.

# §665. Roots with unstable ь‖е vocalism


\* The vocalic realization with ь only in Imv; cf. рек ‹766›, тек ‹941›, жег ‹279›. Contra Sadnik, the root ‹700› потити, потъ, потьнъ is excluded from the root пек ‹660›. From a synchronic point of view, пек and пот are not connected by standard segmental pairings.

§666. Root ‹75› §667. Root ‹190›


\* Only in nonstandard aorist forms.


\* The vocalic realization with ь only in Imv; cf. рек ‹766›, тек ‹941›, пек ‹660›.


\* Only in nonstandard aorist forms.


§668. Root ‹279› §669. Root ‹545›


§670. Root ‹607› §671. Root ‹609›


### §672. Root ‹676› §673. Root ‹766›



\* The vocalic realization with ь only in Imv; cf. жег ‹279›, тек ‹941›, пек ‹660›.

§674. Root ‹810› §675. Root ‹941›



\* The vocalic realization with ь only in Imv; cf. рек ‹766›, жег ‹279›, пек ‹660›.

§676. Root ‹1050›


# Roots with unstable sonant vocalism: standard and nonstandard

## §677. A note on the notion of sonant vocalism

The expression "the vocalism of a formative is sonant" can have both an etymological and a synchronic sense. Etymologically, it refers to the source of the vowel. E.g. the formative брѣг (брѣгъ) has an etymologically sonant vocalism, which follows from external comparison.17 Synchronically, the question is of the series—pure or sonant—of the fundamental vocalic alternations found in the formative. For example, the root ‹535›, with vocalic realizations о and а (мощи, помагати) has pure vocalism, while the root formative ‹570›, with the vocalic realizations ъ and у (мъшица, муха), has sonant vocalism (series H(*u*)).

It follows that synchronically speaking the opposition between pure and sonant vocalism is only defined for formatives with unstable vocalism. Thus, a significant number of etymologically sonant roots are not classified as so-

17 Cf. Russian *берег*, Old High German *berg*, Sanskrit *br ˚ hant*-.

nant synchronically. Such are, for example, сын ‹923› (сынъ), дѫб ‹274› (дѫбъ), пѫт ‹748› (пѫть), пѧт ‹746› (пѧть). Here also belong roots whose etymological sonanthood follows from modern Russian data, namely, such wellknown sources of evidence as pleophony (reflexes such as *torot*, *teret*, *tolot*, *telet*, cf. брѣг ‹59›, Russian *берег*, млѣк ‹558›, Russian *молоко*), and vowel prothesis in liquid sequences (cf. прьст ‹718›, Russian *перст*, врьх ‹126›, Russian *верх*, тлъст ‹961›, Russian *толстый*).18

### Note

The many reflexes of the Proto-Slavic opposition between pure and sonant vocalism in the daughter languages (East Slavic in particular) raises the expectation of such a reflex in OCS, the oldest representative of Slavic. In reality, outside of the synchronic pure ~ sonant opposition, traces of this contrast are not found in OCS itself.19

We encounter the consequences of the etymological pure ~ sonant opposition in OCS in two ways. First, roots that are etymologically sonant show otherwise prohibited consonant combinations. Thus, тлъст ‹961› shows the combination тл, prohibited and repairable outside of sonant roots (see details in § 62). An exception from a synchronic rule thus receives an etymological basis. Second, a distinction between etymologically pure (where Russian has a postposed vocalic element, cf. *крест*: крьстъ) and etymologically sonant (where Russian has a preposed vocalic element, cf. *перст*: прьстъ) combinations with liquids is at the basis of the addendum to Havlík's rule (see details in § 899, *Excursus on* yer *aberrations and Havlík's rules*).

# §678. Unstable roots: pure ~ sonant

Because series and grades in the fundamental alternations intersect, a given root with unstable vocalism cannot always be easily classified as pure or sonant. For example, alloforms of the shape *P*ра*Q* and *P*рѣ*Q*, where *P* and *Q* are segments or segment strings (possibly empty), can be treated under the pure pairing а‖ѣ or the sonant pairing ра‖рѣ (series H(*r*)).

Of course, if the vocalic realizations shown by a given root are not compatible with any sonant series but compatible with the pure series, this root contains pure vocalism (such is, e.g., root ‹535›, alloforms мог and маг). Likewise, if the vocalic realizations shown by a given root are not compatible with the pure series, but compatible with some sonant series, this root contains sonant vocalism. Such is, e.g., the root дух ‹261› (the pure series has no у vocalic realization), and also грѫз‖грѧз ‹207›.

<sup>18</sup> Accordingly, synchronically such roots as град ‹192› (градъ1 'town') and град ‹191› (градъ2 'hail') are morphophonologically equivalent.

<sup>19</sup> The fact that all handbook grammars of OCS are focused on the history of the language, and on comparative Slavic grammar, makes sonant vocalism a necessary part of every such OCS grammar.

The situation is more difficult in case vocalic realizations in a given root are compatible with both pure and sonant vocalism. Other things being equal, this root will be treated as pure; special considerations are required to treat it as sonant.

These special considerations usually amount to observations of adjacent alloforms (see details in § 873, *Commentary on segmental pairings*).20 In the general case, the series which keeps most alloforms as adjacent is preferred, and adjacency is especially important for alloforms in a single paradigm (intraparadigmatic adjacency). The following principle, called the *rule of horizontal and vertical pairings*, is taken into account in intrapardigmatic alloformy: in sonant series intraparadigmatic alloformy is given by adjacent pairings horizontally, and in pure series vertically. Accordingly, preference is given to the series where this principle is not violated.

Consider, for example, roots ‹122›, врьз, врѣз, and ‹323›, зъв, зов, зыв (Table 678.1–2).

Table 678.1. Embedding of the root врьз ‹122› into the table of fundamental alternations


The alloform врьз can have two analyses, as representing the ьр/рь or the ир/рь grade, and so in the sonant embedding table it appears twice. The same is true of the alloform врѣз. In the H(*r*) sonant series embedding, the root ‹122› shows horizontally adjacent alloforms врьзand врѣз, while in the pure series embedding the corresponding alloforms are not adjacent. Given that these alloforms are found within the paradigm (verb °врѣсти 4c\*⤹), the evidence from adjacency is important in assigning this root to the H(*r*) sonant series.

Table 678.2. Embedding of the root зъв ‹323› into the table of fundamental alternations



20 A different consideration leads to treating root ‹891› стрѣг as sonant; see § 767 below).

In the case of root ‹323›, the rule of horizontal and vertical pairings is relevant. The embeddings shown above speaks in favor of treating this root under the H(*u*) sonant series: the alloforms зъв and зов are found within a single paradigm, зъвати 3°\*, Prae зовеши.

Other things being equal, a root is treated as pure. Such are, for example, the roots ‹81›, alloforms вьр (вьрѣти) and вар (варити), ‹221›, alloforms двьр (двьрь), двор (дворъ), and двар҄ (удвар҄ꙗти)21 (Table 678.3–4).

Table 678.3. Embedding of the root ‹81› into the table of fundamental alternations



Table 678.4. Embedding of the root ‹221› into the table of fundamental alternations



§679. On the roots ‹438› крыти and ‹23› бити

The choice of series for these roots is not difficult: крыти and кръвенъ unambiguously point to H(*u*), while разбои and убиица point to H(*j*). However, the identification of alloforms with grades is not obvious. Moreover, different grades can correspond to different morphophonological spellouts, while the phonological forms are identical. For example, graphic Prae биѭ can be derived from two morphophonological representations: бьј=ѫ, the ь grade (as in this grammar), or би*i̯*=ѫ, the и grade (see § 898–899 for details).

§680. On the roots ‹451› въскрьснѫти and ‹855› осльпнѫти

The root ‹451› has two alloforms: крьс (°крьснѫти, °крьсениѥ, °крьсновениѥ, °крьсати), and крѣс (°крѣсити, °крѣсьнъ, °крѣшати, °крѣшениѥ). The root can be embedded into three series: pure, H(*r*), and H(*j*). (Tables 680.1–3, p. 387).

<sup>21</sup> Etymologically both of these roots are sonant. The present grammar considers them anomalous; see § 806.


Table 680.1. Embedding of root ‹451› into the pure series of fundamental alternations

Table 680.2. Embedding of root ‹451› into the H(*r*) sonant series of fundamental alternations


Table 680.3. Embedding of root ‹451› into the H(*j*) sonant series of fundamental alternations


крьс and крѣс are adjacent only in the embedding into H(*r*); this is the solution adopted in the present grammar. However, etymologically this root belongs to H(*j*), showing not adjacent alloforms but alloforms paired by old ablaut (zero grade in нѫ-verbs and athematic forms, о-grade in и-verbs and nominals). See Vasmer, *крес* II.

The same reasoning applies to root ‹855› осльпнѫти.

# §681. The choice of the canonical spellout for roots with ъ or ь

As is well-known, the use of the letters ъ and ь in sources is highly unstable (see § 898–899, *Excursus on* yer *aberrations and Havlík's rules*), which in some cases makes difficult the choice of the canonical spellout for the root. Following etymology, as in Večerka and Sadnik, is not only theoretically questionable but sometimes not straightforward: in many cases, the etymology is not transparent. Lining up the canon to the practice of a single source is also a dead end, because each source is significantly unstable. Often the spellout adopted by Večerka and Sadnik is either not attested in sources at all (e.g. Večerka's headwords грьмѣти and възгрьмѣти: all spellouts have гръм-; cf. ‹196›), or are attested by a minority of uses (thus, Večerka's headword прьсть: the spellouts прьст- 3×, пръст- 7×; see ‹702›). Likewise in Večerka's headword вьдова: this item occurs 13×, but only once as вьд- (Ps Sin 108, 9); other spellouts have ъ, see ‹134›.

In this situation, a practically useful solution seems to follow the published standard. In the PD, the canonical spellout follows the choice in Večerka everywhere except for the following:22 1) мръз ‹566›, not мрьз; 2) мрък ‹567›, not мрьк, мръч, not мрьч, but мрьц, following Večerka; 3) гръм ‹196›, not грьм; 4) смръд ‹861›, not смрьд; 5) пръ ‹702› (пръсть), not прь; 6) дль ‹234› in all entries; 7) тльк ‹962›, not тлък, and also 8) брьселиѥ ‹57›, not бръселиѥ.

### §682. Overview table of roots with sonant vocalism

Roots with sonant vocalism are listed by series. The list by series is preceded by the overview table (Table 682), where a paragraph number with commentary is given for each root.

Note that for each root below, its series is shown (or two alternative series, see § 124), and all alloforms are listed, but alloforms are not identified with individual grades (see § 679, *On the roots ‹438›* крыти *and ‹23›* бити for problems encountered in identifying grades).

There are 102 such roots altogether.


Table 682. List of roots with sonant vocalism

\* Asterisk marks those roots that do not agree with the CVC norm (of the shape CH).

<sup>22</sup> This list includes roots with sonant combinations. Individual departures from Večerka in the question of ъ vs. ь are noted *passim*.


Table 682 (continued). List of roots with sonant vocalism

\* Asterisk marks those roots that do not agree with the CVC norm. Root <516> is of the shape HC. Roots <639> and <334> are of the shape H. Others are of the shape CH.

# *Sonant roots with H(n) vocalism*


‹302› жѧти has been split from ‹153› гънати, contra Sadnik.

§683. Root ‹302› §684. Root ‹310›


звѧ only in звѧщи [звѧ.г.т.и].

# §685. Root ‹384›


# §686. Root ‹550›


The roots ‹‹521›› мьнѣти, памѧть and ‹‹506›› мѣнити, помѧнѫти are unified, contra Sadnik. See Vasmer, *память*, *мнить*, *помянуть*.

Contra Večerka, the spellouts помѣнѫти, въспомѣнѫти are considered aberrant. For the distribution in the sources see Večerka.

### §687. Root ‹664›


§688. Root ‹1074›


Contra Sadnik, all forms with the root alloform кон (cf. коньць, законъ) are not included in this root's membership. Etymologically кон and чѧ show different vocalism grades, and the front (чѧ) and back (кон) grades differ in their initial C position (cf. градъ ‹192› and жрьдь ‹293›).

# *Sonant roots with H(m) vocalism*

### §689. Root ‹268›


The lexeme дымъ is assigned to root ‹268›, contra Sadnik, where дымъ is under root ‹259› (дунѫти).

# §690. Root ‹290›


# §691. Root ‹334›


# *Sonant roots with H(n/m) vocalism*

§692. Root ‹33› §693. Root ‹207›




§696. Root ‹595› §697. Root ‹597›


§694. Root ‹327› §695. Root ‹526›



§698. Root ‹724› §699. Root ‹934›


§700. Root ‹984› §701. Root ‹1006›



# *Sonant roots with H(u) vocalism*

§702. Root ‹36› §703. Root ‹50›


§704. Root ‹63› §705. Root ‹210›


Contra Sadnik, the alloform бл҄уд ‹34› is separated from ‹63›

### §706. Root ‹261›


\* Contra Večerka and Sadnik, here only дъхати 3\* and дыхати 7. See details in § 508.

### §707. Root ‹286›


Sonant vocalism H(*u*) is set up contra etymology: в is the marker of the present, see Meillet, § 218, 265.

### §708. Root ‹323› §709. Root ‹397›


### §710. Root ‹438›


\* Only in PRAE and Imf of the verbs °крыти. See § 449.

### §711. Root ‹470› §712. Root ‹570›


The shapes кы/кыв represent the same vocalism grade, see Vaillant, § 192, 204.


\* Only in PRAE and Imf of the verbs °мыти. See § 449.

### §715. Root ‹670›


### §716. Root ‹688›


### §717. Root ‹773›


\* Only in PRAE and Imf of the verbs °рыти. See § 449.

### §718. Root ‹780› §719. Root ‹781›


The variant р҄юти, р҄ѥвѫ (1× in *Hilandar folios*) is not part of the benchmark corpus of this grammar. See Vaillant, § 217.

### §720. Root ‹844›


Cf. also the aberrant alloform слъ (°слъшати in Mar, Cloz, Ps Sin).

### §713. Root ‹571› §714. Root ‹616›

# §721. Root ‹863›\*


\* Alternatively, this root can be treated as belonging to the pure series. In the pure series, the alloforms снов and сныв show the pairing о‖ы, in the sonant series, the pairing ов‖ыв. While neither of these pairings is adjacent, it is worth noting that the pairing ов‖ыв is observed multiple times (cf. крыв‖кров ‹438›), while for the pairing о‖ы this root would be the sole representative. In Church Slavic, Prae forms like снуѥши are attested.

§722. Root ‹866› §723. Root ‹895›


§724. Root ‹897› §725. Root ‹905›


§726. Root ‹913› §727. Root ‹914›


§728. Root ‹966›


Cf. also the aberrant spellouts with трав- in place of трѣв- Ps Sin (3×).

§729. Root ‹1105›


\* Only PRAE and Imf of the verbs °шити. See § 449.

### Sonant roots with H(j) vocalism

#### 8730. Root <23>


\* Only PRAE, Imf and the n-Part stem of the verbs of ° swTh group; see § 448. Contra Sadnik, the lexeme вичь <26> is separated from root <23>.

#### 8731. Root 495>


\* Only PRAE, Imf and the H-Part stem of the verbs °вити, and in вигалица. The solution is arbitrary. See § 448.

Contra Sadnik, roots << 1078>> (only вналица) and << 1085>> (°вити and others) are unified.

#### ട്ട 732. Root <96> 8733. Root <171>


\* Only PRAE, Imf and the H-Part stem of the verbs °гнити. See § 448.

8734. Root <236>

8735. Root <314>


#### ട്ട 736. Root «491›


\* Only PRAE, Imf and the H-Part stem of the verbs °лити. See § 448.

#### § 737. Root <651>


\* Only PRAE, Imf and the N-Part stem of the verbs °пити, as well as the derivatives пилить, пиганица, пиганьство, пиганьствию. The solution is arbitrary. See § 448.

### 8738. Root <738>

& 739. Root < 769>



§ 741. Root «892›

((пъсн)) in пъснь и пъсньнъ.

### 8740. Root «858›



Contra Sadnik, roots <<897>> строити and <<891>> стръха are unified. See Vasmer, стреха.

# Sonant roots with H(r) vocalism

### 8742. Root <24>


Contra Sadnik, roots << 33>> вьрати and << 63>> вракъ, are unified. See Vasmer, брак II.

#### 8743. Root <42>

8744. Root <58>



Брытъ ш-Part. See § 452.

### 8745. Root <89>


The stems (דצטן) and (דזעט)) are morphophonologically anomalous in затворити, pactreoperm, 3aTreaparth and their derivatives with prothetic T before the influcence of stems with the prefix от, e.g. отворити.

8746. Root <117>

8747. Root <121>



§748. Root ‹122› §749. Root ‹154›



Contra Sadnik, roots ‹‹242›› горѣти and ‹‹257›› грѣхъ are unified. See Vasmer, *грех*.

# §750. Root ‹220›


Cf. дрьколь ‹252› with the opaque stem ⸨дрькол⸩.ь; the treatment (дрь).(кол).ь as a compound of ‹220› and ‹364› is also acceptable.

### §751. Root ‹292› §752. Root ‹295›


Do not confuse ‹292› жрьти 'sacrifice' and ‹295› пожрѣти 'devour, swallow'.

§753. Root ‹309›


### §754. Root ‹322› §755. Root ‹451›


Contra Sadnik, roots ‹‹1134›› зьрѣлъ and ‹‹1148›› зрьно are unified. See Vasmer, *зреть* II, *зерно*.


In this root, the sonant series H(*j*) is established etymologically; see details in § 680.

### §756. Root ‹539›


Cf. also the reduplication in из.⸨мрь.мрѣ⸩.т.и for измрьмьрѫтъ (1× Supr 238, 13), contra Večerka, where we find измрьмьрати.

### §757. Root ‹566› §758. Root ‹567›


Večerka and Sadnik give мрьз in their headwords (see details in § 681).

### §759. Root ‹639›


Not to be confused with ‹12› (разорити, разар҄ꙗти).

### §760. Root ‹658› §761. Root ‹665›


Do not confuse пьрати ‹658› 'fly, soar' and °пьрати ‹665› 'shove, lean, quarrel'.

### §762. Root ‹702›


Večerka and Sadnik give прь in their headwords (see details in § 681). Not to be confused with ‹718› прьстъ 'finger'.


#### §763. Root ‹825› §764. Root ‹861›

Večerka and Sadnik give мрьк in their head-

words (see details in § 681).


Večerka and Sadnik give смрьд in their headwords (see details in § 681). Do not confuse with ‹862› (смрьдъ 'commoner'); although unifying these roots is possible, see Vasmer, *смерд*.

### §765. Root ‹877› §766. Root ‹888›


Spellouts with сръдьц- predominate in sources. See details in § 681.


### §767. Root ‹891›


Do not confuse the verb стрѣщи1 4c 'guard' ‹891› with sonant vocalism with the verb стрѣщи2 4c\*⤹ ‹894›, Prae стригѫ, стрижеши 'shear' with pure vocalism. The verb стрѣщи 4c does not show unstable vocalism in the paradigm. However, in reality, it follows the pattern of влѣщи, брѣщи with sonant vocalism H(*l*) and H(*r*), with the Inf vocalic realization in the PRAE stem. However, for влѣщи and брѣщи the expected vocalic realization рь, ль is attested in OCS, while for стрѣщи the forms стрьѕи, устрьглъ are only found in Church Slavic (see Vaillant, § 213). Prae forms стрѣгѫ, стрѣжеши, are attested in the canon, while н- and ш-Part are not. The interpretation of the root ‹891› contradicts the conventions of § 677, *A note on the notion of sonant vocalism*, but ensures the morphophonological unity of стрѣщи 4c with the 4c\* group of verbs with sonant vocalism, which are: брѣщи, влѣщи, тлѣщи, врѣщи, врѣсти, and чрѣти.

### §768. Root ‹954› §769. Root ‹973›


§770. Root ‹1081›


# *Sonant roots with H(l) vocalism*


Alloform въл only in the lexeme вълати [въл.а.т.и] (spellouts вълаѩ, вълаахѫ). See the comment to the verb вълати 7 (§ 515).


Do not confuse ‹234› дль.г.ъ 'long' and ‹233› длъг.ъ 'debt', длъжьникъ.

§771. Root ‹77› §772. Root ‹105›


Contra Sadnik, roots ‹‹1093›› влѣщи, ‹‹1094›› влькъ and ‹‹1096›› вльчьць are unified. See Vasmer, *волк*, *волчец*.

§773. Root ‹234› §774. Root ‹312›


# §775. Root ‹364› §776. Root ‹516›


# §777. Root ‹553› §778. Root ‹683›


### §779. Root ‹693› §780. Root ‹842›




\* Alloform пла only in compounds (cf. пладьниѥ 'midday').

### §781. Root ‹855› §782. Root ‹856›


In this root, the sonant series H(*j*) is established etymologically; see details in § 680.

### осльпнѫти слѣпъ въсльпати\* въслѣпл҄ѥши (for въсльпати) \* Contra Večerka and Sadnik, who give въслѣпати; see details in § 508. Do not confuse roots ‹856›

(въсльпати 'flow, stream') and ‹855› (слѣпъ 'blind').

### §783. Root ‹883› §784. Root ‹962›


# Pronominal roots

### §785. Overview table

Table 785 lists the roots by number, indicating the paragraphs where these roots are treated in more detail.



#### Table 785. List of personal pronoun roots

8786. Root <7>



Aberrant spellout \*3'% MAR Mk 11, 29.

ВАШЬ [((ВАШ).Ь] BAW with an opaque suffixal extension; morphophonologically

opaque B-initial wordforms that are part of unique lexemes דאו (ва, ваю, васъ, etc.) and азъ (NDu въ) also belong to this root; cf. \$ 805, A note on personal pronoun roots.

R-

### ട്ട 788. Root <136>


Morphophonological anomaly. There is an anomalous variation in a- and th-initial suffixes and terminals. Additionally, there is a confusion of variants with twofold terminals. In the lexeme B&& some terminals are of the soft subtype, contrary to the morphophonological status of the final c. Cf. the declension of the lexeme BLCb (see § 320).

### ട്ട 789. Root <262>

#### 8790. Root <343>



1) Nonstandard alloform û in ad-prepositional wordforms of the lexemes \* n and HME (see § 318).

2) jen and jerep with opaque suffixal extension.

### 8791. Root <350>


### 8792. Root <359>


1)

2) κολ and κοτορ with an opaque suffixal extension; cf. the form κοτεραστο 1× MAR Lk 11, 11 with the aberrant stem ((κοτερ).

3) On the opaque stem ничьж see Vaillant, § 94.

### ട്ട 793. Root <533>


Here also belong morphophonologicall opaque M-initial wordforms as part of the unique lexeme a3' (MA, MEHE, MAN'S, MH, etc.); see § 805, A note on personal pronoun roots.

# 8797. Root «636›

H-

НАШЬ [(НАШ).Б]

Haw with an opaque suffixal extension.

Here also belong morphophonologically

opaque H-initial wordforms as part of the

unique lexeme a3's (на, наю, насъ, etc.);

see § 805, A note on personal pronoun roots.

8796. Root «628»

овамо, овогда and others.

ભા OHB [0H.7]

Cf. the compound orgchya 1x SUPR 286, 18. Cf. the adverbial онамо, онъдє and others.

### \$ 798. Root <797>

\$ 795. Root <624>


1) cen with an opaque suffixal extension.

2) Cf. the compound онъсица 1× SUPR 286, 18.

3) Morphophonological anomaly: variant confusion with twofold terminals.

## §799. Root ‹799› §800. Root ‹803›


Irregular alloformy: св, соб, себ (in себѣ, себе) and с (in the opaque сѧ, си); see below § 805, *A note on personal pronoun roots*.


# §801. Root ‹938› §802. Root ‹943›


тол with an opaque suffixal extension. Irregular alloformy: тв, теб, тоб (in тебѣ, тобоѭ) and т (in the opaque ты, тѧ, ти); see below § 805, *A note on personal pronoun roots*.

### §803. Root ‹965›


## §804. Root ‹1112›


The shape of the root ьн from the form отьнѫдьнъ [от.(ьн).ѫд.ьн.ъ], where ѫд is an adverbial marker.

# §805. A note on personal pronoun roots

The etymology of personal pronouns is not entirely clear; likewise, different authors assign the pronouns to lexemes in different ways. Most wordforms are morphophonologically anomalous, and thus in this grammar have opaque stems (the boundaries between stems and suffixes or terminals are not established). In this grammar, the relation "belong to the same root" is identified with the purely formal feature "have identical initial consonant". Thus, we have 1) "а-initial": ‹7›; 2) "в-initial" combines Sadnik's roots ‹‹1050›› and ‹‹1064››; 3) "м-initial" combines Sadnik's roots ‹‹499››, ‹‹533››, and ‹‹558›; 4) "н-initial": Sadnik's

root <<567>>; 5) "c-initial": Sadnik's root <<786>>; 6) "m-initial" combines Sadnik's roots <<1020>> and <<1022>>.23

# Anomalous roots

### \$806. Overview table

Table 806 lists the roots by number, indicating the paragraphs where these roots are treated in more detail.


### Table 806. List of anomalous roots


<sup>23</sup> The wordforms דג, ד€Ε€, etc. cannot be located in Sadnik's roots dictionary.

#### 8807. Root <9>


are rare; cf. aberrant forms възалк- MAR, възлак-, въслак- ZOGR, ВЪЗЬАЛК- AS, ВЬЗАЛК- MAR, лакомъ SUPR (1х: мимондъю сиї лакомънн SUPR 41, 26).

The root лад <483> ладии has similar aberrant forms: алъд, алд; see Večerka for the distribution of forms. Here also note the form алънии 'fallow deer' by Večerka and альнин by Meyer, reconstructed for the spellout мынин in SUPR: кдина отъ мынии скштинуъ въ лаоднійн SUPR 232, 30-233, 1; there are no other glosses. The lexeme алъмни 'fallow deer' and, accordingly, its root, contra Večerka and following Sadnik, are not included in the PD.

#### ട്ടു 808. Root «16»


The vocalism is sonant H(u). The anomalous alloform with the π νοςalism (cf. Ελλ) is etymologically interpreted as n-infixation in the PRAE stem (cf. chern: cagn <798>; neurn: лагж <482>; °ръсти: °раштя <791>). Cf. also кн, въ, ся in the conjugation of the verb гъгти (see § 543-549).

8809. Root <81>


The pairing of the alloforms вьр вар is not adjacent either in the pure (6 a) or in the sonant (bp||ap) series. Etymologically, sonant vocalism is assumed. See Vasmer, вар II.

8810. Root <85>


Cf. also the aberrant spellout довьльнъ (regularly in sUPR, cf. 371, 19) in place of canonical довольнъ; likwise дъвъя-(ZOGR), дьвьл- (MAR).

§811. Root «107›


BNEX | BABC is a rare example of interlexemic alloformy of the ««|| » grades of velar palatalization; влъсняти (1× CLOZ, м.зъщи влъсняштєї CLOZ 1a, 15-16).

#### §812. Root <111>


The vocalism is sonant H(n), initially ambivalent: в is the consonantal prothesis. Cf. the compound દરતાજપ્રુરમમાદ [(EAAT).0.(x).X.a.H.bj.e] <27> x <111>.

§813. Root <129>


Initially ambivalent: в is the consonantal prothesis. Etymologically sonant H(u) (cf. Russian выпь 'bittern' and вопль 'scream').

### §814. Root <132>


The vocalism is nonstandard H(u). The alloformy is nonstandard: the vocalism alternation งู/รเ is accompanied by a prothetic в word-initially and after vowels: въкняти, навъкняти. Likewise after consonant-final prefixes: cf. извъюкняти.

### 8815. Root <137>


The vocalism is pure and stable. The alloformy is nonstandard: open BS and closed BBT alloforms with the r consonantizer; cf. A in t/tg <1113>, t/tg <1114>, n/n} <329>, ga/gag <216>, дѣ/дежд <225>).

#### 8816. Root <147>


Anomalous extension of the initial C position. See Vasmer, вянуть (with the Indo-European s-mobile).

### 8817. Root «148›


The vocalism is sonant H(n/m), initially ambivalent, в is the consonantal prothesis. Cf. aberrant вжза, съжзъ. Also here cf. съноузынъ <1012> by one of the etymological hypotheses.

### 8818. Root <153>


Anomalous instability of the initial C position: the vocalism alternation by advancement cooccurs with the consonantal alternation by the r | w pairing in the initial C position. Normally, consonant alternations are only tied to the final C position. Accordingly, variation of the initial C position within the same root is impossible. This root is the only exception, because in all other cases such root pairs are split contra etymology (see § 873.6). Etymologically, the vocalism is sonant, as in жѧ/жьн <302> (as in Sadnik); but there is no synchronic basis for finding sonant vocalism in <153>.

§819. Root <159>


The vocalism is sonant H(l); there is reduplication; see § 41. Etymologically, root <159> is related to root <162> гласъ (in which case c in гласъ is a suffixal extension). See Vasmer, голос and гологолить.

### 8820. Root <216>


The vocalism is pure, stable. The alloformy is nonstandard: open да and closed дад alloforms with the A consonantizer, cf. и/ид <329>, дѣ/дєжд <225>, в/ћд <1113>, な/なる <1114>.

### ട്ട 821. Root <221>


The alloform pairing двьр is not adjacent either in the pure (ь) o) or sonant (ьр) ор) series. Etymologically, sonant vocalism H(u) is assumed. See Vasmer, deep.

#### ട്ട 822. Root <225>


The vocalism is pure, unstable (elt). The alloforms are the open gt and closed geg with the consonantizer g; ma by substitutive softening; cf. n/ng <329>, ga/gag <216>, t/th <1113>, t/tg <1114>.

#### ട്ട 823. Root <256>


драхаъ апа драселъ are derivational doublets. драх is a rare example of interlexemic alloformy of the root by the velar palatalization grades «к|ц». See Večerka for distribution in sources.

#### §824. Root ‹276›


Represented only by PRAE stems of the unique verb 'to be' (KCMS): in PRAE forms, the root cannot be parsed out (and the morphophonological spellout is not built); in шт-Part the root is c: c. xurr; cf. (c). Run.bong.o and (iec). T.bcTB.o.

#### §825. Root <318>


The only root without a vowel outside of the pronominal ones; see etymology in Vasmer, змея.

#### 8826. Root <329>


The vocalism is pure and stable. The alloformy is nonstandard: open и and closed ид alloforms with the & consonantizer; cf. t/tg <1113>, t/t2 <1114>, ga/gag <216>, дѣ/дєжд <225>.

### §827. Root ‹482›


An anomalous alloform with the so-called nasal infix in the PRAE stem and the nominal лѧжаꙗ (cf. быти — бѫдѫ ‹16›, сѣсти — сѧдѫ ‹798›, °рѣсти — °рѧщѫ ‹791›).

## §828. Root ‹496›


The series isolation condition is violated: the final C position in different alloforms is occupied by members of different velar palatalization series. These are alloforms of the same root, contrary to the conventions of § 104.

# §829. Root ‹654›


The vocalism is sonant H(*l*). For попелъ and пепелъ with partial reduplication, the morphophonological parse is not established.

### §830. Root ‹763› §831. Root ‹791›


The vowel pairings is not embeddable in any sonant series. In the pure series, the pairing е~а is not adjacent. The vocalism is etymologically sonant H(*r*). See Vasmer, *рать*, *ретивый*.


An anomalous alloform with the so-called nasal infix in the PRAE stem and the nominal сърѧща (cf. быти — бѫдѫ ‹16›, сѣсти сѧдѫ ‹798›, лещи — лѧгѫ ‹482›).

### §832. Root ‹798›


An anomalous alloform with the so-called nasal infix in the PRAE stem (cf. быти — бѫдѫ ‹16›, лещи — лѧгѫ ‹482›, °рѣсти — °рѧщѫ ‹791›). оседълати: contra Sadnik, roots ‹‹788›› оседълати and ‹‹795›› сѣсти are unified; see Vasmer, *седло*, *сидеть*.

### §833. Root ‹838›


скѫпость 'greed' 1× in Euch, проскупьство 1× in Supr.

### §834. Root ‹885›


The vocalism is mixed: sonant H(*n*/*m*) and pure, о~е.

### §835. Root ‹975› §836. Root ‹1010›


is reduplication, cf. из.мрь.мрѣ.т.и ‹539›, гла.гол.ъ ‹159›, see § 41. Etymologically, the root is a borrowing from Greek τάρταρος.

The only root with no consonants in any of its alloforms. Contra Sadnik, where обути is unified with онуща [⸨онущ⸩.а]. See Vasmer, *обуть*, *онуча*. Etymologically, the vocalism is sonant; cf. Church Slavic объвенъ. In OCS ∇обувенъ. See § 509.

### §837. Root ‹1023›


The vocalism is mixed: sonant H(*n*/*m*) and pure а. охѫпити 'grab' for охѫпивъ Supr 1×; хапати for хапьѭщеSupr 1×; похѫпити for похѫпитъ Euch 1×.

### §838. Root ‹1108›


The vocalism is mixed: sonant H(*n*/*m*) in щѧд and pure е in щед. Contra Sadnik, щѧд/щед are not unified with скѫд ‹840›; the latter root has stable vocalism (according to Sadnik, скѫд, щѧд, щед all belong to the same root).

### §839. Root ‹1113›


The vocalism is pure and stable. The alloformy is nonstandard: open ѣ and closed ѣд with the consonantizer д; cf. и/ид ‹329›, да/дад ‹216›, дѣ/деЖ ‹225›, ѣ/ѣд ‹1114›. The root cannot be parsed out in PRAE forms of the unique verb ꙗсти (the morphophonological spellout is not built). Contra Sadnik, the roots ‹‹15›› ꙗсти and ‹‹17›› ꙗдъ are unified.

### §840. Root ‹1114›


The vocalism is pure and stable. The alloformy is nonstandard: open ѣ and closed ѣд with the consonantizer д, (which also occurs elsewhere, cf. ѣ/ѣд ‹1113›, да/дад ‹216›, дѣ/деЖ ‹225›, и/ид ‹329›), and ѣх. Only in forms of the unique verb ꙗти; see § 565–569.

# Suffixes

# §841. General

The catalog shown in Table 841.1 shows the suffixes in the following order: first are listed suffixes with a carrier consonant (§ 844–859), then suffixes without a carrier consonant (§ 860-861).24 Within the first class, suffixes are listed alphabetically by the carrier consonant.


Table 841.1. Overview table of suffixes

According to their CVC schema, suffixes can be standard, which agree with the CVC norm, and have the shape VC, and nonstandard, including finally am-

24 The carrier consonant is the first or only consonant of the suffix.

bivalent ones. All possible CVC schemata are shown in Table 841.2. Standard and nonstandard suffixes can be each other's alloforms.


Table 841.2. Classification of suffixes by CVC schema

# §842. Alloformy of suffixes

Alloformy claims for suffixes have no rigorous basis. The solutions adopted in this book are shown below in tables that show suffix inventories. Although the solutions are arbitrary, let us note some general principles.

1. Two suffixal formatives that show morphophonological variation are treated as alloforms that represent the same suffix. Such are twofold suffixes, e.g. от/ет, and initially- or finally-ambivalent suffixes, e.g. н/ен (initially ambivalent) or ѧт/ѧ (finally-ambivalent).

2. Two suffixal formatives that show standard segmental alloformy of the final consonant are treated as alloforms representing the same suffix. Such are suffixes that show different grades of the same series of the velar palatalization alternation, e.g. ик‖ич‖иц, and suffixes that show different grades of the same series of the substitutive softening alternation, e.g. ел‖ел҄.

3. Two suffixal formatives are treated as alloforms of the same suffix if, in a large proportion of their occurrences, they manifest the same grammatical function. Such are participial suffixes {24} (им, ом/ем) and {50} (ѫщ, ѧщ).

Let us list the suffixes where unification of alloforms does not follow from these general principles.25 These are:


Note that the term "suffix" can, in the general case, refer, first, to the full family, including a nonstandard one, e.g. the suffix (г, иг, ог); secondly, to the pair

<sup>25</sup> Recall that many characteristics of suffixes that could be taken as evidence for or against unification remain unobservable or unknown for OCS; primarily, these are the accentual properties.

or triplet of standard alloforms that are smaller than the full family, e.g. ом/ем or ов/ев; and finally, an alloform by itself, e.g. the suffix ов.

When families are shown, morphophonological variants are listed, but standard segmental ones are not. For this reason, for example, we write (к, ак) and not (к‖ч‖ц, ак‖ач‖ац), or (ел) and not (ел, ел҄).

### §843. On the intersection of suffix families

The alloformy relation adopted here was set up in such a way that different suffix families would not intersect; in other words, so that the shape of the alloform would unambiguously determine the suffix family to which the alloform belongs in all its occurrences. However, in several isolated cases this ideal was not reached. Let us list them.


Luckily, all these cases are such that one of the families is represented by a highly limited group of lexemes, and are given by list in the suffix inventories.26

# Suffixes with a carrier consonant



26 Segmentless substitutive softening adds somewhat to the number of intersections of suffix assortments, and this is a source of some indeterminacy. For example, in the lexeme радоща, it is formally impossible to tell whether ощ represents the suffix ост {36} or the suffix от {42} (the solution adopted is that this occurrence of ощ represents ост {36} and not от {42}). Likewise with порабощати (the solution adopted is that this occurrence of ощ represents от {42} and not ост {36}), or обьщь (the solution adopted is that this occurrence of ьщ represents ьск {37} and not ьт {43}). A few other cases of intersection, without any actual indeterminacy are possible due to consonantal alternations.

Notes on individual suffixes

{1} об: only with roots зъл and ѫтр.

{1}ьб: сѫдьба ‹936› only in Ps Sin. There are no canonical spellouts; spellouts with vocalism of the type сѫдеб- are also absent. Cf. ѡтъ ліца твоего сѫдъба моѣ ізідетъ Ps Sin 16, 2; сѫдобъ твоіхъ раді Ps Sin 47, 13.

### §845. Group 2, carrier consonant в


Notes on individual suffixes

{2} в: 1) often acts as a consonantizer in verbal stems, cf. пиво, избивати, дрьжава, etc.; 2) the ицѣлѣвати model is found in the following words: повелѣвати, съдолѣвати, прѣдолѣвати, одолѣвати, измѫдрѣвати, разумѣвати, проразумѣвати, ицѣлѣвати, оцѣпѣнѣвати.

{3} ав: not to be confused with the suffix sequence а.в, cf. раскопавати [раз.коп.а.в.а.т.и]. The suffix ав occurs in the eight listed items and their derivatives. In the final C position of the preceding formative, the grades are к, C°, and C•.

{4} ив: not to be confused with the suffixal marker of the new ш-Part (type л҄юб.и.въ|, л҄юб.и.въш.и). Outside ш-Part, the sequence и.в is only found in при.въп.и.в.а.т.и ‹129›.

Commentary on individual lexemes дрьжава [дрьж.ѣ.в.а] ‹250› from дрьжати, as



### §846. Group 3, carrier consonant г, ж

Notes on individual suffixes

{5} г: this suffix only in дль.г and звѧ.г and derivatives.

{5} иг, ог: these suffixes in only the two listed lexemes. Cf. the opaque ⸨мъног⸩, ⸨чрьтог⸩, ⸨сапог⸩, ⸨хѫдож⸩, and also ⸨вощаг⸩, ⸨кръчаг⸩. A special group is formed by the following Germanic borrowings: ⸨скълѧѕ⸩.ь, ⸨пѣнѧѕ⸩.ь, ⸨кънѧѕ⸩.ь, ⸨кладѧѕ⸩.ь (and derivatives кънѧжь and пѣнѧжьникъ).

{6} еж only папежь ‹656›, мѧтежь ‹597› and derivatives.

Commentary on individual lexemes звѧщи [звѧ.г.т.и] ‹310› 'relate', cf. звонъ [звон.ъ] with a different grade of H(*n*) vocalism. A hapax occurrence of suffixed lexical

верига [вер.иг.а] ‹89› 'chain' (Supr and As). острогъ [ост.р.ог.ъ] ‹646› 'picket fence'.

### §847. Group 4, carrier consonant д

component among class 4c verbs.


Notes on individual suffixes

{7} ьд: is only found in the listed stems and their derivatives. Cf. обьдо [⸨обьд⸩.о] ‹626›, listed as opaque, even though it is possible to parse as (об).ьд.о for root ‹623›: (об).ьщ.ь, (об).ьщ.ен.ьј.е (or (обь).д.о, cf. Vasmer, *обиход*, and also *обдо*).

Note the morphophonologically unusual spellout страЖѫ, 1× Supr 91, 29–30 for canonical стражьба, 1× Ps Sin 76, 5. In Večerka we find the item стражьда for the above gloss from Supr.

Commentary on individual lexemes бридъкъ [бри.д.ък.ъ] ‹49› 'sharp'. Hapax gloss in Supr: бѣаше же и въздухъ студенъ· и часъ бридъкъ (76, 13–14). The verb брити is not attested; to the same root must be assigned also бритва [(бри).т.в.а] 'razor': ѣко брітва изощрена сътворілъ есі лестъ Ps Sin 51, 4.

прѣдьн҄ь [прѣ.д.ьн҄.ь] ‹719›. Cf. прѣЖьн҄ь [(прѣ).Ж.ьн.ь], see § 862.

### §848. Group 5, carrier consonant з


Notes on individual suffixes


### §849. Group 6, carrier consonant *j*


Notes on individual suffixes

{9} ај: 1) suffixal stems in ај are mixed with expanded PRAE stems in а: cf. раз.(мышл҄).а.т.и, раз.(мышл҄).а*i̯*.еши — раз.(мышл҄).ај.ь (1× Supr with aberrant spellout розмышлꙗи); 2) in stems such as ходатаи, поводатаи, there is the composition ат.ај, where the initial а in ај has no relation to the theme. Such are also исходатаи, позоратаи; cf. also ратаи [(ра).т.ај.ь].

{10} ој/еј is a twofold suffix, part of the set of nominal twofold terminals, cf. ISgf ој.ѫ/еј.ѫ (not included in the counts). In nominal stems the distribution follows the twofold rule; however, еј, contra this rule, is found in сеица [(с).еј.иц.а] (hapax gloss сеици Mk 6, 25 As); here note also the aberrant spellout in Sav: треіцеѭ [(тр).еј.иц.еј.ѫ] (Mt 26, 44) for троицеѭ [(тр).ој.иц.еј.ѫ] (this spellout is also treated as erroneous for третиицеѭ).

{11} ьј — 1) often used in borrowed stems, cf. мрѣкориꙗ [⸨мрѣкор⸩.ьј.а], куциꙗ [(куц).ьј.а] (cf. medieval Greek κουκκιά Pl 'beans'); 2) is found in bicomponential inflections of the standard twofold set of terminals (cf. пѫт.ьј.е); 3) the suffix ьј should be distinguished from combinations of ь and и with epenthetic *i̯*. In particular, in old comparatives we have ь: бол҄ии [бол҄.ь*i̯*.ь], likewise гор҄ии, мьн҄ии etc., and in the adjective прочии [проч.ь*i̯*.ь] (2/a *plenum tantum*).

{12}ѣј forms a substantive stem only in верѣꙗ. In all other cases the stems are adjectival, viz. the new Compar, cf. тачаи [(т).ач.ѣј.ь], старѣишина

[(стар).ѣј.ьш.ин.а]. The combination of the suffix ѣ with epenthetic *i̯* is found in the stem говѣинъ [гов.ѣ*i̯*.ьн.ъ] for говѣти ‹178›.

Commentary on individual lexemes лѧжаꙗ [(лѧж).ај.а] 'broody hen' ‹482› 1× As Mt 23, 37 and беспосагаꙗ [без.⸨посаг⸩.ај.а] 'unmarried' ‹698› 1× Supr 391, 27 are morphophonologically anomalous. First, the suffix ај is supposed to attach to mascu- line substantives. Second, the substantive stem in лѧжаꙗ coincides with the truncated

PRAE stem, which is unusual for nominal stem formation. Cf. сърѧща (see § 865).

ун҄ьшиина [ун҄.ьш.ьј.ин.а] ‹1016›, cf. изун҄ьшина [из.ун҄.ьш.ин.а], both words 1× Supr.


§850. Group 7, carrier consonant к

Notes on individual suffixes

{13} к, ак: the alloform к is found after open roots, cf. зра.к.ъ, just as ц in зрь.ц.а.л.о 1× Supr; ак is found only in pronominal stems. The anomalous shape of the suffix with an initial ѣ in pronominal stems with the basic component вьс, as in вьс.ѣк.ъ. The roots вьс ‹136› and с ‹797› show a morphophonological anomaly.

{14} ик, иц. The alloform иц is used as a marker of nominal stems in feminine lexemes. Otherwise only in корабл҄иць and сиць; see more details in § 864.

{15} ок: these are высокъ, глѫбокъ, грѫстокъ, жестокъ, инокъ [*i̯*ьн.ок.ъ], широкъ, and their derivatives.

{16}ък/ьк. This is a twofold suffix; see details in § 863. The combinations ьк + ьј/ък + ьј are found in the following names of occupations: кръмьчии, корабьчии, шаръчии, самъчии, кън҄игъчии; cf. also the opaque сокачии; see details in Vaillant, § 63, 133. On these suffixes as nominal stem markers see details in § 864.

{17}ык: this suffix is found only in the lexemes listed above and their derivatives. Cf. ⸨мотык⸩.а, ⸨јѧзык⸩.ъ.

Commentary on individual lexemes

вьсакъ. This lexeme has the doublet вьсѣкъ. For distribution in sources see Vaillant, §99. See the suffix {13} above.

дъвакъ. No such headword in Večerka and Sadnik, who register the lexeme кагръличищь for the following glosses: дати жрътвѫ […] дъва кагръличища ли дъва птенъца голѫбина Lk 2, 24 Mar (likewise in Zogr and As). Cf. the same verse in Sav: дати

жрътвѫ […] в грьличища· ли в пьтеньца· голѫбина. See Vaillant, § 109.

сиць. Cf. the derivational doublet сикъ, attested in Church Slavic sources. See Vaillant, § 99.

ближика and ѫжика. These are the only lex- emes where the basic component in front of suffix ик shows C• grade, while having C° grade in the suffixless base. On segmentless substitutive softening see § 864.


§851. Group 8, carrier consonant л

Notes on individual suffixes

{18} л: this is the regular л-Part suffix (л-Part forms are not included in the counts). Outside of л-Part stems, the C-initial suffix л is possible both in V-final stems (cf. дѣло), and in C-final stems (cf. гѫсли [гѫд.т.л.и]).

{19} ел: outside of combinations of the type т+ел҄ (such as учител҄ь), there are slightly over 10 lexemes. On т+ел҄ combinations see § 864 below.

{20}ъл/ьл: this is a twofold suffix; see § 863, *A note on secondary twofold rule*. Cf. opaque скѫдълъ [⸨скѫдъл⸩.ъ] 'sherd' (from Latin *scandula*//*scandella*).

{21}ыл: only in мотыла ‹545›.

{22}ѣл: marks substantival feminine stems of the monovariate declension; otherwise only in кысѣлъ 2/a. Cf. the opaque скрижаль [⸨скрижал⸩.ь].

Commentary on individual lexemes осла [ост.л.а] ‹646›; cf. остръ. число [чит.т.л.о] ‹1076›. гѫсли [гѫд.т.л.и] ‹214›. масло [маз.т.л.о] ‹536›.

мъдьлость [(мъд).ьл.ост.ь] ‹572›. Contra Večerka, who gives the headword мьдлость. The cluster дл, based on the spellout found in Cloz, is prohibited in the canon, cf. вед.л.ъ [велъ]. The shape of the suffix with front vowel is found in such forms as завидьливъ, разумьливъ, свѣтьлъ.


Here, the combination ел+secondary two- fold rule, but without suffix т: the expected *nomen agentis* marker -тел҄ь (and, further, also тел҄ьник.ъ) is supplied by the right edge of the root; cf. властел҄ь [(влад).т.ел҄.ь]. Note that both lexemes are represented by hapax glosses in Supr, respectively: обрѣтел҄ьникъ 186, 21 (spellout without kamora), and приобрѣтел҄ьникъ Supr 160, 24 (kamorated spellout). There are no other substantives ending in ел҄.ьн.ик.ъ. Do not confuse with обрѣтѣль [об.(рѣт).ѣл.ь], приобрѣтѣль.


§852. Group 9, carrier consonant м

Notes on individual suffixes

{23}м: found in a limited set of substantival stems; see the list below under suffix {57} ѧ/ен. All these stems from lexemes with anomalous declension of the type врѣмѧ 0/n; corresponding lexemes are not attested for stems зна.м.ен and пла.м.ен. Outside of this list is only румѣнъ [(руд).м.ѣн.ъ] ‹781›.

{24} им, ом/ем: 1) this is the regular м-Part suffix (м-Part forms are not included in the counts); the alloforms ом/ем are distributed by the twofold rule, the alloforms (ом/ем)~им are distributed by the class of the parent verb; 2) чаѥмьнъ ‹1061› (1× Euch) is the only case where a м-Part stem is expanded by a different suffix (see § 865, *A note on nominal stems and verbal platforms*).

{25} ьм: only found in ꙗрьмъ [(јар).ьм.ъ] ‹347› and derivatives, and in кръчьмл҄ꙗвати [(кръч).ьмл҄.а.в.а.т.и] ‹449›.


### §853. Group 10, carrier consonant н

Notes on individual suffixes

{26} н/ен is the regular initially ambivalent н-Part suffix, cf. възл҄юбл҄.ен.ъ, but призъв.а.н.ъ (н-Part forms are not included in the counts). Frequently, the н-Part stem is found in further derivatives, cf. даꙗниѥ, трьпѣниѥ (see § 865, *A note on nominal stems and verbal platforms*). Outside н-Part, the C-initial variant is possible after C-final stems, cf. (плет).н.иц.а [пленица], (съп).н.ъ [сънъ].

{26} ен: the alloform ен of this suffix is not always easily distinguishable from ен, the alloform of the finally ambivalent suffix ен/ѧ {57}. The majority of ен occurrences that belong to suffix {26} are originally participial stems (there is no clarity only for удѣбенъ ‹271› and студенъ ‹905›); the set of stems that contain the alloform ен of the suffix ен/ѧ {57} is given by list (see lists under the suffix ен/ѧ, § 861).

{26} н: outside н-Part stems is found as part of the verbal marker н.ѫ, cf. постигнѫти. Outside of verbal stems (see § 865, *A note on nominal stems and verbal platforms*) there are slightly under 100 lexemes, cf. пленица, страна, вльна.

{27} ин: in most cases is the marker of subtantival feminine lexemes of the twofold declension, cf. храмина, истина. Otherwise for adjectives, cf. игълинъ, змиинъ, and in nouns referring to persons, воинъ, поганинъ.

{28}ун: only in пѣстунъ.

{29}ън/ьн is a twofold suffix (see § 863, *A note on secondary twofold rule*). See § 864 on forms with the marker -ьн҄ь.

{30}ын҄: this suffix marks morphologically feminine (2/f\*) lexemes in the twofold declension, and is always accompanied by segmentless substitutive

softening of the final н. The kamoration is preserved in further derivatives, as in пустын҄ьникъ (see § 864).

{31}ѣн: 1) in граЖанинъ, жѧтел҄ꙗнинъ and плащаница and their derivatives, the final consonant of the basic component is found in the C• grade by substitutive softening; 2) spellouts with ен and ѣн, and also with ьн compete in some stems. See details in Ch. 24, *Supplement*, § 896, *Lexicographic difficulties in registering lexical aberrations*.


#### §854. Group 11, carrier consonant р

Notes on individual suffixes

{32} р: there are slightly under 50 lexemes, but they are concentrated in a limited set of roots: братръ ‹47›, бъдръ ‹63›, вѣтръ ‹139›, даръ ‹216›, добръ ‹235›, дѫброва ‹274›, мокръ ‹537›, остръ ‹646›, пиръ ‹651›, хытръ ‹1048›, щедръ ‹1108›.

{32} ор/ер: 1) this is a twofold suffix (see § 863, *A note on secondary twofold rule*). In Ps Sin, there is the aberrant spellout седмер- (седмерицеѭ 78, 12; 11, 7); 2) this is a stem extension in the declension of the unique дъщи, мати: дъщер-, матер-; 3) cf. the opaque ⸨четвор⸩.ъ, ⸨јетер⸩.ъ and ⸨котор⸩.ъ (in Mar, there is the spellout котер-: котерааго же отъ васъ оца въспроситъ снъ хлѣба· еда камень подастъ ему Lk 11, 11).

{33}ар҄: this suffix marks masculine (2/m\*) lexemes in the twofold declension, and always shows segmentless substitutive softening of the final р. The kamoration is preserved in further derivatives, e.g. вратар҄ица (see § 864 below).

{34}ыр only in секыра and пастыр҄ь.

Commentary on individual lexemes заматорѣти [за.мат.ор.ѣ.т.и] ‹541›, cf. aberrant spellout with заматер- in Zogr; see also canonical матерьство [мат.ер.ьств.о].

пещера [пек.т.ер.а] ‹660›, the alloform ер contra the standard twofold rule and following the secondary twofold rule, which allows both variants after т (as in мат.ер//мат.ор). See more details in § 863, *A note on secondary twofold rule*.

пастыр҄ь [пас.т.ыр҄.ь] ‹659›, shows the composition of suffix ыр plus the segmentless substitutive softening of the final р (see § 864).


### §855. Group 12, carrier consonant с

Notes on individual suffixes

{35} с: only in гньсь [(гнь).с.ь] ‹171›. Cf. also as part of the opaque stem пѣснь [⸨пѣсн⸩.ь] ‹738›.

{35} ес: only as part of substantive stems that distinguish expanded (with the ес expansion) and syncopated (without a suffixal expansion) workstems in the declension (type слово 0/n and type око 0/n), and their derivatives (cf. словесьнъ). Otherwise only in ложесно (for ложе 2/n).

{36} ст: only in пръсть [пръ.ст.ь] ‹702›, ристати [ри.ст.а.т.и] ‹769› and their derivatives.

{36} ост/ест is a twofold suffix (see§ 863, *A note on secondary twofold rule*).

{37} ьск is a marker of adjectival lexemes (class 2/a). Otherwise only in воиска ‹96› 2/f 'war'.

{38} ств: only with roots бѣг ‹69› (only бѣство), чьт ‹1076› (бечьствовати), and шьд ‹1099› (шьствиѥ).

{38}ъств/ьств: is a twofold suffix (see §863, *A note on secondary twofold rule*).

Commentary on individual lexemes дьньсьн҄ь [дьн.ь.с.ьн҄.ь]. Here, с is effectively a root formative; this adjectival base derives from the phrase дьнь сь+ьн҄.ь (on the adjectival marker -ьн҄ь see § 864). Cf. дьньшьн҄ь.

роЖьство [роЖ.ьств.о] ‹755›. Substitutive softening of the basic component; aberrant spellouts розьство, порозьство (Mar and Cloz) by the alternative version of pairings of substitutive softening. Otherwise, родьство [род.ьств.о]; for distribution in sources see Večerka.

радоща [рад.ощ.а] ‹752›. Substitutive softening used as marker of the substantive lexeme, cf. радость (more details in § 864). Only IPl form is attested in the occurrences, cf.: вьзигра сѧ младѣнищь радощами вь чрѣвѣ моемь Lk 1, 44 Zogr.



This is the regular Inf, Sup, and т-Part suffix, found both after vowels (cf. сѣть, съмрьть) and after consonants (cf. мощь [мог.т.ь], власть [влад.т.ь]). Inf, Sup, and т-Part forms are not included in the counts.

In изѧщьничьскъ the basic component of the suffix ьн, which etymologically contains the т-Part stem (ѩти), shows segmentless substitutive softening.

The combination т.ел҄ has more than 80 occurrences. Altogether there are slightly fewer than 300 lexemes that contain the suffix т. It is not always easy to classify the basic stem as a verbal platform (see § 865, *A note on nominal stems and verbal platforms*), as opposed to a root or nominal stem. Outside of the verbal platforms shown above, there are fewer than 40 lexemes; here are some examples: (сла).д.т.ьн.ъ [сластьнъ], (страд).т.ь [страсть], (та).т.ь [тать], (вра).т.а [врата], за.(вид).т.ь [зависть], (врьт).т.а [врьста], (зла).т.ик.ъ [златикъ], у.(жас).т.ь [ужасть], (чрѣз).т.л.а [чрѣсла], (*i̯*ес).т.ьств.о [ѥстьство], (маз).т.л.ин.а [маслина], (мла).т.ъ [млатъ], (пѫ).т.а [пѫта].

The forms къто [(к).ъ.т.о], чьто [(ч).ь.т.о] ‹359› stand separately. Here т is the pronominal root ‹938›; the form то functions as a cliticized emphatic, cf. Russian *это*, *эта*, but *этот*.


§857. Group 14, carrier consonant т, other than suffix т {39}

Notes on individual suffixes

{40} ат: not to be confused with the sequence а.т, where а is the theme and т is the inifinitive suffix. There are fewer than 20 lexemes with the suffix ат; they

are concentrated in a limited set of roots: бог.ат ‹37›, крил.ат ‹435›, пер.ьн.ат ‹658›, сѫк.ат ‹937›. It is also found in the composition ат+ај: ходатаи, поводатаи, позоратаи. Cf. the opaque stem in печать ‹662›.

{41} ит: not to be confused with the sequence и.т, where и is the theme and т is the infinitive suffix. This suffix marks adjectival lexemes, cf. ꙗдовитъ 2/a. The only substantive lexeme is рыбитвъ 2/m.

{42} от/ет is a twofold suffix; see § 863. Cf. the opaque stems in кл҄ѥвета ‹392›, трепетъ ‹968›.

{43}ът/ьт is a twofold suffix; see § 863. Cf. opaque stems in папрьтъ ‹657›, хрьбьтъ ‹1044›, оцьтъ ‹648›.


§858. Group 15, carrier consonant х, ш

Notes on individual suffixes

{44} х: there are slightly fewer than 70 lexemes with the suffix х, but they are concentrated in a limited set of roots: обѫхати [об.(ѫ).х.а.т.и] ‹111›, грѣхъ [грѣ.х.ъ] ‹154›, прахъ [пра.х.ъ] ‹702›, слухъ [слу.х.ъ] ‹844›, смѣхъ [смѣ.х.ъ] ‹858›, спѣхъ [спѣ.х.ъ] ‹872›, стрѣха [стрѣ.х.а] ‹892›. Apart from these: вънѣшьн҄ь, вьчерашьн҄ь, домашьн҄ь, дьньшьн҄ь, кромѣшьн҄ь, нын҄ꙗшьн҄ь, утрѣшьн҄ь. Cf. aberrant forms with щ: вьчеращьн҄ь, домащьн҄ь, нынѣщьн҄ь. All these forms are anomalous from the point of view of the pRs schema: the vowel that precedes the suffix ш is an inflection.

{44}ъх: only in ветъхъ, ветъшати, обетъшати, обетъшити. Cf. suffix {48}, one of whose alloforms, viz. ъш, coincides with an alloform of this suffix {44}.

{45} их: only in the four lexemes listed above.

{46} ох: only in the three lexemes listed above and derivatives.

{47}ух: only in пастухъ, пастуховъ, гор҄юшьнъ.

{48} (ъш/ьш)/въш: 1) all three suffixes are morphophonological variants of the ш-Part suffix (twofold rule and initial ambivalence); ьш is the suffix in the expanded comparative forms (ш-Part forms and comparatives are not included in the counts). Outside of these forms, the suffix is only found in the superlatives старѣишина, ун҄ьшиина and their derivatives; 2) see § 309 and ff. for the marking of syncopated ш-Part and Compar forms (cf. нес.ъш.и but нес.ъ|, л҄юбл҄.ьш.и but л҄юбл҄.ь|, милова.въш.и but милова.въ|); see more details in § 309 and further.



Notes on individual suffixes

{49} ищ: a marker of substantival lexemes of the twofold neuter (2/n) and masculine (2/m).

{50} ѫщ, ѧщ: 1) this is the regular suffix of expanded щ-Part forms (щ-Part are not included in the counts); distribution follows the class of the parent verb. See § 309 and ff. on the marking of the syncopated щ-Part forms (cf. л҄юб.ѧщ.и but л҄юб.ѧ|, нес.ѫщ.и but нес.ы|, плач.ѫщ.и but плач.ѧ|). Aberrant forms with ѫц, ѧц are found in Kiev, cf. 6b, 1 противѩцꙇхъ; 2) outside of the participial stems, this suffix is only used in the doublet forms тысѫщи// тысѧщи and their derivatives.

# Suffixes without a carrier consonant


§860. Group 17, isolated vowels

Notes on individual suffixes

{51} а: this is the theme vowel that marks verb classes 3, 6, and 7 (and some unique verbs). It is found in infinitives and nominal derivatives with the corresponding verbal platforms (§ 865, *A note on nominal stems and verbal platforms*). Occurrences of this suffix as a theme are not included in the counts.

{52}ѣ: this is the theme vowel that marks verb classes 2 and 7 (and some unique verbs). It is found in infinitives and nominal derivatives with the corresponding verbal platforms (§ 865, *A note on nominal stems and verbal platforms*). Occurrences of this suffix as a theme are not included in the counts.

{53}и: this is the theme vowel that marks verb class 1; also found in the unique verb въпити. It is found in infinitives and nominal derivatives with the corresponding verbal platforms (§ 865, *A note on nominal stems and verbal platforms*). Occurrences of this suffix as a theme are not included in the counts.

{54}ѫ: this is the theme vowel that marks verb class 5. It is only found in verb class 5. Occurrences of this suffix as a theme are not included in the counts.


§861. Group 18, finally ambivalent suffixes

All these suffixes except ѧт in reality have the shape H, and ѧт has the shape HC: the series H(*u*) {55}, {56}, and the series H(*n*) {57}, {58}. All except the suffix ъв/ы represent the same grade of the vocalism. Arbitrarily, these suffixes are treated simply as finally ambivalent, rather than unstable in their vocalism. This solution permits us to limit fundamental alternations to roots. The alloform ен of the suffix ѧ/ен is segmentally identical to the suffix ен (see § 853 above); all occurrences of the suffix ѧ/ен are given below by list.

Suffixes ѧт/ѧ and ѧ/ен have segmentally identical alloforms; all the occurrences of the suffixes are given below by list.

Notes on individual suffixes

{55} (ов/ев)/у: 1) the alloform у is only found in PRAE class 6 verbs; cf. чешуꙗ with an opaque stem; also in compounds, cf. полунощь; 2) ов/ев is distributed by the twofold rule; 3) this suffix is a class 6 verb marker, and a marker of adjectives; 4) the lexeme дьневьнъ violates the twofold rule (on aberrant spellouts of the substantive дьнь see § 401, *Nominal aberrations*).

{56}ъв/ы is distributed by the CVC agreement rule. There are about 30 lexemes with this suffix, concentrated in a limited set of roots. These are: брады ‹45›, букъви ‹64›, жрьны ‹294›, локы ‹503›, любы ‹530›, неплоды ‹673›, свекры ‹805›, смокы ‹860›, хорѫгы ‹1037›, црькы ‹1051›, цѣлы ‹1054›. All these stems form lexemes in the anomalous declension of the type црькы 0/f.

{57}ѧ/ен is distributed by the CVC agreement rule. All lexemes with this suffix are concentrated in a limited gorup of roots. Namely, this is the lexeme имѧ ‹333› and its derivatives, as well as roots that are marked by the suffix sequence м.ен; such are: (брѣ).м.ен/(брѣ).м.ѧ ‹24›, (врѣт).м.ен/(врѣт).м.ѧ ‹117›, (зна).м.ен/[(зна).м.ѧ] ‹319›, (пла).м.ен/[(пла).м.ѧ] ‹654›, (пис).м.ен/(пис).м.ѧ ‹666›, (плед).м.ен/(плед).м.ѧ ‹673›, (сѣ).м.ен/(сѣ).м.ѧ ‹926›, (чит).т.м.ен/(чит).т.м.ѧ ‹1076› (cf. suffix {23} above). Conventionally, the following occurrences of ен have also been included here: ѥлень ‹352›, камень ‹366›, корень ‹414›, младеньць ‹554›, прьстень ‹718›, пътеньць ‹732›, ремень ‹768›, степень ‹885›, and also зеленъ ‹312›.

{58}ѧт/ѧ is distributed by the CVC agreement rule. It is only found in the declension of unique lexemes in the отрочѧ 0/n group, viz.: агнѧ ‹3›, жрѣбѧ ‹297›, кл҄юсѧ ‹393›, козьлѧ ‹400›, овьчѧ ‹627›, осьлѧ ‹643›, отрочѧ ‹766›, and also in those listed in the table above and their derivatives.

# §862. On the distribution of alloforms of polyvariate suffixes

Let us examine three cases, in order: 1) alloforms *P* and *Q* are morphophonological variants, e.g. н/ен; 2) alloforms *P* and *Q* are related by *standard segmental alloformy*, e.g. ик‖иц; 3) alloforms *P* and *Q* stand in a relation of nonstandard alloformy, e.g. (иг, ог).

In the cases 1 and 2, *P* and *Q* have definite segmental similarities and differences (in the case of twofold variance, they must have appropriate initial vowels; in case of ambivalence, they must have appropriate CVC schemata); in case 3, segmental similarities and differences are in the general case not regulated.

1. Alloforms *P* and *Q* are morphophonological variants. Although the distribution rules are included in the statement of alloformy itself (the twofold rule for twofold formatives, the CVC agreement rule for CVC-ambivalent ones), each particular pair of alloforms may depart from the expected distribution. For example, the twofold rule is in some cases violated in the distribution of the suffix ов/ев (cf. дьневьнъ instead of the expected \*дьновьнъ; архиереовъ instead of the expected \*архиереѥвъ); the CVC agreement rule is violated in the distribution of the suffix н/ен (cf. гроз.н.ъ instead of the expected \*гроз.ен.ъ).

2. Alloforms *P* and *Q* are related by standard segmental alloformy. In this case, distribution rules are not part of the statement of the alloformy, which only covers the segmental pairings in the final C position. For example, we have ык‖ыц‖ыч (cf. владыка, владыцѣ, владычьствиѥ), ик‖ич‖иц (cf. мѫченикъ, мѫченичьскъ, гѫсѣница), ел‖ел҄ (веселъ, учител҄ь).

The distribution of alloforms of polyvariate suffixes assumes a key difference between inflection and derivation: in the former case, forms are built by an algorithm which determines the transition from a paradigmatic call (the input to the algorithm) to the wordform that answers the paradigmatic call (the output or result). There is nothing alike in derivation.

In inflection, in cases of morphophonological variation, the variant choice of a given formative is determined by the preceding formative that is part of the morphological skeleton. In cases where the preceding formative is itself polyvariate, the skeleton fixes one specific alloform; cf. град.ѣ (the workstem in the skeleton is град, not граЖ); кльн.ъш.и (the workstem in the skeleton is кльн, not клѧ). In case of segmental alloformy, the segmental shape of a given occurrence of a formative is determined by segmental replacement rules: the final consonantal segment is replaced in the formative that stands last in the workstem. Thus, the morphological skeleton of the form влад.ыц.ѣ contains the workstem влад.ык (and not влад.ыч or влад.ыц); the last consonant of the last alloform of the workstem is subject to the replacement к→ц, i.e. the alloform ык is replaced by the alloform ыц.

In derivation there are no rules of synthesis of derived stems. Thus, observations on the distribution of polyvariate formatives are limited to statements of allowed and prohibited adjacencies, i.e. by syntagmatics. Thus, in a derived stem like, тѧж.ест.ь or тѧг.ост.ь, we have no basis for taking as the starting alloform тѧг rather than тѧж, or vice versa. In the derived stem кон.ьч.ьн.ъ or роЖ.ьств.о, we have no basis for setting up the starting stem as кон.ьц or кон.ьк or кон.ьч, or, respectively, род or роЖ. A question such as whether тѧг was replaced by тѧж before the suffix ост/ест, or, conversely, whether ест/ост takes the variant ост after тѧг, in the context of derivation is not posed and not discussed. Likewise, the question is not posed why (добл҄).ьств.о shows a substitutively softened consonant, C• grade, while по.(доб).ьств.ьј.е shows the consonant in the C° grade. Cf. such full derivational doublets as (чу*i̯*).ьств.о and (чу).в.ьств.о, (прѣ).д.ьн҄.ь and (прѣ).Ж.ьн҄.ь, (роЖ).ьств.о and (род).ьств.о. Lexeme pairs such as these realize alternative derivational models. The choice between them is not treated in this book.

Of course, paradigmatic rules that regulate segmental alloformy remain in force in derived stems: we have двьрьник.а (GSg for двьрьникъ), двьрьниц.ѣ (LSg for двьрьникъ), etc. However, the question of why двьрьникъ in its starting form shows the suffix alloform ик, while двьрьница shows the alloform иц, is not posed. Likewise, rules that regulate selection of morphophonological variants within a paradigm remain in force: cf. от.роч.ѧ, but отроч.ѧт.а, имѧ, but им.ен.и. However, the CVC agreement rule cannot explain why in зла.к.ъ the root is represented by the V-final variant, while in зел.ен.ъ by the C-final variant.

### §863. A note on secondary twofold rule

The standard twofold rule requires the overslash (back) variant after C°. However, among twofold suffixes, this condition is obscured, cf. ор/ер: заматорѣти and заматерѣти, десѧторъ, but матерьство, and some others. In this grammar, the twofold rule for suffixes is extended to allow underslash (front) variants after C°. The corresponding distribution of variants is called *secondary twofold rule*, and suffixes that demonstrate such a distribution are called *secondary twofold suffixes*. Such a treatment significantly reduces the suffix inventory, unifying suffixes that differ in their initial vowels and that are distributed by the secondary twofold rule.

### §864. Suffixes as markers of nominal stems

Most suffixes can function as markers of nominal stems. Each suffix usually corresponds to a particular paradigmatic class, with a distinction between soft and hard subtypes in case of twofold declension, cf. лѫк.ав.ъ but пьс.ьј.ь. In case of substantives, gender can vary, cf. шип.ък.ъ 2/m and тет.ък.а 2/f. In the following two cases, the picture is somewhat more complicated.

A note on morphological gender (двьрьникъ — двьрьница)

In this pair of lexemes, the stems contains the same string of formatives, and differ only in the grade of the final consonant of the final suffix. In such pairs, the difference in grades of the final consonant of the basic component regularly corresponds to a difference in morphological gender:27 the grade к belongs to 2/m and the grade ц belongs to 2/f. Cf.: багърѣница, блѫдьница, гѫсѣница, грѣшьница, десница, златица, колесьница, мьздьница, ѫсобица, срачица, четворица, шуица; but блѫдьникъ, съборьникъ, вратьникъ, въгодьникъ, градьникъ, грѣшьникъ, понедѣл҄ьникъ, златикъ, молитвьникъ, обьщьникъ, праздьникъ, тъчьникъ. Otherwise only for корабл҄иць 2/m and сиць 2/p.28

A note on segmentless substitutive softening (коза — кожа)

In this pair of lexemes, the stems, while not different in the formatives they contain, differ only in the grade of the final C of the last formative (root in this case). In such cases, a difference in grade corresponds to a difference in inflectional subtype: коза 2/f belongs to the hard subtype, and кожа 2/f to the soft subtype. The segmental differences of these alloforms are related to the pairing by the substitutive softening alternation. Accordingly, the stem кож can be represented as коз+substitutive softening, as if the sequence коз acted as a basic component, and replacement by substitutive softening as a suffix. This stem-formation effect in this book is called *segmentless substitutive softening*. It is found where an occurrence of a substitutively softened formative is motivated neither by paradigmatic conditions nor by syntagmatic requirements of a following suffixal formative. Note that there are no suffixes that require the grade

<sup>27</sup> The suffixes ик(ъ) and иц(а) are often thought to belong to different families, and the suffix иц(а), as well as ьц(ь), is seen to undergo the so-called third palatalization. See, for example, Xaburgaev, § 122–123, 180–182.

<sup>28</sup> There is one more suffix (ък/ьк, ъц/ьц), where the pairing к‖ц is observed in the final C position in markers of nominal stems. The distribution of the alloforms of this suffix as markers of nominal stems is as follows: ьк is only found in adjectives (such are мальчькъ, тѧжькъ and гор҄ькъ); ьц is only found in substantives (cf. рожьць, слъньце, овьца; slightly over 50); ък is found both in substantives (cf. тетъка, шипъкъ), and in adjectives (cf. мѧкъкъ; slightly fewer than 50); ъц is impossible as a separate nominal stem marker.

C• of the preceding suffixal formative by substitutive softening. Suffixes that can combine with a substitutively softened basic component also show stems with a basic component ending in a simple consonant, cf. (граЖ).ѣн.ин.ъ, but (млад).ѣн.ьц.ь and similar. Segmentless substitutive softening often acts as a marker of nominal stems. Cf. among substantives: духъ 2/m and душа 2/f ‹261›, свѣтъ 2/m and свѣща 2/f ‹810›; among adjectives: удобъ and добл҄ь ‹235›; also among suffixal formatives: гор.ьн.ъ and гор.ьн҄.ь ‹186›. Often we find substitutive softening in suffixal lexemes that have no lexical correlates with a C° grade in the final position. Such are, first, all substantive lexemes with the marker тел҄.ь: cf. учител҄ь, мѫчител҄ь, etc.;29 second, a large proportion of adjectives with the marker ьн҄.ь: cf. ближьн҄ь, вечерьн҄ь, вышьн҄ь, врьховьн҄ь, and many others. The result of segmentless substitutive softening is usually preserved in further derivatives, cf. the preserved kamoration in such derived stems as родител҄ѥвъ, родител҄ьница, властел҄ьскъ [(влад).т.ел҄.ьск.ъ], мѫчител҄ьство, etc.

Note that segmentless substitutive softening in some cases distinguishes pairs of lexemes that are regularly opposed in meaning. First, substitutive softening distinguishes possessive adjectives from possessor nouns, cf. ѥлень and ѥлен҄ь, вельбѫдъ and вельбѫЖь, пророкъ and пророчь, отьць and отьчь, and many others; second, old comparatives from their parent adjectives, cf. грѫбъ and грѫбл҄ь (грѫбл҄.ьш.и), худъ and хуЖь (хуЖ.ьш.и), лихъ and лишь (лиш.ьш.и).30

In other cases simple vs. substitutively softened basic components are in variation, as in родьство and роЖьство, земьскъ and земл҄ьскъ, or бъдръ and бъЖр҄ь; in such doublets, sometimes only a single variant is included in the canon (so, бъЖр҄ь is aberrant and бъдръ is canonical), or both variants are canonical (as in родьство and роЖьство, земл҄ьскъ and земьскъ).

### §865. A note on nominal stems and verbal platforms

Deverbal stems are a special category within nominal stems. Such stems can be represented bicomponentially: [verbal platform] + [nominal suffixes], e.g. [род.и.т]+[ел҄], [трьп.ѣ.н]+[ьј]. Nominal suffixes are V-initial in their CVC shape. Accordingly, the verbal platform is expected to be C-final. This shape is ensured, first, by truncated stems in case of closed roots, and second, by stems of nominal forms, infinitive/supine or participles. щ- and ш-Part stems (active participles) rarely participate in forming verbal platforms for nominal stems (cf., however, сѫщиѥ, сѫщьство, and сѫщьнъ, щ-Part); other stems actively participate in nominal derivation. These are derivational patterns in -аниѥ, -ениѥ (the


н-Part platform), -тие; -тел҄ь, -ател҄ь, -ител҄ь (Inf platform); -ало, -ило (л-Part platform), and others. With this general background, some specific notes are in order.


# CHAPTER 24 **Supplement**

### §866. Trubeckoj's nonstandard phonology

Indeterminacy in the choice of the phonological inventory is nearly always present in descriptions of currently spoken languages. This problem becomes more serious when we are facing a dead language where only written data are available. Specific spellouts may have indeterminate phonological analysis, or more generally, the indeterminacy may lie in the inventory of phonemes, let alone the phonetic interpretation. Whether such ambiguity is motivated or not is a property of the language itself. This issue was first examined for OCS by Trubeckoj (1954), who showed that OCS can be analyzed with two equivalent phonemic inventories: one containing the *yod*, which we here call "standard phonological inventory", and one without *yod*.1 Another choice has to do with treatment of nasal vowels.2 Trubeckoj treats the nasal vowels *ę* and *ǫ* as combinations of the form *e*+� and *o*+�, where � is a consonantal segment.3

Thus, there are two competing phonologies. The first one, which we here call *standard* or *yodful*, establishes the vowel inventory as in Table 23 (p. 15) and the consonant inventory as in Table 27 (p. 16), which we can notate as C∪{j}. This vowel inventory is also shown in Figure 4, left. The second phonology, which


FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

we shall call *nonstandard* or *yodless* (with splitting of the nasal vowels), is based on the vowel inventory shown in Figure 4, right, and the consonant inventory C∪{�}. The consonant inventory of the yodless phonology differs from that of the standard phonology by the absence of /j/ and presence of /�/.

Figure 4. Yodful (left) and yodless (right) vowel inventories.

Yodless phonological representations can be derived from yodful ones by the following rules, applied in order:


For example: *ǫže*⇒*o*�*že* (ѫже 'rope'), *jǫže*⇒*jo*�*že*⇒*ö*�*že* (ѭжеASgf for \*и), *zeml'ǫ*⇒*zeml'o*� (земл҄ѭ ASg for земл҄ꙗ), *vyjǫ*⇒*vyjo*�⇒*vyö*� (выѭ ASg for выꙗ); *l'ubl'ǫ*⇒*l'ubl'o*� (л҄юбл҄ѭ 1SgPrae for л҄юбити); *jęti*⇒*je*�*ti*⇒*e*�*ti* (ѩти Inf), *zeml'ę*⇒*zeml'e*� (земл҄ѩ GSg for земл҄ꙗ); *jako*⇒*äko* (ꙗко), *kon'a*⇒*kon'a* (кон҄ꙗ GSg for кон҄ь), *kraja*⇒*kraä* (краꙗ GSg for краи); *krajemь*⇒*kraemь* (краѥмь ISg for краи); *kraju*⇒*kraü* (краю DSg for краи), *jugъ*⇒*ügъ* (югъ), *uxo*⇒*uxo* (ухо).

Clearly the yodless and yodful phonologies have the same sets of phonological contrasts, although in many cases the phonemic composition of words differs between the two analyses.

Yodless phonology with nasal splitting perfectly agrees with the graphics of early Glagolitic sources, and was designed as a representation of those sources by Trubeckoj. The earliest Glagolitic alphabet contains vowel letters4 that unambiguously match the 11 vowel phonemes of the yodless phonology:


Glagolitic has special letters for *ü* and *ö*, while Cyrillic nasal letters ѫ, ѭ and ѧ are represented in Glagolitic by digraphs:

<sup>4</sup> More precisely, vowel graphemes; cf. the three symbols for /i/, and the digraph for /y/.


Early Glagolitic lacks a letter corresponding to the phoneme /j/, as well as letters that match Cyrillic iotated ones.

Trubeckoj used his construction as an argument for the historical primacy of Glagolitic, which was controversial at the time. He showed that it was the primary alphabet designed by St. Cyril specifically for the Slavic language.

Trubeckoj also notes that the yodful/yodless distinction might have been relevant for Slavic dialects at the time Slavic writing was designed, and that the dialect that underlies OCS was yodless. Over time, OCS changed toward a yodful phonology, and the graphics of Cyrillic sources reflect this shift.

# §867. On word-initial vowels

When a paradigmatic wordform begins with a vowel, its initial formative is nonstandard. Indeed, the CVC norm requires both prefixes and roots to be C-initial. Phonology allows only the vowel и to be word-initial. In the mapping from morphophonological to phonological representations, all initial front vowels except и receive an epenthetic *i̯*. Thus, vowel-initial forms with vowels other than и are anomalous. The Tables 867.1–5 below compare initial V with initial jV. They include all relevant material.5


Table 867.1. Initial #а~#ја, #ѣ~#јѣ

\* Also #j.а with the pronominal root j ‹343›.

5 The anomalous, initially ambivalent root ъп/въп ‹129› is a special case (see § 813). There are no other root or prefixal formatives with initial ъ or ы. On initial ь see below in Table 867.3.

The combination jit is allowed in mph representations, but can only occur at formative boundaries, cf. cmpoj.th.aure lmf. This is due to the fact that the choice between mph representations with word-initial #j's vs. #ja is conventionally made in favor of ja. For example, for <349> we have jach and not jtkey; for <345> jan and not jtm (rana), etc. Cf. also the roots jac <348> (погасъ), ап <10> (вънєзаапъ) and ap <12> (разарати), which are not attested word-initially. Also note the aberrant forms \$3% (MAR) for a3%, \$11544 (CLOZ) for armsus, and also BBHE3ATHE (Mk 13, 36 ZOGR).


Table 867.2. Initial #0 ~ #jo, #e ~ #je


\* Also #j.€ with the pronominal root j <343>.

Also, the prefixal formatives or/oms and on belong to the same class.

Among extraparadigmatic formatives note the potentially correctable anomalous formative ६, which does not take epenthetic i, cf. ६८६.

Table 867.3. Initial #b ~ #jb, #n ~ #jr


\* Only #j.b and #j.n combinations with the pronominal root j <343>.

\*\* Root <339> is attested only with a prefixal formative.

Also, the prefixal formative n3 belongs to the same category.

Pairwise distinction between morphophonologicalis, jb, jn, and jn, which are all realized phonologically as /n/, are not always well-supported. For ex-

ample, for root bm <34> (the mph representation without j, and with b, rather than n, is supported by the paradigms of the verb ผลาห and its derivatives (BB3ATH, BB36MEWH; BB36MATH, BB36MAREUH). However, in other cases the question cannot be so easily settled. For example, for root иг <330> (игълинъ), the choice between isr, jur, or jur cannot be reliably motivated. This grammar makes the following arbitrary choice: the morphophonological representations ib, jb, and jn are marked, i.e. are assigned to particular formatives, only if that choice is supported by evidence. In all other cases, the spellout n is selected.

Table 867.4. Initial #oy ~ #joy, #x ~ #jx


Cf. nonstandard alloforms вък <132>, вон <111>, вяз <148>. The roots of <1010>, 013 <1012>, x <111>, xv <1121> are only found with prefixal formatives.

\* Only #j.x with the pronominal root j <343>.

The prefixal formatives of and x also belong to this class.

Note also the aberrant spellout ютр- (ZOGR, MAR, SAV, PS SIN, EUCH), ПАНЖЧИН- (SUPR).

Table 867.5. Initial #A~#jA


The root arp <1116> is attested only with a prefixal formative.

\* Also #j.a with the pronominal root j <343>.

### § 868. On wordform boundaries and loose and tight formative adjacencies

1. In separating an OCS text into linear units it is sometimes necessary to consider not just ordinary wordforms corresponding to single lexemes, but also clitic groups. A clitic group in the trivial case consists of a single wordform, but can also consist of a string of wordforms, each corresponding to its own lexeme. In the majority of cases, a nontrivial clitic group contains one ordinary wordform, called the clitic group's head, and accompanying clitics. The following sequences illustrate:


Less commonly a clitic group can consist of a string of clitics, cf. въ+\*и+же — вън҄ьже).

Membership in the class of clitics is a lexical property, and as such should be given by a list. It is a property of wordforms, not lexemes: some wordforms of a lexeme may be clitics while others may be full words. Most clitics are extraparadigmatic forms (conjunctions, particles, and prepositions), but some nominal and verbal wordforms are also clitics.6

2. Clitics have certain properties that distinguish them from ordinary forms. These properties manifest themselves in both syntactic and morphophonological behavior. From comparative Slavic grammar we know that clitics have particular accentual properties. Clitics form a certain linear unity with their head. Rewrite rules that may apply at boundaries within this unity are similar to those seen at formative boundaries within a wordform. For example, we have из+рѫкы ⇒ издрѫкы, just as из+рещи⇒издрещи; из+чрѣва⇒ичрѣва, just as из+чисти ⇒ ичисти. Such combinations are treated in the same way as the boundaries between prefixes and following formatives (i.e. by the general mph⇒ph/norm rules).

Combinations of a preposition with the pronoun \*и, which in this case takes an initial н҄, in canonical OCS do not undergo any segmental rewrite rules (cf. из него Mk 1, 25 Mar). Aberrant combinations are possible where зн҄ is mapped by an aberrant cluster rule (see § 77), cf. из+н҄ѥго ⇒ из.н҄ѥго ⇒ ижн҄ѥго (cf. ꙇжн҄его Mk 1, 26 Zogr, иж него Mk 1, 26 Mar), also без+н҄ѥго (cf. бежн҄его Jn 1, 3 Zogr, бежнего Jn 1, 3 As) etc.

3. Boundaries between clitics and heads can also ignore Havlík's rule, which generate *yer* aberrations. So, canonical рабъ+тъ can be realized as работъ (from the string рабътъ), or раб тъ (from the isolated string рабъ).

4. Because some rules in OCS distinguish two types of adjacencies (tight and loose), the strength of boundaries between formatives ranges along the following scale: tight adjacency, loose adjacency, clitic adjacency, word adjacency (space between words).

5. The distinction between loose and tight adjacencies within a paradigmatic wordform lies within the domain of paradigmatics. In this book, each

<sup>6</sup> OCS clitics are not well-studied. There may be insufficient OCS data to establish precisely the set of clitics. In this book, the set is not fixed, because there is no need to do so in view of the benchmark tasks of the grammar.

adjacency is defined as either loose or tight. The results of a perhaps arbitrary decision are shown in Ch. 4, § 69–71, *Rules of the first cycle*. Observed facts that contradict it are treated as aberrations.

# §869. On the law of the velars

1. This law (Table 869) prohibits velar + front vowel combinations both in phonological and morphophonological representations, in the latter case within as well as across formatives.

Table 869. The law of the velars


Among borrowings, formatives may show violations of this prohibition, but such violations cannot occur across formative boundaries.

Here are some examples.


For lexemes from the benchmark list, there is a fixed canonical spellout of the root formative that is shown in the dictionary. There are 13 lexemes in the PD and the benchmark list with prohibited combinations with velars; they are архиереи, архиереискъ, архиереовъ ‹11›; геона, геоньскъ ‹157›, декѧбръ ‹226›, кесар҄ь, кесар҄ѥвъ ‹373›, китъ, китовъ ‹374›, херовимъ, херовимьскъ ‹1026›; хитонъ ‹1027›. In general, spellouts in the sources are unstable. All non-canonical spellouts are aberrant. Among attested spellouts of the forms of the lexemes геона (22 glosses), and геоньскъ (3 glosses), not a single one in the sources is canonical. For remaining lexemes, aberrant forms occur alongside canonical ones (see details in Večerka). Cyrillic transliterations of Glagolitic sources, as well as in Cyrillic sources, may include the letters к҄ and г҄. These letters are used in Večerka's dictionary in names of headwords. Spellouts of lexemes outside the benchmark list cannot be distinguished as canonical vs. aberrant. The spellout закьхѣе Lk 19, 5 Mar is noteworthy (see other spellouts of this lexemes and their distributions across sources in Večerka).

2. Prohibited combinations occur in morphological skeleta, but all such combinations are repaired by boundary adjustment rules. Indeed, workstems ending in к, г, х can take terminals beginning with front vowels. In the nominal declension, for example, we have враг+ѣ (LSg), враг+и (NPl), and many other examples, where prohibited combinations are repaired by boundary adjustment rules: враѕ=ѣ, враѕ=и. In conjugation we have мог+еши (Prae), мог+ѣте (Imv), мог+ѣахъ (Imf). After boundary rules we have мож=еши (Prae), моѕ=ѣте (Imv), мож=ѣахъ (Imf). It is important to note that while boundary rules ensure compliance with the segmental law of the velars, those rules themselves are not segmentally decidable: which of several ways of removing prohibited velar combinations should apply to a particular form cannot be determined on the basis of the segmental context. The same potentially prohibited combination (in these examples, г+ѣ), can result in different outcomes: мог+ѣахъ⇒мож.ѣахъ (and later можаахъ), but мог+ѣте⇒моѕ.ѣте (and then моѕѣте).

3. Boundary adjustment rules prescribe the application of an alternation (or, more pedantically, a replacement by an alternation). Alternations differ from other mechanisms of segmental rewrite rules in that they are not representable as segmentally decidable rules. All segmentally decidable rules are included in the set of rules mapping morphophonological into phonological representations (mph ⇒ ph/norm). However, a different view of alternations is common, where alternations include the distribution of their grades as obligatory components (cf. Zaliznjak 1967, § 6.23–24). Under this approach, which is very well-established in linguistics, alternations as a mechanism of synchronic grammar almost perfectly copy diachronic sound changes. However, while making the connection between synchronic objects and their diachronic analogs is valuable, the connection is not clear when such diachronic analogs are included in the synchronic grammar without necessary restructuring.

4. As is well-known, the synchronic opposition (мог+ѣахъ⇒мож.ѣахъ) ~ (мог+ѣте⇒ моѕ.ѣте) results from different sources of the segment ѣ: from PIE \**ē* vs. PIE \**oi*, *ai*. In the historical grammar, this is the opposition of two dynamic processes, first and second palatalization. In order to preserve segmental decidability of the corresponding rules, for a single phonetic segment ѣ some authors assign two morphophonological precursors, ѣ1 and ѣ2. This leads to underlying representations of the type мог+ѣ1ахъ and мог+ѣ2те (cf. Lunt 1974, *Epilogue*, § 4.2). In this book, such a move is not allowed in principle: the inventory of atomic units in phonological and morphophonological representations is the same. One advantage of the present approach is that it makes it unnecessary to distinguish ѣ1 and ѣ2 where there is no synchronic necessity to do so, e.g. in closed roots, бѣ1л in бѣлъ, but пѣ2н in пѣна, making such choices a matter of etymology.

5. It is noteworthy that the synchronic law of the velars is not explicitly stated in grammars of OCS, neither by Vaillant nor Diels; only Lunt supplies a small paragraph in the section on phoneme combinations (Lunt 2001, §2.412–2.4121,

and Lunt 1974, §2.411–2.4111). Instructional grammars also ignore this law, e.g. Xaburgaev, Remneva, Gasparov.

6. The law of the velars is violated in aberrant spellouts that show the confusion of the *yers*. E.g. the following are attested: As ѩзкь мои Lk 16, 24; кь намъ Lk 16, 26; кьто Lk 10, 36; врагь члкь то сътворі Mt 13, 28; likewise кьнѧзу Supr 218, 9–10; скрьгьщѧ Mk 9, 18 Sav.

### §870. On partial neutralization of vowel advancement contrasts

Vowel advancement, represented by the four phoneme pairs (а~ѣ, о~е, ъ~ь, and ы~и), is contrastive in phonological representations after simple consonants (п, б, т, д, с, з, л, н, р, м, в): cf. басъ and бѣсъ, возъ and везъ, etc. In other contexts the contrast is almost entirely eliminated and the distribution is complementary. The opposition цѣ~ца is a special case (cf. тѧжьцѣ, but отьца).

Table 870 shows the situation after mph-soft (row 1) and velars (row 2); for comparison, simple Cs are shown in row 3.

Table 870. Contrast and complementary distribution of vowels by advancement


For example, in the context after velars (row 2) we have only back members of the pairs: а, о, ъ, and ы. After mph-soft consonants, we have the front members е, ь, and и, and the back one for а~ѣ. In other words, for а~ѣ, the simple rule (front after mph-soft, back elsewhere) does not apply. This rule also does not apply word-initially and after vowels (see Table 51 on p. 34).

In morphophonological representations, vowels are opposed in advancement both after mph-soft consonants and word-initially, but these are contexts of *neutralization*, i.e. contexts where a morphophonological opposition is eliminated in the phonology. In this connection, note alternative spellouts with ѣ after mph-soft consonants (e.g. чѣсъ, мол҄ѣаше or молѣаше), relatively frequent both in Glagolitic and Cyrillic sources (of course, these spellouts are aberrant).

The prohibition against word-initial and postvocalic ъ and ы in morphophonological representations is ensured by alloformy. For example, the nonstandard root formative ъп has the alloform въп that occurs word-initially, cf. въз.(ъп).и.т.и, but (въп).и.т.и. The suffix of ш-participles ьш/ъш has the alloform въш that occurs postvocalically.

The fact that the opposition of vowels by advancement is in some way correlated with the oppositions between sets of consonants that are or are not compatible with front vowels, is largely behind the morphophonological mechanism of the twofold rule. See also § 866, *Trubeckoj's nonstandard phonology*.

# §871. On the combinations бн, пн, and мн

These clusters are absolutely allowed, cf. об.(нов).и.т.и ⇒ обновити, об.(сльп).н.ѫ.т.и⇒осльпнѫти. However, in some cases the labial is removed, i.e. п.н, б.н, and м.н combinations are mapped in canonical spellouts by the c1c2⇒c2 rule, cf. (тим).н.а⇒тина (see § 75).

For нѫ verbs, the picture is rather chaotic, and the data are sparse. Some нѫ verbs do not have stems with the suffix н in the sources. For such verbs, all the attested forms are PRAE of class 3, or forms of class 4 (Aor, л-Part, ш-Part, н-Part). For example, there is the hapax gloss охръмѫ (root aorist Ps Sin 17, 46). All the data are shown in Table 871: these are all class 5 verbs with final labials.

The c1c2⇒c2 column shows the infinitive without the labial if such a stem is attested. The "Stable" column shows the infinitive with a labial followed by н, if such a stem is attested. The "Unknown" column shows a hypothetical infinitive, if there is no stem with the н suffix in the sources at all.

Because the basic phonological form of the verb is uncertain, its spellout in the dictionary is also difficult to choose. In the paradigmatic dictionary we choose the same forms as in the following table.


Table 871. Spellouts of нѫ verb stems in sources and dictionaries

Illustrations

КАНЖТИ - ЗЕМЛЪ ПОТОБАСЕ СВА ИВО НЕСА КАНЖШИА: ОТЪ ЛИЦА БА СІНАННА PS SIN 67, 9.

ОСЛЬПНЯТИ - ТО КТО СЛЕПЪ НЪ ОТРОЦИ МОИ И ГЛОУСИ' НЪ ВЛАДЖШТИ ИМИ' И ОСЛЬПНЯША РОБИ БОЖИЇ SUPR 323, 4-6.

ОУСЪНЖТИ - Призърг оуслъши мы гі бже мог просветт очи мог еда когда ОУСЪНЖ ВО СЪМОЪТІ PS SIN 12, 4; ВЪ МИОЪ ВЪКОИТЪ ОУСЪНЖ И ПОЧИЯ ВЪКЪ ТЪКЪ ТЪК ГІ ЄДІНОГО НА ОУПЬВАНЬЕ ВЬСЕЛІЛЪ МНА ЕСІ PS SIN 4, 9; СЖШТОУ БО ВЕЛИКОУОУМОУ ОТъцоу въ хъвзянъ: и молитвя твораштоу къ когу, въздовма са и оусънж MANTI SUPR 275, 20-22.

ИСТОПНЖТИ - И ОВО ИХЪ ИЗБИША ОВО ЖЕ ПРОГНАША: МНОЖАНШИИ ЖЕ НУТЬ ВЪ ДОУНАВЪ БъЖАШТЕ ИСТОПНЯША SUPR 197, 8-11.

разгънжти, съгънжти - сf. в въздаша емоу къмигън исания продока и разгънжвъ кънигъ! обръте мъсто идеже въ написано: дуъ гиъ на мыть. егоже ради помаза ма: Благовъститъ ништиниъ посъла ма: ссцълитъ съкроушенъвна СОДЦЕМЬ: | ... | съгънжвъ кънигъ! ВЪДАВЪ СЛОУБЪ СЕДЕ Lk 4, 17-20 МАR (likewise in AS and SAV).

ГЪИЯЖТИ — ДЪЛАНТЕ НЕ ЕРАШНО ГЪВВИЖШТЕЄ НЪ ЕФАШИО ПРЪБЪВВАЊИЦИЕС ВЪ ЖІВОТЪ ВЪЧНЕАМЪ Jn 6, 27 AS (cf. ГЪІБЛЯШТЕЄ in ZOGR and MAR).

погъюняти - вко не възможъю естъ пркоу погъюняти кромъ нерсма Lk 13, 33 MAR; мыните ли фко галильане сис гръшьнъыше паче въсъхъ галилъванъ БЪЩА" БКО ТАКО ПОСТРАДАША: НИ' ГЛУЖ ВАМЪ НЪ АШТЕ НЕ ПОКАЕТЕ СА: С ВСН ТАКОЖДЕ ПОГЪІБНЕТЕ Lk 13, 2-3 ZOGR.

ПООЗАВНЯТИ — WT'L СМОКОВЬНИЦА ЖЕ НАОУЧИТЕ СА ПОНТ'ВЧИ' ЕГДА ОГЖЕ В'Б ЕЊА ЕЖДЕТЪ МЛАДА И ЛИСТВИЕ ПООЗАВНЕТЪ: ВЪСТЕ ТВКО БЛИЗЪ ЕСТЪ ЖАТВА МЕ 24, 32 ZOGR (likewise in MAR, AS, SAV); егда прознавия гръшьніці "БКО Тр'ЕВА" " ВЪЗНІКЖ ВЬСІ ТВОРИШТИ БЕЗАКОННІЕ PS SIN 91, 8.

#### §872. On combinations with 3-final prefixes

All combinations of the form (c, 3) + (4, %, шт, жд, ц, s) are phonologically prohibited. General cluster rewrite rules require c1c2 > c0 for 3x (i.e. 3ж = жд), and c1c2 = c2 in other cases (deletion of the first consonant), see § 74, 75. This is seen in examples like из.жи.т.и > иждити, из.ц ѣ л.а.т.н > ицъ лати, из.чист.и.т.и > ичистити, раз.шир.и.т.и > раширити. However, some prefixal lexemes are represented by aberrant forms, where the prohibited combination can either be preserved (ct. вєцьньнъ alongside canonical вєцьньнъ for EE3. LEN.bH.B, or EECHETHE alongside canonical EENECTHIE), or the cluster is repaired by c1c2 → c0, or there may be some other segmental aberrations, such as epenthesis of 3. Some lexemes show only aberrant forms, which creates some lexicographic difficulties.

Table 872 on p. 442 compiles all the data.'

<sup>7</sup> Data on the combination 3.ж is not included, since there are no attested aberrations from the main rule c1c2 > c0.


Table 872. Forms in the dictionary and sources

1) In Večerka's dictionary, the headword бещиньница is for the hapax gloss бещініцѧ Cloz 2a, 24–25.

2) 1× aberrant spellout in Supr: бесчьстьвьнѫѭ 512, 7.

3) бесꙿчьстиꙗ in Supr 74, 25–26.

4) Also aberrant spellouts їстѣлениѣ Ps Sin 37, 4 and 37, 8, and истѣлитъ Supr 115, 6.

5) Večerka mistakenly notes only ищез- and изꙿчез-; but there is also ичезѫ Ps Sin 101, 4. For изчезнѫти Supr 487, 3 shows a spellout with an apostrophe (*payerok*): изꙿчезе 487, 3.

The c1c2⇒c2 column shows forms with the prefix-final з deleted, if such a form is attested (the mapping according to c1c2⇒c2 follows the general rule). The column c1c2⇒c0 shows mappings by the c1c2⇒c0 shema if such a spellout is attested. The c1c2 column shows attested forms preserving the prefix-final consonant (this mapping does not follow the general rule).

Uncertainty about the phonological shape of a word causes difficulties in selecting the starting form of a lexeme for inclusion in the dictionary. Here, the paradigmatic dictionary shows spellouts as seen in the following table, viz. the dictionary shows the form from the leftmost column—so to speak, the best of the attested forms.

A circle marks lexemes whose dictionary form is aberrant.

### §873. On the types of segmental pairings8

### 1. Quasisegments and segmental pairings

Some sequences of segments are functionally equivalent to single segments. Such are, first, combinations of vowels with sonants in the sonant series of fundamental alternations (thus, ъв‖ыв behaves as ъ‖ы, ор‖ра as о‖а, etc.). Second, these are clusters in consonant alternations (thus, ск‖щ behaves as к‖ч, м‖мл҄ as з‖ж, etc.). Such sequences are called *quasisegments*. The same string of segments can be a free segment sequence in some occurrences and a quasisegment in others. For example, ск in исказити is a free sequence, while in искати it is a quasisegment. ра in рана is a free sequence, but a quasisegment in брати. To determine the nature of each string it is necessary to turn to morphophonological analysis and the alternations that are found in the relevant formative.

Each pair (*x*, *y*), where *x*, *y* are segments or quasisegments, is called a *segmental pairing*. For example, (ѧ, ѣ), (х, к), (в, вл҄), (ра, рь) are segmental pairings, while the pair ре and рь is not a segmental pairings, because there is no quasisegment ре.

### 2. Phonological pairings

Some but not all segmental pairings are *phonological pairings*. For example, the vowels ъ‖ы, о‖а, е‖ѣ, and ь‖и form phonological pairings (ph pairings), while the vowels а and е, or а and у do not. Phonological pairings are related to the accepted phonological classification of segments, but is not formally determined by it. The membership of the set of phonological pairings is given by a list. The simplest type of phonological pairings is one-dimensional, whose members are in opposition by a single phonological feature. The notion of phonological pairing is similar to Trubeckoj's notion of correlation, but is not identical to it. Ph pairing is a symmetrical relation: if *a*‖*b*, then *b*‖*a*. The most important phonological pairings are noted in the statements of inventories (§ 25, 31).

<sup>8</sup> This discussion can be seen as a development of Trubeckoj's discussion of oppositions; see Trubeckoj 1939, Ch. III.

### 3. Morphophonological pairings

Some but not all segmental pairings are called *morphophonological pairings* (mph pairings). Namely, *a* and *b* form a morphophonological pairing if *a* and *b* participate in some alternation and belong to one series of that alternation, and *a* and *b* are segments or quasisegments. For example, к‖г form a ph pairing, but not a mph pairing; п‖пл҄ forms a mph pairing, but not a ph pairing. Likewise, the pairings ц‖ч and ѕ‖ж do not form phonological pairings, while е‖ѣ forms both a ph and a mph pairing.

The mph pairing relation is symmetrical: if *a*‖*b*, then *b*‖*a*. It is also transitive; if *a*‖*b* and *b*‖*c*, then *a*‖*c*. All mph pairings are explicitly given by alternations. Morphophonological pairings may also called *pairings by a given alternation*.

### 4. Adjacency

Some morphophonological pairings are *adjacent* and others are not. A given mph pairing *a*‖*b* is adjacent by a given alternation *A* if *a* and *b* form adjacent grades of the alternation, *a* being the leading and *b* being the following grade. For each alternation, the adjacency relation is given explicitely by list, see Ch. 5. For example, the mph pairings о‖а and ь‖е are adjacent by the fundamental vowel alternations (о>а and ь>е), while the mph pairings о‖е and а‖ь are not adjacent.

The adjacency relation is antisymmetrical, i.e. if *a*>*b*, then it is not the case that *b*>*a*. All adjacent pairings are explicitly given by alternations, because the statement of alternations includes information on grade adjacency.

### 5. A note on notation

The parenthesized notation (*x*, *y*) is used to denote any segmental pairing. The sign ‖ is used for any phonological or morphophonological pairing, and also for the formatives containing it. So, we write к‖ч, о‖е, and also рок‖роч, зла‖зел. It is possible to show pairings transitively: ь‖е‖о. The sign ~ is used whenever it is necessary to indicate that the pairing is non-adjacent, or the formatives are not segmentally adjacent, e.g. о~е, зла~зел. The sign > is used whenever it is necessary to indicate adjacency: к>ч, ь>е.

### 6. Alloformy and pairings

Let *G* and *H* be some formatives of the same position class, and (*x*, *y*) a segmental pairing.

The formatives *G* and *H* are *comparable by the segmental pairing* (*x*, *y*) *in segmental position t* if *G* and *H* can be written as *PxQ* and *PyQ*, where *P* and *Q* are some (possibly empty) segment strings, and *x* and *y* replace the segmental position *t* in the formatives *G* and *H*, respectively.

For example, the formatives ног and рог are comparable by the pairing (н, р) in the initial C segmental position. The formatives сѧд and сѣд are comparable by the segmental pairing (ѧ, ѣ) in the medial V segmental position. The formatives ик and иц are comparable by the segmental pairing (к, ц) in the final C segmental position.

If two formatives *G* and *H* are comparable by a morphophonological segmental pairing *x*‖*y*, we say that *G* and *H* are *comparable by the alternation A* if the segmental pairing *x*‖*y* is a morphophonological pairing by the alternation *A*, and the segmental position that is tied to alternation *A* is the same as the segmental position by which the formatives are compared.

If two formatives *G* and *H* are comparable by the morphophonological alternation *x*‖*y*, we say that they are *adjacent by alternation A* (or simply *segmentally adjacent*) if the segmental pairing *x*‖*y* is a morphophonological pairing that is adjacent by the alternation *A*.

The alloformy relation, which determines the boundaries of formative families where all members are each others' alloforms, is not explicitly given in the present grammar (see § 640 for prefixes, § 642–840 and RD for roots, and § 841–861 for suffixes). Below we consider several properties of the alloformy relation. These properties can be understood in two ways: first, as criteria that help the grammar writer, and second, as properties of the proposed alloformy relation.

### The comparability-by-alternation condition

This condition favors setting up alloformy of formatives that are comparable by some alternation. However, in some cases alloformy can be established for formatives that are segmentally comparable in some segmental position by a segmental pairing that is not morphophonological. For example, this is the case with the anomalous root formatives сѧд and сѣд, or лѧг and лег (these roots show intraparadigmatic alloformy).

### The series isolation condition

This condition does not allow a family of formatives to contain formatives comparable by the segmental pairing (*x*, *y*) if *x* and *y* participate in alternation *A* but represent different series of that alternation. The only violation of this condition is found in root ‹496›, with the alloforms лих‖лиш and лѣк, which represent different series of the velar palatalization alternations (the series х and к). It is an anomalous root.

The condition on agreement between segmental adjacency and segmental position

This condition does not allow one family of formatives to contain formatives *PxQ* and *PyQ* with the pairing *x*‖*y* by alternation *A* if the segmental position occupied by segments *x* and *y* is not the segmental position tied to the alternation *A*. According to this condition, a single etymological family, containing the root formatives град and жрьд, is separated in this grammar into two different roots (‹192› and ‹293›, respectively). Likewise for roots каз ‹362› and (чез, чѣз) ‹1065›, ход ‹1021› and (шьд, шид) ‹1099›, скѫд ‹840› and (щед, щѧд) ‹1108›, etc. The only violation of this condition is found in root ‹153› гънати, женѫ (see § 818).

#### The segmental adjacency condition

This condition prefers alloformy between adjacent formatives, but does not prohibit a family from containing formatives that are not segmentally adjacent. This criterion applies also in determining the series of the fundamental alternations for specific formatives: according to this criterion, for example, the root

‹451› крьс is treated as representing the series H(*r*), contrary to etymology that points to H(*j*) (more details in § 678, 680). However, in several cases a single family contains alloforms that are not segmentally adjacent.

Note that the replacements that apply in paradigmatics are linked with the adjacency relation. Accordingly, intraparadigmatic alloformy that is generated by segmental replacement rules can only be segmentally adjacent alloformy. Because synchronic grammar seeks to reduce segmental alloformy that is generated by segmental pairings not found in paradigmatics, it seeks to reduce segmental alloformy generated by non-segmentally adjacent pairings.

§874. On verbs with unstable root vocalism

1. The behavior of verbs that show unstable root vocalism in the paradigm at first appears so peculiar that the distribution of vocalic realizations seems impossible to state except by a list. In reality, there are several principles at work here, which determine a fairly rigid order.

2. Whether a verb synchronically shows unstable root vocalism in its paradigm is a lexical property. Even if it is known that the root contains unstable vocalism, it is impossible to predict from some properties of this root whether a verb containing it has unstable vocalism. For example, the root рек ‹766› has unstable vocalism, but none of the verbs with this root do so in their paradigms (cf. °рѣкати 7, °рещи 4, and °рицати 3).9

Verbs with unstable root vocalism in their paradigms are possible only in the плакати 3 and нести 4 classes.

3. The verbal paradigm does not contain more than two grades of fundamental alternations, and these grades must be adjacent. For example, we have:


4. The grades represented in the verbal paradigm form either horizontal or vertical pairings, according to the rule of horizontal and vertical pairings. This rule states that verbs with pure vocalism must have vertical pairings, while verbs with sonant vocalism must have horizontal ones, as shown in Table 874.1.

<sup>9</sup> The alternation in the vocalism observed in the Imv forms of the рещи group of verbs (Imv рьцѣте), and in old aorists of the type нѣсъ and рѣхъ, is outside of the systematic principles stated above. In this grammar we do not consider the pairings between these forms as resulting from the unstable root vocalism paradigmatic effect.



The following verbs violate this principle: 1) the verb стрѣщи, стригѫ from the group чисти 4c\*⤹ from root ‹894› with anomalous vocalism: pure vocalism shows the horizontal pairing и‖ѣ in the paradigm; 2) verbs from the group крыти 4h\*⩨⤸ show the ъ‖ы grade pairing by the H(*u*) series. Here, «Inf» shows the heavy grade, and «Prae» shows the light grade, but by vertical and not by horizontal pairing, as expected for sonant vocalism. Note that no series from any horizontal pairing contains the grade ы.

5. Verbs with unstable vocalism show the opposition between two workstems: the workstem of the PRAE «Prae» system, and the workstem of the INF-AOR «Inf» system. The former is represented in all subparadigms of the PRAE system, and the latter in the forms Inf and Sup. For other subparadigms, the distribution of workstems is determined by allotment rules that may be distorted by the expansion of alien workstem effect. This effect does not destroy the unity of the workstem in the PRAE system, but can affect subparadigms of other systems, interfering with their unity.

It is sufficient to compare the workstems of «Prae» and «Inf» in considering the vocalic realizations. Here, the cross-class rule is in effect (Table 874.2).


Table 874.2 The distribution of heavy and light grades

Thus, knowing the vocalic realization in «Inf» and the alternation series found in a given root, one may predict the vocalic realization of the PRAE forms («Prae»). For other subparadigms, the expansion of alien workstem effect applies.

# §875. On Jakobson's law

1. In this grammar the principle that prohibits clusters and hiatuses at boundaries is called Jakobson's law. In other words, this law requires CVC agreement of the end of a formative with the beginning of the next formative: either C-final + V-initial, or V-final + C-initial. This law *pro*scribes rather than *pre*scribes, just as the law of the velars. It does not contain any instructions on what repairs might be undertaken in case an underlying form contains any precursor to the prohibited combination.

2. CVC agreement is widely represented in languages as a factor for selecting alloforms that differ by their initial C vs. V, or final C vs. V (CVC-ambivalent alloforms). These rules call for the selection of a C-final alloform before a V-initial one, and a V-final one before a C-initial one. These rules, for example, govern the selection of aorist terminals охъ/хъ, оховѣ/ховѣ, etc. However, CVC-ambivalence is not a property of all formatives, while all formative boundaries are subject to Jakobson's law.

CVC-ambivalence is simply one of the repair strategies for violations of Jakobson's law. In other cases, the threat of hiatus is avoided by consonant epenthesis (e.g. epenthetic *i̯*, cf. зна+еши⇒знаѥши), while the threat of a cluster is avoided by cluster simplification (cf. греб+ти⇒грети).

3. Jakobson's law is a morphophonological principle that organizes morphophonology as well as morphology. In morphology, it is manifested in the opposition of truncated (C-final) and expanded (V-final) basic stems in the conjugation. This opposition materially expresses the opposition between systems: the PRAE system uses C-final stems and V-initial terminals; the INF-AOR system uses V-final stems and C-initial terminals; and the IMF system contains initially ambivalent terminals. This is the set of facts considered by Jakobson in his famous 1948 article.

§876. On the morphological composition of stems and wordforms

1. The morphological composition describes the organization of strings of formatives—wordforms and stems. The simplest representation of morphological composition is the pRs schema; however, the pRs schema does not show constituent composition. Although this book does not attempt a full description of constituency that reflects the entire derivational history of a stem or a wordform, in some cases it is necessary to use some information on constituent composition. Accordingly, we consider some special cases of partial constituent compositions.

2. In the paradigmatics, the most important information on constituency is given by the parse of a wordform into two components: the stem and the terminal. Two representations reflect this partition: a deeper one, the *morphological skeleton*, and one closer to the surface, the *inflectional spellout*.


The stem that appears in the morphological skeleton is called the *workstem* of the wordform.

3. In stem formation, the most important information on constituent structure involves the selection of the *basic component*, i.e. the part of the stem that precedes a given suffix. For example, if the wordform is въ.(сел҄).ен.ьск.ъ, then въ.(сел҄).ен is the basic component for the suffix ьск, while въ.(сел҄) is the basic component for the suffix ен.

4. For verbs, paradigmatic mechanisms are tightly intertwined with the mechanisms of stem formation, which requires considering more than two-component constituent structures.

A *lexical component* is a basic component that precedes a theme, or a class suffix, where one is present (in classes 5 or 6). The lexical component is found inside workstems.


Examples such as these show that the *expanded basic stem* is just the basic component that precedes the infinitive suffix т, while the *truncated basic stem* is the basic component that precedes the theme.

5. Accounting for the morphological structure of nominal forms of the verb (infinitive, supine, and participles) requires access both to workstems of the corresponding nominal wordforms, i.e. fragments that precede a terminal, as well as to basic components by the suffixes of the nominal forms, i.e. fragments that precede suffixes of the participle and т of the infinitive and supine. Table 876 (p. 450) shows workstems as parts of morphological skeleta, and basic components for some forms.

It is easy to see that the basic component of nominal wordforms coincides with the workstem of finite forms: при.нес is the basic component in [(при.нес)+т]+[и], and the workstem in the morphological skeleton [при.нес]+[ѫ].


Table 876. Morphological skeleta and basic components

# **Excursus on the grammatical regularity of canonical wordforms**

### §877. General

Canonical wordforms exist on a scale of grammatical regularity. Grammatical regularity takes into account paradigmatic, morphophonological, and phonological properties of wordforms. Regularity on each of these axes is orthogonal to regularity on the others.

In the general case, a paradigmatic call corresponds to only one canonical wordform. However, this is not true in cases of so-called doublets, which are equally regular forms, cf. GLSg (слово) — словесе//словеси. Secondly, in some cases a paradigmatic call may correspond to two canonical wordforms, of which one is primary, and another is secondary. In this case, the primary forms have a higher measure of regularity, cf. for 1SgAor (ити), the primary form идохъ is more regular than the secondary form идъ.

The opposition between canonical and alternative spellouts is defined for forms from sources. Alternative spellouts are possible for any form, regardless of its degree of grammatical regularity (on the classification of alternative spellouts see § 886–897, *Excursus on aberrations*).

# Paradigmatic regularity and morphological anomalies

§878. A measure of grammatical regularity of paradigmatic classes

Both verbs and nominals have paradigmatic classes where the measure of regularity is defined. The most regular are standard classes—regular verbs of the main paradigmatic classes, and standard inflectional classes of nominals. Less

regular are marginal subclasses of the standard classes—irregular verbs and deformations of standard inflectional classes of nominals. Finally, the least regular class contains unique lexemes. Here is the distribution of these classes.


Paradigmatic behavior of irregular verbs is described in terms of a given set of paradigmatic effects, which distort the paradigmatic standard of one of the main classes. Paradigmatic behavior of unique verbs cannot be represented as such a systematic departure, even though among unique verbs certain paradigmatic effects can still be observed.

Nominal lexemes in marginal paradigmatic classes show certain deformations of the paradigmatic standard of one of the standard declensional classes, which distinguish their paradigmatic behavior both from the main standard and from the behavior of unique nominal lexemes. Deformations of nominal paradigms are described in terms of a given set of paradigmatic effects. Paradigmatic behavior of unique nominals is not representable in terms of systematic departures from the standard of a main declension type, even though among unique nominals certain paradigmatic effects can still be observed.

In this way we can define a hierarchy of grammatical regularity of a lexeme as a whole. On this axis, unique lexemes are the least regular.

# §879. Paradigmatic regularity of a single wordform

A measure of regularity of a given wordform can be established along two independent parameters. One parameter determines the grammatical status of the wordform in the paradigm of its lexeme; here we can distinguish *primary* and *secondary* wordforms. Another parameter concerns the morphological composition of this wordform. Here, we distinguish morphologically normal and *morphologically anomalous* wordforms, or *morphological anomalies*.

# §880. Primary and secondary wordforms

The notion of secondary wordforms covers those which are in some sense supernumerary in the paradigms of their lexemes. Some are extra wordforms in the sense that they have no special place in the free paradigm. A limited number of lexemes have free paradigms that are extended by such additional wordforms. Others are extra in the sense that they occupy a paradigmatic cell that is already occupied by an eponymous primary wordform; these are "substitute" wordforms. For example, vocative forms for nominals, or the conditional of the verb быти do not form part of a standard free paradigm of their grammatical class, leaving secondary additional forms. The secondary form вѣдѣ, 1SgPrae (вѣдѣти) coexists with the eponymous primary form вѣмь, being its doublet. The same is true of so-called nonstandard aorists. Table 880 lists all groups of secondary forms.


Table 880. The inventory of secondary forms

### §881. Morphologically anomalous forms

Forms that stand outside of the general rules of paradigmatic synthesis are considered morphologically anomalous. These forms are not built step by step. For this reason, generally speaking, for such forms there is no defined morphological structure, inflectional spellout, or even morphophonological representation. Anomalous forms may belong to paradigms of the main paradigmatic classes, as well as irregular verbs or deformed nominal paradigms, or to unique lexemes.

Let us list the groups of all anomalous forms.

Among *verb forms*, the following are morphologically anomalous:


Among *nominal forms*, the following are morphologically anomalous.

1) Syncopated forms of щ- and ш-participles (see more details in § 309). They have an arbitrary morphophonological representation that ends in a fragment that is neither a suffix nor a terminal, cf. the syncopated forms л҄юбл҄.ь|, нес.ы| and the expanded forms л҄юбл҄.ьш=и, нес.ѫщ=и. The symbol | shows that the form lacks a terminal.

2) The following forms of unique lexemes: господѣ(GSg) and господю(DSg) of the lexeme господь 0/m; the forms очима and ушима of the unique lexemes око 0/n and ухо 0/n; the form къто of the lexeme къто 0/p; the forms чьто, чесо, чьсо, чесого, чесомь, чесому, чьсому of the lexeme чьто 0/p, and all the forms of the lexemes азъ 0/s, ты 0/s and сѧ 0/s. For all these wordforms no parse into formatives is established; otherwise only к=ъ.то and ч=ь.то.10

# Segmental regularity and segmental anomalies

§882. Phonological and morphophonological regularity

Violations of phoneme syntagmatics can be *phonological* or *segmental anomalies*. *Morphophonological anomalies* are relatively rare cases of violations of the distribution of ambivalent formatives (CVC agreement violations), and of the distribution rules of twofold formatives (violations of twofold rule). Likewise, such are forms that contain anomalous and pronominal roots, opaque stems, forms with anomalous pRs structures (e.g. isolated reduplications; see § 41), and syncopated forms щ- and ш-participles. Also, morphophonological anomalies include forms that contain an anomalous version of some formative (e.g. the prefix in възъꙗрити).11

### §883. Violations of the CVC agreement rule

The only CVC-ambivalent prefixes are въ/вън and съ/сън. The prefixes от and отъ are not morphophonological variants. On violations of the CVC agreement rule, see *Prefixes*, § 641.3°.

With labile roots, departures from CVC agreement are found in verbs with the C-final stem arrest paradigmatic effect (in the irregular verb groups пл҄ьвати 3h\*⩨ and крыти 4h\*⩨⤸). In class 4 verbs, the destribution of C- and V-final versions of the truncated workstem before initially ambivalent suffixes and terminals is given by the distribution rules for workstems.

The following suffixes are CVC-ambivalent. Initially ambivalent: {26} н/ен, {35} с/ес, {44} х/ъх, {48} (ъш/ьш)/въш. Finally ambivalent: {55} (ов/ев)/у, {56} ъв/ы, {57} ѧ/ен, {58} ѧ/ѧт. See details on departures from CVC agreement for the suffix {26} н/ен in *Suffixes*, § 853. The suffix {55} (ов/ев)/у as a class marker for class 6 миловати verbs shows the V-final variant before V-initial

11 Morphophonological anomalies are not systematically marked in this book.

<sup>10</sup> Other departures from the paradigmatic standards of the main inflectional classes in nominal declension are described as deformations of that standard (cf., for example, the class 2/p\* вьсь and сиць).

terminals and suffixes in violation of the CVC agreement rule due to the C-final stem arrest paradigmatic effect.

Among initially ambivalent terminals, the only departures from CVC agreement are found in nonstandard aorists (the old sigmatic 2 aorist, type рѣхъ).

### §884. Violations of the twofold rule

The following suffixes are twofold: {10} ој/еј, {16} ък/ьк, {20} ъл/ьл, {24} ом/ем, {29} ън/ьн, {32} ор/ер, {36} ост/ест, {38} ъств/ьств, {42} от/ет, {43} ът/ьт, {48} ъш/ьш, {55} ов/ев. Departures from the twofold rule found in the bechmark list lexemes are described for suffixes in terms of secondary twofold rule (see § 863), and for nominal terminals in terms of deformations of main declension types.

Note separately the forms сеица [с.еј.иц.а] and дьневьнъ [дьн.ев.ьн.ъ], which would be morphophonologically anomalous outside of the application of the secondary twofold rule.

Violations of the twofold rule are also possible among borrowed lexemes. For example, the lexeme архиереи has the form архиереомъ (емъ is expected by the twofold rule). See § 89.

### §885. Phonological anomalies

Phonological anomalies are forms that contain prohibited combinations of phonemes, and forms that violate the CVC norm of the word. Because C-final wordforms are not found in canonical OCS, the only possible violators are the following: 1) V-initial forms; 2) forms with a hiatus; 3) forms with prohibited clusters; 4) forms violating the law of the velars.

Group 1 (V-initial). The only vowel allowed word-initially is the phoneme и; other vowel phonemes are prohibited there. Thus, all attested vowel-initial forms are phonologically anomalous. Nonstandard formatives that violate the CVC norm of their positional class are carriers of those anomalies: V-initial prefixes and V-initial roots. These are, for example, обѣдъ, отъвести; овъ, азъ, агньць, окуса. Full lists are found in *Prefixes*, § 640; *Roots*, § 654 (V-initial roots), § 785 (pronominal roots), and § 806 (anomalous roots). See also § 867, *On word-initial vowels*.

Group 2 (Hiatus). All hiatuses except Vи are prohibited. Accordingly, attested forms containing hiatus are phonologically anomalous. Hiatuses are found in the following groups of forms. (1) In loans inside root formatives. In the benchmark corpus these are ‹6› (аеръ), ‹11› (архиереи), and ‹157› (геона). (2) At boundaries between V-final prefix and V-initial root. Cf. научити, паѫчина. (3) At boundaries between V-final root or suffix and V-initial terminal or suffix. Cf. архиереомъ, архиереовъ. Note that such forms are possible only among loans, and that in such cases their terminal or suffix shows the hard subtype, in violation of twofold rule. (4) In the personal forms of the imperfect, cf. трьпѣаше

[трьп.ѣ=аше], плакааше [плак.а=аше], несѣаше [нес=ѣ.аше]. Note that there are aberrant forms where the hiatus is eliminated, cf. паѭчина, трьпѣꙗше (the so-called iotated imperfect), трьпѣше (the so-called contracted imperfect).

Group 3 (Prohibited clusters). All attested forms with prohibited clusters contain it inside the root. Accordingly, syntagmatic anomalies are represented by roots. Prohibited clusters are found in two groups of roots: these are, first, roots that contain prohibited goups of the type C+liquid (л or р), as e.g. нравъ ‹615› or власъ ‹106›, and second, several isolated roots (cf. седм ‹814›). The roots of the first group are etymologically sonant. Full lists are found in Ch. 4, § 62, *Lists of syntagmatic anomalies*.

Group 4 (velar law violations). Such forms are possible only in loans, cf. китъ, архиереи, etc. See more details in § 869, *On the law of the velars*.

# **Excursus on aberrations**

§886. Alternative spellouts and their canonical analogs

Our investigation starts with two spellouts. The first, call it *actual* (α\*), is a concrete spellout in a particular source. The other, (α) is its canonical analog. These two forms are spellouts of particular wordforms, and the grammatical interpretation of the canonical wordform α is transferred to the actual form α\*. Segment-by-segment comparison of α and α\* can establish whether these strings are identical or not.12 In the first case we say that the string is a canonical spellout, or that the actual string α\* is an *alternative* spellout of the wordform α.

### §887. Aberrant forms and aberrant derivations

Among alternative spellouts there is a class of *aberrant* spellouts or forms. Aberrant forms correspond to *aberrant derivations*, which are modified canonical derivations. The modification results from the application of *aberrations*, or operations that act on particular expressions of canonical spellouts.

Let the spellout α\* have α as its canonical analog. Let χ(*Α*) be the paradigmatic call of the canonical form α, and Δ be its canonical paradigmatic derivation. In other words, the derivation Δ has χ(*Α*) as its input and α as its output, and is constructed according to the rules of paradigmatic synthesis. The paradigmatic call of the aberrant form α\* is identical to that of its canonical analog. The derivation Δ\*, which takes as its input χ(*Α*) and produces α\* as its output, is the aberrant derivation of the form α\* if it can be represented as the result of applying aberrations (δ1, δ2, …, δk) to the canonical derivation ∆.

<sup>12</sup> If the source is Glagolitic, then α\* is a string of Glagolitic letters, while α is a string of Cyrillic letters. Because the Glagolitic ⇔ Cyrillic transliteration is fixed (see § 132), we can consider α\* to be a string of Cyrillic letters as well (so-called expanded Cyrillic).

# §888. Objects substituted by aberrations

A given aberration of the form Λ⇒Λ\* substitutes the expression Λ by Λ\* in the canonical derivation. Aberrations can substitute various objects: 1) segments (such aberrations are called atomic); 2) segmental rewrite rules, or segmental pairings in alternations (such aberrations are called modifying aberrations); finally, 3) workstem construction rules, or selection rules for terminals or suffixes of personal verb forms (such aberrations are called paradigmatic).13

# §889. Domain and modality of application of aberrations

In the general case, aberrations are free, i.e. their domain of application is not restricted. The existence of the aberration Λ⇒Λ\* assumes that every occurrence of the object Λ in any derivation Δ can be replaced by the object Λ\*. For example, the atomic aberration ѕ⇒з can be possible for any occurrence of the phoneme ѕ (e.g. ѕѣло and the aberrant зѣло, мъноѕи and the aberrant мънози, etc.).

Secondly, aberrations are in the general case *optional*, i.e. have the modality of permission rather than requirement. For example, in the same source we may find canonical forms alongside aberrant ones (e.g. in Mar, alongside canonical мъноѕи there are aberrant forms мноѕи, мънози, мнози, etc.).

In fact, however, some aberrations are restricted to particular sources. For example, the aberration ѕ⇒з is almost absent from As and Ps Sin. In other cases it turns out that in a particular source an aberration applies in all eligible cases. For example, in Sav, the aberration ѕ⇒з applies always (there are no spellouts with ѕ at all). Also, Sav has no canonical imperfects, only contracted ones. A separate case are so-called *lexical aberrations* (see below), whose domain of application is limited to particular lexemes.

# §890. A general overview of alternative spellouts

As indicated above, every alternative spellout α\* has a canonical analog α, and the grammatical interpretation of the canonical wordform α is transferred to the actual spellout α\*. A separate case are *morphologically strange spellouts*, whose canonical analog cannot be found. Such spellouts are not even called "alternative" (see below). All spellouts that are different from canonical ones are sorted as follows:

<sup>13</sup> Formally, atomic aberrations substitute trivial segmental rewrite rules that are part of the mph ⇒ ph/norm. For example, the aberration ѕ⇒з is the replacement of the trivial rule ѕ⇒ѕ by the alternative rule ѕ⇒з.


*Graphically alternative spellouts* differ from the rest in that they have the same phonological representation as their canonical analogs. All other alternative spellouts differ phonologically from canonical analogs. Graphically alternative spellouts are generated by alternative graphic rules, i.e. alternative rewrite rules from phonological to graphic spellouts, the so-called *basic writing systems*  of a given source.

*Erroneous spellouts* are not generated by any mechanism. They are considered technical errors unrelated to grammar.

*Morphologically strangeforms* are those that lack a canonical analog, and thus their grammatical properties can only be approximated. They are treated as random morphological errors. Bad morphology does not necessarily imply the action of some aberration; likewise, setting up a separate parent lexeme for such a form is also not always justified. Such are, for example, the forms несътръпѣтиꙗ Supr 377, 30–378, 1, визжьMar Jn 20, 27 (for виЖь), растрьзоваѥтъSupr 350, 28 (for which Večerka supplies the entry растрьѕовати), съгъбалъ Euch 35b, 2 (for which Večerka supplies the entry съгъбати), обрѣмьꙗѧ Supr 280, 1 (for this hapax gloss Večerka supplies an entry with two starting forms обрѣмлꙗти and обрѣменꙗти). Such morphologically corrupt or strange forms are sporadically noted in this book in comments under specific lexemes. Formally, being the remainder, they should be classified as erroneous forms, but they differ grammatically from ordinary mistakes, for example in that they have no canonical analog. Cf. the mistakes рарабъ Ps Sin 78, 2 for рабъ, брѣгру Mt 8, 32 Sav for брѣгу, икованыѩ Ps Sin 89, 12 for окованыѩ.

In some cases the same aberrant form may be the result of different aberrant derivations, and the difference may be that in one derivation, the form is interpreted as segmentally aberrant, while in another as paradigmatically aberrant. For example, the form пѫтемь (for canonical пѫтьмь) can be interpreted as paradigmatically aberrant (intrusion of the nonstandard terminal емь), or as segmentally aberrant (strengthening of ь).

### §891. Types of aberrations and aberrant forms

All aberrations and aberrant forms fall into three types: segmental, paradigmatic, and lexical.


Segmental aberrations apply to the paradigmatic derivation that generates the canonical analog of the corresponding aberrant form, replacing particular segmental rewrite rules in that derivation. The aberration consists of replacing some general segmental rewrite rule that applies in the canonical form by an alternative analog. Paradigmatic aberrations apply to the paradigmatic derivation that generates the canonical analog of the corresponding aberrant form, replacing either a workstem formation rule, or a terminal selection rule.14

In the general case, a segmentally aberrant form and its canonical analog have identical morphological skeleta, and for some types of segmental aberrations, identical morphophonological spellouts as well. On the other hand, paradigmatically aberrant forms and their canonical analogs have different morphological skeleta.15 Here are some examples.


Examples 1-2 illustrate atomic segmental aberrations; examples 3-6 are examples of modifying aberrations. In 3-4, canonical cluster rewrite rules from the mph > ph/normsystem are replaced with alternative ones.16 In 5-6, canon-

<sup>14</sup> In sources that show an aberration of the type x=x\*, one can always expect hypercorrection, viz. an aberration of the form x\* ⇒ x. For example, in AS Lk 8, 29, we find the aberrant spellout with ь, желѣзьнъними for canonical жєлѣзнъюми.

<sup>15</sup> In the general case, an aberrant derivation can result from the application of several aberrations. An aberration δ; can apply to an expression that has undergone another aberration δ;, which can be distinct from δ;, or identical to it.

<sup>16</sup> In examples 3–4, the starting forms of the lexemes are aberrant. In these forms (not only starting forms, but obligue ones as well), the aberrant structure is the workstem. In example 3, it is the combination of the basic component with the suffix (πο.ΜοΓ + Τ); in example 4, it is

ical alternation pairings are replaced by alternative ones, for boundary adjustment rules in 5 and workstem formation in 6.

### §892. Lexical aberrations

In segmental and paradigmatic aberrations, the aberrant form and its canonical analog belong to the same lexeme whose name is its starting form; call it *A*. Both the canonical derivation α and the aberrant derivation α\* start from this starting form. Accordingly, the canonical and aberrant derivations have the same input, viz. the expression χ(*A*), which is the paradigmatic call of the wordform α of the lexeme *A*. In lexical aberrations, the input to the derivation differs: the canonical derivation of the canonical form α starts from some input *A*, while the actual form α\* starts from a different input, call it *A*\*. For example, this is the case in рибѣ (α\*) for риба (*Α*\*), whose canonical analogs are рыбѣ (α) for рыба (*Α*); обѣцалъ (α\*) for обѣцати (*Α*\*), whose canonical analogs are обѣщалъ (α) for обѣщати (*Α*), земьѭ (α\*) for земьꙗ (*Α*\*), whose canonical analogs are земл҄ѭ (α) for земл҄ꙗ (*Α*).

Because in this grammar inputs are not constructed by rule, but are retrieved from the dictionary of lexemes, the notion of aberration is not directly applicable to them. Within the limits of this notion, we must either recognize spellouts as erroneous, or introduce doublet lexemes into the dictionary (e.g. обѣщати// обѣцати, or земл҄ꙗ//земьꙗ), or, as is done here, establish a separate class of lexemic or lexical aberrations. Lexemes are the objects substituted by these aberrations. One lexeme together with its whole set of wordforms is considered canonical, and the other aberrant. For example, we have the canonical lexeme рыба (рыба, рыбѫ, рыбы, …, рыбами), and the aberrant lexeme риба (риба, рибѫ, рибы, …, рибами). Of course, because the corpus of texts is limited, not all wordforms are attested in both canonical and aberrant lexemes. It is also understood that aberrant lexemes are set up only if they are attested by at least one gloss, while canonical ones may be established also when they are only attested by aberrant forms. For example, we have 2× attestations of the aberrant lexeme риба and 84× attestations of the canonical рыба.

the combination of a prefixless stem with a prefix (из+чит). Because this book does not deal with word formation, formally the notion of aberration is not applicable to these examples. What is aberrant in the examples considered here are boundary adjustment rules that must apply to the morphophonological representations of workstems retrieved from the dictionary (PD). Thus, the aberrations for oblique forms can be derived without any formal difficulties, which arise only for starting forms. Note that there is a similar problem when a formative that is part of a workstem is aberrant. Here we introduce the notion of lexical aberration (see below in § 892). However, in this particular case, that notion does not preclude a certain formal contradiction.

# §893. Localizing lexical aberrations


Comparing the starting forms of canonical and aberrant lexemes localizes the aberrancy in one of the formatives. Below are some examples.

Clearly, in some cases the aberrant lexeme shows an alternative alloform that is present in the assortment of the corresponding formative. For example, the aberrant lexeme отъврѣсти (cf. canonical отъврѣщи). In other cases, the aberrant lexeme contains an exclusive aberrant alloform that is not represented in the canonical assortment. For example, this is the case with the aberrant lexemes бѣзъврѣменьнъ and розбити. In a significant number of cases exclusive aberrant alloforms can be represented as members of an alternative assortment of the formative, generated both by canonical as well as alternative alternations.

For example:


In these examples, alternative assortments are generated by the alternative pairings by the substitutive softening alternation (see § 117).

However, in other cases, e.g. рыб~риб, exclusive alloforms cannot be generated by segmental alloformy either in its canonical or in its alternative variant. In this class of exclusive alloforms that cannot be represented by segmental alloformy are, among others, all those where the initial C varies. Cf. туЖ (canonical) and стуЖ and щуЖ ‹988›, щит (canonical) and шчит ‹1110›.

# §894. A note on forms aberrant by the substitutive softening alternation

Aberrations that show alternative pairings by the substitutive softening alternations are unevenly attested in sources. Namely, alternative pairings for dentals are found not in all sources (attested in Kiev, Ps Sin, Mar, Supr, and Cloz). Kiev never shows canonical pairings, while the others show occasional alternative pairings alongside the majority that are canonical. Alternative pairings for labials are found in all sources except Kiev. They are found especially frequently in Supr.

The possible types of aberrations are shown in Table 894.17


Table 894. Aberration types

\* See below in § 895.

Lexical aberrations of types 4, 5, and 6 appear systematically in particular sources, and as such are not always marked in the lists of commented lexemes.

<sup>17</sup> Note that the classic interpretation of spellouts aberrant by the alternative pairings in the substitutive softening alternation for labials simply treats them as lacking epenthetic *l*, i.e. as an atomic aberration of the form *l'*⇒0, cf. пристѫпл҄ь⇒пристѫпь (Mt 8, 19 Zogr), погубл҄ѥниѥ ⇒ погубьѥниѥ (Supr 215, 27–28), капл҄ю⇒капью (Supr 499, 29), земл҄и⇒земи (Ps Sin 54, 21), etc. (cf. Vaillant, § 39). However, Vaillant himself notes, regarding this phenomenon, that «il ne paraît pas d'origine proprement phonétique: les groupes du type пл҄ sont conservés régulièrement à l'initiale, où l'on a toujours пл҄ьвати 'cracher', бл҄юсти 'garder', etc., et la chute de л épenthétique n'intéresse que les désinences flexionnelles et les formations dérivées. Il s'agit d'une modification de l'alternance du type п/пл҄ simplifiée en п/п'».

#### Illustrations

1 - не оставия вась сиръ: придж к вамь In 14, 18 as 30a (сf. не оставлья BACK CHOL TOIAN KE BAN'S In 14, 18 AS 93d); JEANNE CAABA YOUTH BHATTH' И ЦЕСАРЕСТВИЮ НЕБЕСЬНОЮ: ЗЪЛО ЛЮБЫЖ ВЫШЬНЯЖ ЧЕСТЪ: МЖКЪІ СА БОНЖ FEWHECKLIA SUPR 88, 3-6.

2 - въсжда твоего гі насъицени просимъ тъм: отъ въсъхъ противнациуъ сна намъ съпаси нъи кіеv ба, 22-6b, 2; марта же мльвъше о мнозъ слоужъвъ Lk 10, 40 SAV (cf. in the same verse: и мар'та млъватьаше: о мнозъ слоужьев ZOGR); гр ОТРОКЪ МИ ЛЕЖИТЪ ВЪ Храминъ: ОСЛАЕЄНЪ ЖИЛАМИ" ЛЮТЪ ГАКО МЖЧА СА МЕ 8, 6 SAV; ТАКО ПОНЕМЛГАШЕ СА И ЛЮБЫРШЕ: ГАКОЖЕ ОЕШТИ ЛИЦЕМЪ КЪ ЛИЦОУ: И ОУСТЪВ КЪ ОУСТОМЪ ГЛАГОЛАШЕ ГАКОЖЕ ДРОУГЪ КЪ СВОЕМОУ ДООУГОУ SUPR 383, 26-29.

3 - видъ ниа сеа иджира къ севъ. и гла се агньць сжи вьземал грузуы МИРА In 1, 29 SAV; НЕ ТОЛЬМА ПОНИЕМЬЕТЪ ОУСТА ПРОДОЧЬСКА: ГАКОЖЕ КООТЪКЪНИХЪ и тихъинуъ чловъкъ SUPR 383, 19-21; cf. also for каняти 5: съмідетъ чеко дождь на роуно ї теко каплъ капяшть на земля PS SIN 71, 6.

4, 7 - ЕКОЖЕ НЪИ ЕСІ НЕБЕСЬСКЪНЬА ПЩИА НАСЪИТИЛЪ' ТАКОЗЕ ЖЕ С ЖИВОТЪ НАШЬ СИЛОЖ ТВОЕН ОУТВОЬДІ КІЕV 4b, 3-6; а дрУзии очченици кораскицемь nongoma' ne E'blud E0 ganeye ot b 3emba Jn 21, 8 SAV (cf. in the same verse: a Дроузи оченнци коравлицемь придж: не въша во далече отъ земля ZOGR; likewise: на лици всел земла Lk 21, 35 SAV); конь же тъ приведе въ манастъюют МЬЛЕТЬ' НА КРЪМЫЙ ПОНУОДАШТИИХЪ ТОУ СТРАННИКЪ SUPR 565, 3-5 (сf. ПОТОМ ЖЕ ГРАДЬЦЬ МАЛЪ НА БОВЗЪ СЬТВООНВЪ: И ТЪ ВЪ РЪДЪКЪМА ЧАСЪ! ДЪНИ ДЪЛАА: OT'S TOTO KOLANAR CEE'S HM SALLE SUPR 519, 5-7).

5 - ЛИЦЕ ДЕВИЧЕ ХОДЪ ДЕВИЧЪ: ОЧИ ДЕВИЧИ: ОСКЛАВЪКНИЮ ДЕВИЧЕ: БЕСЕДА ДЕВИЧА: А ЖТРОВА НЕ ДЕВИЧА SUPR 240, 11-13; НЕСТЪ ВО ТАЇНО ЕЖЕ НЕ ГАВИТЪ са: ни оутаено еже не кждетъ познано: и въ тавение не придетъ Lk 8, 17 SAV; Въ КОНЕЦЪ ПЛОМЪ ПЪСНІ ОСВЕЛЦЕ(НЬВ) ДОМОУ ДАВА РЅ SIN 29, 1; ТВОВ СВРАТАЋ Въсемогъ вже: 'Еже се нън премлелу: на раздръшение' и на очишчение намъ БЯДЖ KIEV 5а, 13-16.

6 - да исправі нъи в очисти не нашихъ делъ раді нъ овъта твоего раді сже еси объщелъ налъ клет 3b, 8-12; ни азъ есмъ съзъдавъии сеее: ни азъ ПОГОУБЫЛЬЖ СЕЕЄ НЪ ІЄСТЪ ГОСПОДЬ СЪТВОДИВЪНИ МА И ТЪ ОБНОВИТЪ МА СИЛОНЖ CBOFER' H T'S MA H3EAB'STAT'S H XDAHHT' HENOSONSHA OTT'S SUPR 209, 17-21; лице гавыаютъ печаль SUPR 474, 26; cf. also in the n-Partstem: милостівать OE'ELLENHE KIEV 2a, 8-9.

7 — дьни же съвъшоу розьства иродова: пласа дъшти продитьдина по сръдъ и оутоди продови Mt 14, 6 мак; соутоува пакъі гла жізнь: соцтоуво розьство (here in the manuscript the letters ждι are written above the word) • Въ коупъ 1 ПОДОЗЪСТВОУ" І СЛЫШІ ВЪ РЕЧІ" СОУГОУВО ДОЗЬСТВОУ ВЕШТИ" І ВВЕСПЛЕШТИ ЧЮДЕСА" анклъ марні" матери хвъ: рожъство его влаговъствова: и аклъ: мари магдалъни: порожденье его сже отъ гроба: влаговъствотвова CLOZ 13b, 40-14a, 7.

8 - 1 съдравие намъ дазь и доушы наших и телеса очисті клеу 4а, 11-13; ПОДАЗЬ НАМЪ ПООСИМЪ ТЊА ВЪСЕМОГЪИІ ЕЖЕ ЕЛАЖЕНЪНЊ РАДІ МЖЧЕНИЦИА ТВОЕНА фелицітъи въкоупьняют молитвя кіеv 2а, 12-16; сf. принеси пръста твоє (го съмо в визжь ряць мон Jn 20, 27 маг.

9 — мѫченꙇкъ твоіхъ гі чьсті чьстѩце молімъ тѩ просѩце· да ѣкоже ѩ есі славоѭ твоеѭ небесьско[о]ѭ утврьділъ· такозе же ꙇ нъи мілостіѭ твоеѭ прꙇмі Kiev 6b, 4–10; ходатаѩцю блаженуму клименту мѫченику твоему Kiev 1b, 11–13.

10 — твоѣ свѩтаѣ вьсемогы бже· ѣже се нъи пріемлемъ· на раздрѣшение· ꙇ на очишчение намъ бѫдѫ· а тъи самъ помоцьѭ твоеѭ вѣчьноѭ зашчіті нъи Kiev 5a, 13–18; цѣсарьствѣ нашемь гі мілостьѭ твоеѭ прізьрі· ꙇ не отъдазь нашего тузꙇмъ· ꙇ не обраті насъ въ плѣнъ народомъ поганьскъимъKiev 4b, 8–13.

# §895. Synchronic aberrations and diachronic changes

Note that aberrations in synchronic grammar on the one hand, and dialectal differences understood to result from differences in sound changes on the other, while analogous in content, cannot be reduced to one another.

For example, historical grammar sets up the following sound changes:


In the synchronic grammar, the analog of this system of sound changes is the system of aberrant pairings by the substitutive softening alternation.


Cf. насытити, насыщен=, but насыцен= in Kiev; очистити, очищен=, but очишчен= in Kiev. There are no examples of the ск‖шч pairing in Kiev (they could be expected, e.g., in the Prae forms of the verb искати). However, in Kiev we find the Imv form зашчити, for the canonical infinitive защитити. From an etymological point of view, this form shows a Kiev-specific development of \**skj* (or \**sk* + front vowel) into *šč*;18 however, in the synchronic grammar there is no aberrant pairing ск‖шч, because the position here is the initial C, where no alternations take place. The aberrant form Imv зашчити (for canonical защити) results from a lexical aberration (зашчитити for canonical защитити).

# §896. Lexicographic difficulties in attesting lexical aberrations

Because the corpus is rather limited, some lexemes represented by only a few glosses may be found only in aberrant forms. It is difficult to find a good lexicographic solution in such cases. Reconstructing canonical forms for inclusion in the dictionary, even if such a reconstruction seems obvious, runs the risk of inventing nonexistent lexemes. Nonetheless, in a few cases such a reconstruc-

<sup>18</sup> See the etymological details in Van Wijk, § 18.

tion is performed in this work. For example, PD contains the fictional canonical lexeme рабиискъ for the hapax aberrant form робиискоѭ Supr 562, 12. In other cases PD shows the aberrant form. All such cases are discussed in comments under specific lexemes in the overview of paradigmatic classes.

In the general case, partial phonological differences between eponymous wordforms with the same meanings can occur either when one of the forms is canonical and the other aberrant, or when both are canonical. In the latter case the forms are doublets. For example, ави сѧ Zogr and ѣви сѧMar (in Mk 9, 4) are forms of doublet lexemes авити and ꙗвити. A special case are derivational doublets, where the lexemes differ in partially identical synonymous suffixes (cf. младеньць and младѣньць), or in alloforms of the basic components (cf. родьство and роЖьство). It is hardly appropriate to develop formal criteria to distinguish doublets from aberrations, but the factual boundary between the two categories is explicitly drawn in this book. In extreme cases the distinction is quite clear and obvious (compare, on the one hand, авити and ꙗвити, with more than 200 occurrences with a relatively even distribution across sources, and on the other hand, розга and разга, with 16 occurrences in the Gospels and Ps Sin, out of which разга occurs once in Zogr and Mar each). On the other hand, in many intermediate cases the decisions are relatively arbitrary.

Let us consider one example in more detail. Suffixal forms with the suffixes ѣн {31}, ен {57}, and ьн {29} in some cases create competing lexemes. On the membership of stems with ен {57} see § 861. Note that spellouts with ен can be treated as containing not the suffix ен {57}, but as resulting of the strengthening of ь in the suffix ьн {29}. This is the interpretation of only the form мѣденъ Ps Sin 17, 35.

In other pairs of parallel forms some spellouts could be treated as aberrant, but a general overview of the cases does not seem to yield to a reasonable and nontrivial solution. This book treats all competing lexemes with these suffixes as independent canonical lexemes, i.e. doublets, because it does not appear that the distinction between the suffixes ѣн {31}, ен {57}, and ьн {29} corresponds to a meaning difference.

Table 896 shows all the relevant material.

PD shows one, two, or three lexemes, depending on the attestation of forms: one occurrence for багърѣница, младьнъ etc.; two occurrences for дрѣвѣнъ and дрѣвьнъ, пътѣньць and пътеньць etc.; three occurrences for младѣньць, младьньць, and младеньць.


Table 896. Competing uses of the suffixes ѣн, ьн, and ен


# Table 896 (continued). Competing uses of the suffixes ቴዞ, ын, and єн

1) Cf. камень 1/m, каменне 2/n, каменьнъ 2/a, окаменьние 2/n.

2) Cf. пламень 1/m, пламеньнъ 2/a.

# §897. The task of describing aberrations and the task of describing sources

Although phonological effects encountered in sources are fairly diverse, they turn out to be representable as the result of a rather limited set of aberrations. Description of sources and description of aberrations are distinct undertakings. Description of sources is not a goal of the present work. Rather, our goal is to show that the canonical OCS proposed in this book satisfies the conditions formulated at the outset: namely, that the mapping between the canonical language to the observed sources is ensured by the writing systems as given, and by a limited list of aberrations. This book shows that the proposed list of aberrations is adequate, by showing how each aberration corresponds to aberrant spellouts observed in the texts of the sources. The completeness of the list is not shown explicitly; in a sense, completeness was not one of the goals. The convention is that any form that is non-canonical and attested in some source, if it cannot be represented as resulting from one of the fixed aberrations, it is treated as erroneous. For example, there are no aberrations that could represent the effects ѫ⇒о or ѧ⇒е, which are fairly frequently observed in Ps Sin (see Ch. 7, *Survey of the sources*, § 198). Formally this means that all such spellouts are errors.19 However, this book does not apply the aberrant ~ erroneous distinction overly pedantically: some alternative spellouts are considered aberrant even when the available aberrations are insufficient to generate them.

This spellout of aberrant forms is not intendend as a reconstruction of the mental actions of the creators of the manuscrips. Also, the above remarks apply to the grammar of aberrations as proposed here. Both the list of aberrations and their classification in a different grammar might be different.

# **Excursus on** *yer* **aberrations and Havlík's rules**

# §898. General remarks

The following reasons necessitate a detailed look at *yer* aberrations undertaken in the present excursus. First, these are the most frequent and salient aberrations. Second, these aberrations are the subject of many specialized works, and a treatment of the *yers* forms a necessary part of handbooks and textbooks. In

<sup>19</sup> Of course, the factual boundary between aberrant and erroneous forms can hardly be drawn objectively. For example, for the canonical starting form довольнъ, along with the canonical forms, we have the spellouts довьльн- Supr, дъвъл҄ьн- Zogr, дьвьльн- Mar. Formally, aberrant forms of the previx are representable as resulting from the atomic *yer* aberration (hypercorrection), while the aberrant form of the root results from the aberrant distribution of root alloforms. However, it is also possible that these forms result from a poor grasp of the morphological composition of the stem, and as a result, an approximate reproduction of the form with a random distribution of vowels, generating a morphologically strange form (as in Russian *удвольтворительно* for *удовлетворительно* or English misanalysis *donzerly* for *dawn's early* in *dawn's early light*).

this domain the simplicity of the observed data is combined with a rather nontrivial grammatical interpretation of the facts.

The underpinnings of *yer* theory is simple: a single object (in some cases the phoneme ъ, in other cases the phoneme ь) has three alternative realizations: strengthening (realization of ъ as о and of ь as е); deletion (realization of both ъ and ь as 0); and confusion (realization of ъ as ь, and of ь as ъ). Correspondingly, the theory aims to uncover the conditions on this triple alternative.

The nontrivial grammatical interpretation arises for two reasons. First, the status of the participants of the observed changes is unclear: do these segmental units belong to graphics, phonology, or morphophonology in the case of synchronic grammar, or to OCS or some etymological image of OCS in the case of diachronic and comparative-historical grammar? Second, the boundaries within which the behavior of these units is observed are not clearly set: is the observation limited to OCS data, or does it go beyond it both in time and in terms of dialects? The generalization on the triple alternative known in Slavic studies as *Havlík's rule* was first formulated as a sound law that applied in the change from older to newer forms of the language. Although the data of OCS sources are not in obvious contradiction with this rule, it can hardly be attributed to the synchronic grammar of OCS, or of single sources. Note that both the discovery of Havlík's rule and its subsequent verification rely on material which goes far beyond OCS, including, on the one hand, sources where *yer* aberrations are represented much more broadly than in OCS, and on the other, sources that altogether lack *yer* aberrations (such as the Ostromir Gospels). Thus, it turns out that in the scheme "old *A*⇒new*B*", for OCS, the source (*A*) is buried in etymology, while the outcome (*B*) belongs to the younger dialectal neighbors, in particular to Old Russian and other Church Slavic data that are not direct descendants of OCS.

#### The objects of comparison

While considering *yer* aberrations, we compare (1) a spellout in the source text, which is treated as a particular wordform; (2) the canonical morphophonological representation of that wordform; (3) the normalized spellout of that canonical wordform. Thus, for example, the spellouts день Mar Mk 9, 31, денъ Mar Mt 26, 17, and дьнь Mar Mk 14, 12 (all NASg) are compared with (2) the morphophonological representation дьн.ь, and (3) the normalized spellout дьнь. The spellouts дьнииMar Lk 17, 22, дьнеиMar Lk 20, 1, and дънеиMar Lk 8, 22 (all GPl) are compared with (2) the morphophonological representation дьн.ьј.ь, and (3) the normalized spellout дьнии. The spellout дьнье Mar Mk 2, 20 (NPl) is compared with (2) the morphophonological representation дьн.ьј.е, and (3) the normalized spellout дьниѥ. The main object of study is the individual occurrence of ъ in the morphophonological representation of the wordform. Each such occurrence is evaluated as having a standard or aberrant reflection in the

spellout of the source,20 and, correspondingly, the terms "standard" ~ "aberrant" apply to such occurrences. Table 898 shows some examples.



Fall, strengthening, confusion, and regression of the *yers*

The following types of *yer aberrations* are possible.

1. In place of the *yer* occurrence in the morphophonological representation we find no vowel at all. For example, in the spellout дниі (for дьн.ьј.ь) Sav Mt 11, 12, there is no vowel in place of the first *yer* occurrence. In such cases we speak of the *fall of the yers* aberration.

2. In place of the *yer* occurrence in the morphophonological representation we find the letter о in place of morphophonological ъ, and the letter е in place of morphophonological ь. This is the case in the spellout дънеи Mar Lk 8, 22 for the second *yer* occurrence, and in the spellout день Mar Mk 9, 31 for the first *yer* occurrence. In such cases we speak of the *yer strengthening* aberration.

3. In place of the *yer* occurrence in the morphophonological representation we find the letter ъ in place of morphophonological ь, or the letter ь in place of the morphophonological ъ. For example, this is the case in the spellout дънеи Mar Lk 8, 22 for the first *yer* occurrence. In such cases we speak of the *yer confusion* aberration.

4. In place of the *yer* occurrence in the morphophonological representation we find the letter ъ in place of the expected normalized ы, or the letter ь in place of the expected normalized и, where the *yer* occurrence is adjacent in the mph form to *j* or *i̯*. Such *yer* occurrences surface as /y/ (ы) for /ъ/, and as /i/ (и) for /ь/ by the mph⇒ph rules. For example, this is the case in the spellout дьнье Mar Mk 2, 20 for the second *yer* occurrence, compared with the normalized дьниѥ. In such cases we speak of the *yer regression* aberration.21

A note on the term "hidden yers"

In the mapping from morphophonological to the phonological and graphic (normalized) representations, ъ and ь are realized as ы and и, respectively,

21 Spellouts with *yer* regression after *j* or *i̯* are not attested: aberrant spellouts observe the graphic prohibition against ъ and ь after vowels. This applies to spellouts like \*краь for крај.ь, normalized краи, or \*достоьнъ for до.сто*i̯*.ьн.ъ, normalized достоинъ.

<sup>20</sup> Because the mapping from graphics to phonology is unambiguous, it is immaterial whether the spellout is in letters or phonemes.

if they are adjacent to the phoneme *j* or the symbol *i̯*. Cf. нов.ъ*i̯*.ь⇒новыи, etc. In pedagogical practice such instances of the letters ы and и are called *tense* or *hidden yers*.

Havlík's rule

Havlík's rule contains two subrules. Havlík-1 (H1 below) defines *yer* occurrences as *strong* (ъ̟) or *weak* (ъ̠). Havlík-2 (H2 below) predicts the expected aberration—fall, strengthening, or confusion. In the present treatment, possible regression is treated as part of H2.

Here is the rule H1.


The rule is clearly recursive. It must be applied starting from the rightmost occurrence of ъ or ь in the morphophonological representation.

For example, we have в з.м. for възьмъ, в з.м.еши for възьмеши; при.*i̯*м. for приимъ; свин.ј. for свинии GPl, свин.ј.и for свинии NLDSg; крст. for крьстъ, крст.а for крьста; двр.ц.а for двьрьца, двр.н.ик. for двьрьникъ; дн. for дьнь, дн.ј.е for дьниѥ, с .бр.а.т.и for събьрати; бј.еши for биѥши, у.бј.и for убии Imv.

### Note

The relevant linear unit is the clitic group, or the extended wordform, i.e. the wordform together with the adjacent clitics. Thus, Havlík-1 applies to strings such as въ кръви (a possible aberrant spellout is во кръві Ps Sin 29, 10), въ тъ часъ (a possible aberrant spellout is во тъ часъ Lk 12, 12 Mar), отъ съвѣдѣнии (a possible aberrant spellout is ото съвѣдѣнеї Ps Sin 118, 152), дьнь сь (a possible aberrant spellout is днесъ Ps Sin 2, 7), лежитъ сь (a possible aberrant spellout is лежітось Lk 2, 34 As), избавитъ и (a possible aberrant spellout is ізбавітоі Ps Sin 21, 9), убиимъ и (a possible aberrant spellout is убиѣмои Mt 21, 38 As, with a paradigmatically aberrant Imv terminal), прѣдамь и (a possible aberrant spellout is прѣдамеи Mt 26, 15 Mar), рабъ тъ (a possible aberrant spellout is работъ Mt 18, 26 Mar, As), etc. (see § 868, *On wordform boundaries and loose and tight formative adjacencies*).

The rule Havlík-2 states that strong *yers* (i.e. the phonemes ъ and ь in corresponding occurrences) may be strengthened or confused, but may not fall; weak *yers* may fall or be confused, but not strengthened:22


<sup>22</sup> This is the form of the rule as formulated by Antonín Havlík, and often the term "Havlík's rule" covers only this statement.

Usually Havlík's rule on *yer* confusion is stated differently: only weak *yers* are claimed to be subject to confusion (cf. Lunt 1974, § 2.523–2.524; Lunt 2001, § 2.624 and ff.). However, the sources show many examples with *yer* confusion in strong positions. Cf. вьзъми Mk 2, 9 Mar; бръвъно Mt 7, 4 Zogr; въшъдъшю Mk 9; 28 Sav, опъръ сꙙ Supr 558, 28.

### *Yer* aberrations in OCS sources

Like all other aberrations, *yer* aberrations show possible modifications of canonical spellouts. In other words, Havlík's rules allow certain realizations, rather than require them. Note that in general, OCS sources show canonical spellouts with *yers* more often than aberrant ones. Not all aberrant spellouts follow Havlík's rule. For example, the form всь (for вьсь), frequent in Sav, shows the fall of a strong *yer*.

### Standard and anomalous *yer* aberrations

Specific *yer* aberrations that follow the Havlík-2 rule are called *standard yer aberrations*; those that do not follow the rule are called *anomalous yer aberrations*.

The following are examples of spellouts that show standard *yer* aberrations. Fall (for weak occurrences of *yers*): мноѕи (for мъноѕи), кто (for къто), злоба (for зълоба), дом (for домъ); пожри (for пожьр҄и), дни (for дьни), вси (for вьси), подобно (for подобьно), пси (for пьси), дам (for дамь).

Strengthening (for strong occurrences of *yers*): сотъ (for сътъ), вопл҄ь (for въпл҄ь), доЖъ (for дъЖь), золъ (for зълъ), кровъ (for кръвь), плоть (for плъть), жерьцемъ (for жьрьцемъ), отеръши (for отьръши), юної (for юн.ъ*i̯*.ь, canonical юныи) Ps Sin 118, 9; шедъ (for шьдъ), весь (for вьсь), наченъ (for начьнъ), въземъ (for възьмъ), сѫчець (for сѫчьць), горешъ (for гор҄ьшь); сътвореи (for съ.твор҄.ь*i̯*.ь, canonical сътвор҄ии), велеи (for вел.ьј.ь, canonical велии), дьнеи (for дьн.ьј.ь, canonical дьнии); сънемъ (for съньмъ NASg: архиереи и старьци· и кънижъникы и весь сънемъ Mk 15, 1 Mar), and сонъмы (for съньмы APl: прѣдадѧтъ бо вы на сонъмы· ꙇ на съньмищихъ вашихъ убиѭтъ вы Mt 10, 17 Mar).

Confusion (for weak occurrences of *yers*): изъмѫ (for изьмѫ), въсе (for вьсе), на зъдѣ (for на зьдѣ), приставъникъ (for приставьникъ); посьлеши (for посъл҄ѥши), азь (for азъ), сътъретъ (for сътьр҄ѥтъ), кьто (for къто), рабь (for рабъ).

Confusion (for strong occurrences of *yers*): вьзьми (for възьми), осълъ (for осьлъ), възъмъ (for възьмъ), пришъдъ (for пришьдъ).

*Yer* confusion can give rise to prohibited combinations of velar with front vowels, cf. кьнижьници Mt 23, 27 Mar, отъгьнавъ Supr 179, 17.

Regression (for hidden *yers*): пьетъ (for пьј.етъ, canonical пиѥтъ), дьнье (for дьн.ьј.е, canonical дьниѥ), абье (for аб.ьј.е, canonical абиѥ), сьѩ (for с.ьј.ѧ, canonical сиѩ), невѣрью (for не.вѣр.ьј.у, canonical невѣрию); also змъи сь Ps Sin 103, 26 (for зм.ьј.ь с.ь, canonical змии сь NSg), which shows regression and confusion of a weak hidden *yer*.

Sources show many examples of *yer* aberrations (viz. fall, strengthening, and confusion) that contradict Havlík's distributional rule. Below are some examples that show anomalous *yer* aberrations.

Anomalous fall (for strong occurrences of *yers*): тъгда приде съ н҄ими ис· вь всь нарицаемѫѭ· ћенꙿсимани Mt 26, 36 Zogr for вьсь 1/f 'village, town'; и абие всь градъ изиде вь сърѣтение ісу Mt 8, 34 Sav for вьсь 2/p\* (and also in Sav more than 10×); бѣ бо великъ днь въ тѫ сѫботѫ Jn 19, 31 Sav; трьми дньми Mt 27, 40 Sav; Mk 15, 29 Sav; дьнь сь зимнъ Mt 16, 3 Zogr; ꙇѣко быстъ дьнь събърашѧ сѧ старьци людьсции· ꙇ архиереи и кънижьници· ꙇ вѣсѧ и на сънмь свои Lk 22, 66 Mar.

Anomalous strengthening (for weak occurrences of *yers*): приде вь витаниѭ· ꙇдеже бѣ лазаръ умерои Jn 12, 1 Mar (for у.мьр.ъ*i̯*.ь, canonical умьрыи); ꙇ се изношаахѫ умерошъ снъ иночѧдъ матери своеи· ꙇ та бѣ въдова Lk 7, 12 Mar; тъи лі едінъ прішелецъ есі въ иерлмъ Lk 24, 18 As; ꙇ шедъши домови обрѣте отроковицѫ лежѧщѫ на одрѣ и бѣсъ ишедъшъ Mk 7, 30 Mar.

Standard *yer* aberrations in sources

The picture described above applies to OCS as a whole. However, each source shows its own peculiarities, not all of which are fully understood.

In different sources aberrant spellouts are distributed differently among lexemes and wordforms. One can thus assume that in different sources different aberrant spellouts could have the status of alternative variants. Cf. the distribution of forms of мъногъ and къто described in the Introduction.

The source Sav is unique in that it shows no strengthenings (cf. Ščepkin, pp. 94–113), and frequently shows forms with fall of strong *yers*, such as всь.

### §899. An addendum to Havlík's rules for *yers* with liquids

The distinction between free and bound *yer* occurrences

Classical *yer* theory distinguishes two types of *yer* occurrences, which are called *free* and *bound* in the discussion below. In order for a *yer* to be bound, the following three conditions must be met.

1. The observed *yer* occurrence belongs to a root that etymologically shows sonant vocalism of the H(*r*) or H(*l*) series. For example, this condition is met for the wordforms гръд.ъ ‹197›, мрь.т.в.ъ ‹539›, прьст.ъ ‹718›, and съ.бьр.а.т.и ‹24›, but not for the wordforms крьст.ъ, крьст.а ‹452›, and двьр.ьц.а ‹221›.

2. The vocalic realization of the given alloform is V-final (i.e. has the form ръ, рь, лъ, ль, and not ър, ьр, ъл, ьл). For example, this condition is met for the wordforms гръд.ъ, мрь.т.в.ъ, and прьст.ъ, but not for the wordforms у.мьр.ѫ, съ.бьр.а.т.и.

3. The observed *yer* occurrence must be followed by a consonant. For example, this condition is met for the wordforms гръд.ъ, мрь.т.в.ъ, прьст.ъ, and об.трь.т.и, but not for the wordform об.трь.

### The behavior of bound *yers*

According to classical *yer* theory, bound *yers* cannot fall or strengthen, but show only confusion. Such occurrences are not classified as weak or strong. In the string of syllables that determines the weak vs. strong status of free *yers*, such

occurrences behave as ordinary (or full) vowels (cf. for example Vaillant, § 20; Lunt 1974, § 2.53–2.531).

Havlík's rule with the bound *yer* correction

Statements of the rules that take into account the bound vs. free *yer* distinction are shown below, even though no tools for making this distinction are provided in the present grammar.


Here "yes" means 'such a replacement is possible', and "no" means 'such a replacement is impossible'.

In other words, bound *yer* occurrences are not subject to the weak vs. strong distinction, and for the purposes of this distinction are treated as ordinary vowels. For example, in съмрьть, the first *yer* is free and weak (leading to possible fall, смрьть), while in събьрати the first *yer* is free and strong, because the second *yer* is free and weak (leading to possible strengthening, собьрати).

The fact that the first *yer* in съмрьт is weak is shown by spellouts like во съмръті Ps Sin 12, 4 (for в с мрьти, not в с мрти; see Vaillant, § 24), and ото съмръти Ps Sin 114, 7 (for от с мрьти, not от с мрти; see Vaillant, § 20);23 cf. also въ сꙿмрьти мѣсто Supr 489, 16 (for с мрьти, not с мрти).24

Bound *yers*, showing neither fall nor strengthening, are frequently subject to confusion: пръста (for прьста), утвръди (for утврьди), мрътвъ (for мрьтвъ), пръвѣе (for прьвѣѥ); грьтань (for грътань), крьмити (for кръмити), скрьбите (for скръбите), влъкъ (for влькъ), наплъньше (for напльн҄ьше), мльва (for млъва).

Note that all spellouts with *yer* aberrations are subject to Havlík's rules without the bound *yer* correction (i.e. the rules H1 and H2). The three examples listed above (во съмръті, ото съмръти, and сꙿмрьти) are within the range of errors due to anomalous *yer* aberrations.25

<sup>23</sup> The ms. has ото съ[съ]мръти, repeating the prefix съ across lines.

<sup>24</sup> The ms. shows въ сꙿмрьти: с is added above the line.

<sup>25</sup> Cf.: «[…] le vieux slave possédait un *r* et un *l* voyelles […]. Il les note ръ, лъ (ou рь, ль, simple variante): съмръть 'mort', плънъ 'plein'; mais le ъ qui suit р et л est graphique ou représente le léger accompagnement vocalique qui facilite l'émission de *r* et *l* voyelles, et ce n'est pas un jer réel. On le voit par deux faits: d'une part il n'est jamais vocalisé, tandis que les véritables groupes ръ, рь, лъ, ль connaissent la vocalisation des jers forts […]; d'autre part, pour la répartition des jers forts et des jers faibles […], le ръ de съмръть est traité comme voyelle pleine, d'où gén. сꙿмрьти Supr 48916, ото съмръти 'de la mort' Ps Sin CXIV, 7, alors que le ъ

Note that Havlík's theory—both in the simplified version without bound *yer* correction, and in the more sophisticated version with bound *yer* correction is based on comparative grammar of Slavic languages, where East Slavic data have the leading role.

Aberrant spellouts corresponding to canonical -ии-, -ыи-

As is well known, many sources show forms with a single и instead of expected ии, or with a single ы instead of expected ыи. Expected graphic ии/ыи combinations are possible only across boundaries, where the first formative ends in ыj/ы, ъj/ъ, иj/и, or ьj/ь, and the second formative begins with и/jи or ь/jь. For example, we have при+иЖити (Inf, canonical прииЖити), гньј+и (Imv, canonical гнии), бьј+и (Imv, canonical бии), нов.ъ*i̯*.ь (NSgmPlen, canonical новыи), нищ.ь*i̯*.ь (NSgmPlen, canonical нищии), у.биј + ьц.а (canonical убиица), у.биј+ьств.о (canonical убииство), у.мъ*i̯*+и (Imv, canonical умыи), нов.ъ*i̯*.имъ (DPlPlen, canonical новыимъ), нов.ъ*i̯*.ихъ (LPlPlen, canonical новыихъ); also избавитъ+и (3SgPrae +ASgm(\*и), canonical избавитъи),26 etc.

The phonological and morphophonological interpretation of contracted spellouts is ambiguous. Phonologically we could assume the lack of a distinction between /i/ and /ii/. From a morphophonological point of view, several spellouts could show the effects of *yer* aberrations. For example, the spellout новы for новыи (нов.ъ*i̯*.ь) could be interpreted as the fall of the last weak *yer*. Taking into account that in several sources spellouts with the letter ы are graphically indistinguishable from the sequence ъ+и, the effect can also be seen as regression of the penultimate *yer*. The safest interpretation seems to be a graphic one: every phonological /ii/ can be written as и; every phonological /yi/ as ы. This treatment is supported by such infrequent forms as на срдциихъ вашиихъ Lk 21, 14 As (for canonical на срьдьцихъ вашихъ). However, such spellouts are also consistent with the phonological treatment of this phenomenon.

For most wordforms, sources show both canonical and aberrant spellouts, even though in some cases a certain standard is fixed. For example, при+ити (also in oblique forms, e.g. при+идѫ etc.) is always written as прити, while for при+иЖити we only have forms like прииЖити.

Given the ambiguity of the grammatical interpretations of forms in this class, all such forms are treated separately in the analysis of the sources.

de кръвь est traité comme jer faible dans во кръві Ps Sin XXIX, 10» (Vaillant, §20). Given the generally unstable spellouts with *yers* observed in OCS, it seems unconvincing to maintain the distinction between free and bound *yer* occurrences in OCS itself.

<sup>26</sup> See more in § 868, *On wordform boundaries and loose and tight formative adjacencies*.

# **Excursus on the description of phoneme syntagmatics**

## §900. General

Syntagmatics studies allowed and prohibited adjacencies of units of a lower rank inside units of the immediately higher rank. For phonemes, there are two units of an immediately higher rank: formatives and wordforms. Because a wordform has two types of representations—morphophonological and phonological ones—the question of phoneme adjacencies can belong to either of these representations. In the case of morphophonological representations, adjacent phonemes can be within a formative or across a formative boundary. Thus, we have the syntagmatic situations shown in Figure 5.

Figure 5. Four syntagmatic situations

In the general case, the syntagmatic situations S1–S4 are pairwise distinct. For example, a formative can begin with the vowel ъ, cf. бук.ъв.и (situation S1), but the wordform in neither ph nor mph representation cannot begin with the vowel ъ (situations S2–S4).

To describe phoneme syntagmatics means the following: for each phoneme combination (including unit-initial and unit-final positions), and for each syntagmatic situation, one must establish a syntagmatic evaluation, which in the simplest case is binary: "allowed" vs. "prohibited".

Fortunately, for OCS it is fairly simple to establish a syntagmatic evaluation for two immediately adjacent phonemes, because the syntagmatics of longer combinations is established by the recursive rule of *syntagmatic depth reduction*, namely:

*xyz* is allowed if and only if *xy* is allowed and *yz* is allowed; *xyz* is prohibited if and only if *xy* is prohibited or *yz* is prohibited.

Thus, the full picture of phoneme syntagmatics can be represented by four tables of the following form.


Here, *a*, *b*,… are phonemes or the symbol # (the space, or the left/right boundary),27 and the table cells contain syntagmatic evaluations (usually "allowed" or "prohibited").

Of course, such a way of representing syntagmatics is not very useful, and grammars usually use more compact and easy-to-use formats of presenting data. First, some rows or columns turn out to be equivalent, and this makes the tables shorter and more general. For example, the rows for к, г, and х are equivalent in all four tables. Second, some situations may be equivalent or almost equivalent. Third, syntagmatic evaluations ("allowed" vs. "prohibited") are to some extent elastic, and this permits us to unify some equivalent rows or columns. The issue is that some attested combinations are syntagmatic anomalies (cf. декѧбръ given the prohibited combination кѧ), while unattested ones can be treated arbitrarily as allowed or prohibited.

Now let us turn to how such a reduction in syntagmatic data is achieved in practice.

### §901. Equivalence of rows and columns

The view of syntagmatics as the application of two ordered syntagmatic filters the CVC schema filter, and particular filters in each allowed schema— gives a more general picture of syntagmatic data. Indeed, the OCS CVC filter prohibits a large number of theoretically conceivable combinations of phonemes, leaving only some groups of combinations for the particular filters. In the particular filters, the equivalence of rows and columns in most cases can be described using phonological features. For example, front vowels are syntagmatically equivalent, and opposed to back vowels (a special case is the phoneme /i/ word-initially and after vowels).

<sup>27</sup> From a syntagmatic point of view, the symbol # that marks the so-called loose formative adjacency (see § 69) is equivalent to the absolute beginning or end of a wordform.

# §902. Equivalence of syntagmatic situations: statement of the problem

It turns out that the syntagmatic situations S1, S2, S3, and S4 are largely identical. In other words, it turns out that if some combination is allowed in one situation, it is also allowed in the other, and if some combination is prohibited in one situation, it is also prohibited in the other.

§903. Equivalence of syntagmatic situations: the syntagmatics of edge conditions and internal conditions

Note that the syntagmatic situation S1 is opposed to the situations S2–S4 in terms of the rank of the superordinate unit. In terms of syntagmatic possibilities, the situation S1 is opposed to the situations S2–S4 only in edge conditions. The syntagmatics of edge conditions deals with the availability of a given phoneme in the initial or final position of the superordinate unit, i.e. the possibility of edge combinations of the type *x*# or #*x*. For both types of superordinate units the syntagmatics of edge and internal conditions should be considered separately.

For edge conditions, syntagmatics is limited to the evaluation of the CVC schema. Allowed and prohibited CVC schemata for isolated formatives and formative strings that make up a wordform are described by the CVC norm, which is established for each formative class (see § 36–37). What is allowed for some class of formatives may be prohibited for wordforms (e.g. C-final roots and suffixes are allowed, while C-final wordforms are prohibited).

For internal combinations, the situation S1 (within a formative) is equivalent to the situation S3 (within a formative inside a wordform).

Thus, situation S1 does not need to be considered as a separate syntagmatic situation.

### §904. Morphophonology and phonology

Syntagmatic situations S2 and S3 belong to the morphophonology, while situation S4 belongs to the phonology. It is natural to expect that phonology allows fewer combinations than morphophonology. In other words, if some combination *xy* is allowed in the phonology (i.e. in S4)—see row 1 below—it is a fortiori allowed in the morphophonology (i.e. in S2 and S3). If some combination *xy* is prohibited in the morphophonology (i.e. in S2 and S3), it is a fortiori prohibited in the phonology (i.e. in S4)—see row 4 below. This is borne out by the facts.

The following is also true: if some combination *xy* is allowed inside a formative (the situation S3), then the combination *x*.*y* is allowed across a boundary (the situation S2), and if some combination *xy* is prohibited in the phonology (the situation S4), it is prohibited inside a formative (the situation S3). This claim is called *Kuryłowicz's thesis* below.28

<sup>28</sup> See Kuryłowicz 1948 and Kuryłowicz 1952.

However, among combinations that are prohibited in the phonology (i.e. in S4) and allowed in the morphophonology (i.e. in S2 or S3), some combinations are allowed across boundaries (situation S2), but not inside formatives (situation S3), in agreement with Kuryłowicz's thesis (see row 3 below), while others are allowed both across boundaries (situation S2) and inside formatives (situation S3), contra Kuryłowicz's thesis (see row 2 below). The possible combinations are shown in Tables 904.1–3.


Table 904.1. Possible combinations of evaluations in different syntagmatic situations

Facts corresponding to row 2 contradict Kuryłowicz's thesis, or rather some more general principles that underly that thesis. They are special circumstances that are represented by a few isolated cases. Outside of these special situations, the more general principle holds, which requires that in the mapping from mph to ph representations, formatives inside a wordform change their phonological shape only at boundaries.

These special circumstances (row 2) correspond to cases that have the following nonstandard morphophonological properties. First, all these cases have to do with the morphophonological representations of roots that undergo a change in their vocalism (e.g. жег.хъ⇒жѣг.хъ in the aorist). Second, the prohibition of the corresponding cases has graphic, not phonological content. All such special cases involve the allowance in the situation S2 of combinations of shibilants or j with ѣ, ы, or ъ—combinations that are prohibited in the phonology. Thus, in the mapping between mph and ph, these combinations are replaced with combinations of the form shibilant or j + а, и, or ь. Indeed, the spellouts жѣ and жа, (as well as жъ and жь, and жы and жи) are phonologically equivalent, but the normalization prescribes only one spellout: жа, жь, and жи, but not жѣ, жъ, жы.

Note that removing row 2 from the table makes the S3 and S4 columns equivalent, see Table 904.2 on p. 478.

Conversely, excluding column S3 makes rows 2 and 3 equivalent, see Table 904.3 on p. 478.


### Table 904.2. Table 904.1 without row 2

### Table 904.3. Table 904.1 without column S3


Thus, the syntagmatic situation S3, unlike the situations S2 and S4, is superfluous, outside of the special cases of row 2. There are only two situations that are distinct from each other: the morphophonological one (with the boundary (*x*.*y*) or without (*xy*), shown as *x*(.)*y* in the Table 904.3), and the phonological one, and special cases are treated as such.

### §905. The description format adopted here

Given the above considerations, in describing syntagmatics we will henceforth use tables that distinguish only two rows: mph and ph.

Moreover, the ph row can be understood to cover the syntagmatic situations S4 and S3 without distinguishing the type of representation to which it applies, except for special cases (see Table 904.1, row 2), which are marked in the comprehensive syntagmatic tables with a plus sign and a circle in the mph row. At the same time, the mph row can be understood to cover the situations S2 and S3 without distinguishing combinations with a boundary and without. In other words, a plus in the mph row can be understood as allowing the combination in the mph representation of the wordform. Indeed, if some combination, e.g. кт, is allowed in the mph representation across formative boundaries (cf. рек.т.и), it is allowed in mph representations without any specification as to

the presence of the boundary. The fact that it is impossible inside a formative (there are no formatives of the type ктV or Vкт), follows from the fact that кт is prohibited in phonological representations, and does not belong to the set of isolated special cases.

Accordingly, in qualifying some combination as allowed or prohibited in the mph row, the presence of a boundary is not indicated. If we say that *xy* is morphophonologically allowed, this means that either *x*.*y* or *xy* is allowed; if *xy* is prohibited, then both *x*.*y* and *xy* are prohibited.

# §906. On the agreement between syntagmatics and the mph⇒ph/norm rules

A combination *xy* that is allowed in mph and prohibited in ph is called *repairable* if it is removed in the mapping between morphophonological and phonological representations, i.e. if mph⇒ph/norm rules contain a rule of the form *xy*⇒*R*, where *R* is a single phoneme or a sequence of phonemes (cf. кт⇒щ, рек.т.и⇒рещи, or дт⇒ст, пад.т.и⇒пасти). Note that the rules may also remove combinations that are prohibited in morphophonological representations. For example, the morphophonological representation зна.еши contains a prohibited hiatus ае, which is repaired by mph⇒ph/norm rules.

The mph⇒ph/norm rules must agree with the syntagmatics in that all combinations that are allowed in mph and prohibited in ph must be repairable. The converse is also true: the rules must not remove combinations that are allowed in ph, and, of course, should not introduce combinations prohibited in ph.

In terms of their logical status, these statements are conditions on the correctness and perfection of the grammar, and may be false in particular cases that are empirical violations of grammatical conditions. This grammar of OCS contains such empirical violations: first, some instances of ph-prohibited combinations are not removed (e.g. in седмъ, дм is a prohibited combination, but the same combination дм is removed in e.g. плед.м.ѧ; likewise, несѣаше contains a prohibited hiatus, ѣа, but the same hiatus is removed in e.g. сѣꙗти [сѣ.а.т.и]). Second, some allowed combinations in some instances are removed (see the example об.нѣм.ѣ.т.и [онѣмѣти]). Of course, all such strange cases are listed individually.

### §907. An extended evaluation system; anomalous cases

Along with the syntagmatic evaluation in term of "allowed" vs. "prohibited", each combination can be characterized along the "checked" vs. "unchecked" scale (see Table 907 on p. 480). The combination *xy* is checked if it is represented in the benchmark corpus, and unchecked otherwise. All four cases are possible.


Table 907. Possible combinations of syntagmatic evaluations and checkedness

Of course, grammar writers seek to reduce the size of classes 3 and 2, but it is inevitable that these classes end up non-empty. The separation line on which the grammar writer focuses is the one dividing 1 and 3 from 2 and 4; the invisible distinction between 3 and 4 can be drawn on grammatical rather than empirical grounds. The boundary between 1 and 2 is the distinction between the norm and *anomalies*. Anomalies violate certain declared grammatical rules that are represented in the corpus, in the canonical set of forms. For example, we have canonical китъ, китовъ, which violates the law of the velars.

### §908. A note on checkedness and uncheckedness

As a consequence of Kuryłowicz's thesis, any phoneme combination allowed in S3 is allowed in S2. That is, any combination that is allowed inside a formative is also allowed across boundaries. Kuryłowicz's thesis is true as long as the syntagmatic evaluation "allowed" is not split into two, "allowed and checked" and "allowed and unchecked". Indeed, given the limited segmental diversity of formatives it often turns out that some combination *xy* is checked in S3 but not in S2. For example, the combinations с.к and к.в are not checked in S2, while they are allowed and checked in S3 and S4 (cf. дъска, квасъ). All such combinations are considered allowed in S2.

Combinations that are not checked in S2, S3, and S4 can also be allowed. For example, among the combinations (п, б, м) + (т, д), only п.т, б.т, and б.д are checked (in S2), while all six combinations are considered syntagmatically allowed in mph and syntagmatically prohibited in ph. All such combinations must be repairable. They are repaired by the rule that in all such combinations deletes the first consonant (cf. об.тек.т.и⇒отещи, and об.дѣ.т.и⇒одѣти). Thus, among mph⇒ph/norm rules, there are many that are in some sense fictional.

### §909. A note on the source of syntagmatic anomalies

Violators of syntagmatic prohibitions that are checked in the benchmark corpus of wordforms are syntagmatic anomalies. However, in the narrow sense, syntagmatic anomalies are only those words that contain phonologically prohibited strings of phonemes in phonological representations. For example, the wordform знаѥши /znaješi/ does not contain phonologically prohibited strings of phonemes, although the morphophonological representation зна.еши contains a morphophonologically and phonologically prohibited hiatus. The wordforms научити, седморъ, and китъ contain phonologically prohibited strings of phonemes in phonological representations. Henceforth, this narrowing clarification is left out, since we shall not consider any other syntagmatic anomalies.

A syntagmatic anomaly can be carried by the wordform as a whole—as in the case of научити—or by a formative, as in the case of китъ, седморъ, агньць. The sources of syntagmatic anomalies should always be sought in the formatives that make up a wordform. These are either nonstandard formatives that violate edge conditions (i.e. the CVC norm of their class), as in научити (the V-initial root уч) or агньць (the phonologically prohibited initial V), or formatives that contain phonologically prohibited strings of phonemes (i.e. syntagmatically anomalous formatives), as in the case of китъ (the phonologically prohibited string ки), or седморъ (the phonologically prohibited string дм).

Syntagmatic anomalies are either *potentially correctable* or *uncorrectable*. Correctable are those anomalies that contain in their phonological representation phonologically prohibited strings containing repairable phoneme combinations. Among the examples listed above, седморъ is a potentially correctable anomaly: it contains the repairable string дм (cf. руд.м.ѣн.ъ⇒румѣнъ). On the other hand, научити, агньць, and китъ are uncorrectable anomalies, as these forms contain absolutely prohibited phoneme strings (ау, #а, and ки), for which there are no repair rules.

This distinction has technically significant consequences. Potentially correctable anomalies must not be subjected to the mph⇒ph/norm rules: if they were to undergo those rules, they would lose their prohibited phoneme strings, change phonologically, and cease to be anomalous. To prevent this, such forms must not be fed as inputs to mph⇒ph/norm rules, and accordingly, such potentially correctable anomalies are given by a special list (see § 62). Of course, there is no technical need to create lists of uncorrectable anomalies.29 Let us note that among natural uncorrectable anomalies are 1) all wordforms containing as a first formative one that begins with a non-front vowel (such as обидѣти, отъдати, агньць, овьца, ухо, угасати, ѫза, etc.); 2) all wordforms containing absolutely prohibited clusters (e.g. нравъ, шлѣмъ, црькы, чрьвь, пожрѣти, жласти, ѕвѣзда, цвѣтъ, влькъ, врагъ, etc.); 3) all wordforms containing prohibited hiatuses (i.e. VV but not Vи) in their phonological representations; these are all wordforms with a V-final prefix and a V-initial (but not и-initial) following formative (root or prefix), such as научити, прѣобидѣти, приобрѣсти, etc., and all personal forms of the imperfect, such as несѣаше, трьпѣаше, плакааше, etc.

<sup>29</sup> They are uncorrectable for the very reason that there are no segmental rewrite rules that could change the phonological shapes of their formatives.

# **Excursus on the contamination and competition between paradigmatic classes**

### §910. General

Even though the distribution of wordforms into lexemes in this work is assumed to be given and is not formally subject of discussion, even though this procedure is generally quite transparent and familiar, and even though for the majority of cases the result of this procedure is fixed in the main lexicographic sources, still in some cases the grammar writer is faced with fairly nontrivial issues. Before turning to some of these questions in this excursus, note the following important consideration. Existing OCS dictionaries are designed to help readers of texts, and are not grammatical; each gloss that is attested in the corpus of texts must be reflected in the dictionary. The dictionaries do not care how the reader should arrive at the needed lexicographic entry. All such dictionaries do not concern themselves with whether some gloss will correspond to a separate dictionary entry, or a cross-reference, or whether a single gloss can be found under different lexemes in several different entries, whether the headword contains an attested base form of a word or a reconstructed one, and what the principles of reconstruction are. In this sense, with perhaps a few exceptions, Večerka's dictionary adequately meets its goals. However, many goals of a grammatical dictionary remain unsolved. In particular, the questions of distribution of wordforms among lexemes are either unaddressed in Večerka's dictionary or addressed inconsistently.

The task of separating wordforms into lexemes in the present case is made more difficult by two factors. First, because the data of the corpus is quite limited, most lexemes are not represented by a full set of their wordforms, and often by just a few forms. Such lexemes are called *incomplete*. For example, for съхнѫти we have only сьхнѣашеSupr 345, 11, for извѧзати we have only извꙙзавъшеSupr 79, 22–23, for умастити only умастілъ Ps Sin 22, 5 and умастіті Ps Sin 103, 15. Secondly, the corpus material contains both canonical and aberrant forms, and often classifying a given form as canonical or aberrant depends precisely on the distribution of the observed forms among lexemes.

### §911. Preliminary assumptions and auxiliary concepts

In the discussion below we abstract away from all aberrations except paradigmatic class aberrations and the aberrant ш-Part of the л҄юбивъ type. In other words, we consider: 1) aberrant class 7 PRAE system forms within class 3 lexemes, cf. увѧзаѭтъ; 2) aberrant class 3 PRAE system forms within class 5 lexemes, cf. угльбл҄ѭ; 3) aberrant ш-Part, н-Part, л-Part forms, and personal class 4 aorist forms within class 5 lexemes, cf. угасошѧ. Accordingly, examples from sources may show normalized spellouts replacing actually observed ones. Such spellouts show the effects under discussion but do not copy pedantically ev-

ery peculiarity of the forms actually attested in sources. For example, we write увѧзаѭтъ for увѩзаѭтъ Ps Sin 9, 23; угльбл҄ѭ for углъбѭ Ps Sin 68, 15; угасошѧ for угасошꙙ Supr 110, 15, etc.

Let us call a verbal wordform *paradigmatically determinate* if it can belong to a lexeme of only one paradigmatic class, and cannot belong to any lexeme of any other paradigmatic class. Otherwise, we will call the form *paradigmatically indeterminate*. For example, the form съхнѣаше can belong only to class 5 (the infinitive съхнѫти is reconstructable), while the wordform извѧзавъше can belong to both class 3 (извѧзати, извѧжеши), and to class 7 (извѧзати, извѧзаѥши). Paradigmatically determinate wordforms are also called *diagnostic*, and paradigmatically indeterminate ones *non-diagnostic*.

Let us call a *megalexeme* the set of verbal wordforms with the same lexical component. For example, the wordforms застѫпаѥтъ, застѫпаѭща, and застѫпи are wordforms of the same megalexeme (these wordforms represent two paradigmatic classes, застѫпити 1 and застѫпати 7). Likewise for the wordforms причѧщаѥмо, причѧщаѥши, причѧсти, причѧстихъ, причѧщьши (classes 1 and 7). Attributing wordforms to the same megalexeme requires their lexical component to be represented by the same string of formatives, even though in different wordforms the root formative may be represented by different alloforms. For example, the wordforms затъкнѫти and затыкаѭщѧ belong to the same megalexeme (the lexical components are за.тък and за.тык); likewise for дыхааше, дыхаѭщу, душетъ (the lexical components are дых and душ); likewise for задушивъ and задъхнѫти (the lexical components are за.душ and за.дъх). On the other hand, the forms бѣситъ сѧ and бѣсьнуѥтъ сѧ belong to different megalexemes: their lexical components are бѣс and бѣс.ьн, respectively.

Let us use the term *megafamily* to refer to a set of megalexemes sharing a root, if their lexical components differ only in the presence or absence of a prefix. The Table 911 (p. 484) shows the megafamily дъх ‹261›.

In the Table 911, all wordforms are somehow distributed between lexemes and classes. Let us call such a representation of a megafamily its *reduced distribution*. We use the term *simple membership* of a megalexeme or megafamily to describe the set of wordforms representing that megalexeme or megafamily.

Then the task of lexicography is to create a reduced distribution for each simple membership. In other words, each megafamily should be represented as distributed across lexemes and paradigmatic classes.


Table 911. The megafamily дъх ‹261›

ш- and щ-Part are represented in the table by their base forms.

See § 508 for class 3\* verbs дъхати, душетъ. Here and below, for simplification, PRAE forms with the base душ are treated as representative of class 3\*, while the aberrant form дышетъ is ignored altogether.

### §912. Constructing distributed megafamilies

Let us assign a class label to each grammatically characterized form. If the form is paradigmatically determinate, then its label is the number of the corresponding class. For example, задушивъ has the label 1, while издъхнѫ has the label 5. If the wordform is paradigmatically indeterminate, its label is a list of paradigmatic classes compatible with it. For example, дыхааше has the label 3/7. If the wordform is compatible with several classes and aberrant in one of them, then the class number where it is aberrant is shown in parentheses. For example, въздыхаѥте has the label (3)/7.

Let us create a table (see Table 912.1, p. 485) where rows contain paradigmatic classes and columns contain prefixes. In the cells we place the labeled forms from some megafamily, so that 1) a cell for a prefix and a class contains all corresponding forms, and 2) paradigmatically indeterminate forms are placed in the table several times—in all columns corresponding to some class symbol in the label. Let us call the resulting table the *unreduced distribution* of the megafamily.


Table 912.1. Unreduced distribution of the дъх megafamily ‹261›

The unreduced distribution table can be abbreviated by leaving out forms with identical labels in each group. Here and below we consider such abbreviated tables. The Table 912.2 is the abbreviated version of the unreduced distribution table of the дъх ‹261› megafamily.

Table 912.2. Abbreviated table of the unreduced distribution of the дъх megafamily ‹261›


Some columns of this table contain forms that are diagnostic for their class, while others do not. Columns that contain forms diagnostic for their class are called *independent*, as opposed to *dependent* columns. Independent columns represent paradigmatic classes that cannot be eliminated from a megafamily (in Table 912.2 these are the columns for classes 1, 3, and 5). At the same time, dependent columns can in the general case be eliminated, and their forms redistributed among independent columns. Eliminating a dependent column means removing some redundant lexemes from a megafamily. For example, in

Table 912.2, the dependent column for class 4 can be removed, and the form дъше 4/(5) remains in class 5, as aberrant for the class 4 aorist дъхнѫти. This eliminates a class 4 lexeme from this megafamily.

In constructing the reduced distribution, one must consider not individual megalexemes, but the entire megafamily. In creating the reduced distribution of a megafamily, the following criteria are suggested.

Minimizing the number of paradigmatic classes of a megafamily

Other things being equal, a reduced distribution with the smallest number of paradigmatic classes for a given megafamily is preferred.

Table 912.3 shows the unreduced megafamily ник ‹610› and its reductions according to Večerka's dictionary, and the dictionary in the present work (PD).


Table 912.3. The ник ‹610› megafamily

Classes 3 and 4 are dependent, while class 5 is independent. For this reason the reduced distribution needs only to show class 5.

Technically, this criterion requires eliminating dependent columns. However, in some megafamilies, different columns may be eliminated. Below this situation is described as *class competition*. Class competition is resolved according to the following criterion.

### Competing class hierarchy

If among two competing classes one can be seen as relatively new (open), while another as relatively old (closed), then, other things being equal, the paradigmatically indeterminate lexeme is preferably attributed to the open class. Open classes are 1, 5, and 7.

As an example, consider two megafamilies where classes 3 and 7 compete.



клик ‹383›

In these megafamilies, class 5 is represented by diagnostic forms (καινλιώλ 5, въскликняша 5), and should be present in the reduced distribution of these megafamilies. On the other hand, classes 3 and 7 are represented only by non-diagnostic forms. Thus, there is competition between 3 and 7. By the criterion of minimizing the number of paradigmatic classes of a megafamily, one of the competing classes must be excluded. By the competing class hierarchy criterion, class 3 is excluded as being closed, while all present forms by class 3 with substitutively softened stem are treated as aberrant class 5 forms, while other forms (with an expanded stem) as canonical class 7 forms.

As these examples show, while excluding all paradigmatically indeterminate columns is impossible, there should be no more than one such column.


Let us show one more megafamily with class competition.

Note that the criteria formulated above do not resolve all difficulties that arise with competing classes, and sometimes the dictionary author must make arbitrary decisions that are not motivated by precisely defined factors; cf., for example, the megafamily Hb3 <609>.

### § 913. Competition between classes 3 and 7, and 3~7 correlative pairs

Class 3 has no paradigmatically determinate forms at all: all PRAE system forms in the general case can be treated as aberrant class 5 forms, and all system IMF and INF-AOR forms as class 7 forms. According to the criteria formulated above, there should be no megalamilies with lexemes of both classes 3 and 7. This is not the case, however. Lexemes of classes 3 and 7 in one megatamily are possible if the representatives of those classes differ in root vocalism. The majori-

ty of such pairs are those where class 3 is represented by an irregular verb with unstable root vocalism, and in class 7 the root occurs in a lengthened grade relative to class 3 infinitive vocalism (see § 26). These cases are the following: ‹24› бьрати ≈ бирати, ‹220› дьрати ≈ дирати, ‹261› дъхати ≈ дыхати, ‹315› зьдати ≈ зидати, ‹323› зъвати ≈ зывати, ‹665› пьрати ≈ пирати, ‹666› пьсати ≈ писати, ‹883› стьлати ≈ стилати, ‹334› въз.ьмати ≈ въз.имати, ‹288› жьдати ≈ жидати. Other correlative pairs have a somewhat different structure: ‹973› трьѕати ≈ тръгати, ‹992› тъкати ≈ тыкати, ‹500› лобъзати ≈ лобызати, ‹512› лъгати ≈ лыгати, ‹545› метати ≈ мѣтати, ‹916› сълати ≈ сылати, ‹863› основати ≈ оснывати.

Table 913.1 on p. 489–490 shows the summary of the 3\*≈7 correlative pairs of the megafamily ‹334›.30

In all other cases class 3 and class 7 lexemes cannot be combined in one megafamily. In this case, competition between classes 3 and 7 is resolved as follows.

Class 3 PRAE forms with substitutively softened truncated stems are treated as diagnostic, as long as the megafamily does not contain diagnostic class 5 forms (with the suffix н). If the megafamily has diagnostic class 3 forms, then paradigmatically indeterminate 3/7 forms in any lexemes are assigned to class 3, and no class 7 lexeme is established. If there are no diagnostic class 3 forms, then paradigmatically indeterminate 3/7 and 3/5 forms are assigned to classes 7 and 5, respectively.

<sup>30</sup> This megafamily contains class 0 verbs (имѣти, имамь, and aberrant forms like имѣѭтъ), class 3\* verbs (възьмати [въз.ьм.а.т.и], въземл҄ѭ), class 4h verbs (възѧти, възьмѫ [въз.ьм.ѫ]), and class 7 verbs (възимати [въз.им.а.т.и], възимаѭ).


Table 913.1. Class 3 and 7 verbs of the megafamily หмати <334>


Table 913.1 (continued). Class 3 and 7 verbs of the megafamily имати <334>

As an example, consider Table 913.2 containing a table fragment for the вѧз ‹148› megafamily, which shows competition between classes 3 and 7. All PRAE class 3 forms are diagnostic, because there are no diagnostic class 5 forms, and forms labeled 4/(5) (only in the увѧз- megalexeme) are treated as class 4 forms, because that class contains the diagnostic INF-AOR system forms.31


Table 913.2. Lexeme type combinations in one megafamily

\* Večerka distinguishes увѧзати13 'bind, tie around' (увꙙзанъ Supr and увꙙзано Supr), and увѧзати2 7 'get bogged down, mixed up', the latter for the hapax gloss увѩзаѭтъ Ps Sin 9, 23.

Lexemes whose class 3 PRAE forms are not attested (such are, for example, the families °дѣлати, °гнѣвати, °копати), including incomplete lexemes where PRAE forms are not attested at all (°ласкати, °шѧтати), belong to the open class 7.

There are 85 verbs in class 3. The following seven verbs are a special case, in that an additional piece of evidence for their attribution to class 3 is the attestation of class 3 PRAE forms in Church Slavic manuscripts: изваꙗти, позобати, пискати, плѧсати, уръвати, основати, and шьпътати.32

Here are these verbs:

31 The megalexeme is увѧсти, represented by the following four forms: увѧстъ, увѧзошѧ, увѧзе, увѧзъши. In Večerka's dictionary these forms are distributed among the lexemes увѧзнѫти 5 and увѧсти 4.

32 Vaillant's grammar also classifies all of these verbs as class 3 on the basis of Church Slavic evidence, as does Lunt, except for the verb плѧсати.


# **Excursus on the morphology of personal forms of the imperfect**

### §914. Hiatus in imperfect forms

The systematically preserved hiatus in standard imperfect forms suggests loose paradigmatic adjacency between a string that we may call *Imf platform*, and the terminal. Then Imf forms have the skeleton of the type [нес.ѣ]+аше, and not нес+[ѣаше]. Such a morphological skeleton assumes a two-step synthesis of the form. On the first step, the Imf platform is built, which is identical to the expanded stem for verbs that have an expanded stem ending in а or ѣ, and for all other verbs is derived by adding the special appended ѣ to the truncated stem.33 For example, we have, on the one hand, platforms трьпѣ, плака, милова, дѣла, and on the other hand, л҄убл҄+ѣ (⇒л҄юбл҄ꙗ), двигн+ѣ (⇒двигнѣ), нес+ѣ (⇒несѣ), мог+ѣ (⇒можа), кльн+ѣ (⇒кльнѣ), зна+ѣ (⇒знаꙗ), сѣ+ѣ (⇒сѣꙗ), чу+ѣ (⇒чуꙗ). The addition of the appended ѣ results in the removal of the phonological hiatus according to the general rules (by adding an epenthetic *i̯*). On the second step, the resulting Imf platform is concatenated with the personal terminals of the imperfect of the аше type (they all begin with а). This addition of personal terminals is what results in loose adjacency, and is not accompanied by *i̯*epenthesis. We thus have: трьпѣ#аше, плака#аше, милова#аше, дѣла#аше; л҄юбл҄ꙗ#аше, двигнѣ#аше, несѣ#аше, можа#аше, кльнѣ#аше, знаꙗ#аше (знаꙗаше), сѣꙗ#аше (сѣꙗаше), чуꙗ#аше (чуꙗаше).

<sup>33</sup> In class 1 verbs the truncated stem has substitutive softening of the final C.

Within this interpretation, the participles of the imperfect system are separated from personal forms: in the participles the suffixes are attached not to the special Imf platform, but to ordinary workstems—to expanded ones, if they end in а or ѣ, or to truncated ones otherwise, observing the CVC agreement rule so long as the participial suffixes are initially ambivalent. For example, we have the following participles: л҄юбл҄.ьши, нес.ъши, кльн.ъши, but зна.въши, сѣ.въши, чу.въши, as well as трьпѣ.въши, плака.въши, etc.34

Note that the rules for building personal Imf forms proposed above are equivalent to those listed in the main part of this grammar, where the choice of ѣашеor аше-type terminals is made according to a morphological balance principle: [long stems + short terminals] or [short stems + long terminals] (see § 459). However, the rules listed here are not equivalent to imperfect formation rules proposed by Vaillant, § 159, Van Wijk, § 59, and Lunt 1974, § 9.1–6. There, Imf forms are built directly from the Inf form by means of a rule of "vowel agreement between infinitive and imperfect". Here is how Vaillant formulates this rule: «il est en -аахъ en regard de tous les verbes en inf. -ати, aor. -ахъ, quelle que soit leur flexion de présent: знаахъ 'je connaissais', inf. знати (prés. знаѥ-); глаголаахъ 'je parlais', inf. глаголати (prés. глагол҄е-); вѣроваахъ 'je croyais', inf. вѣровати (prés. вѣруѥ-) […]. Il est de même en -ꙗахъ en regard des verbes en inf. -ꙗти: авл҄ꙗахъ 'je manifestais', inf. авл҄ꙗти (prés. авл҄ꙗѥ-); стоꙗахъ 'je me tenais', inf. стоꙗти (prés. стои-). Il est en -ѣахъ en regard des verbes en inf. -ѣти, aor. -ѣхъ: умѣахъ 'je savais', inf. умѣти (prés. умѣѥ-) […]».

### §915. On the imperfects of the verbs знати and съмѣти

For the verb знати we find Imf знаꙗше in Sav and знааше in As (Mt 1, 25). The form знааше in As is easily interpretable as aberrant for знаꙗше, with the "loss of intervocalic *j*" aberration. Considering that in Sav all Imf forms are contracted, these spellouts can be considered contractions of знаꙗаше. This interpretation is supported by the form съмѣꙗше — Sav Mt 22, 46 (in Mar and As there is the aorist съмѣ, not found in Zogr), which is a contraction of съмѣꙗаше. It is easy to see that the forms знаꙗаше and съмѣꙗаше cannot be derived using the classical rule of vowel agreement between the infinitive and the imperfect.

In Sav, however, we find the form съмѣше (Jn 21, 12), which in the present grammar is treated as aberrant for съмѣꙗаше [съ.мѣ=ѣаше]. It is formed directly by a proportion like умѣти : умѣше (contracted imperfect to stan-

34 As is well known, personal imperfect forms are an innovation from a historical point of view, whose etymology remains obscure. In this sense, separating personal forms of the imperfect from participles agrees with historical facts. Note that the Imf platform, which is not directly comparable either with the infinitive or with the present stem, could be treated as the first component of an analytical construction, where the "conjugated element -ах" (as Meillet called it) plays the role of an auxiliary verb. If we accept the hypothesis that the iotated imperfects are older, in this conjugated element (now ꙗх.ъ, not ах.ъ) we could see the root aorist from ꙗх- (Inf ꙗти, ꙗхати 0; see § 565–569); cf. Meillet, § 293–297. As for the Imf platform, it is tempting to compare it with Russian *l*-less participles (cf. Zaliznjak 2004, § 3.39).

dard ум.ѣ=аше), съмѣти : съмѣше, which assumes treating the initial string съм as root: съм.ѣ.т.и by class 7 just like ум.ѣ.т.и 7. Analyzing съмѣти with the root мѣ is supported by the verb измѣти ‹586› from the same family (ізмѣтъ сѩ — 3SgAor Ps Sin 72, 21, cf. also ізмѣніе — Ps Sin 88, 52).

# CHAPTER 25 **Summary**

§916. Step-by-step construction of nominal wordforms1

0. The called wordform is specified by its paradigmatic call, an expression of the form *K*(*L*), where *K* is the name of some paradigmatic cell (the called property), and *L* is the name of some lexeme (the called lexeme).

1. Normally the called lexeme is given explicitly, but in the following four cases it is given indirectly, and the starting form must be constructed before proceeding further:

If *L* is a participle that is specified by a reference to the parent verb, then starting forms are built using verb synthesis rules; see § 918. Once the starting forms are built, the participles must be assigned a paradigmatic index: 2/a\* for щ- and ш-Part, and 2/a for all other participles.

If *L* is a supine that is specified by a reference to the parent verb, then the starting form is built using verb synthesis rules, and the supine terminal is ъ/ь (2-base); see also § 88.

If *L* is a comparative that is specified by a reference to the parent adjective, then starting forms are built using comparative formation rules, see § 919 and ff. Once the starting forms are constructed, the comparatives must be assigned the paradigmatic index 2/a\*\*.

1 As noted, any wordform that can be constructed by rules can also be constructed using profiles of type representatives from the given inflectional class. See profiles of nominal lexemes in Ch. 11.

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

If *L* is a pronoun with a clitic, then the graphic strings corresponding to clitics must be detached, resulting in the basic pronoun that is found in the paradigmatic dictionary. After the called wordform is built, the clitics should be reattached. Clitics that are subject to this procedure are: prefixes ни and нѣ, and postfixal clitics же, Же, жьдо, л҄юбо (see details in § 316).

2. If the called lexeme is given explicitly, it must be located in the paradigmatic dictionary. Note its paradigmatic index *I* and the morphophonological representation of the starting form.

3. If the index *I* has the shape 0/*x*, the lexeme is unique, and the called form can be found in the full paradigm of the lexeme, or built using a sample lexeme. See Ch. 12, *Unique nominal lexemes*, Table 357. Otherwise proceed further.

4. Build the workstem by discarding the inflection from the mph representation of the starting form. If the string ends in a vowel, add the symbol *i̯*.

5. If the called form is secondary, namely vocative (Voc) or personal dative (D2), see construction rules in § 355, 356; if the called form is an adprepositional form of the pronominal lexemes \*и or иже, construction follows the paradigms given in § 318.2

6. For all other (primary) forms, find the standard terminal set that corresponds to the paradigmatic index of the called lexeme using Table 302. See § 289 for the catalog of nominal terminals.

7. In the terminal set, find the cell that corresponds to the called form. If the cell has several variants, find the needed variant using the twofold rule (§86) depending on the right edge of the workstem.


8. For marginal subclasses, introduce the necessary corrections:

9. Build the morphological skeleton of the form workstem + selected terminal.

10. Apply boundary adjustment rules, using § 288.

11. Using mph⇒ph/norm rules (see § 63 and ff), rewrite the morphophonological representation into a phonological one.

2 See the full list of secondary forms for nominals and verbs in § 880.

§917. Step-by-step construction of verbal wordforms3

0. The called wordform is specified by its paradigmatic call, an expression of the form *K*(*L*), where *K* is the name of some paradigmatic cell (the called property), and *L* is the name of some lexeme (the called lexeme).

1. Find the called lexeme *L* in the paradigmatic dictionary. Note its paradigmatic index *I* (the verb class is the first digit of the index) and the morphophonological representation of the starting form.

2. If the called form is a secondary aorist, construction follows the rules in § 477–480. Otherwise proceed.

3. If the index *I* contains the symbol ∇, the paradigm contains anomalous forms. Check the form *K*(*L*) against lists of anomalous forms in § 463 (PRAE), § 466 (IMF), or § 881, and if found, build it according to instructions found there. Otherwise proceed.

4. If the index *I* begins with 0, the lexeme is unique, and the called form can be found in the full paradigm or profile of the corresponding prefixless verb. See Ch. 21, *Unique verbs*, Table 516. Otherwise proceed.

5. Determine whether the called lexeme is a regular or irregular verb. If the index contains the symbols \*, °, •, h, ⩨, ⤸, or ⤹, then the verb is irregular, and regular otherwise.

6. Use the called property *K* to determine the system to which the called form belongs, and whether the called form is nominal (infinitive, supine, participle) or finite (otherwise). See § 405, Table 405.

7. Determine the basic stem (i.e. determine whether the expanded or truncated basic stem is needed) for the given system within the given class, using Table 427 (basic stem allotment rules).

8. Find the basic stem using the morphophonological representation of the lexeme using the rules in § 431.

9. Build the workstem for the subparadigm; see § 432 if the verb is regular. If the verb is irregular, use Table 440 (*Workstems of irregular verbs*) and the lists of key forms of irregular verbs, § 434.

<sup>3</sup> As noted, any wordform that can be constructed by rules can also be constructed using profiles of type representatives of the corresponding class and subclass. See profiles of regular verbs in § 429, and profiles of irregular verbs in § 454.

10. Find the standard terminal or suffix set for the called subparadigm using the terminal catalog in § 455. The PRAE system has several sets; choose the set by conjugation class: *i*-conjugation uses classes 1 and 2, *e*-conjugation uses other classes.

11. In the terminal set, find the cell that corresponds to the called wordform. If there are several variants, determine the correct variant: using the morphological balance rule (§ 459) if the called form is a personal imperfect form; otherwise, using the CVC agreement rule (§ 93) or the twofold rule (§ 86).

12. Build the morphological skeleton of the form workstem + selected terminal of a finite form or suffix of a nominal form.

13. Apply boundary adjustment rules using § 462. The result is a morphophonological representation of the called form.

14. Using mph⇒ph/norm rules (see § 63 and ff), rewrite the morphophonological representation into a phonological one.

15. The result is the called form in case the called form is finite. For nominal form, the result is the stem, and the form must be completed using nominal form rules, see § 916 above.

§918. An overview of participles

Here are all the participles:


Participle formation rules

The following suffixes are added to the workstem of the corresponding system:


\* The slash separates twofold rule variants.


\* The slash separates twofold rule variants. The tilde separates CVC agreement rule variants. \*\* Before the suffix ен, the phonemes к, г, х are replaced by ч, ж, ш; ов.ен is only found in class 5 verbs.


Cluster rewrite rules apply for C-final stems, e.g. ΜΕΤ.Λ > ΜΕΛ, ΒΑΔ.Τ > ΒΑΥΤ; note that for the irregular verbs in the groups мръти 4h\* (\$ 451), влъшти 4c\* (\$ 452), and чисти 4c\* (§ 453) the workstem of ¿- and ™-Part has the vocalic realization as in «Prae», distinct from the vocalic realization in «Inf».

# Peculiarities in the declension of шт- and ш-participles (class 2/a\*)

### 1. Distribution of syncopated and expanded stems:


2. Special terminals:


Sample:


# Comparative construction rules

# §919. Two types of comparatives: new and old

Comparatives are distinguished as old and new, which are two wordformation strategies (see below), where the new one is productive, while the old one is used in a limited set of adjectives. Because comparatives are represented by a small number of occurrences, it is difficult to establish with certainty their distribution in canonical OCS. Most likely, for certain adjectives only the old model is found. In some cases the comparative is only found in adverbial forms, as e.g. coyie (1x SUPR 394, 11, the old comparative, but the new comparative coyven 9×, including 4× adjectivally). In some cases the sources show an old and a new comparative for the same adjectives; arbitrarily, we treat new comparatives as aberrant in such cases (see § 396).

# §920. Construction of starting forms: markers4

As noted above, for comparatives two starting forms are built: syncopated NSgmBrev, and expanded NSgfBrev. These forms result from attaching the following markers to the parent adjective workstem, as shown in Table 920. On *comparativum tantum* lexemes see § 281.


Table 920. Comparative markers

In case the parent A-lexeme is a suffixed adjective with suffixes ок (глѫб.ок.ъ), ък (слад.ък.ъ), ьк (гор҄.ьк.ъ), the workstem for the old comparative is suffixless, cf. глѫб.ок.ъ — глѫбл҄ь, глѫбл҄ьши.

### §921. Construction of starting forms: boundary adjustment rules

In the new comparative, к→ч replacement by the velar palatalization alternation. In the old comparative, C→C• replacement by the substitutive softening alternation.

# Examples

Old comparative. грѫбъ: грѫбл҄.ь, грѫбл҄.ьш.и; слад.ък.ъ: слад + ь ⇒ слаЖ.ь ⇒ слаЖь, слаЖ.ьш.и ⇒ слаЖьши.

New comparative. новъ: новѣи, новѣиши; гор҄ькъ: гор҄ьк + ѣj.ь ⇒ гор҄.ьч.ѣj.ь ⇒ гор҄ьчаи, гор҄.ьч.ѣj.ьш.и ⇒ гор҄ьчаиши.

### §922. Old comparatives

Lists of old comparatives attested in OCS sources are given in Table 922 on p. 501, with a note on authors of OCS grammars. The PD lists as separate lexemes *comparativum tantum* (noted by + in Table 922), while the symbol [+] notes that the PD only lists the parent lexeme.

For глѫбл҄ь cf.: глѫбьшаꙗ Supr 464, 4; for наивѧщь: наивꙙщеSupr 201, 20; for прѣвышь: прѣвышьи Supr 469, 1; for наитрѣбл҄ь: наитрѣбьши Supr 339, 30; for тврьЖь Vaillant notes the form отврьзыCloz 11b, 14 (see Vaillant, §89).

Diels, as well as and Lunt 1974, note the lexeme рачии for рачъшѫѭ Cloz 2b, 23. Večerka treats this form as ш-Part for рачити, and Vaillant apparently does the same. In the PD, following Večerka, this form is treated as ш-Part.

Cf. also such adverbial forms as паче, ниже, драже (cf. Euch 44a, 15) and others.

<sup>4</sup> I.e. sequences that consist of a suffix and a terminal, a single suffix, or a single terminal.


Table 922. Old comparative forms

### § 923. A note on comparative spellouts in sources

Due to graphic/phonological aberrations, spellouts ending in -nn, -- nn are often replaced by abbreviated spellouts with -и, -ъп. For new comparative forms, this results in the replacement of the canonical (NASgmPlen) ending in -ни (cf. HOB'EHH) by an aberrant form ending in -n (cf. HOB'EH). Likewise, for old comparative forms this results in the replacement of the canonical (NASgmPlen) ending in -ни (cf. гржвлин) by an aberrant form ending in -n (cf. гржки). For new comparative forms this eliminates the Brev ~Plen opposition in direct cases in masculine singular: there is just one wordform of the type мовъи. Most likely, old comparatives also eliminate this opposition, where canonical (NASgmBrev) of the type гржель are not attested.

For these reasons, grammars note forms like грждин as canonical for (NASgmBrev) (i.e. the equivalence of long and short forms is postulated for these cells). Dictionaries follow the same principle. In this grammar, starting forms for comparativum tantum are given following the traditional approach (as in Večerka's dictionary), but the grammar sets up canonical paradigms that do distinguish short and long forms in these cells.

So in Vaillant, § 82, and Lunt 1974, § 4.19.

The contamination of spellouts in -ни/-и and -ъии/-ъв in sources results in the corresponding effects in direct cases of the singular in ш- and шт-participles, and due to standard segmental changes, the forms (NSgmPlen) | wr-Part) (cf. HEC.BIg. - HECBIH) and (NSgmPlen) | w-Part (cf. HEC.B.b - HEC'SIM) coincide for the corresponding verb classes. The choice of canonical forms in this grammar follows the standard approach for participles.6

### §924. Illustrations

егда нечистъи дуъ" сзидетъ отъ чка: пръходитъ сквозъ вездънаа мъстат ишта поков: и не овретаю глетъ: възврашти са въ домъ моготъ некудоуже 13ИДЪ" 1 ПОНШЬДЪ ОБРЪТАЕТЪ" ПОМЕТЕН'' ОУКРАШЕНЪ' ТЪГДА 1ДЕТ''' 1 ПОСМЕТ'' ДОУГЪРУ" ГОРЬШЬ СЕБЕ СЕДМЬ' І ВЪШЬДЪШЕ ЖИВЖТ" ТОУ' І БЪІВАЖТ"В ПОСЛЕДЬЯ ВА чкоу томоу горьши пръвъюхъ Lk 11, 24-26 ZOGR;

Егда крепъкъ огоряжь са уранитъ свои дворъ въ миръ сятъ имънить его, а понеже кръплен его нашедъ побъдитъ и и въсъ оряжиъ его отъметъ на нъже оупъвааше и користь его раздаатъ Lk 11, 21-22 мая;

МЕДЪ БО КАПЛЕТЪ ОТЪ ОУСТЬНОУ ЖЕН'Ъ! ЕЛЖДЫНЪ!" ГАЖЕ ВЬ ВРЪМА ПОМАЖЕТ''В ГОВТАНЬ ТВОН: НЪ ПОСЛЕЖДЕ ГООБЧАЕ ЗЛЬЧИ ОБОДШТЕШИ: И ОСТОВК ПАЧЕ МЕЧА обождоу остра SUPR 350, 30-351, 4;

ЕЛАЖЖ "WEHФА И НИКОДИМА" БЪКТЕ БО ПРЕЖДЕ ХЕРОВИМЪ ХЕРОВИМЪНША БОГА Bb CEE'S HOCALITA SUPR 458, 4-5.

<sup>6</sup> So in Vaillant, § 83-84, and Lunt 1974, § 4.19.

# CHAPTER 26 **Chrestomathy**

# **Normalized texts**

The texts given below are slightly modified normalizations created for pedagogical reasons by Andrej Zaliznjak for his classes in OCS held at Moscow State University in the 1970s (therefore, quoted as MSU70 here and in the footnotes). Changes to Zaliznjak's original constructions bring the texts into line with the canon described in this book.

The most salient regular difference between these texts and MSU70 has to do with the writing system. MSU70 does not use the letters ꙗ and ѥ; accordingly, MSU70 uses ѣ in place of /a/ after /j/ and after kamorated consonants (cf. ѣко, моѣ, земл҄ѣ vs. ꙗко, моꙗ, земл҄ꙗ using the normalization adopted here; see § 33–34). There are some other minor differences in the writing system. In rendering borrowings that are not found in PD, the spellouts below follow those in MSU70. Such spellouts inevitably contain letters from the expanded Cyrillic alphabet; see § 131–132. Cf. below spellouts like ћетьсимани, каиꙗфѣ and others; cf. голъгота, преторъ, евреискы.

In some cases MSU70 texts contains forms that are here treated as aberrant, as e.g. съблажн҄ѭ for canonical съблазн҄ѭ. Most such forms are noted in the commentary.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

1.

Mt 26, 31–75

[31] тъгда глагола имъ исусъ· вьси вы съблазните сѧ о мьнѣ въ сиѭ нощь· пьсано бо ѥстъ поражѫ пастыр҄ꙗ и разидѫтъ сѧ овьцѧ стада· [32] по въскрьсновении же моѥмь вар҄ꙗѭ вы въ галилеи· [33] отъвѣщавъ же петръ рече ѥму· аще и вьси съблазнѧтъ сѧ о тебѣ· азъ николиже не съблазн҄ѭ(a) сѧ· [34] рече ѥму исусъ· аминь глагол҄ѭ тебѣ· ꙗко въ сиѭ нощь прѣЖе даже кокотъ не възгласитъ· три краты отъврьжеши сѧ мене· [35] глагола ѥму петръ· аще ми сѧ кл҄ючитъ съ тобоѭ умрѣти· не отъврьгѫ сѧ тебе· такоЖе и вьси ученици рѣшѧ· [36] тъгда приде исусъ въ вьсь наричемѫѭ(b) ћетьсимани· и глагола ученикомъ сѣдите ту· доидеже шьдъ помол҄ѭ сѧ тамо· [37] и поимъ петра· и оба сына зеведеова· начѧтъ скръбѣти и тѫжити· [38] тъгда глагола имъ исусъ· прискръбьна ѥстъ душа моꙗ до съмрьти· пожидѣте сьде и бъдите съ мъноѭ· [39] и прѣшьдъ мало паде ниць молѧ сѧ и глагол҄ѩ· отьче мои аще възможьно ѥстъ да мимоидетъ отъ мене чаша си· обаче не ꙗко азъ хощѫ· нъ ꙗкоже ты· [40] и въставъ отъ молитвы приде къ ученикомъ и обрѣте ѩ съпѧщѧ· и глагола петрови· тако ли не възможе ѥдиного часа побъдѣти съ мъноѭ· [41] бъдите и молите сѧ да не вънидете въ напасть· духъ бо бъдръ а плъть немощьна· [42] пакы въторицеѭ шьдъ помоли сѧ глагол҄ѩ· отьче мои аще не възможетъ чаша си мимоити отъ мене· аще не пиѭ ѥѩ· бѫди вол҄ꙗ твоꙗ· [43] и пришьдъ пакы обрѣте ѩ съпѧщѧ· бѣсте бо очи имъ тѧжьцѣ· [44] и оставл҄ь ѩ пакы шьдъ помоли сѧ третиицеѭ·(c) тъЖе слово рекъ· [45] тъгда приде къ ученикомъ· и глагола имъ съпите прочеѥ и почиваѥте· се приближи сѧ година· и сынъ чловѣчьскыи прѣдаѥтъ сѧ въ рѫкы грѣшьникомъ· [46] въстанѣте идѣмъ· се приближи сѧ прѣдаѩи мѧ· [47] и ѥще глагол҄ѭщу ѥму· се июда ѥдинъ отъ обою на десѧте приде· и съ н҄имь народъ мъногъ· съ орѫжии и дрькольми· отъ архиереи и старьць л҄юдьскыихъ· [48] прѣдаѩи же и дастъ имъ знамениѥ глагол҄ѩ· ѥгоже аще лобъжѫ тъ ѥстъ имѣте и· [49] и абиѥ пристѫпл҄ь къ исусови рече ѥму· радуи сѧ равьви· и облобыза и· [50] исусъ же рече ѥму· друже на н҄ѥже ѥси пришьлъ твори· тъгда пристѫпл҄ьше възложишѧ рѫцѣ на исуса и ѩсѧ и· [51] и се ѥдинъ отъ сѫщиихъ съ исусомь· простьръ рѫкѫ извлѣче ножь свои· и удар҄ь раба архиереова· урѣза ѥму ухо· [52] тъгда глагола ѥму исусъ· възврати ножь свои въ своѥ мѣсто· вьси бо приимъшеи ножь ножемь погыбнѫтъ· [53] ли мьнитъ ти сѧ ꙗко не могѫ нын҄ꙗ умолити отьца моѥго· и приставитъ мьнѣ вѧще нежели дъва на десѧте лећеона анћелъ [54] како же убо събѫдѫтъ сѧ кън҄игы·

<sup>(</sup>a) MSU70 съблажн҄ѭ. (b) MSU70 нарицаѥмѫѭ, i.e. by class 7; likewise below. (c) MSU70 третицеѭ.

гако тако подовааше въ ги: [55] въ тъ часъ рече исоусъ народомъ: тако НА РАЗБОННИКА ЛИ ИЗИДЕТЕ СЪ ФОЖЖИЕМЬ И ДОБКОЛЬМИ ЊАТИ МА: ПО ВЪСА дьни съдъахъ при васъ въ црькъве и не њете мене. | 56 се же въсе бълстъ" ДА СЪЕЖДЖТЪ СА КЪНИГЪІ ПРОДОЧЬСКЪМ»: ТЪГДА ОУЧЕНИЦИ ВЬСИ ОСТАВЛЬШЕ и бъжаша: |57 они же имъше исоуса въса къ кангафъ архнереови: ИДЕЖЕ КЪНЖЕНИЦИ И СТАРЬЦИ СЪБЪРАША СА: 58 ПЕТРЪ ЖЕ ИДЪАШЕ по немь издалече. до двора архиереова: и въшьдъ вънжтрь съдъяше съ слоугами: видъти коньчиня: |59 архнерен же и старьци: и съньмъ ВЬСЬ ИСКААХЯ ЛЪЖА СЪВЪДЪТЕЛЬСТВА НА ИСОУСА ГАКО ДА ОУБИНЖТЪ И [60] и не овретя: мъногомъ лъжемъ съвъдътелямъ пристяпльшемъ" последь же пристипающа дъва лъжа съведчетелга | 61 ръсте съ рече МОГЖ РАЗОЏТТИ ЦОЬКЪВЬ БОЖИИЖ: И ТОЬМИ ДЬНЬМИ СЪЗЬДАТИ НА: | 62 И ВЪСТАВЪ Архиерен рече ѥмод: ничесоже ли не отъвърштавањши: чьто сни на та съвъдътельствориятъ: 6) [63] исоусъ же мльчааше: и отъвъштавъ архиерен рече кмог. заклинање та вогомь живъима суда речешн намъ' аште тъ есн христосъ същъ вожни. [64] глагола немор исоусъ тъп рече обаче глагольк вамъ отъ сель оузьрите съща чловъчьскаюто (1) СЪДАШТА О ДЕСНЯЮЖ СИЛЪГ И ГОЖДЖШТА НА ОБЛАЦЪХЪ НЕБЕСЬСКЪНУЋЪ 65 | ТЪГДА АРХНЕРЕН РАСТЪЬЅА рИЗЪІ СВОЊ ГЛАГОЛНА" ГАКО ВЛАСВИМИЯ» ОЕЧЕ ЧЬТО КШТЕ ТОВЕОУКИЪ СЪВЪДЪТЕЛЬ: СЕ НЪИНА СЛЪНШАСТЕ ВЛАСВИМНИЯ ИГО" [66] чьто са вамъ мынитъ: они же отъвъштавъше ръша: повиньнъ IECT"В СъМръти" | 67 | T"ВГДА ЗАПЛЪВАША ЛИЦЕ IEГО" И ПАКОСТИ КАМОУ ДЪГАША" ОВИ ЖЕ ЗА ЛАНИТИ ОСДАОНША [68] ГЛАГОЛИЯШТЕ: ПРОДЬЦИ НАМЪ ХОНСТЕ къто естъ оударии та: [69] петръ же вънъ съдъаше на дворъ: н пристяпи къ немог еднна рабъются глагодияшти. и тър въ съ исоусомъ галиленскъюмь: [70] онъ же отъврьже са пръдъ въсъми глагодна: не ВЪМЬ ЧЬТО ГЛАГОЛЕШИ: 771 ИШЬДЪШОУ ЖЕ ЕМОУ ВЪ ВРАТА: ОУЗЬОВ И дрогтага: и глагола ниъ тоу и сь въ чловъкъ съ исоусомъ назадаминомъ: [72] и пакър отъврьже са съ клатвоех: тако не знаия чловъка: [73] не по МЪНОՏЪ ЖЕ ПОИСТЯПЛЬШЕ СТОЊШТЕН ОВША ПЕТООВИ' ВЪ ИСТИНЯ И ТЪГ отъ ннуъ еси: нео и бесъда твога авъ та творитъ: [74] тъгда начатъ ротити са и клати са гако не знаеж чловъка: и авни кокотъ възгласи" [75] и поманж петръ глаголъ исоусовъ нже рече ключъ гако пръжде даже КОКОТЪ НЕ ВЪЗГЛАСИТЪ: ТОН КОАТЪР ОТЪВФЪЖЕШИ СА МЕНЕ: И ИШЬДЪ ВЪНЪ плака са горько

<sup>(</sup>a) Wordforms of the lexeme съвъдътельство without kamora in MSU70.

<sup>(</sup>b) Wordforms of the lexeme съвѣдътъ?ьствовати without kamora in MSU70

<sup>(</sup>c) MSU70 живъмь: aberrant spellout for Plen (see § 394); likewise below.

<sup>(</sup>d) MSU70 чловъчьскааго: aberrant spellout for Plen (see § 394); likewise below.

# 2.

### Mt 27, 1-5

[1] оутроу (3) же съвъшоу: съвътъ сътвориша въси архнерен и старьци дюдьстин(6) на исоуса тако оувити на [2] и съвазавъще и въсл н предаша и поньтьскоучемом (6) пилатоу ићемоноу. [3] тъгда видувавъ ИЮДА ПОЪДАВЪІН ЕГО" ГАКО ОСЖДИША Н' РАСКАГАВЪ СА ВЪЗВОАТИ ТЪН ДЕСАТИ сьоеврьникъ: архнереомъ и старьцемъ: [4] глагольм: съговшихъ поъдавъ кръвь неповиньня: они же ръша: чьто естъ намъ: тъв офзьриши: [5] и ПОВОЫ" СЪРЕБОО ВЪ ЦОЬКЪВЕ ОТИДЕ" И ОШЬДЪ ВЪЗВЪСН СА:

### Jn 18, 29-40

29 ИЗИДЕ ЖЕ ПИЛАТЪ КЪ ЙИМЪ ВЪНЪ: И рече: КЖИХ ОВУЧЬ ПОННОСИТЕ НА чловъка сего: [30] отъвъшташа же и овша кмол: аште не ви къму съ зълодън: не вниъ предали его тевъ: [31] рече же ниъ пилатъ: понитете Н ВЪ!' И ПО ЗАКОНОУ ВАШЕМОУ СЖДИТЕ КМОУ' ОВША ЖЕ ЕМОУ НЮДЕН' НЕ достоитъ намъ оувити никогоже: [32] да слово исоусово съеждетъ сл. кже рече клеплья конеж съмрьтия хотъемше оумрети. [33] въниде ЖЕ ПАКЪІ ПИЛАТЪ ВЪ ПОЕТООЪ' И ВЪЗЪВА ИСОУСА И ОЕЧЕ ЕЛИОУ' ТЪБ ЛИ ЈЕСН цѣсадь нюденскъ [34] отъвъшта кмоу исоусъ: о себъ ли тъ глаголеши се ли ини тебъ ръша о мьюъ: [35] отъвъшта пилатъ исда азъ июден IECME: POAZ TBOH H ADXHEDEN ПОВДАША ТА МЫНЪ' ЧЬТО ЈЕСИ СЪТВООНАЪ' 36] отъвъшта нсоусъ: цесарьство монентъ отъ сего мира: аште отъ СЕГО МИРА ЕН БЪЛЛО ЦЕСАОЬСТВО МОИ: СЛОУГЪР ОУБО МОЊА ПОДВИЗАЛУЫ СА БИША: ДА НЕ ПОЕДАНЪ БИМЬ ИЮДЕОМЪ: НЪМА ЖЕ ЦЪСАДЪСТВО МОЕ НЪСТЪ отъ сждоу: | 37 | рече же юмоу пилатъ: оуво цесарь ли есн тъвъ отъвъшта нсоусъ тър глаголюши гако цъсарь ксмь азъ азъ на се роднуъ сах и на се придъ въ вьсь миръ. да съвъдътельство, истинъ. въсткъ иже IECT"B OT"B HCTHN'SI ПОСЛОУШАЮТ" ГЛАСА МОНЕГО" | 38 ГЛАГОЛА КЕМОУ ПИЛАТ"В" ЧЬТО ЕЕСТЪ ИСТИНА" И СЕ ОЕКЪ ПАКЪР ИЗИДЕ КЪ ИЮДЕОМЪ' И ГЛАГОЛА НАУЪ" АЗЪ НИ ЕДИНОВА ВИНЪ НЕ ОБРЪТАНЖ ВЪ НЕМЬ: |39 ЕСТЪ ЖЕ ОБЪЧАИ ВАМЪ ДА IEДHHOID BAM'S OTTEROYWTH HA ПАСХЖ' YOUT'ETE AN OVEO ДА OT'SПОУШТ'Я вамъ цесарга нюденска: |40 възъпнша же въси глагойскште: не сего НЪ ВАРААВЖ: БЪ ЖЕ ВАРААВА РАЗЕОННИКЪ.

Jn 19, 1-15

[1] тъгда же пилать покать исоуса: и вни: [2] и вомни съплетъше въвньць ОТЪ ТОЬНИГА" ВЪЗЛОЖИША НА ГЛАВЖ ΈΝΟΥ" И ВЪ РИЗЖ ПРАПОЖДЪНЖ ОБЛЪША и: |3| и глаголаахж: радоун са цесарю нюденскъ: н кнгадж и по ланитама: 4 | ИЗИДЕ ЖЕ ПАКЪІ ПИЛАТЪ ВЪНЪ: И ГЛАГОЛА ИМЪ' СЕ ИЗВОЖДЖ И ВАМЪ ВЪНЪЪ да разоуменете гако въ немь винъ не овречата ни кампова. | 5 изиде

<sup>(</sup>а) MSU70 ютроу.

<sup>(</sup>b) MSU70 âюдьсции (see § 111).

<sup>(</sup>c) MSU70 поньтьскоүмму; the headword in Večerka lacks ь: понтьскъ, although the sequence нт is a prohibited cluster (see § 59).

же исоусъ вънъ носа трьновъ въньць и прапождьня ризж: и глагола ИМЪ СЕ ЧЛОВЪКЪ' | 6 | ЕГДА ЖЕ ВИДЪША И АРХНЕОЕН И СЛОУГЪ!" ВЪЗЪПИША ГЛАГОЛЕЖШТЕ: ПООПЬНИ И ПРОПЬНИ И: ГЛАГОЛА ИМЪ ПИЛАТТЪ: ПОНАТЕТЕ ВЪ! Н пропенете: а.зъ бо не обръстаня въ немь винъ! 7 отъвъшташа кмоу НЮДЕН' МЪГ ЗАКОНЪ ИМАМЪ И ПО ЗАКОНОУ НАШЕМОУ ДЛЪЖЬНЪ ЕСТЪ ОУМОВТИ' ГАКО СЪЩЪ БОЖИИ ТВОДИТЪ СА: | 8 | ЕГДА ЖЕ СЛЪЩИ ПИЛАТЪ СЕ СЛОВО" паче оувога са |9| и въниде въ преторъ пакъг и глагола исоусови" отъ кждоу кси тъю исоцсъ же отъвъта не сътвори клюу. [10 глагола њиоц ПИЛАТ'В' МЬНЪ ЛИ НЕ ГЛАГОЛЕШИ' НЕ ВЪСН ЛИ ГАКО ВЛАСТЪ ИМАМЬ ПООПАТИ ТА: И ВЛАСТЬ ИМАМЬ ПОУСТИТИ ТА: |11| ОТВВЪШТА ИСОУСЪ: НЕ ИМАШИ ОБЛАСТИ НА МЬНЪ НИКОЮЊЖЕ: АШТЕ НЕ ЕН ТИ ДАНО СЪ ВЪЩЕ: СЕГО ДАДИ Предавъти ма тевъ болин гръхъ иматъ. |12 отъ того пилатъ искааше ПОУСТИТИ И НЮДЕН ЖЕ ВЪПИГААХЯ ГЛАГОЛЕЖШТЕ: АШТЕ СЕГО ПОУСТИШИ НЪСН ДООУГЪ КЕСАДЕВИ: ВЪСЪКЪ ИЖЕ СА ТВОДИТЪ ЦЕСАОЬ: ПООТИВИТЪ СА кесариевн: |13 | пилатъ же слъщавъ та словеса: изведе вънъ нсоуса" н съде на сядништи: на месть наричемъемь литостротъ: євренскъ же гаваата (6) [14] въ же парасќевъкин пасть. С година въ гако шестапа и глагола июдеомъ се цесафь вашь: |15| они же въпигаахж: възьми ВЪЗЬМИ Пропьни и глагола имъ пилатъ. цесары ли вашего пропьня: отъвъшташа архнерен: не имамъ цъсарга: тъкъмо кесарга:

#### Mt 27, 23-36

23 икемонъ же рече ниъ: чьто зъло сътвори: они же излиха въпитахъ глагодияште: да пропатъ сждетъ: [24] видъвъ же пилатъ гако ничесоже ОУСПЕКТЪ' НЪ ПАЧЕ МЛЪВА БЪВАЕТЪ' ПОНИМЪ ВОДЖ ОУМЪ ОЖЦЪ ПОВДЪ народомь глагольмъ неповныхъ есль отъ кръвна) сего правьдьнаюточе) във оузърите: [25] и отъвъштавъще въсн дюдне ръша: кръвь кго на насъ н на чадъуъ нашихъ: [26] тъгда отъпоусти имъ вараавя: исоуса же бивъ пръдастъ илъ да и пропьнятъ: [27] тъгда воини ићемонови: пръимъше исоуса на сждиншти: съвераша на нь вьсх спиря: 28 и съвлькъше и хламидовъ чрьвленом одъща и [29] и съплетъше въкъщь отъ трьника ВЪЗЛОЖНША НА ГЛАВЖ ЕГО: И ТОЬСТЪ ВЪ ДЕСНИЦЯ ИГО: И ПОКЛОНЬШЕ СА НА кольног предъ нимь: рягаахъ са кмоу глагольжште: радоун са цесарю нюденскъ: |30 и планнявъше на нь: прикася трьстъ: и внгаахж и по главъ: [3] и кгда порягаша са кмоу съвлеша съ него уламидж: и облъщая и въ ризъл свою: и въсл и на пропатик: [32] исходяште же овръчти ЧЛОВЪКА ЌУрИНЕНСКА: ИМЕНЕМЬ СНМОНА: СЕМОУ ЗАДЪЩА ПОНЕСТИ КОБСТЪ ИГО"

<sup>(</sup>a) See Footnote (b) on p. 504.

<sup>(</sup>b) Večerka гаввата, see distribution by sources ibid.

<sup>(</sup>c) MSU70 пасць (see § 111).

<sup>(</sup>d) MSU70 кръвє (see § 403).

<sup>(</sup>e) See Footnote (d) on p. 505.

[33] и пришьдъше на мѣсто наричемоѥ(a) голъгота· ѥже ѥстъ наричемо(b) краниѥво мѣсто· [34] дашѧ ѥму пити оцьтъ съ злъчиѭ(c) съмѣшьнъ· и въкушь не хотѣаше пити· [35] пропьнъше же и· раздѣлишѧ ризы ѥго· метѫще(d) жрѣбиѩ· [36] и сѣдъше стрѣжаахѫ и ту·

Jn 19, 19–22

[19] напьса же и титьлъ пилатъ· и положи и на крьстъ· бѣ же напьсано· исусъ назареи цѣсарь июдеискъ· [20] сего же титьла мъноѕи чисѧ отъ июдеи· ꙗко близъ бѣ мѣсто града· идеже пропѧсѧ исуса· и бѣ напьсано евреискы грьчьскы и латиньскы· [21] глаголаахѫ же пилатови архиереи июдеистии·(e) не пиши цѣсарь июдеискъ· нъ ꙗко самъ рече· цѣсарь ѥсмь июдеискъ· [22] отъвѣща пилатъ· ѥже пьсахъ пьсахъ·

Mt 19, 38–52

[38] тъгда пропѧсѧ съ н҄имь дъва разбоиника· ѥдиного о деснѫѭ· и ѥдиного о шуѭѭ· [39] мимоходѧщеи же хул҄ꙗахѫ и· покываѭще главами своими· [40] и глагол҄ѭще· ува· разар҄ꙗѩи црькъвь· и трьми дьньми съзидаѩ ѭ·съпаси себе· ащесынъ божии ѥси·сълѣзи съ крьста· [41] такоЖе же и архиереи рѫгаѭще сѧ· съ кън҄ижьникы и старьци и фарисеи глаголаахѫ· [42] ины съпасе· себе ли не можетъ съпасти· аще цѣсар҄ь издраил҄ѥвъ ѥстъ· да сълѣзетъ нын҄ꙗ съ крьста· и вѣрѫ имемъ ѥму· [43] упъва на бога да избавитъ и нын҄ꙗ· аще хощетъ ѥму· рече бо ꙗко божии сынъ ѥсмь· [44] тоЖе же и разбоиника пропѧтаꙗ съ н҄имь поношаашете ѥму· [45] отъ шестыѩ же годины тьма быстъ по вьсеи земл҄и· до девѧтыѩ годины· [46] при девѧтѣи же годинѣ· възъпи исусъ гласомь велиѥмь глагол҄ѩ· ели ели лема савахтани·(f) ѥже ѥстъ боже мои боже мои· въскѫѭ мѧ ѥси оставилъ· [47] ѥтери же отъ стоѩщиихъ(g) ту· слышавъше глаголаахѫ· ꙗко илиѭ зоветъ· [48] и абиѥ текъ ѥдинъ отъ н҄ихъ· и приимъ гѫбѫ· испльн҄ь оцьта· и възньзнѫвъ(h) на трьсть· напаꙗаше и· [49] а прочии глаголаахѫ· остани да видимъ· аще придетъ илиꙗ съпастъ ѥго· другыи же приимъ копиѥ· прободе ѥму ребра· и изиде вода и кръвь· [50] исусъ же пакы възъпивъ гласомь велиѥмь· испусти духъ· [51] и се катапетазма црькъвьнаꙗ раздьра сѧ· съ вышьн҄ꙗѥго(i) краꙗ до нижьн҄ꙗѥго(j) на дъвоѥ· и земл҄ꙗ


потрасе са: и каменне распаде са: [52] и грови отвръса са: и мънога ТЪЛЕСА ПОЧИВАЊШТИНУЪ СВАТЪНУЋ ВЪСТАША:

# Texts from the original manuscripts

### Psalterium Sinaiticum

#### Ps 90

хвала пънгъ дава: [1] живял въ помошті въшьнъго.(4) въ кровъ Ба НЕСЬНАЕГО ВЪДВОРІТЪ СЊУ [2] РЕЧЕТЪ БОУ ЗАСТЯПЬНИКЪ МОІ ЄСІ ТЪП' І привъжште мое въ мог оупъваня на нь. [3] ско тъ 13вавитъ мил отъ съті ловьчьа: і отъ словеси мыхтєжьна: [4] плештєма своїма ос'єніт'ї тьа' ї подъ криль его надъєши ска: [5] цитьмь обідєть ръснота єго' не одвоїшн сых отъ страха ноштьнаего: от релъ (6) л (е) тьяштья въ день: [6] отъ ВЕШТІ ВО ТЪМЪ ПОЪХОДІАШТЊАЊА: ОТО СЪОБАШТЊА І ДЕМОНА ПОЛОУДЬИЪЕГО [7] падетъ отъ странъ твоєм тъкнашті: ї тъма о десняж твона къ тевъ не пристяпитъ: [8] окаче очима съмотриші ї въздание гръшьнікомъ оузьоши. [9] фко тър се гі оупъвание мое вышьмъго положилъ есі пристежиштє твоє: [10] не прідєть къ тєвть зъло ї рана не пристяпитъ тълес твоемь: [1] тко албломъ заповъстъ о тевъ: съхоаниті ть во вьсъхъ пятєхъ твоїхъ [12] на ржкахъ возъмять ть еда когда претъкнешн о камень ногж твоїж [13] на аспідж і васильскаа НАСТУПШІ. І ПОПЕРЕШІ ЛЬВА І ЗЛІВ' [14] ІКС НА МІЖ ОУПЪВА ІЗБАВЛЖ І' покръиж и теко позна има мое: [15] възоветъ ко мић и оуслъпших и съ НІМЬ ЄСМЬ ВЪ СКОВЕІ ИЗЪМЖ І ПООСЛАВЛІЖ І: [16] ДЛЪГОТЖ ДЬНЕІ НСПЛЪНЖ I' I' ABAIR EMOY CHINIE MOE.

Ps 101

[1] молитва нишаго егда очныетъ поведъ глуъ прольетъ молитвъ своеж [2] гг оуслънш молитвя мож: и въпль мо къ тебъ да придетъ: [3] не ОТЪВРАТІ ЛИЦА ТВОЄГО' ОТЪ МНЕ ВЪ НЪЖЕ ДЕНЪ ТЯЖЖ ПОЛКЛОНІ КО МНЪ оухо твоє: въ нъже денъ призовя тья: њудо оуслувши мњ: [4] иде ичезж вко дъилъ дьни мог ї кості мова тько соушило сосъхя ста: [5] повьєнъ Бъю в съю съно и-съще срдце моє вко завъюхъ съмъсти хлъвъ мон [6] WT'T' FAACA BB3 BIXAND'S MOED ПРИЛЬПЕ КОСТЬ МОВ ПЛЪТИ МОЕГ [7] оуподовихъ съ нењсътт поустувить (6) въруъ ъко ношънъ вранъ на НЪКОНЦИ: [8] ЗАБЪДЪХЪ ГЕРИУВ ЕКО ПЪТЩА СОЕРАШИЪ СВА НА ЗЪДЪ: [9] ВЪСЪ денъ поношаахъ ми враби мог и увальниции сњ многих клънъахъ сма [10] иде попелъ тько хлъвъ всъ и пітъе моє съ плачемъ раствар'в утъ

<sup>(</sup>a) For въщитьго (Sever'janov).

<sup>(</sup>b) For or or o струлъ (Sever'janov).

<sup>(</sup>c) See § 188.

<sup>(</sup>d) For поустъпныны (Sever' janov).

[11] ѡтъ ліца гнѣва твоего и ѣрості твоеѩ· ѣко възнесъ нізъвръже мѩ· [12] дънье моі ѣко сѣнъ уклонішѩ сѩ· и азъ ѣко сѣно исохъ· [13] ты же гі вь вѣкъ прѣбываеші· и памѩ-т-воѣ(a) вь родъ і родъ· [14] ты вьскресъ помилуеші сиона· ѣко врѣмѩ помиловаті· ѣко приде врѣмѩ· [15] ѣко благоволішѩ рабі твоі камень-е-го·(b) и пръстъ его ущедрѩтъ· [16] и убоѩтъ сѩ ѩзыці имені гнѣ· и вьсі црі земьніі славы твоеѩ· [17] ѣко съзіЖетъ гъ сиона· и ѣвітъ сѩ вь славѣ своеі· [18] прізьрѣ на молітво(c) съмѣреныїхъ· и не унічьжі моленьѣ ихъ· [19] да напишѫтъ сѩ си вь родъ інъ· ї людье зиЖеми вьсхвалѩтъ гѣ· [20] ѣко прініче съ высоты стыѩ своеѩ· гъ съ нси на землѭ прізьрѣ· [21] услышаті вьздыханіе окованыхъ· раздрѣшіті сны умръщвеныхъ· [22] възвѣстіті вь сионѣ їмѩ гне· и хвалы его въ їимѣ· [23] егда сънъмѫтъ сѩ людье вькупѣ· и црі работаті гю· [24] отъвѣща ему на пѫті крѣпості своеѩ· умаленье днеі моіхъ звѣсті(d) мнѣ· [25] не вьзведі мне вь прѣполовленье дънеі моіхъ· въ роды родъ лѣта твоѣ· [26] въ начѩтокъ ты гі землѭ осънова· и дѣла рѫку твоею сѫтъ неса· [27] та погыбнѫтъ ты же прѣбываеші· и вьсѣ ѣко різа обетъшаѭтъ· и ѣко одѣало съвьеши ѣ ї измѣнѩтъ сѩ· [28] ты же самъ еси· и лѣта твоѣ исконьчѣѭтъ(e) сѩ· [29] снове рабъ твоіхъ веселѩтъ(f) сѩ· и сѣмѩ іхъ вь вѣкъ исправітъ сѩ·

Ps 103

псалмъ дадовъ о тварі вьсего мира· [1] блгси дше моѣ гѣ· гі бже мои вьзвелічілъ сѩ еси ѕѣло· вы-исповѣданъе· и вьллѣпотѫ облѣче сѩ одѣѩи сѩ свѣтомъ ѣко и різоѭ· [2] пропинаѩи неба ѣко и кожѫ· покрываѩ водамі прѣвыспрънѣ его· [3] полагаѩи облакы вы‑исхоЖенье свое· ходѩи на крилу вѣтрьню· [4] творѩ анг҄лы своѩ дхы· и слугы своѩ огнь палѩщь· [5] оснываѩи землѭ на твръді своеі· не прѣклонитъ сѩ въ вѣкъ вѣку· [6] бездъна ѣко и різа одѣние еѩ· на горахъ станѫтъ въды·(g) [7] ѡтъ запрѣщеньѣ твоего побѣгнѫтъ· отъ гласа грома твоего устрашѩтъ сѩ·[8] вьсходѩтъ горы нізъходѩтъ въ полѣ· въ мѣсто еже есі осъновалъ імо·(h) [9] прѣдѣлъ положи егоже не прѣідѫтъ· ни обратѩтъ сѩ покрытъ землѩ· [10] посылаѩ источьнікы во дьбрехъ· посрѣдѣ горъ проідѫтъ воды· [11] напоѩтъ вьсѩ звѣрі сілъныѩ·(i) жидѫтъ аонагрі(j) въ жѩЖѫ своѭ· [12] на ты пьтіцѩ нбныѩ прівітаѭтъ· отъ


срѣды камьньѣ дадѩтъ гласъ· [13] напаѣѩ горы отъ прѣвъспръніхъ своіхъ· отъ плода дѣлъ твоіхъ насытітъ сѩ землѣ· [14] прозѩбаѩи пажитъ скотомъ ї травѫ на служьбѫ члкмъ· їзвесті хлѣбъ отъ землѩ· [15] и вино вьзвеселитъ срдце члку· умастіті ліце олѣимъ· и хлѣбъ срдце члку укрѣпитъ· [16] насытѩтъ сѩ дрѣва польскаа· кедры ліванъскы-ѩ-же есі насадилъ· [17] ту пьтіцѩ угнѣздѩтъ сѩ· еродово жіліще обладаетъ іми· [18] горы высокыѩ еленемъ· каменъ прібѣжіще заѩцемъ· [19] сътворілъ есі лунѫ вь врѣмна слънъце позна западъ свої· [20] положилъ есі тьмѫ ї быстъ нощъ· въ неже прѣідѫтъ вьсі звѣрье лѫжьніі· [21] скумені рікаѭще вьсхытеіте·(a) ї испросіте у ба піщѫ себѣ· [22] въсіѣ слънъце и собьрашѩ сѩ· и вь ложихъ своіхъ лѩгѫтъ· [23] изіде члкъ на дѣло свое· и на дѣланьѣ своѣ до вечера· [24] ѣко възвеличишѩ сѩ дѣла твоѣ гі· всѣ прѣмѫдростьѭ створи· исплъні сѩ землѣ тварі твоеѩ· [25] се море велікое пространое· ту гаді и імъже нѣстъ чісла животънаа малаа съ велікыми· [26] ту кораблі прѣплаваѭтъ· змъи сь їже съзъда рѫгаті сѩ ему· [27] вьсѣ отъ тебе чаѭтъ· да дасі пищѫ імъ въ благо врѣмѩ· [28] давъшю тебѣ імъ съберѫтъ· отъвръзъшю же тебѣ рѫкѫ вьсѣчъскаа· їсплънѩтъ сѩ благості· [29] отъвращьшю же тебѣ лице възмѩтѫтъ сѩ· отымеші дхъ іхъ ї ищезнѫтъ· и въ пръстъ своѭ възвратъ(b) сѩ· [30] посълеші дхъ своі съзіЖѫтъ сѩ· и обновіші ліце землі· [31] бѫді слава гнѣ вь вѣкъ· възвеселітъ сѩ гъ о дѣлѣхъ своихъ· [32] прізіраѩ на землѭ творѩ ѭ трѩсті сѩ· прікасаѩи сѩ горахъ въскурѩтъ сѩ· [33] въспоѭ гві въ жівотѣ моемъ· поѭ бу моему доідеже есмъ· [34] да насладитъ сѩ ему бесѣда моѣ· азъ же възвеселѭ сѩ о гі· [35] исконъчаѭтъ сѩ грѣшьніці отъ землѩ· и безаконніці· ѣко не быті имъ·

# *Codex Suprasliensis*

From from the vita of St. Paul the Simple (Supr 171, 15–173, 26) въ ѥдинъ убо отъ дьнии· зѣло лютѣ бѣсꙙщъ сꙙ юноша приведенъ быстъ къ блаженууму антонию· възьрѣвъ же великыи антонии на юношѫ· глагола водꙙщиимъ и· нѣстъ се моѥ дѣло· о семъ бо чину бѣсовьстѣѣмь начꙙльнѣѣмь· нѣсмъ сꙙ ѥще съподобилъ благодѣти· нъ сии даръ· паула ѥстъ прѣпростааго· шъдъ же великыи антонии къ искусънууму паулу· веды и ты· и глагола ѥму отьче пауле· иЖени бѣсъ сии отъ чловѣка сего· да цѣлъ идетъ въ свои 172. домъ· и да прославитъ бога· глагола ѥму паулъ· а ты что· рече ѥмꙈ антонии· нѣсмъ празденъ азъ ино дѣло имамъ· и оставивъ ту отрока великыи антонии· и възꙿврати сꙙ въ своѭ хызинѫ· въставъ же безлобьныи старьць· и помоливъ сꙙ· и призвавъ бѣсꙙщааго сꙙ глагола· рече отьцъ

<sup>(</sup>a) For вьсхытіте (Sever'janov). (b) For възвратѩтъ (Sever'janov).

антониї нзиди изъ чловъка: въсъ же съ хоулоги въпнише глагода. Не НЗІДЖ ЗЪЛЪІН СТАРЧЕ АЖКАВЪІН' ВЪЗЬМЪ ЖЕ МИЛОТНИЯ СВОЕЖ БИІЯШЕ И ПО ХОЪБЪТОУ ГЛАГОЛА ИЗИДИ РЕЧЕ ОТЪЦЬ АНТОНИИ: БЪСЪ ЖЕ ХОУЛЪ ВЪШТААШЕ НА АНТОНИГА Н НА ПАВЪЛА" ГЛАГОДА ЕФЛЬШЬМИ' ГЛАДИВАЊ СИПА' НЕНАСЪЩЕТЕНАГА" ГАЖЕ НИКОЛИЖЕ СВОИМЪ НЪСТА СЪЩТА" КЖЕЖ ОБЪШТИНЖ ИМАТА СЪ МНОЕЖ: ЧТО НЪГ НЪГ МЖЧИТА: ГЛАГОЛА ПАВЪЛЪ: ИЗАЋЗЕШИ ЛИ: НАН ИДЖ КЪ ХОІСТОСОЧ' И ГОРЕ ТЕБЪ ИМАТЪ СЪТВОИТИ' ПОХОУЛИ ЖЕ И ХОИСТОСА сверъпъти тъ бъсъ выпиа: не изндж: разгитвавъ же са павълъ на въсъ' И ИШЕДЪ ИЗЪ ХЪЗНИЪІ СВОЕА ВЪ ПОВПЛАДЬНЫЕ" ЕГУП ТЪСТИИ ЗНОЕВЕ ОУБО' НИЧИМЪЖЕ СЖТЪ ХОУЖДЪШИ: ВАВУЛОНЬСКЪМА 173. ОНОА ПЕШТЪНИЦА: ставъ оцво на камени сватьи старьцъ на знон: молгаше са вогоу глагода сице: тъ видиши їс хоє распаттии при пянтьствиз пилать: гако не ималуъ сьлъсти съ камене сего: ни нмамъ гасти ни пити домьдеже ОУМЪРЖ: АШТЕ МЕНЕ НЕ ОУСЛЪШИЧЪ) И ИЖДЕНЕШИ ЕЪСА СЕГО ОТЪ ЧЛОВЪКА СЕГО И СВОЕОДЬ СЕГО СЪТВООНШИ ОТ В ДОУХА СЕГО НЕЧИСТАВГО ТЕШТЕ ЖЕ глагодяштог препростогогию и сьмереногогическог паулог "соусовор" Пръжде съконьчания молитвъ: възьпи гласомъ бъсъ идж идж: н ИСХОЖДА НОУЖДЕЮ И МЯКОВ ЖЕНОМЪ КЕМЪ ОТЪХОЖДЖ ОТЪ ЧЛОВЪКА ОУЖЕ К' ТОМОУ НЕ ПРИЕЛИЖЖ СА КЪ СЪМЪРЕНОУОУМОУ И ПОВЕПРОСТОУОУМОУ ПАУЛОУ' ЖЕНЕТЪ БО МА И НЕ ВЪДЪ КАМО ИДЖ' И АБИЮ ИЗИДЕ БЪСТЪ' И Пръложи са въ змин великъ: гако седмь десатъ лакъттъ: и иде пльза къ чръмъногочног порог.

From the vita of Isaac of Dalmatia (Supr 191, 21-193, 29)

ТЪГДА ЦЪСАОЬ ОУАЛЪ' И ТЪ СЪБЪРАВЪ ВОА СВОА: ГОТОВЫЩШЕ СА ИЗИТИ НА СЖПОСТАТЪ! И БЪСТЪ ИСХОДАШТУ ЕМОУ НА ПОЛЕ ГЛАДАТЪ ВОН своихъ разгоръвъ са стъиниъ доухомъ: акъ ннъ даннилъ о соцсанъ: СВАТЪН ОТЪЦЪ НАШ ИСАЙНУ ПОИСТЖПИ КЪ ЗЪЛОЧЬСТВЪНОУФУМОУ ГЛАГОЛА цъсароу отврьзи цръкви правовърънъютъ и оуправитъ ти господъ ПЖТЪ 192. ТВОИ ПОВДЬ ТОБОИЖ: ОНЪ ЖЕ ВИДЕВЪ И ВЪ ТАКОМЪ НИШТИ образъ: и въ прътиштихъ старца: преосидъ и: и не отъвъшта кмоу НИ ЕДНОГО СЛОВЕСЕ: НЕ БО БЪ ТЪГДА ТОУ НИ СЛЕДА ЧОВНООНЗЪЧЬСКА: ТАЧЕ пакъы на оутрига нсходаштоу кмоу: ста предь нимъ и рече: цасароу: (4) ОТВРЬЗИ ЦАЪКВИ ПРАВОВЪРЪНЪНУЪ: И ОДОЛЪЕШИ ВРАГОМЪ ТВОИМЪ И ВЪЗВРАТИШИ СА СЪ МИРОМЪ: ЦЕСАОЬ ЖЕ СЛОВЕСИ СИЛЖ ПОЧОУВЪ' ЕЖЕ рече стъп тако обратнши са съ миромъ: съвътовааше съ съвътникъв СВОИМИ' ДА ОТВОЬЗЖТЬ ЦАЬКЪВИ' ПРЕПОСИТЪ ЖЕ И ПРОЧИИ' ИЖЕ БЪЛУЖ ОТЪ ариевъ зълъча въръг паче съвратише цъсаръ. и нарягавъше са кмог И ОУСМИРАВЪШЕ СА СЛОВЕСЕМЪ ЈЕГО: ШИБААХЖ И ХОВЗАНЪР ПОСЛОУШАВЪ ЖЕ ИХЪ ЦЕСАРЬ: ПРЕОБИДЕВЪ И ПОИДЕ: ВЪГОДЪНИКЪ ЖЕ БОЖИИ БЕСПОВСТАНИ

<sup>(</sup>a) BaBY/10Hb | ck'hia: transition between folios 172 and 173.

<sup>(</sup>b) For or or or or or or on (Meyer).

<sup>(</sup>c) For canonical нжжденя.

<sup>(</sup>d) For canonical цѣсаĵю.

МОЛИШЕ БОГА. ДА ВЪСКООВ СЪТВООИТЪ ПОМОШТЪ О ПРАВЕН ВЪРЪ: ПО ДЪВОЮ ЖЕ ДЬНИЮ ОУЧИНИВЪ ЦЕСАРЬ ВОА СВОА. ИСХОЖДААШЕ НА ДАТЪТ ЕЛАЖЕНЪІН ЖЕ ПОВСТАВЪ И НА ПЖТИ ПОСТЪПИ КЪ НЕМОУ" И ИМЪ ЗА ФУЗДЖ конть кго ставыаше и глагода отврьзи цръкви правовъръчънуъ: онъ ЖЕ ПО ОБЪРНАЮ СВОЕМОУ НЕПОВКЛОНЕНЪ 193. ПОВЕТЫСТЪ СЖШТИИ ЖЕ ЕЛИЗЪ ЦЪСАРА: БНГАХЯ И ДА ОТЪСТЖПИТЪ: И НЕ МОЖААХЖ ОТЪТЪТЪТН ИГО: БЪІСТЪ ЖЕ ОТЪТОГО ПЛИШТЪ' И СЪТЕКОША СА МНОЗИЧАЯ И РАЗЛИЧЬНО КЪЖДО ЕНГАХХ Н' ОВИ ПОЖТИКАТЪ ОВИ ЕНЧИ' А ДРОУЗИИ ЖЬЗЛИЕМЪ' Н ЕДВА NEKOLO CEETE EOTY ПОПОУСТИВЪШОУ: ВЪЗМОГОША ОТЪТОЪГНЖТИ ОЖЦЪ ИГО ОТЪ ОУЗДЪН КОНЬНЪКА ЦЕСАОД: ТЪГДА ЦЕСАРЪ ОУАЛЪ ВЪЗВЪСНВЪ СА ВЬЗЬОВВЪ СЕМО ОНАМО: ВИДЪ МЕСТО НЕКАКО. АКЪІ ВАПЖ СЯШТУ НСАКШЯ: И ТИНЪІ СМОЬДАШТЯ ПЛЬНЯ: И ЛЪСЪ ЧАСТЪ ВЪ НЕИ: И ТОБНИЕ ЗЪЛО ЛЮТО и кжпинии: въ неже мъсто аштє въпадєть какъ любо скотъ: то к томоу живо не излъзетъ: съмотривъ оуво цесто то: и оувъдъвъ тако сътъ съмрьтъна естъ: повелъ въвръшти и тоу; и иде платъмъ своимъ Въвръженог же въвъшь довьюмог исповъдников хоовоч неавном. въ съмрьтъняю пжчиня: пръбъютъ цълъ: не приняз никакого же (6) зъла авние во силотя господа нашего "с уса" пришъдъша дъва аггела нздраста (1) н нс тинъ тол и изведъша и постависта (4) на пяти и рекоста ключ миръ тъвъ кръпи са и възмжжан.

#### From the vita of St. John Climacus (Supr 275, 10-276, 23)

ІЕДНОЕХ ОУБО ОТЪ ДЫНИ. ПОВЕЛЪ ОТЪЦЬ їЖАНЪ МИСЕОУ ИТИ НЪ НА КОЕ МЪСТО И КОПАВШОУ ООВЪ СЪВРАТИТИ ВОДЖ И ПОНВЕСТИ НА НАПАРАНИЕ ЗЕЛИЮ° ДОШЪДЪШОУ ЖЕ ЕМОУ НАРЕЧЕНААГО МЪСТА" ПОВЕЛЪНОЮ БЕЗ ЛЪНОСТИ ТВОРЕАШЕ: ЗНОЮ ЖЕ ВЪ ПОВПЛАДЬНИИ БЪВЪШОУ: ВЪЗЛЕЖЕ МШУСИИ ПОДЪ КАМЪКОМЪ ВЕЛИКОМЪ' И ОУСЪПЪ ПОЧИВААШЕ' НЕ ХОТАН ЖЕ ЗЪЛА ДАБОМЪ СВОИМЪ ГОСПОДЬ ВАОИ ПО ОБ'БИЧАЮ' СЖШТОУ БО ВЕЛИКОУФУМОУ ОТ БЦОУ ВЪ ХЪВИНЪ' И МОЛИТЪЖ ТВООДШТОУ КЪ БОГУ. ВЪЗДАЪМА СА И ОУСЪНЖ МАЛЪГ Н ВИДЪ НЕКОГА ВЪ СВАТИТЕЛЬСТЪ ЛЪПОТЪ ВЪЗЕФУЖДАНЖШТА И И ПОНОСАШТА IEMOY O CLHE H ГЛАГОЛЯШТА: "WANE КАКО БЕС ПЕЧАЛИ СЪПИШИ" А МWHCH ВЬ ВЕЛИЦЪ БЪДЪ ПРЪБЪВААТЪ: АБИЮ ЖЕ ВЪСКОЧИВЪ ЗА ОУЧЕНИКА СВОЕГО молитъж творгаше: таче пришъдъщод кемоу вечер, въпрашааше его, кда ТИ СА ЧТО ЗЪЛО СЪТВООН (6) 276. ОНЪ ЖЕ РЕЧЕ ГАКО КАМЪР ВЕЛИКЪ МА ХОТЪ ПОГНЕСТИ" ПОЧИВАЊЕШТОУ МИ ВЪ ПРЪПЛАДЬНИЕ ПОДЬ НИМЪ" АШТЕ НЕ БЪЛ МЕНЕ НАПОАСНО ВЪЗЪВАЛЪ' ДА ОШТОУТИВЪ ОТЪСКОЧИХЪ' СЪМЪРЕНЪВИ ЖЕ ОТЪЦЬ ПО ИСТИНЪ' НИЧЕСОЖЕ ОТЪ ВИДЪНЪНЪ СЪПОВЪДАВЪ ТАННЪ ЖЕ БЛАГААГО ЕОГА ПОХВАЛИ' БЪТАШЕ ЖЕ И ОБРАЗЪ' И ЦЪЛИТЕЛЬ ВОЪДОМЪ НЕВИДНАТЪ' ГАКОЖЕ НЕКЪГДА НЕКТО ИСААЌНИ ИМЕНЕМЬ' ТАЖЕСТИНЖ ПЛЬТОЛЮБИВААГО

<sup>(</sup>a) The sequence но, left out at the end of the line, are restored by the editor.

<sup>(</sup>b) The letter €, left out at the end of the line, is restored by the editor.

<sup>(</sup>c) The letter a, left out at the end of the line, is restored by the editor.

<sup>(</sup>d) The letter a, left out at the end of the line, is restored by the editor.

<sup>(</sup>e) ch | reoph: transition between folios 275 and 276.

Бъса велми съдръжниъ: и въ отъчаании Бъвъ къ великоуоумоу семоу отъцоу притече и съ сльзами и выпльмъ исповъдааше свои връдъ ратъ: Блаженъи же чоудивъ са въръ кго: на молитвъ станевъ рече W Дроуже: И ТАКО МОЛИТВЪНАА СЛОВЕСА СЪКОНЬЧАВААХЖ СА: ЕШТЕ НИЦОУ ЛЕЖАШТОУ БОЛАШТОУОУМОУ' ГОСПОДЬ ТВОРАН ВОЙЖ ВЪГОДЬНИКОМЪ СВОИМЪ' ТЪ ОУСЛЪЩАВЪ И РАБА СВОИЕГО МОЛИТВЪГ И ФЕЛЬГЪЧА ЕРАТЪ ТОМОУ ОАТЪЪ и благодарьствиста (а) га ва-

From the vita of St. James the Faster (Supr 514, 6-517, 25)

БЪСТЪ НЪКЪК ОТ ЪШЪЛЬЦЪ: ВЪ ВЕСНЪТЕМЬ ГРАДЪ: ПООФИОНШИЪ НАОНЦАЕМЪ: ИМЕНЕМЬ НАКWBЪ: И СЬ МАЛОВРЕМЕНЬНААГО СЕГО ЖИТЪГ СОУЧЕТ НААГО ОТ"ВВОЪГЪ СА" ЖИВЕ НЪ ВЪ КОЕН ПЕШТЕРЪ Еї" ЛЪТ"В" ТОЛИКО ЖЕ АЛУЧЪВОВ Н БЛАГЪНИИ ДЕЛЪГ ПОСПЕШИВЪ: ЧЬСТЪНЪ АВИ СА И ОУГОДЪМЪ БОУ ГАКОЖЕ КАМОУ И ДАРЪ НА БЪСЪ! ПОЛОУЧИТИ: И МНОГЪ! ИНЪ ЛЪЧЬБЪ!" ХСА НАШЕГО ТВООИТИ "MEHEME" BLCH WE MANA WHTHIN JHBAWTE CA K HEMOY CLEHOAAXX CA" HY WE МНОЖАНШИН" БЕЗАКОНЫНЪНУЪ САМАРАНЬ БЪТАХЯ ВЪРЪГ" АЖЕ ВИДА К НЕМОУ ПРИХОДАШТА БЖИЇ ЧЛОВЪКЪ' НА МНОЗЪ ОГЛАШАА ОТЪ БЖИНУЪ КНИГЪ' НА ИСТИНЬНЖЖ ВЪОЖ ФЕОАШТААШЕ: НЪ НА ООДЪ ЧЛОВЪЧЪСКЪМ ИСКОНИ ВОНЖАН ДНГАВОЛЪ' И НАНПАЧЕ НА РОБЪР СПАСА НАШЕГО ХСА' ВИДА СЕБЕ ОТЪ МНОГЪ! ЕЛАГЪН ДЪТЪЛИ МЖЖА И ПРАВЪДНААГО ЖИТИГА НА МНОЗЪ ПООГОНИМА ВЪСТА НА НЬ ХОТАИ ПООГНАТИ ОТЪ ПРЪЖДЕНАВЕЧЕНААГО МЪСТА: ВЪЛЪЗЪ БО ВЪ ЕДНОГО ОТЪ САМАРАНЪ: ИСКОНИ И ВЪННЯ. ИСТИНЪ: ОТЪВРАШТАЊЕТНИХЪ СА ПОНГОТОВА КГО СЪЕДАТИ ВЬСА ПОСЛОУШАНЯШТАА ИГО' ДРОУГЪР И срьдоводях и рабъю на лагания стааго мжжа. да и оуловивъ възможетъ отъ странъ тъхъ прогнати: съеравъше же са вьси въ жилиште жьрьца своюто и много мъюливъше и проказълъвъще - 515. коньчънъ издинъ АША СЪВЪТЪ' И ПОНЗВАВЪШЕ ЕЄСТОУДНЯ ЖЕНЯ И БЛЖДНУХ' ДАША ЕН ДВА ДЕСАТИ ЗЛАТИЦЪ' И ИНО ТОЛИКО ЖДЕ ОБЪШТАВАЊШТЕ ДАТИ ЕН' АШТЕ ВЪЗМОЖЕТЪ ЗАПАТИ ЕЖИЮ РАБОУ НАКWBOY' ГАКОЖЕ ИМЪ ТОЖ ВИНОНЖ' ВЪЗМОШТІ ОТЪ ЗЕМЬА СВОЕА СЪ СТОУДОМЪ МЖЖА ОТЪГНАТІ' ТЪЕМИ ОБЪШТАНИИ ПОБЪЖДЕНА ЖЕНА: ИДЕТЪ К НЕМОУ: ПОЗДЪ ЗЪЛО НОШТИНЖ: И ТЛЬКНЖВЪШИ ВЪ ДВЪРН МОЛГААШЕ ЕГО ВЪВЕСТИ НЖ: ОНОМОУ ЖЕ НЕ ХОТАШТОУ" И НА МНОЗЪ ТО СЪТВООНТИ ОТЪЛАГАНЖШТОУ" ПОВЕЋВААШЕ EE CTOYAA TABKALUTH' H CB MHOFOIM MONHTBOHK MONAWTH HOMATH HA' И МАЛЪІ ОУВРЪЗЪ И ВИДЪВЪ ИЯ МНЕАШЕ МЪЧЬТОУ БЪГТИ И ПОБКОЬСТИВЪ СЕБЕ ЗАТВООНВЪ ЗАКЛЮЧИ ДВЪДИ' И ВЪЗВРАТИВЪ СА И СТАВЪ НА ВЪСТОКЪ' ПОНЛЕЖНО МОЛИТВЖ БОГОВИ ПОННОШААШЕ: МНОГОУ ЖЕ ЧАСОУ МИНЖВЪШОУ: Н НОШТИ ОУЖЕ ПРЕПОЛОВАШТИ СА: НЕ ПОВСТАНЪВШЕ ТЛЪКЖШТ! И ВЕЛИКОМЪ ГЛАСОМЪ ВЬПЫЖШТИ' ПОМИЛОУН МА РАБЕ БЖИН ЕДА КАКО ЗВЪРЬМИ ИЗЪДЕНА БЖДЖ ПОВДЪ ГЛАВОЙ ТВОЕНЖ: ПОМЪКЛИВЪ ЖЕ ПРАВЪДЪНЪЩЪ: И ВЪДЪ! ВЪ МЕСТЪХЪ ТЪХЪ ЗВЪРНИ МНОЖЪСТВО ВЪ РАЗМЪЩЛЕНЫА ВЪПАДЪ" ОТВОЪЗЪ ДВЬОН И ГЛАГОЛА ЕН' ОТЪКЖДОУ ПОНШЛА ЕСН СЪМО: КОГО ЛИ ИШТЕШИ: ОНА ЖЕ рече' отъ манастъю'в њемъ сьде Близъ сжиталго' и посла ма нгоуменью

<sup>(</sup>a) Aberant form for 3DuAor(благодарьствити).

ДОНЕСТИ ПРОСВООЖ ВЪ СЫЖ ВЬСЬ' И ВЪЗВРАТИВЪШИ МИ СА И ИДЖШТИ ВЪ манастирь: омрькохъ на мѣстъ семь: да молж та чаче бжин помильн МА И ПОИИМИ" ДА НЕ ЕЖДЖ ЗВЪРБМИ ИЗЪДЕНА: 516. ТЪГДА ЖЕ ОУЖЕ оумилосовдивъ са выведе на и предъставивъ ки водж и улъвъ. Вълъзъ САМЪ ВЪ ЖТОВНЯЯ КЛЕТЪКЖ ЗАКЛЮЧИ СА: ОСТАВИВЪ ЖЕНЖ ВЪ ВЫВШИНИ КЛЪТ ЦЪ: ОНА ЖЕ ГАДШИ ПОМЛЬЧА ВЪ МАЛЪ ЧАСЪ: И ПОТОМЪ НАЧА КОИЧАТИ: И ПЛАЧЖШТИ ПОНВОЬЖЕ СЕБЕ КЪ ДВЬОЕМЪ: И СЪ ГООВКАМИ СЛЪЗАМИ ЗОВЪАШЕ стааго и повклонивъ са двърьцами: и видъвъ их одрьжимъ и окъю въ МНОЗЪ БОЛЕЗНИ СЖШТЯ ВЪ НЕДООУМЕНИИ БЪІСТЪ: ЧТО СИ ЕН ЕЖДЕТЪ НЛИ Ч'ТО ЕН СЪТВООИТЪ ВЪПРАША КА" ОНА ЖЕ ГЛА КЕМОУ ПОНЗОН НА МА Н пръкрьсти маг гако срьдучьног волъзных одръжима ксмъ: то слъщавъ ИЗЛЪЗЪ И АБИК НАЛОЖИ КОАДЖ ВЕЛИКЯ: И ЛЪВЖИХ СВОИЯ ОЖКЖ ВЪЗЛОЖН НА ОГНЬ ДЕСНЖИХ МАЖА ОТЪ МАСЛА СТЪИНУЪ: ГОВАШЕ ИЯ ТЕПЛОСТИНЖ ОЖКЪІ СВОЕА: И ПОВЕКОЬШТАА ЕЯ ЧАСТО ПО ПРЫСЕМЬ НЕ ПОВСТААШЕ: ОНА ЖЕ СВОИМЪ СТОУДОМЪ ТЪКПАШТИ" И ХОТАШТИ КГО ОУЛОВИТИ" И НА похотънью зъло стааго привести: гла кмод. модж ти са мажи ми средьце на дльзъ: да мн престанетъ одръждития ма волъзки: опъ же по слитии въ немъ простости. творъ'ше ки аже на врачех: коупно ЖЕ И Проказьства ЛЖКАВААГО ВЪДЪ! И БОЛ СА: ЕДА КОЛИ ОТЪ МНОГААГО к нен попеченька: весьмовтьнямъ пользую нанесетъ. до двою или до трии часъ: тако без милости лъвжих ржкж къ огню придъя трыгъаше ДОНЕЛИЖЕ ОУДОВЕ РЖЧЫНУЪ ПРЪСТЪ ИЗГООВЪШЕ ОТ "ВПАДОША" ТО ЖЕ сотонныескъмимъ къзнемъ противью са творъаше: пакоже клюу отъ ОГНУВ ПОНЕЗИВАНИЕТИЯ БОЛЪЗНЫЖ НЕСЕТОВПИМЖЕ 517. ЗЪЛФУОУМОУ помъюлог не наити на средьце: она же преславное то видъввъши и в сегъ Бъявши: видѣъще в со рякж сталго отъ огна очже въсх изгоръвзящих Въсплакав'ши горцъ и въздъхняв'ши припаде къ ногама стааго' н ржама своима бижшти са въ прьси выпнаше: оу горе мынъ окаамън и отъмнентьи: од горе мынъ тако сьмъдь еслу огных въчнааго: од горе МНЪ Гако ЖНШТЕ (6) ЕСМЪ ДНАВОЙЕ: СТВИ ЖЕ ОУЖАСТЪНЪ ПРОТИВЖ ТОМОУ Бъютъ рече к нен: въстани жено: и съ многож ноужденя въставивъ их отъ ЗЕМЬА: И ПОИЛЕЖНО МОЛИТВЯ СЪТВООНВЪ: ГЛА ЕН ПОВЪЖДЪ МИ ЧТО ТЪ ТЪТ ВЪ она же неколи пришедши в са гакоже въчис дъло: по въсемог съказа IEMOY" ГАВИВЪШИ ЕМОУ ЛЖКАВЬНЪТНУЋ САМАРЕНЪ" НАНПАЧЕ ЖЕ СОТОНИНО ПООУЧЕНИЕ НА ПРАВЬДНААГО И ВЪСТАНИЕ: Н АБИЮ ВЕЛМИ ВЪЗДЪХНЖВЪ ДАЕЪ ЕОЖНИ' И МНОГО СЪ СЛАВОСЛОВИИМЪ И СЪ СЛЪЗАМИ' БЛАГОДАОНВЪ БА ОГЛАСИ 1 - ДАВЪ ЖЕ ЕН ПООСВОРЪГ И ПОСЪЛА ИЖ КЪ СТОУОУМОУ АЛЕЗАНДРОУ ЕППОДЪ ДОШЪДЪШИ ЖЕ ЦРЬКВИ" ВЬСА ПО СЛЕДОВАНИЮ БВИ ПОВЖДЕ ПОТОМ' ЖЕ Н ПРЕПОДОВНОГОУМОГ МЖЖОУ ИСПОВЪДА" ГАКОЖЕ И ТЪ МНОГО ОГЛАСНВЪ НЕ: Н ОБОВТЪ Г» ПОИЛЕЖЬНО° О ЗЪЛЪЩУЪ ГАЖЕ СЪТВООИ КАНЖШТЪ СА' ДАСТЪ ТЕН БЕСЪМОЬТЪР БАНЖ.

<sup>(</sup>a) See § 471-472.

<sup>(</sup>b) In SUPR жиштє 2x for canonical жилиштє. Here in the ms. ли is a late addition above the line (Sever'janov). See § 892-893, and also § 865.

From the vita of Konon of Isauria (Supr 43, 9-45, 1)

ТАКО КМОУ ЖИТИК БОГЪ ПОЧЬТЕ ГАКОЖЕ И ПО СЪМОВТИ ТЕГО СЛАВЪНЪНША сътвори: юште же къ симъ чоудесемъ и се въстъ, въдовици кожи оувозъ снъ иночадъ кдначе съсъ!" игоже носдшти въ пазоусъ мати имоу" въ ВРЪМА ЖАТВЪ ШЪДЪШИ ДА СЪБЕРЕТЪ КЛАСЪІ НА ПИШТИ СЕБЪ" ПОЛОЖИ НА ЗЕМИ СНА СВОЕГО И КЪ ЖАТЕЙВНЕМЪ ПОНСТЪПИВЪШИ' И БЕОЖШТИ КЛАСЪГ ВъНЕЗААПЪ ВЛЪКЪ ПОНШЕДЪ ОБРЕТЕ ПОВОЪЖЕНЪ ДЪТИШТЪ НИКОМОУЖЕ същето од него и въскъютивъ и отиде. жатейть же оуслъшавъше ПЛАЧА МЛАДЕНИШТА" И МАТИ ПОЧОУВЪШИ ОБРАТИ СА: И РАЗОУМЪВЪШИ СВОЕГО Зъла въскрича съ въсъми: жатейсне же ови клицаахъ ови течаахж ВЬ СЛЕДЪ ВЛЬКА: И НЕ ПОСТИГЪШЕ ИГО ВРАТИША СА МАТИ ЖЕ ИГО И НЕ ИСПОЧНВЪШИ ВЪЗДЪВЪШИ РЖЦЪ И ТЕКЖШТИ ИДЕ ПЛАЧЖШТИ СА ДО СТААГО МЖЖА ПОВЕДАТИ КМОУ СВОЕЖ БЪДЖ И ДОШЬДЬШИ МАНАСТЪЮВ' ВЬСКОНЧА МАТИ ДЕТИШТА ВЕЛ'МИ ГЛАГОЛЖШТИ' СТВИИ КОНОНЕ ДАЖДЬ МИ сна монго онъ же оуслъшавъ гласа плачоу въскочи къ двъремъ по ОБЪРЧАЮ: ОНА ЖЕ ВИДЪВЪШІ И: И ПОУВАТИВЪШИ НОЗЪ КГО ПЛАКА СА ГООРКО. ОНЪ ЖЕ ВЪСТАВИВЪ ЊЕ ВЪПРАШААШЕ: Ч ТО ЕСТЪ ВИНА" ОНОИ ЖЕ ОТЪ ПЕЧАЛИ НЕ МОГЖШТИ ИСПОВЪДАТИ: СЪПОВЪДАША ИЖЕ ЕЪША СЪ ЙЕЖ ПОНШЪЛИ: И ОУСЛЪЩАВЪ БЪВЪШЕЕ: И ПОПЕЧЕ СА СЪ МАТЕОВЪЖ ДЪТИШТА: И СТАВЪ ПОН В съуть и въздъвъ на въсотъ очи ї ряць помоли са и призъвавъ сточе ТЪНИМЕНЬНОЕ ИМА: ПОВЕЛЪ БЪСОМЪ ИТИ И ОБРАТИТИ ВЛЬКА СЪ ДЪТНШТЕМЬ ЦЕЛОМЪ' ОНИ ЖЕ ПОНЗВАНИ БЪВЪЩЕ И НЕВИДИМИ СКОРО ПОИДОША' И ПОСТИГОША И ДАЛЕЧЕ ЗЪЛУ ЕШТЕ МЛАДЕНИШТЪ ВЪ ОУСТЪВЪ НОСАШТЪ" ДА ови поувативъше за оуши влъчаахъ н' дроузии же съзажда 47 пориваахъ А ДРОУЗИЇ ЕНГАХЯ' И ТАКО ВЕДОША ВЛЬКА КЪ НОГАМА СТОУОУМОУ КОНОНОД' и падъ положи дътиштъ предъ въсъми: пакоже въсъмъ въскричати гласомъ велнімъ: и отъ многъю радости просльзивъшемъ са славити БОГА О ТАКОН СИЛЪ ДАНЕН КМОУ" И ВЬЗЕМЪ ДЪТНШТА ОТЪ ЗВЪОВ ДАСТЪ матери и молитвя сътворивъ 45. отъпоусти въсъхъ съ влъкомъ:

From the vita of St. Artemon of Laodicea (Supr 221, 14-225, 30)

СЛЪЩАВЪ ЖЕ КОМИСЪ ТАКО ВЪСИ КОВСТИГАНИ СЪЕРАЛИ СА СЖТЪ ВЪ ЦРЪКЪН коупно съ епискоупомъ: повелъ воиномъ своимъ оседълатти кмоу койь цъсарьскъ да въшъдъ въ оржжи () своимъ въ црькъве: кръвьмъ Пролиганию сътворитъ кръстипаномъ: иметъ же епискоупа сисинния: н съ арфельомъ презвутеромъ: нзлъзъ же из града и съ воинъ: въяше БО ЦОЪКЪІ ПАТИ ПЪПЬОИШТЪ ОТЪ ГОАДА" И ПОНЕЛИЖАЊЖШТОУ СА ИМОУ КЪ Цръкъвн: пко въ дала едного пъпришта" въстрасе са и огнь ВЕЛИКЪ НАЧА ЖЕШТИ И ПАКОЖЕ СЪПАДЪШОУ СА ЕМОУ СЪ КОНА ОУМИРАТИ" Принесъше же одръ възложивъше и на нь принесоша въ претторъ" полоуношти ( 222. же бъвъшоу" повелъ комисъ свъшта принести"

<sup>(</sup>a) For ch зажда; see \$ 336.

<sup>(</sup>b) For canonical opzwur (Sever'janov).

<sup>(</sup>c) полоү | ношти: transition between folios 221 and 222.

и свътити по въсемог претором. и предървати съ нимъ вонномъ кго: н ГЛАГОЛА КЪ ДОМЕСТИКОМЪ СВОИМЪ: ВАСНЬ 3/ КОЬСТИПАНИ ПРОКЛАША МА' И БОГЪ ИХЪ МЖЧИТЪ МА ГЛАГОЛАША К НЕМОУ ДОМЕСТИЦИ: ДОБРОЧЪСТИВИН БОЗИ И ОСВЕШТАНИЕ БОГЪИНА АРТЕМИДЪР ТИ ТА ИМЖТЪ СЪНАВЪДЪТИ" лютъ же кмоу стражджштоу и не могжштоу тръпъти волъзни' повелъ ВОННОМЪ СВОИМЪ ШЪДЪШЕ ВЪ ЦОБКЪВЕ: И РЕШТИ ПИСКОУПОЧ сискомъ Гако БОГЪ Кръстиганескъ Великъ ЕСТЪ ПОМОЛИ СА ОУБО ЗА МА КЪ БОГОЧ ДА ВЪСТАНЯ И ИЗБЯДЯ МЖКЪ СЕНА И ПОСТАВЬЯ ТИ ЗЛАТЪ ОБОЛЗЪ ВЪ ГОАДЪ СЕМЬ: ШЕДЪШЕ ЖЕ ВОННИ ГЛАГОЛАША ЕПИСКОУПОУ ЕЖЕ СЛЪШАША: ОТЪВЪШТАВЪ ЖЕ ЕПИСКОУПЪ ГЛАГОЛА ИМЪ ШЪДЪШЕ ПОВЪДИТЕ КОМИТОУ ВАШЕМОУ" ЗЛАТО ТВОЕ И СЬОЄЕРО ТВОЕ ДА БЯДЕТЪ СЪ ТОБОЕЖ' АШТЕ ли да въроучеши въ господ нашъ її хо повидеши вользуни тол ЛЮТТЫА: ТОМОУ ЖЕ РЕКЪШОУ ВЪРОУНА ТЪЧЫЯ ДА ИЗБЕЖДЖ СЕГО: АБИНЕ ЖЕ СЪДРАВЪ БЪСТЪ' И ПОВЕЛЪ ВОИНОМЪ СВОИМЪ ВЪПОДШТИ КОЪКЪИГЯ ИМЪ' да ндетъ съ ними въ кесаринскъми 223. градъ изшедъшоу же IEMOY H3 ГРАДА ЛАОДИÈИНСКААГО И ШЪДЪШОУ ТОИИ ПОПОНШТЪ СЪОБЪТЕ И рабъ божии артемонъ: идъ1 из лова: дивная животъ оцлавьяа словомъ ховомъ: вьслъдьствоваста же въследъ кго дъва елени: н шесть дивнихъ козъ: ижє идоша къ єписковчою сисниню: имъщие во ЕПИСКОУПЪ СИСИНИИ ВРЪТ ПОГОАДЪ КРАСЬНЪ' ПОВЕЛЪ ЖЕ КОМИСИИ СТАВИТИ Кръкъютъ: и въпрашааше рава кожна ар темона: како звърн сма(в) оулови: мынђаше бо комис, тако питањин сжт"в и кротъци: глагола же К НЕМОГ РАБЪ БОЖИН АРТЕМОНЪ: СЛОВЕСЕМЬ ХОНСТОСОМЪ СИ ОУЛОВЪЖНИ Бъща комись рече старче тако ми есть разоумъти тако крьстиванъ ІЄСН' АРТЕМОНЪ РЕЧЕ' АЗЪ ОТЪ МЛАДЕНЬСТВА КОЬСТИГАНЪ ЈЕСМЪ' ПОВЕЛЪ ЖЕ АБИЮ КОМИСЪ ВЕОНГЪІ ЕМОУ НАЛОЖИВЪШЕ НА ВЪНЖ: ПОЪДАТИ И ДВЪВМА воннома: и въслъдовати кмоу дожи ї до ќесаринска града: Т"БГДА ПОВЕЛЪ СВАТЪК АРТЕМОНЪ' ЕЛЕНЬМА И ДИВИНМЪ КОЗАМ' ИТИ КЪ ЕПИСКОУПОУ КЪ ВРЪТПОГОАДОУ' ОНИ ЖЕ СЪ СПЕХОМЪ ШЪДЪШЕ СТАША БЛИЗЪ ГРАДА' ВИДЪВЪ же а епискоупъ (8) 224. глагола къ вратароу" отъкждоу пондоша СИА ДИВИА КОЗЪР И ЕЛЕНИ I АБНЕ ЖЕ ЕДИНЪ ОТТЪ КЛЕНИЮ ОТВОЪЗЪ ОУСТА свога рече чловечскомъ гласомъ: рабъ кожин артемонъ атъ бъюстъ" НЕЧИСТЪВИМЪ КОМИСОМЪ: Н ИДЪ! ВЪ ЌЕСАРНИСКЪИН ГРАДЪ: СЬВАЗАНЪ ПОЪДАНЪ Бъютъ дьвъма воннома намъ повелъ прити къ врът поградоу семочъ оужасенъ же бъвъ епископъ: о чловъчьстъ гласъ елени: и не имъ въръг о стъемь артемонъ: повелъ вратароу призъвати днакона филеа: н

<sup>(</sup>a) Extrapardigmatic; 'perhaps, maybe'.

<sup>(</sup>b) For επικκογπογ. Večerka has a separate headword, πιςκογπτ.

<sup>(0)</sup> For cen by the supr basic writing system, see §228. Cf. the footnote 17 in Sever'janov's edition: «the parchment is silent on the fluke that led the scribe to iotate the A».

<sup>(</sup>d)

<sup>(</sup>e) For canonical сных.

<sup>(</sup>f) For χρης тосовомъ (Sever'janov).

<sup>(</sup>g) Enn|ckoynz: transition between folios 223 and 224.

пришъдъшу ѥму глагола къ н҄ему епискупъ· диꙗконе филеа· видиши дивиѧѧ сиѧ козы и ѥлени· стоѧщꙙ близъ града· глагола ѥму диꙗкъ· еи господи виЖѫ· дивьно чудо ти имамъ съповѣдати филеа· ѥдинъ бо отъ ѥлению чловѣчьскомъ гласомъ глагола· ꙗко презвутеръ и рабъ божии артемонъ· ѧтъ бывъ нечистымъ комисомъ· идетъ въ к҄есариѭ· и ужастъ объдръжитъ мꙙ о чловѣчьстѣ гласѣ ѥлене· и аще то тако ѥстъ· то грꙙди вьзьми благословьѥниѥ· и иди въ к҄есариѭ· съ ѥдинѣмꙿ слугоѭ· и испытаи аще то ѥстъ 225. истина· велика бо печаль въ мьнѣ ѥстъ· възьмъ же диꙗкъ благословьѥниѥ отъ епискупа· и помоливъ сꙙ идѣаше пѫтьмь· шъдъ убо въ к҄есариѭ· искаꙿше раба господьнꙗ артемона· бѣаше же свꙙтыи вь темници· обрѣтъ же и диꙗкъ глагола къ н҄ему· рабе бога вышън҄ѣꙿго· и пастуше словесънъихъ овецъ· како ны остави ѥдины· и въпалъ ѥси въ рѫцѣ томител҄и· и никꙿтоже отъ насъ увѣдѣ· развѣ ѥдинъ отъ дивиихъ животъ глагола епискупу· ꙗко рабъ божии· артемонъ· ѧтъ бывъ комисомъ нечистыимъ иде въ к҄есариѭ· не вѣровавъ же епискупъ· того ради посъла мꙙ сѣмо· аще истина ѥстъ ѥже слыша· великѫ бо печаль иматъ о тебѣ· глагола кꙿ н҄ему свꙙтыи артемонъ· рабе божии филеа· шъдъ вьзвѣсти владыцѣ моѥму свꙙтууму епискуп · да молитъ сꙙ за мꙙ· да побѣЖѫ нечистааго· и хръстоненавидьнаꙿго примышл҄ѣѧ· и бѫдѫ съпричꙙстьникъ христосу· ѥще бо нѣстъ въпрашаниѥ о мьнѣ было· шъдъ убо вьзвѣсти епискупу· ꙗко рабъ божии артемѡнъ вь тьмници сѣдитъ· и молитъ сꙙ да сꙙ молиши за н҄ь·

PART IV

Dictionaries

# **How to use the dictionaries**

Two dictionaries: their names and typical purposes

There are two dictionaries: paradigmatic (PD) and root (RD). The wordlists of these dictionaries coincide. In the PD, the unit (and the headword) is the lexeme; in the RD, it is the root (and the headword is the root's number). In the PD, each lexeme is supplied with a paradigmatic index, the morphophonological spellout of its starting form, and the number of its root. In the RD, each root is supplied with all the lexemes of the wordlist that contain it, and for each wordform the shape of its root formative is noted.

The paradigmatic dictionary should be used to determine if a lexeme belongs to the controlling lists of lexemes (§ 3), to find its paradigmatic index (which is necessary to build the paradigm), and to determine the morphophonological spellout of its starting form, and to find the lexeme's root and its number.

The root dictionary should be used to find the set of lexemes with a given root (among those from the controlling list), the alloforms of the root, and its number.

### On the wordlists of the dictionaries

The wordlists are based on the wordlists of Večerka's and Sadnik's dictionaries. However, the wordlists here have been narrowed: first, in terms of the corpus of sources, and second, by filtering out certain grammatical parameters.

The benchmark source corpus includes only the following seven sources: Zogr, Mar, As, Ps Sin, Kiev, Sav and Supr. Accordingly, items that are not found in any of these sources may be excluded from the wordlist.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

The wordlist only contains paradigmatic lexemes (V— verbs, S — substantives, and A — adjectives): thus, items such as абиѥ, издрѧдь, нъ, ни, доидеже are excluded from the wordlist.1

The wordlists does not includes compounds: thus, such items as добротворити, доброчьстиѥ, козьлогласованиѥ are excluded from the basic wordlist.

Toponyms, anthroponyms, and some other Greek and Latin borrowings are also not included, so, such items as дамаскъ, емаусъ, давыдъ, козма, акротомъ, алгуи, алсосъ, газофилак҄иꙗ, енк҄ениꙗ.2

# Paradigmatic dictionary

# Spellouts and their order

Each word is represented by its starting, or dictionary form (which is the headword of the dictionary entry). For substantives (S) this is NSg, for adjectives (A) this is NSgmBrev, for verbs (V) this is the infinitive. All words are given in the normalization.

The order of words in the PD is inverse-alphabetic (see the alphabetic order in § 16). For homonymous entries, the headword is the starting form with a superscript index, cf. вести1 (root вед ‹75›) and вести2 (root вез ‹84›). To the extent possible, these indices agree with those in Večerka.

# Structure of the entry

The dictionary entry, in the general case, contains four fields: 1) root number, 2) morphophonological spellout, 3) dictionary form (headword), and 4) pardigmatic index. Cf.:


The *root number* corresponds to the numbering in the root catalog, given in its entirety in the RD. In the morphophonological spellout, *parentheses* surround the root, and *double parentheses* frame an opaque root (an opaque stem; see § 628). Paradigmatic indices show the inflectional class; see details for *nom-*

<sup>1</sup> See details in § 276, 303 on participles and deadjectival adverbs. Some individual cases of departure from Večerka's lexicographic decisions are noted *passim*, e.g. in the lists of commented lexemes that are omitted from this translation (§ 324, 324, 336 and others; see also *Index of words*). Some compounds with не are also excluded, such as неслава (Večerka has слава and неслава), неврѣдъ (Večerka has врѣдъ and неврѣдъ).

<sup>2</sup> See details in § 636.

*inals* in § 302, and for *verbs* in § 416 and § 419 (regular and unique verbs), and § 454 (irregular verbs).

# Structure of the paradigmatic indices

The index begins with an Arabic numeral. The digit 0 in verbs and nominals means that the lexeme is unique (cf. § 357 for nominals and § 516 for verbs). The index of nominals besides 0 can begin with 1 (monovariate, *i*-declension) or 2 (twofold declension). In verbs, digits from 1 to 7 indicate paradigmatic class. Further symbols after the digit refer to the next levels of classification. For nominals, the symbols *a* and *p* indicate an adjective, while *m*, *n*, *f* and *s* indicate a substantive.

# Other symbols

The symbol // indicates a doublet lexeme, cf. for example:


The symbol ∇ indicates morphological anomalies in the paradigm, cf. for example:


For other marks and symbols in extenders of paradigmatic indices see *Abbreviations and symbols*.

# Root dictionary

# Root spellouts and their order

The headword of the entry is the root number. Roots are arranged by number. Numbers are assigned alphabetically, namely: alloforms of the same root are sorted alphabetically, and then root numbers are assigned by the alphabetic order of the first alloforms of each foot. For example, the root with the alloforms (бор, бра) is given the number ‹42›, the root with the alloform бот the number ‹43›, and the root with the alloform брад the number ‹44›. Roots with coinciding alloform sets are numbered arbitrarily, cf. брад ‹44› and брад ‹45›.

### Dictionary entry structure

In the general case the entry contains the following three fields: 1) root number (headword), 2) the segmental shape of the alloform, 3) the list of lexemes where

the given root is represented by the given alloform. Cf.:


References to paragraph numbers are given for roots with commentary in the main part of the book.

The segmental shape of the alloform corresponds to the morphophonological spellout, and in some cases may fail to coincide with any graphic substring of the spellouts of corresponding lexemes. Cf. for the root ‹23› with the alloform бој, for разбои, убоиство etc.

Lexemes that represent a given alloform are given in the dictionary in ordinary alphabetic order. The symbol // after a lexeme indicates the presence of a doublet. Cf. ѥдинъ// and ѥдьнъ//. A superscript index at the end of a lexeme distinguishes homonyms, cf. вести1 (root вед ‹75›) and вести2 (root вез ‹84›).

# **Paradigmatic Dictionary**


First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9




































































































































# **Root dictionary**


First name Last name, affiliation, affiliation country, email, ORCID FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice) Anna Polivanova, Lomonosov Moscow State University, Russian Federation, akpolivanova@yandex.ru, 0000-0003-2303-7159

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 Lev Blumenfeld, Carleton University, Canada, lev.blumenfeld@carleton.ca, 0000-0002-3335-7503 Artemij Keidan, Sapienza University of Rome, Italy, artemij.keidan@uniroma1.it, 0000-0003-4944-0675 Referee List (DOI 10.36253/fup\_referee\_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, Old Church Slavic. Grammar and Dictionaries, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9






ꙗ

л҄ѥниѥ

2

2

2

изводити наводити низъводити отъводити поводатаи приводити проводити прѣдъизводити прѣпроводити разводити воЖ воЖь ‹76› ва j ваꙗниѥ изваꙗти ‹77› (see § 62, 771) вал валити възвалити извалити отъвалити привалити вал҄ ва л҄ꙗти отъва л҄ꙗти отъва л҄ѥниѥ вль вльна 1 вльнити вльньнъ вльн҄ꙗти вльн҄ѥниѥ въл вълати ‹78› вап вапа ‹79› вап повапьнити ‹80› вар варити 1 прѣдъварити вар҄ вар҄ꙗти прѣдъвар҄ꙗти ‹81› (see § 678, 809) вар варити 2 варъ възварити разварити вьр вьрѣти прѣвъзвьрѣти съвьрѣти ‹82› (see § 787) ⸨ваш⸩ вашь ‹83› ведр ведро ‹84› (see § 661) вез весло вести извести привеслати привести прѣвести воз возъ ‹85› (see § 810) вел велѣниѥ велѣти повелѣвати повелѣниѥ повелѣти вол волити вольнъ доволъ довольнъ изволити неповольнъ вол҄ во л҄ изво вьл довьльство довьлѣти ‹86› вел велии велииство великота великъ величаниѥ величати величиѥ величити величьствиѥ величьство възвеличати възвеличити прѣвеликъ ‹87› ⸨вельбѫд ⸩ вельбѫдъ


възвѣсити

‹98› вид видовати видъ видьнъ видьць видѣниѥ видѣти възавидѣти възненавидѣти завида завидьливъ завидѣти зависть завистьливъ завистьникъ невидѣниѥ незавидьливъ незавидьнъ независтьнъ ненавидѣниѥ ненавидѣньнъ ненавидѣти ненавистиѥ ненависть ненавистьникъ ненавистьнъ обида обидьливъ обидьливьство обидѣти позавидѣти привидѣниѥ прѣобидѣти уобидѣти ‹99› вин винар҄ь вино виньнъ2 ‹100› ⸨виноград⸩виноградъ виноградьнъ ‹101› (see § 663) вис висѣниѥ висѣти вѣс въвѣсити низъвѣсити обѣсити повѣсити прѣвѣса съвѣсити вѣш обѣшати ‹102› вит витати обитати обитѣль обитѣльнъ привитати прѣвитати ‹103› (see § 62) влаг влага ‹104› (see § 62) влад владыка владычица владычьн҄ь владычьскъ владычьствиѥ владычьство властел҄ь властел҄ьскъ власти власть обладаниѥ обладати область областьнъ съвласти ‹105› (see § 62, 772) влак облакъ влач влачити извлачити облачити облачьнъ облачьць привлачити развлачити съвлачити вльк влькъ вльч вльчьць влѣк влѣщи въвлѣщи

обильнъ

извлѣщи облѣщи отъвлѣщи повлѣщи привлѣщи прѣоблѣщи съвлѣщи съоблѣщи ‹106› (see § 62) влас власъ власѣнъ ‹107› (see § 62, 811) влъс влъснѫти влъх влъхвованиѥ влъхвовьнъ влъхвъ влъш влъшьба влъшьбьнъ влъшьскъ влъшьствиѥ влъшьство ‹108› (see § 62) вльн вльна 2 ‹109› вод безводьнъ вода водьнъ наводиѥ наводьнити ‹110› вол воловьнъ волъ ‹111› (see § 812) вон҄ вон҄ꙗ вон҄ꙗти обон҄ꙗти ѫ обѫхати ‹112› воск воскъ ‹113› ⸨вощаг ⸩ вощага ‹114› (see § 62) враб врабии

‹115› (see § 62) враг врагъ враж вражениѥ вражии вражьда вражьдовати вражьскъ ‹116› (see § 62) вран вранъ ‹117› (see § 62, 746) врат вратити вратъ 1 вратъкъ възвратити обратити отъвратити прѣвратити развратити съвратити вращ възвращати възвращениѥ обращати обращениѥ отъвращати прѣвращати развращати развращениѥ съвращати увращати врьт врьтѣти врѣт безврѣменьнъ врѣменьнъ врѣмѧ ‹118› (see § 62) врач врачевати врачевъ врачевьнъ врачевьскъ врачь врачьба врачьбьнъ изврачевати ‹119› (see § 62) врьб врьбиѥ врьбьница




засъвѣдѣтел҄ьствовати извѣдѣти извѣстити извѣстовати извѣсть извѣстьнъ извѣщаниѥ извѣщати2 извѣщениѥ исповѣданиѥ исповѣдати исповѣдовати исповѣдь исповѣдьникъ исповѣдьница исповѣдѣти навѣдѣти невѣста невѣсть невѣстьникъ невѣстьство недовѣдѣти несъвѣда несъвѣдьнъ несъвѣстьнъ повѣданиѥ повѣдати повѣдовати повѣдь повѣдѣти повѣсть повѣстьнъ проповѣданиѥ проповѣдател҄ь проповѣдати проповѣдовати проповѣдь проповѣдьникъ проповѣдѣти проувѣдѣти съвѣдь съвѣдѣниѥ съвѣдѣтел҄ь съвѣдѣтел҄ьствиѥ съвѣдѣтел҄ьство съвѣдѣтел҄ьствовати съвѣдѣти съвѣстовати съвѣсть



грѣхъ грѣшьникъ грѣшьница грѣшьничь грѣшьнъ погрѣшениѥ погрѣшити прѣгрѣшениѥ прѣгрѣшити съгрѣвати съгрѣти съгрѣшати съгрѣшениѥ съгрѣшити ‹155› гас гасити гаснѫти погасити угасати угасити угаснѫти гаш угашениѥ ‹156› (see § 657) гваЖ пригваЖати гвоЖ пригвоЖениѥ гвозд гвоздии гвоздиинъ гвоздь пригвоздити ‹157› ⸨геон⸩ геона геоньскъ ‹158› глав възглавьница глава главизна главьнъ главл҄ оглав л҄ѥниѥ ‹159› (see § 819, 835) ⸨гла.гол⸩възглаголати глаголаниѥ глаголати глаголъ изглаголати неизглаголаньнъ оглаголаниѥ оглаголати проглаголати съглаголати ‹160› глад гладивъ гладъ ‹161› глад гладити гладъкъ загладити изгладити погладити глаЖ заглаЖати ‹162› (see § 819) глас възгласити гласити гласовати гласъ огласити пригласити прогласити съгласьнъ глаш въглашениѥ възглашати възглашениѥ глашати оглашати оглашениѥ приглашати ‹163› глин глиньнъ ‹164› глум глумити глумъ непоглумьнъ поглумити глумл҄ глум л҄ѥниѥ ‹165› глух глухъ ‹166› глът поглътити глъщ поглъщати ‹167› гльб угль ( б

)нѫти


подъгориѥ ‹187› ⸨господ ⸩ господа господинъ господин҄ь господын҄и господь господьн҄ь господьскъ господьствиѥ господьство ⸨госпоЖ⸩ госпоЖа ‹188› гост гостиница гостиньникъ гостиньница гостиньць гостити гость гостьникъ ‹189› ⸨готов ⸩ готовати готовити готовъ приготовати приготовити съготовати съготовити уготованиѥ уготовати уготовити ‹190› (see § 661, 667) граб грабити разграбити грабл҄ граб л҄ѥниѥ разграб л҄ѥниѥ греб гребениѥ грети погребениѥ погребите л҄ ь погрети гриб погрибати гроб гробище гробъ гробьнъ грѣб погрѣбаниѥ погрѣбати ‹191› град градъ 2 ‹192› град безградьникъ възградити градити градъ 1 градьникъ градьнъ градьскъ градьць заградити заградьнъ ограда оградити приградъ прѣградити граЖ възграЖати възграЖениѥ граЖанинъ граЖениѥ граЖь ограЖениѥ ‹193› гран гранъ ‹194› гроз гроза грозьнъ ‹195› гроз гроздъ грозновиѥ грознъ ‹196› (see § 664, 681) гром громовъ громъ громьнъ гръм възгръмѣти гръмѣти ‹197› гръд възгръдѣниѥ гръдость гръдъ гръдын҄и


дарьствовати дате л҄ ь дати издаꙗти издати неуподар҄ѥниѥ одарити отъдаꙗти отъданиѥ отъдати подаꙗти подавати подар҄ꙗти подате л҄ ь подати придаꙗти придавати придати продаꙗти проданиѥ продати прѣдаꙗниѥ прѣдаꙗти прѣдавьникъ прѣданиѥ прѣданьникъ прѣдате л҄ ь прѣдати раздаꙗти раздавати раздавьникъ дад подадите л҄ ь прѣдадите л҄ѥвъ ‹217› дав давити подавити удавити давл҄ подавл҄ꙗти удавл҄ѥниѥ удавл҄ѥнина ‹218› дав давьнъ

‹219›

дал дальн҄ь

подалиѥ удалити дал҄ удал҄ꙗти


‹220› (see § 750) дар ударити дар҄ удар҄ꙗниѥ удар҄ѥниѥ дир въздирати дира раздирати

дор раздоръ дьр въздьрати дьраниѥ дьрати издьрати одьрати прѣдьрати раздьрати

‹221› (see § 678, 821) двар҄ въдвар҄ꙗти удвар҄ꙗти

двор въдворити дворъ


‹235›







‹297› (see § 62) жрѣб жрѣбьць жрѣбѧ ‹298› жуп жупище ‹299› ⸨жупан ⸩ жупанъ ‹300› ⸨жупел ⸩жупелъ ‹301› жьзл жьзлиѥ жьзлъ ‹302› (see § 683, 818) жѧ жѧтва жѧтвьнъ жѧтел҄ꙗнинъ жѧтел҄ь жѧти пожѧти ‹303› жѧд въЖѧдати жѧданиѥ жѧдати жѧдьнъ жѧдѣти жѧЖ жѧЖа ‹304› жѧл жѧло ‹305› (see § 62) ѕвѣзд ѕвѣзда ‹306› ѕѣл ѕѣло ‹307› заЖ заЖь ‹308› ⸨заjѧц ⸩ заѩць ‹309› (see § 62, 753) зар озарити зар҄ зар҄ꙗ// оза р҄ꙗти зир възирати

зазирати







клѧтва клѧти проклѧти ‹385› клок клокотати ‹386› клосн оклоснити ‹387› кльц кльцати ‹388› кльч кльчьтати кльчьтъ ‹389› клѣст съклѣстити клѣщ съклѣщати ‹390› клѣт клѣтъка клѣть ‹391› клѧч клѧчати ‹392› ⸨кл҄евет ⸩ кл҄ѥвета кл҄ѥветар҄ь кл҄ѥветати кл҄ѥветьникъ кл҄ѥветьнъ окл҄ѥветавати окл҄ѥветаниѥ окл҄ѥветати ‹393› кл҄ус кл҄юсѧ ‹394› кл҄уч закл҄ючити кл҄ючар҄ь кл҄ючати кл҄ючити кл҄ючь прикл҄ючаи прикл҄ючити съкл҄ючити ‹395› коб кобь

‹396› ⸨кобыл ⸩ кобыла ‹397› (see § 709) ков ковар҄ьство ковъ ковьникъ къ къзнь кы кызнь кызньникъ ‹398› ков ковати наковало наковально оковати оковъ поковати ‹399› ⸨ковьчег ⸩ ковьчегъ ⸨ковьчеж ⸩ ковьчежьць ‹400› кож кожа кожьнъ коз коза козьлищь козьлъ козьлѧ козьл҄ь ‹401› (see § 653) ко j покои покоинъ покоити прѣпокоити ‹402› кок кокотъ кокошь ‹403› кол колесьница колесьничьнъ коло ‹404› ⸨колѣб ⸩ въсколѣбати колѣбаниѥ колѣбати
















<612> няжд НЖЖДА ног нога няждынкъ ногътъ няждыгр нож ножень ПОНЯЖДАТИ подъножик ПОННЖЖДАТИ ((пазнєг)) ПАЗНЕГЪТЪ <622> <613> HIRB મિલાજીન (ноздр)) ноздри ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ผู้ <614> <623> (see § 655) ношт ношть രങ്ങ ФБЫШТЄВАТИ ноштынъ ФВРШТЕНИК ФБНОШТЬНИЦА ОБЫШТИК ОБНОШТЫНЪ ФБЫШТИНА ФБЫШТИТИ <615> (see § 62) ФБЫШТЬ НОАВ НОАВЪ ФБЫШТЫНИКЪ <616> (see § 714) ФБЫШТЫНЪ HOY O нзноурити ФБЫШТЬСТВО првизноұрити ПОНОБЫШТАВАТИ HOVQ нзнограсник ПОНОБЫШТАТИ Приобыштити ныр пронъривъ пронърне <624> (see § 795) пронърити രങ്ങ оба проныръ ОБОДАКЪ пронъюрсство овон <617> <625> ны ОУНЫВАТИ ((ОБЛАШ)) ОВЛАШЬ 0/14/9/14/16/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14/14 ФБЛАШЬСКЪ of Historicher in the ОУНЫТИ <626> (обьд)) обьдо <618> મજાી нъмнашьнь <627> OB овых <619> овынь ныр НЫРИШТЄ ОВЫЦА <620> овычь ньм НЪМОВАТИ овыча нѣмость <628> (see § 796) ньмъ ов овъ онѣмቴፕи ньмл онъмлати <629> ((овошт)) овоштє <621> ОВОШТЫНЪ НЯДИТИ няд ПОНЯДИТИ <630> принждити ((огав)) огавни

<631> (see § 655) осьла orfi orific ОСЬЛАТИНЪ orginal осыль <632> <644> одръ ((оскръд))оскръдъ одр одрыць <645> <633> осм осмъ ((ојьм)) онми осмь онтьскъ <646> онмьство ост нзострити <634> наострити ок БЕЗОКОВАТИ осла око острик БєЗОЧЬСТВО ભ острити оческих острогъ острость <635> остръ ((олов)) ОЛОВО остьнъ оловънъ <647> (see § 655) <636> (see § 797) OT отъщь ભા ОНУ ОТЪЧИНА <637> отъчь ((оногшт)) отъчьствию ОНОУШТА отъчьство <638> првотыць ((опаш )) ОПАШЬ <648> <639> (see § 759) ((оцьт)) оцьтъ op орало оцетьнъ орати оцетвиъ рало ρα <649> ратан ПАД ВЪПАДАТИ <640> ВЪПАДЕНИЕ op орьлъ ВЪПАСТИ орьль ДОПАСТИ <641> ЗАПАДЪ ЗАПАДЬНЪ ((оржд)) орждик ЗАПАДЬСКЪ <642> ИСПАДАТИ ((орхж)) въоржити ИСПАСТИ оржжию НАПАДАТИ оржжынкъ НАПАДЕНИЕ огоряжити НАПАСТИ напасть <643> НАПАСТЪСТВОВАТИ oc осьлъ НИЗЪПАСТИ осьаьскъ

отъпадати отъпадениѥ отъпасти паданиѥ падати падениѥ пасти 1 подъпадати припадати припасти пропадь пропасть распадати распалина распалиньнъ распасти съпадати съпасти 1 упасти 1 ‹650› ⸨пазух ⸩ пазуха ‹651› (see § 737) па j напаꙗниѥ напаꙗти упаꙗти пи испити пиво пивьца пиръ пити питиѥ питии упивати упити упитиѥ по j напоити поити 2 упоити пь j пиꙗница пиꙗнъ пиꙗньствиѥ пиꙗньство ‹652› пак пакость пакостьникъ ‹653› пал палица ‹654› (see § 829) пал запалити опалити палити подъпалити попалити пал҄ опал҄ꙗти пал҄ѥниѥ попал҄ꙗти распал҄ѥниѥ ⸨пепел ⸩ пепелъ// упепелити пла въспланѫти пламень пламеньнъ пламѣнъ пол полѣти ⸨попел⸩ попелъ// ‹655› пан паница паничица ‹656› пап папа папежь ‹657› ⸨папрьт ⸩ папрьтъ ‹658› (see § 760) пар парити пер въперити перьнатъ пьр пьрати ‹659› пас опасениѥ опасивъ опасъ опасьнъ паства паствина паствиньнъ пасти 2 пастуховъ пастухъ пастыр҄ь





простын҄и прѣпростъ прощ прощениѥ ‹710› при j неприꙗзнинъ неприꙗзнь неприꙗзньскъ неприꙗзньство приꙗзнивъ приꙗзнь приꙗтел҄ь приꙗти 1 ‹711› присн присность приснъ ‹712› ⸨притран ⸩ притранъ ‹713› (see § 653) про прокъ прокыи прочии ‹714› прот противити противл҄ꙗти противл҄ѥниѥ противъ противьникъ противьничь противьнъ противьство сѫпротивити сѫпротивьникъ сѫпротивьнъ ‹715› прът прътищь прътъ ‹716› прьв прьвъ прьвѣньць ‹717› прьс прьси ‹718› (see § 762) прьст прьстень ‹721› ‹727›

прьстъ ‹719› (see § 653) прѣ прѣдьн҄ь прѣЖьникъ прѣЖьн҄ь ‹720› ⸨прѣгын҄ ⸩ прѣгын҄ꙗ прѣм прѣмьдивъ прѣмьн҄ь ‹722› прѣсн опрѣснъкъ опрѣснъчьскъ ‹723› прѣт въспрѣтити запрѣтити прѣтити прѣщ въспрѣщати въспрѣщениѥ запрѣщати запрѣщениѥ прѣщениѥ ‹724› (see § 698) прѧг въпрѧщи припрѧщи распрѧщи съпрѧщи прѫг прѫгъ сѫпрѫгъ прѫж прѫжати ‹725› прѧд прѧсти ‹726› прѧд въспрѧнѫти прѧт въспрѧтати прѧтати съпрѧтати ‹728› прѫд прапрѫда// прапрѫдъ// прапрѫдьнъ// прѣпрѫда//





рыло рыти ‹774› рог рогъ рож рожанъ рожьць ‹775› ⸨рогоз ⸩ рогозина ‹776› роЖ роЖиѥ розг розга ‹777› ⸨розвьн ⸩розвьнъ ‹778› рос роса росити ‹779› рот ротити ‹780› (see § 718) ру въздрути рути ‹781› (see § 719) руд руда румѣнъ ръЖ ръЖа ‹782› руш раздрушати раздрушениѥ раздрушити рушити ‹783› ръз ръзати ‹784› ръп бездръпътьнъ обръпътити поръпътати ръпътаниѥ ръпътати ръпътъ ‹785› рыб рыба рыбар҄ь рыбитвъ рыбица ‹786› рыд въздрыдати рыданиѥ рыдати ‹787› рьв въздрьвьновати порьвьновати рьвьнивъ рьвьниѥ рьвьнител҄ь рьвьновати рьвьность ‹788› рѣд рѣдъкъ ‹789› рѣп рѣпиѥ ‹790› рѣсн върѣснити рѣснота рѣснотивьнъ ‹791› (see § 831) рѣт изобрѣсти изобрѣтати изобрѣтениѥ обрѣсти обрѣтател҄ь обрѣтати обрѣтел҄ьникъ обрѣтениѥ обрѣтѣль приобрѣсти приобрѣтати приобрѣтел҄ьникъ приобрѣтениѥ приобрѣтѣль сърѣсти сърѣтаниѥ сърѣтати сърѣтениѥ рѧщ сърѧща ‹792› рѣш издрѣшениѥ




скврьн҄ѥниѥ ‹828› склаб осклабити склабити склабл҄ осклабл҄ѥниѥ ‹829› (see § 62) ⸨сковрад⸩ сковрада ‹830› скоп скопити скопьць ‹831› скоп заскопиѥ ‹832› скор наискорѣи скорость скорота скоръ ‹833› скот скотии скотъ скотьнъ ‹834› ⸨скрижал ⸩ скрижаль ‹835› скрин скриница скрин҄ скрин҄ꙗ ‹836› скръб бескръбьнъ въскръбѣти оскръбити оскръбѣти поскръбѣти прискръбьнъ прѣскръбьнъ скръбити скръбь скръбьнъ скръбѣти скръбл҄ оскръбл҄ꙗти ‹837› скрьж поскрьжьтати скрьжьтаниѥ скрьжьтати скрьжьтъ ‹838› (see § 833) скуп проскупьство скѫп скѫпость ‹839› ⸨скълѧѕ ⸩ скълѧѕь ‹840› скѫд неоскѫдѣньнъ нескѫдьнъ оскѫдити оскѫдѣти скѫдость скѫдъ скѫдѣти ‹841› ⸨скѫдъл ⸩ скѫдълъ скѫдъльниковъ скѫдъльникъ скѫдъльничь ‹842› (see § 780) сла въсладити насладити наслаЖати наслаЖениѥ сладость сладъкъ сланъ сласть сластьнъ слатина услаЖати сол осолити солило солищьскъ соль сольнъ ‹843› слаб неослабьнъ ослаба ослабити ослабѣти раслабити раслабѣти слабость слабъ


наслѣдиѥ наслѣдити наслѣдованиѥ наслѣдовати наслѣдьникъ наслѣдьствовати неислѣдьнъ повъслѣдовати послѣдовати послѣдъкъ послѣдьн҄ь послѣдьствовати слѣдити слѣдованиѥ слѣдъ ‹858› (see § 740) сми j въсмиꙗти посмиꙗти просмиꙗти смиꙗти усмиꙗти смѣ смѣхъ ‹859› смик смикати ‹860› смок смокъвиниѥ смокъвинъ смокъвьница смокъвьничьнъ смокъвьнъ смокы ‹861› (see § 681, 764) смрад осмрадити смрадъ смрадьнъ смраЖ просмраЖати смръд въсмръдѣти смръдѣти ‹862› (see § 764) смрьд смрьдъ ‹863› (see § 721) снов основаниѥ основати сныв оснывати ‹864› сноп снопъ ‹865› снѣг снѣгъ снѣж оснѣжити ‹866› (see § 722) сов совати су исунѫти ‹867› сок сокъ ‹868› ⸨сокач ⸩ сокачии ‹869› соп сопьць ‹870› сох посоха ‹871› спод сподъ ‹872› спѣ непоспѣшьнъ поспѣти поспѣховати поспѣхъ поспѣшати поспѣшениѥ поспѣшити поспѣшьникъ поспѣшьствовати приспѣти прѣспѣвати прѣспѣти спѣти спѣхъ спѣшениѥ спѣшити спѣшьнъ съпоспѣшьникъ успѣвати успѣти успѣхъ успѣшениѥ успѣшьнъ

‹873› спѫд спѫдъ ‹874› (see § 62) срам бесрамиѥ бесрамъкъ бесрамьнъ осрамити посрамити срамити срамота срамъ срамьнъ срамл҄ осрамл҄ꙗти осрамл҄ѥниѥ посрамл҄ꙗти срамл҄ꙗти усрамл҄ꙗти ‹875› (see § 62) срач срачица ‹876› (see § 62) срьб срьбаниѥ ‹877› (see § 62, 765) срьд осрьдованиѥ срьдьце срьдьчьнъ усрьдиѥ усрьдьнъ срѣд посрѣдьн҄ь срѣда срѣдьн҄ь ‹878› (see § 62) срьп срьпъ ‹879› (see § 657) ста въстаꙗти въставати въставител҄ь въставити въставл҄ꙗти въставл҄ѥниѥ въстанивъ въстаниѥ въстаньнъ въстати достати

настаꙗти

наставити наставл҄ꙗти наставл҄ѥниѥ наставьникъ настати недостатъковати недостатъкъ недостатъчьнъ непрѣстаньнъ остаꙗти оставити оставл҄ꙗти оставл҄ѥниѥ останъкъ остати остатиѥ остатъкъ остатъчьнъ поставити поставл҄ꙗти поставл҄ѥниѥ приставити пристав л҄ꙗти приставл҄ѥниѥ приставьникъ пристанище пристати прѣдъпоставити прѣдъставити прѣдъстати прѣстаꙗти прѣставити прѣставл҄ꙗти прѣставл҄ѥниѥ прѣстати стаꙗти ставило ставити ставл҄ꙗти ставл҄ѥниѥ стадо станиѥ станъ стати съвъставити съставити съставл҄ꙗти съставл҄ѥниѥ съставъ

съставьникъ състати сѫпостатъ уиставл҄ѥниѥ уставити уставл҄ꙗти уставл҄ѥниѥ уставъ уставьнъ устати сто достоꙗниѥ достоꙗти достоинъ достоиньнъ достоиньство застоꙗти настоꙗниѥ настоꙗти непостоинъ остоꙗниѥ остоꙗти отъстоꙗти постоꙗниѥ постоꙗти прѣдъстоꙗти прѣдъстолиѥ прѣстоꙗниѥ прѣстоꙗти прѣстолъ растоꙗти стоꙗниѥ стоꙗти столъ съпрѣстольнъ състоꙗниѥ състоꙗти състольникъ устоꙗти

‹880›

стар прѣстарѣти старица старость старъ старьць старьчь старѣишина старѣишиньскъ старѣишиньство

старѣишиньствовати състарѣти ‹881› ⸨стегн ⸩ стегно ‹882› стеж стежеръ ‹883› (see § 783) стел҄ постел҄ꙗ стел҄ꙗ стил постилати стьл настьлати подъстьлати постьлати стьлати ‹884› стен въстенати стенаниѥ стенати ‹885› (see § 834) степ степень степеньнъ стоп стопа стѫп въстѫпати застѫпаниѥ застѫпати застѫпити застѫпъ застѫпьникъ истѫпити настѫпати настѫпити непостѫпьнъ непристѫпьнъ остѫпати остѫпити 1 остѫпити 2 остѫпьнъ пристѫпати пристѫпити прѣстѫпати прѣстѫпити прѣстѫпьникъ прѣстѫпьнъ растѫпити стѫпаниѥ стѫпати

стѫпити състѫпити устѫпати устѫпити стѫпл҄ застѫпл҄ѥниѥ пристѫпл҄ѥниѥ прѣстѫпл҄ѥниѥ ‹886› (see § 659) стиг постигнѫти стиж непостижьнъ постижениѥ стьг стьгна стьѕ стьѕа ‹887› (see § 62) стлъп стлъпъ стлъпьникъ ‹888› (see § 766) стра пространиѥ пространити пространъ пространьство прѣстран҄ꙗти страна страньникъ страньнъ стрѣ прострѣти распрострѣти ‹889› страб устрабити ‹890› страд бестрастиѥ бестрастьнъ непостраданьнъ острастити постраданиѥ пострадати прѣстрадати страданиѥ страдати страдьба страсть страстьникъ страстьничьскъ страстьнъ стра стро ‹893› ‹896›

‹891› (see § 663, 767) страж стража стражь стражьба стрѣг стрѣщи 1 ‹892› (see § 741) j устраꙗти j нестроѥниѥ пристроити строѥниѥ строи строинъ строител҄ь строител҄ьство строити състроити устроѥниѥ устрои устроити стрѣ стрѣха страх страхованиѥ страхъ страш пострашениѥ пострашити пристрашьнъ страшивъ страшити страшьнъ устрашати устрашениѥ устрашити ‹894› (see § 663, 767) стриг постригати стриж пострижениѥ стриѕ постриѕати стрѣг пострѣщи стрѣщи 2 ‹895› (see § 723) стров островъ стру струꙗ строп отъстропити

стропъ





тесати

‹950› тет тетъка ‹951› тик тикъ ‹952› тим тимѣниѥ тимѣно тина тинавъ тиньнъ ‹953› тин тинь ‹954› (see § 768) тир въстирати истирати отирати прѣтирати трь отрьти прѣтрьти сътрьти трьти ‹955› тис тиса ‹956› (see § 663) тиск истискати сътискати утискати тѣщ потѣщити тѣщити утѣщати ‹957› (see § 663) тих тихость тихота тихъ тиш отишати отишиѥ тишина тишити утишити тѣх утѣха тѣш тѣшити утѣшати утѣшениѥ утѣшити утѣшьнъ

‹958› (see § 62) тлап въстлапити ‹959› (see § 62) тлък протлъковати тлъкованиѥ тлъкъ ‹960› (see § 62) тлъп тлъпа ‹961› (see § 62) тлъст отлъстѣти утлъстѣти ‹962› (see § 62, 681, 784) тльк тлькнѫти тлѣк сътлѣщи тлѣщи ‹963› тол утолити ‹964› том томител҄ь 1 томител҄ь 2 томити утомити томл҄ томл҄ѥниѥ ‹965› (see § 803) тр третии третиица триѥ трои троинъ троица троичьнъ троичьскъ ‹966› (see § 728) трав отравити травити тров отровениѥ тру натрути трѣв трѣва трѣвьнъ ‹967› трат тратати ‹968› ⸨трепет ⸩въстрепетати

( п )нѫти

л҄ꙗти

трепетати трепетъ трепетьнъ ‹969› ⸨тризн ⸩ тризна ‹970› трош растрошити трошити ‹971› труд потрудити трудити трудъ трудьнъ утрудити труЖ труЖаниѥ труЖати утруЖати ‹972› труп трупиѥ трупъ ‹973› (see § 769) тръг въстръгати въстръгнѫти истръганиѥ истръгати истръгнѫти отътръгнѫти потръгнѫти протръгнѫти прѣтръгнѫти растръгати растръгнѫти трьѕ въстрьѕати отътрьѕати потрьѕати протрьѕати прѣтрьѕати растрьѕаниѥ растрьѕати трьѕаниѥ трьѕати ‹974› тръг тръгъ тръж тръжище тръжьникъ тръжьство ‹975› (see § 835) ⸨тръ.тър⸩ вътрътърати ‹976› тръх тръхъть ‹977› трьн трьниѥ трьновъ трьнъ трьнѣнъ ‹978› трьп потрьпѣти прѣтрьпѣниѥ прѣтрьпѣти сътрьпѣниѥ сътрьпѣти трьпѣливъ трьпѣниѥ трьпѣти утрь ‹979› трьст трьстиѥ трьсть ‹980› трѣб истрѣбити отътрѣбити потрѣбити прѣтрѣбивъ сътрѣбити трѣбл҄ истрѣбл҄ꙗти истрѣбл҄ѥниѥ потрѣб ‹981› трѣб потрѣба потрѣбованиѥ потрѣбьнъ трѣба трѣбище трѣбованиѥ трѣбовати трѣбьникъ трѣбьнъ трѣбл҄ наитрѣбл҄ии



TX46HB оусныхыв <1010> (see § 836) <1018> ОБОУ ТИ ov 01/07 <1011> оудъ OYA одукт <1012> (see § 655, 817) ្យ​យោ 0/3 CBHO/3PHP <1013> 0/34 0/34A <1014> oy Tp ((ογκροπ)) оукропъ <1015> oym BE30YMHE BE30/MPHHKB BE30/MPHB Въразоумити H30YMETH недоогичник <1020> нєраЗоץ'мичьнъ OYX or Xo нєраЗоУмьничьнъ ியா поразоумъти ПООУМИТИ ПООРАЗОУМЪВАТИ разоумивъ разоумичьнъ ХАЖД разоумъ разоумьливъ разоумьнъ разоумъвати ХОД разомменны разоумъти съразоумъник ОУМЪ 0/WPH/P оумъти ογλλ BE30YMAG <1016> ુલ મુની изођућьшина ્બિમાલ ਼ਮੀਲਾ ਸ нсуодити оунувшинна нсуодиштє нсуодъ <1017> нсуодьнъ orem of of carpers

HAOYCTHTH пооустити посустьнъ OYCTA ОУСТИТИ ОУСТЬНА ПООУШТАНИК ПООУШТАТИ пооуштеник <1019> (see § 655) ЗАФУТОВНЫ оутриц οΥ Τρο оутрыннца ОУТОВНИЕВАТИ оутрынь OV TOTH 01/ TO BILLAN Въночшити ЗАФУШАНИК ЗАОУШЕНИЮ ЗАФУШИТИ <1021> (see § 657) нсуаждати ПРИХАЖДАТИ Пръхаждати ҚАЖДАТИ въсходити въсходъ ВЪХОДИТИ въходъ ВЪХОДЕНЪ ДОХОДИТИ ЗАХОДИТИ ЗАХОДЪ ИСХОДАТАГАТИ нсуодатан ИСХОДАТАННИКЪ


похвала похвалити

находити


хлѫп въсхлѫпати


хулити хульникъ хульнъ хул҄ въсхул҄ꙗти похул҄ꙗти похул҄ѥниѥ хул҄ѥниѥ ‹1047› хыз хызина хызъ ‹1048› хыт въсхытати въсхытити исхытити похытати похытити расхытити съвъсхытити хытрость хытростьнъ хытръ хытрьць хыщ въсхыщати въсхыщениѥ въсхыщьникъ расхыщати хыщениѥ хыщьникъ ‹1049› ⸨хѫдож ⸩ хѫдожьникъ хѫдожьствиѥ хѫдожьство ‹1050› (see § 62, 676) цвит отъцвисти процвисти процвитати цвисти цвѣт цвѣтило цвѣтъ цвѣтьнъ цвѣтьць

цвѣтьчанъ

‹1046›

хул охулити

похулити хула

‹1037›

хлѫпати

⸨хръзан

‹1044› ⸨хрьбьт

‹1045›

худ охудити

⸩ хръзанъ

⸩ хрьбьтъ

охудѣти ухудѣти худость худъ


цѣсар҄ѥвати цѣсар҄ѥвъ цѣсар҄ити цѣсар҄ица цѣсар҄ь цѣсар҄ь цѣсар҄ьскъ цѣсар҄ьствиѥ цѣсар҄ьство цѣсар҄ьствовати ‹1059› цѣст оцѣстити цѣстити цѣщ оцѣщати оцѣщениѥ ‹1060› цѧт цѧта ‹1061› ча дочаꙗньнъ отъчаꙗниѥ отъчаꙗти причаꙗти чаꙗниѥ чаꙗти чаѥмьнъ ‹1062› чар чари ‹1063› час часъ часьнъ ‹1064› чаш чаша ‹1065› (see § 658) чез ичезнѫти чѣз ичазати ‹1066› чел чело ‹1067› ⸨чел҄ад⸩ чел҄ꙗдь ‹1068› ⸨чел҄уст ⸩ чел҄юсть

‹1069› чес чесати ⸨чешу⸩ чешуꙗ ‹1070› чет причетати съчетаниѥ съчетати ‹1071› (see § 62) ⸨четвор⸩ четворица четворъ ⸨четврь⸩ четврьтъ четврьтъкъ ⸨четыр⸩ четыре ‹1072› (see § 653) чи испочити почивати почити ‹1073› чин бесчиниѥ бещиница бещиньнъ въчинити причинити съчинити учинити чинити чинъ чин҄ въчин҄ꙗти съчин҄ѥниѥ учин҄ѥниѥ чин҄ѥниѥ ‹1074› (see § 688) чин зачинати начинаниѥ начинати учинаниѥ чѧ безначѧльнъ въчѧло въчѧти зачѧло зачѧти зачѧтиѥ начѧло начѧльникъ начѧльнъ начѧти начѧтиѥ начѧтъкъ събезначѧльнъ учѧти ‹1075› чист ичистити нечистота нечистъ очистити прѣчистъ чистина чистител҄ь чистител҄ьство чистити чистость чистота чистъ чищ очищати очищениѥ ‹1076› (see § 659) чит бечисльнъ бечисменьнъ въчисти изъчитати ичисти неищисльнъ отъчисти почисти почитаниѥ почитати причисти причитати ращисти число числьнъ чисмѧ чисти чьт бесчьствиѥ бесчьствьнъ бечьствовати бечьстиѥ бечьстьнъ нечьствовати нечьстивъ нечьстиѥ нечьстовати нечьсть

нечьстьнъ почьтениѥ причьтьникъ прѣчьсть прѣчьстьнъ чьстивъ чьстити чьсть чьстьнъ чьтениѥ чьтьць ‹1077› (see § 62) ⸨чловѣк⸩ чловѣкъ ⸨чловѣч⸩ въчловѣчениѥ въчловѣчити чловѣчинъ чловѣчь чловѣчьнъ чловѣчьскъ чловѣчьство ‹1078› (see § 62) чрьв очрьвити чрьвити чрьвь чрьвл҄ чрьвл҄ѥнь ‹1079› (see § 62) чрьм чрьмьновати чрьмьнъ ‹1080› (see § 62) чрьн очрьнѣти чрьнило чрьница чрьнъ чрьньць чрьньчьскъ ‹1081› (see § 62, 770) чрьп ичрьпати начрьпати почрьпало почрьпальникъ почрьпати чрьпати чрѣп почрѣти ‹1082› (see § 62) чрьт начрьтаниѥ

начрьтати чрьта чрьтати ‹1083› (see § 62) ⸨чрьтог⸩ чрьтогъ ‹1084› (see § 62) чрьщ очрьще ‹1085› (see § 62) чрѣв чрѣво чрѣвьнъ ‹1086› (see § 62) чрѣв чрѣвии ‹1087› (see § 62) чрѣд чрѣда чрѣЖ чрѣЖениѥ ‹1088› (see § 62) чрѣз чрѣсла ‹1089› (see § 62) чрѣн чрѣновьнъ ‹1090› чу безъчувьствьнъ въсчудити почудити почути прѣчудити учути чувьствиѥ чувьство чувьствовати чувьствьнъ чудесьнъ чудити чудо чудьнъ чуиство чути ‹1091› ⸨чьван⸩ чьванъ ‹1092› чѧд бещѧдиѥ бещѧдьнъ ищѧдиѥ чѧдо

чѧдь чѧдьце ‹1093› чѧст причѧстиѥ причѧстити причѧстьникъ причѧстьнъ съпричѧстьникъ учѧстиѥ учѧстьнъ чѧсть чѧстьнъ чѧщ причѧщаниѥ причѧщати причѧщениѥ ‹1094› чѧст чѧстъ ‹1095› шар шаръ шаръчии ‹1096› шест шестъ шесть ‹1097› ши шиꙗ ‹1098› шиб прошибати шибати ‹1099› (see § 659) шид ушидь шьд въшьствиѥ дошьстиѥ ишьствиѥ нашьствиѥ ошьльць ошьстиѥ пришьльствиѥ пришьльство пришьльствовати пришьльць пришьствиѥ пришьстиѥ съшьствиѥ шьствиѥ шьствовати шьстиѥ ‹1100› шип шипъкъ шипъчанъ ‹1101› шир раширити уширити широкъ широта шир҄ рашир҄ꙗти ‹1102› (see § 62) шлѣм шлѣмъ ‹1103› шуj шуи шуица ‹1104› шум въшумѣти шумъ шумьнъ ‹1105› (see § 729) шы съшити шити ‹1106› шьп шьпътаниѥ шьпътати ‹1107› шѧт шѧтаниѥ шѧтати ‹1108› (see § 838) щед прѣщедръ ущедрити щедрота щедръ щѧд пощѧдѣти щѧдѣти ‹1109› щет отъщетити щещ отъщещати ‹1110› щит защитител҄ь защитити

защитьникъ щитъ щищ защищати защищениѥ ‹1111› щут ощутити ‹1112› (see § 655, 804) ьн въиньнъ инакъ инокость инокъ иночьскъ иночьство инъ отьнѫдьнъ ‹1113› (see § 839) ѣ ꙗто ѣд ꙗдениѥ ꙗдовитъ ꙗдъ ꙗдь ꙗдьца ꙗсли ꙗсти изѣдениѥ изѣсти обѣданиѥ обѣдовати обѣдъ обѣсти отъꙗдъ поꙗдати поꙗсти сънѣдати сънѣдениѥ сънѣдь сънѣдьнъ сънѣсти уꙗдениѥ ‹1114› (see § 840) ѣ ꙗти въꙗти възѣти приꙗти2

прѣꙗти

‹1115› (see § 655) ѣдр ꙗдра ꙗдро ‹1116› ѧтр обѧтрити ѧтр҄ обѧтр҄ꙗти ѧщр҄ обѧщр҄ѥниѥ ‹1117› ѫг ѫгълъ ѫгъльнъ ‹1118› ѫгл ѫгль ‹1119› ѫд ѫдица ‹1120› ѫтр вънѫтрьн҄ь ѫтроба ѫтробьнъ ѫтрьн҄ь ‹1121› (see § 655) ѫч паѫчина

Indices

# **Index of words**

The index includes lexemes and wordforms (sometimes stems, e.g. бьран- or °дути), both aberrant and canonical. The order is direct-alphabetic. Numbers refer to paragraphs. Upright (not italic) numbers indicate a lexeme (or group of lexemes, cf. °дути); boldface selects those mentions in the book where the lexeme is discussed as a sample lexeme from a paradigmatic class. In all other cases numbers are italic. Aberrant forms are shown in the spellout of a particular source.

Canonical lexemes are necessarily included in the index if they are mentioned in the dictionaries of commented lexemes (omitted in this translation) (under the overview of paradigmatic classes, Chs. 11 and 20), or in the commentary on individual lexemes (in the overview of suffixes, Ch. 23). Also included are all unique lexemes. Other lexemes are included selectively. Also selectively are included canonical wordforms (including secondary or anomalous forms or forms listed for comparison with aberrant ones).

Aberrant forms are necessarily included in the index where they are listed in the book as illustrating a particular aberration. Other aberrant forms are included selectively.


FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9





















,








# **Subject index**

References are to paragraph numbers. Roman numerals indicate paragraphs of the *Introduction*, whose numbering is separate from the general numbering.

1-simplex · 289 2-base · 289 2-combi · 289 2-duplex · 289 2-pron · 289 aberrant: ~ (alternative) pairings · 111, 117 ~ derivation · X, 887 ~ spellouts (forms) · 129, 887, 890 paradigmatically ~ spellouts (forms) · 129 ~ cluster transformation · 76 segmentally ~ spellouts (forms) · 129 aberration (operation) · 887 atomic ~ · 139, 140 lexemic (lexical) ~ · 139, 892–893 modifying ~ · 141 paradigmatic ~ · 891 segmental ~ · 129, 891 *yer* ~ · 140 ad-prepositional forms of the pronoun \*и · 318

adjacency: ~ relation of grades · 100 ~ relation of formatives · 873.6 segmental ~ condition · 873.6 adjective: short (indefinite) ~ · 269 uninflectable ~ · 323 long (definite) ~ · 269 agreementless numbered lexemes · 382 agreementless numberless lexemes · 379, 382 alloforms (variants) of the formative · 13 distribution of ~ · 83 adjacent ~ · 873.6 alloformy: intraparadigmatic/interlexemic ~ · 98 ~ relation · 13 segmental ~ · 82 standard types of ~ · 82 ~ of the suffixes · 842 allographs · 132 alternation · 98, 99

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

replacement by ~ (replacements rules by ~) · 100 velar palatalization ~ · 105 substitutive softening ~ · 112 ~ series · 99 free/limited ~ · 102 ~ segmental position · 101 pairing by a given ~ see *pairing*, *morphophonological* segmental position tied to an ~ · 101 ~ grades · 99–100 series isolation condition ~ · 104 ~ undergoers · 99 comparable by ~ · 873.6 fundamental vowel ~ · 119 alternative: ~ spellouts · 129 ~ correspondences · 111, 117 see also *aberrant* ambivalent · 39 analysis-by-synthesis · 247fn anomalies: morphophonological ~ · 882 syntagmatic (segmental, phonological) ~ · 882, 907–909 anomalous: ~ replacement ц→ч, →ж · 108, 115 ~ roots · 643 morphologically ~ forms · 881 aorist: root ~ · 477 old sigmatic 1 ~ · 478 old sigmatic 2 ~ · 479 assortment of the formative see *formative* athematic verbs · 417 basic component · 876 basic correspondence · 130, 134 basic stem see *verb stem* basic writing systems · 134 benchmark corpus · I benchmark list of lexemes, wordforms · 3, 257 bicomponential terminals · 295 boundary adjustment · 251–252, 260 ~ rules · 260 ~ rules for verbs · 462 ~ rules for nominals · 288 bundle · 406

C-declension (consonant declension) · 391 c-simplex · 289 canon · IV canonical OCS · IV category values · 243 Čistovič's thesis · 246fn clitics · 868 clusters · 59, 73 combinations with sonants · 677, 681 comparative · 279, 919–924 compositionality see *principle of compositionality* conditional · 546 conjugation: *e*-conjugation · 458 *i*-conjugation · 458 CVC ambivalence · 39 CVC schema · 36, 38 CVC-ambivalent suffixes · 95, 883 CVC norm: ~ of the formative · 36 ~ of the word · 42 D2 see *personal dative* declension (type): basic twofold ~ · 299 C-declension · 391 *i*-declension · 299 monovariate ~ · 299 pronominal ~ · 299 twofold ~ · 299 *u*-declension · 391 ~ type deformations · 263, 301 ~ and nominal inflectional classes · 299 definite adjective see *adjective* deformation: ~ of nominal terminals · 392 ~ of the declension type · 263, 301 degrees of comparison see *comparative* domain of an alternation · 102 double substitutive softening · 118 doublet wordforms (lexemes) · VIII, 877 effect: morphophonological ~ (double substitutive softening) · 118 paradigmatic ~ see *paradigmatic effect* epenthetic *i̯*· 6, 70

expanded/syncopated: ~ workstem · 285, 286 ~ form · 286 extraparadigmatic: ~ lexemes · 249fn ~ forms · 249 fall of the *yers* see *aberration*, *yer* ~ finite forms of the verb see *representations* first palatalization · 110, 869.4 fork · 130 segmentally decidable ~ · 130fn form see *spellout* formative family see *formative* formative · 6 ~ autonomous uses as prepositions · 637 ~ comparable by the alternation · 873.6 alloforms of ~ · 13 assortment (family) of ~ · 78 autonomous ~ · 36fn family (assortment) ~ · 78 finally ambivalent ~ · 39 hard/soft subtype of the twofold ~ · 85 initially ambivalent ~ · 39 monovariate ~ · 13, 78 nonstandard ~ · 38 overslash/underslash variant of the twofold ~ · 85 polyvariate ~ · 13, 78 positional classes of the ~ · 6, 36 segmental positions of the ~ · 37 segmentally adjacent ~ · 873.6 variant ~ 13fn free paradigm · 243 geminates · 61 gloss · 2.3 grade of the alternation · 99–100 heavy/light ~ · 125 adjacent ~ · 100 grammatical classes · 249 hard/soft variety: ~ of the twofold declension· 299 ~ of the twofold formatives · 85 *i*-declension · 299 imperfect: contracted ~ · 467 iotated ~ · 471 present ~ · 469 indefinite adjective see *adjective*

inflection see *terminal* inflectional class of the nominal · 299 *j*-present · 465 Jakobson's law · 43, 875 kamora · 17 kamorated: ~ letters · 19 ~ phonemes · 30 Kuryłowicz's thesis · 904 *l-epentheticum* · 117fn labile · 95, 128 law of the velars · 56, 869 lexical component of the verb · 422, 876.4 long adjective see *adjective* loose formative adjacency · 53 megafamily · 911 megalexeme · 911 morphological balance rule see *rule* morphological composition · 250, 876 ~ of a word (pRs-schema) · 40 ~ of nominal forms · 283 ~ of verbal forms · 409 morphological gender · 267 morphological skeleton · 251 morphoneme · 6fn morphophonological: ~ variants · 84 ~ variation · 82 ~ soft consonants · 28 neutralization · 870 new CVC ambivalence · 97 new ш-participle · 473 nominal forms of the verb see *representations* nonstandard aorist · 476 normalization · 1, 4 opaque: ~ parse · 628 ~ root · 628 ~ stem · 628 pairing: adjacent ~ · 873.4 alternative ~ in the substitutive softening · 117 alternative ~ in the velar palatalization · 111 morphophonological ~ (by a given alternation) · 873.3

phonological ~ · 873.2 phonological consonant ~ · 31 phonological vowel ~ · 25 rule of horizontal and vertical ~ · 678, 874.4 segmental ~ · 873 vertical/horizontal ~ · 125 paradigmatic: ~ address · 245 ~ call · 245 ~ cell · 243 ~ class · 255 ~ classes of the verb · 416 ~ construction · 242, 246 ~ index · 257 ~ lexemes · 249 ~ standard · 255, 429 ~ standard of the verb · 429 paradigmatic effect · 263 absence of substitutive softening ~ · 437 alien terminals in the paradigm ~ · 301, 386 alien workstem expansion ~ · 433, 438fn C-final stem arrest ~ · 433 labileness in the paradigm ~ · 433, 437 substitutive softening ~ · 433 syncopated stems in the paradigm ~ · 301 unstable root vocalism ~ · 437 parent: ~ lexeme of the comparative · 279 ~ verb · 275 participle: л-participle (л-Part) · 253, 407 м-participle (м-Part) · 407 н-participle (н-Part) · 407 т-participle (т-Part) · 407, 483 ш-participle (ш-Part) · 407 щ-participle (щ-Part) · 407 parts of speech · 249fn personal dative (D2) · 282, 356 positional classes of formatives · 6, 36 primary forms see *secondary forms* principle of compositionality · 130 profile · 256 pRs schema · 6, 40

pure/sonant vocalism · 121 quasisegments · 99, 873 reduplication · 41 representation (spellout): graphic ~ · 1 morphophonological ~ · 1, 6 phonological ~ · 1, 5 phonological ~ of alternative spellouts · 136 representations: nominal ~ (short and long adjectives) · 273 verbal ~ (finite and nominal) · 405 root: anomalous ~ · 643 closed ~ · 38 labile ~ · 128 opaque ~ · 628 open ~ · 38 pronominal ~ · 643 unique ~ · 643 rule: ~ of horizontal and vertical correspondences pairings · 678, 874.4 grade replacement ~ · 100 ~ of syntagmatic depth reduction · 48 basic stem allotment ~ · 427 boundary adjustment ~ for nominals · 288 boundary adjustment ~ for verbs · 462 cross-class ~ · 874.5 CVC agreement ~ · 93 Havlík's ~ · 898 ~ of morphological balance · 459 twofold ~ · 86 second palatalization · 110, 869.4 secondary forms (wordforms) · 262, 880 ~ of the nominals · 282 ~ of the verb · 408 segment · 8 segmental rewrite rules · 869.3–4 segmentally decidable rules · 65, 130fn, 869.3 segmentless substitutive softening · 864 series isolation condition · 104 short adjective see *adjective* simple consonants · 30 sonant vocalism · 121

sonants · 30 spellout (representation): aberrant ~ · 129, 890 alternative ~ · VIII, 129, 886 corrupt (erroneous) ~ · 15, 129, 890 graphically alternative ~ · 129, 890 inflectional ~ · 251 inflectional ~ of comparatives · 281 inflectional ~ of participles · 278 morphologically strange ~ · 890 paradigmatically aberrant ~ · 129 segmentally aberrant ~ · 129 starting forms of comparatives, щ- and ш-participles · 285 stem: expanded/syncopated ~ of the nominals · 285–286 workstem · 252, 258 see also *verb stem* subparadigm · 243 substitutive softening see *alternation* suffix: ~ alloformy · 842 CVC ambivalent ~ · 95, 883 standard/nonstandard ~ · 841 twofold ~ · 90, 884 supine: ~ for velar-final stems · 88 ~ terminal · 407 syncopated workstem, form see *expanded*/*syncopated* syntactic gender · 267 syntagmatic depth reduction · 48, 900 system: PRAE, IMF, INF-AOR · 405 tense *yer* · 898 terminal (inflection) · 6, 36 bicomponential/monocomponential ~ · 293–295 decomposition of verbal ~ · 295–298 decomposition of verbal ~ · 611 hard and soft varieties of ~ · 299 infinitive ~ · 407 monovariate ~ · 292 ~ as paradigmatic variant · 259 supine ~ · 407 twofold ~ · 85 twofold ~ of the nominals · 292 twofold ~ of the verb · 458

terminal set · 259 twofold/monovariate ~ · 292 standard/nonstandard ~ · 292 third palatalization · 110 tight formative adjacency · 53 twofold: ~ declension type · 299 ~ formatives · 85 ~ rule · 86 ~ suffixes · 90, 884 ~ terminal sets · 292 ~ terminals (inflection) · 85 ~ variance · 84–86 nominal ~ terminals (inflection) · 292 new ~ variance · 91 secondary ~ variance · 863 verbal ~ terminals (inflection) · 458 *u*-declension · 391 u-simplex · 289 unique: ~ lexemes · 255 ~ nominal lexemes · 302, 357 ~ roots · 643 ~ verbs · 417 unstable vocalism see *paradigmatic effect* variants of the formative see *alloforms of the formative* velar palatalization see *alternation* verb class competition · 912, 913 verb family · 421 verb stem: basic stem/workstem ~ · 413, 431 basic expanded/truncated ~ · 413, 431 basic ~ allotment rules · 427 expanded/truncated workstem ~ · 428 verb classes see *paradigmatic classes* verbal platform · 865 verbs: irregular ~ · 419, 434, 454 regular ~ · 419, 429 vocalic realization · 123 vocalism: pure ~ · 121 sonant ~ · 121 unstable ~ see *paradigmatic effect* vocative · 282, 355

wordform · 2, 248 workstem see *stem yer* confusion, fall, regression, strengthening · 898 see also *aberration*, *yer* ~

# **Bibliography**

# Studies

### Čistovič

L. A. Čistovič, A. V. Vencov, and M. P. Granstrem. *Fiziologija reči. Vosprijatie reči čelovekom*. Leningrad: Nauka, 1976.

### Diels

P. Diels. *Altkirchenslavische Grammatik. Mit einer Auswahl von Texten und einem Wörterbuch*. Heidelberg: C. Winter, 19632.

### Durnovo 1929

N. N. Durnovo. "Mysli i predpoloženija o proisxoždenii staroslavjanskogo jayzka i slavjanskix alfavitov", *Byzantinoslavica* 1, 1929: 48–85.

#### Durnovo 2000

N. N. Durnovo. *Izbrannye raboty po istorii russkogo jazyka*. Moskva: Jazyki slavjanskoj kul'tury, 2000.

### Dybo 1981

V. A. Dybo. *Slavjanskaja akcentologija. Opyt rekonstrukcii sistemy akcentnyx paradigm v praslavjanskom*. Moskva: Nauka, 1981.

#### Dybo 2000

V. A. Dybo. *Morfonologizirovannye paradigmatičeskie akcentnye sistemy. Tipologija i genezis*. Vol. I. Moskva: Jazyki russkoj kul'tury, 2000.

#### Gasparov

B. Gasparov. *Old Church Slavonic*. München: Lincom Europa, 2001.

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

### Havlík

A. Havlík. "K otázce jerové v staré češtině", *Listy filologické*, 16 (1889): 45-51.

### Jakobson

R. O. Jakobson. "Russian conjugation", *Word* 4, 1948: 155–167.

### Kraveckij

A. G. Kraveckij and A. A. Pletneva. *Istorija cerkovnoslavjanskogo jazyka v Rossii (konec XIX–XX v.)*. Moskva: Jazyki russkoj kul'tury, 2001.

### Kuryłowicz 1948

J. Kuryłowicz. "Contribution à la theorie de la syllabe", *Biuletyn Polskiego Towarzystwa Językoznawczego* 8 (1948): 80–114.

### Kuryłowicz 1952

J. Kuryłowicz. "Uwagi o polskich grupach spółgłoskowych", *Biuletyn Polskiego Towarzystwa Językoznawczego* 11 (1952): 54–68.

### Leskien

A. Leskien. *Handbuch der altbulgarischen (alkirchenslavischen) Sprache. Grammatik. Texte. Glossar*. Weimar: Böhlau, 1871.

### Lunt 1974

H. G. Lunt. *Old Church Slavonic Grammar*. The Hague, Paris: Mouton, 19746.

### Lunt 2001

H. G. Lunt. *Old Church Slavonic Grammar*. Berlin, New York: Mouton de Gruyter, 20017.

### Meillet

A. Meillet. *Le slave commun, seconde édition revue et augmentée avec le concours de A. Vaillant*. Paris: Champion, 19342.

### Plungjan

V. A. Plungjan. *Obščaya morfologija: vvedenie v problematiku. Učebnoe posobie*. Moskva: Ėditorial URSS, 2000.

### Polivanova 2001

A. K. Polivanova. "Formal'naja paradigmatika i klassy slov v russkom jazyke", in W. Gruszczyński (ed.), *Nie bez znaczenia. Prace ofiarowane Profesorowi Zygmuntowi Saloniemu z okazji jubileuszu 15 000 dni pracy naukowej*. Białystok: Wydawnictwo Uniwersytetu w Białymstoku, 2001, pp. 211–228.

### Polivanova 2008

A. K. Polivanova. "Principy postroenija segmentnoj grammatiki", in *Obščee i russkoe jazykoznanie*. *Izbrannye raboty*. Moskva: RGGU, 2008, pp. 210–303.

#### Remneva

M. L. Remneva. *Staroslavjanskij jazyk. Učebnoe posobie*. Moskva: Akademičeskij proekt, 20042.

### Ščepkin

V. N. Ščepkin. *Rassuždenie o jazyke Savvinoj knigi*. Sanktpeterburg: Izdanie Imperatorskoj Akademii Nauk, 1899.

### Trubeckoj 1954

N. S. Trubetzkoy. *Altkirchenslavische Grammatik. Schrift-, Laut- und Formensystem*. Wien: Rohrer, 1954.

### Trubeckoj 1939

N. S. Trubetzkoy. *Grundzüge der Phonologie*. Prague, Leipzig: Harrassowitz, 1939.

### Uspenskij

B. Uspenskij. *Principles of structural typology*. The Hague: Mouton, 1968.

#### Vaillant

A. Vaillant. *Manuel du vieux slave*. Paris: Institut d'études slaves, 1948.

### Van Wijk

N. Van Wijk. *Geschichte der altkirchenslavischen Sprache*. Berlin, Leipzig: W. de Gruyter, 1931.

### Xaburgaev

G. A. Xaburgaev. *Staroslavjanskij jazyk*. Moskva: Prosveščenie, 1974.

### Zaliznjak 1967

A. A. Zaliznjak. *«Russkoe imennoe slovoizmenenie» s priloženiem izbrannyx rabot po sovremennomu russkomu jazyku i obščemu jazykoznaniju*. Moskva: Jazyki slavjanskoj kul'tury, 20022 [first ed. 1967].

#### Zaliznjak 1985

A. A. Zaliznjak. *Ot praslavyanskoj akcentuacii k russkoj*. Moskva: Nauka, 1985.

### Zaliznjak 2004

A. A. Zaliznjak. *Drevnenovgorodskij dialekt*. Moskva: Jazyki slavjanskoj kul'tury, 2004.

# Sources

#### *Codex Assemanius* (As)

J. Kurz (ed.), *Codex Vaticanus 3. Slavicus glagoliticus*. Vol. II. Praha: Nakladatelství Ceskoslovenské Akademie Věd, 1955.

#### *Codex Marianus (*Mar)

V. Jagić (ed.). *Quattuor Evangeliorum versionis palaeoslovenicae codex Marianus glagoliticus, characteribus cyrillicis transcriptum*. Berlin: Weidmann, 1883.

*Codex Suprasliensis* (Supr*)*

S. Sever'janov (ed.). *Suprasl'skaja rukopis'*. Sanktpeterburg: Izdanie Imperatorskoj Akademii Nauk, 1904.

J. Zaimov and M. Capaldo (eds.), *Suprasălski ili Retkov sbornik*. Sofija: Bălgarskata Akademija na Naukite, 1982–1983.

### *Codex Zographensis (*Zogr)

V. Jagić (ed.), *Quattuor evangeliorum codex glagoliticus olim Zographensis nunc Petropolitanus*, Berlin: Weidmann, 1879.

### *Enina Apostle*

K. Mirčev and X. Kodov (eds.). *Eninski apostol. Starobalgarski pametnik ot XI vek*. Sofija: Bălgarskata Akademija na Naukite, 1965.

*Euchologium Sinaiticum (*Euch)

R. Nahtigal (ed.). *Euchologium Sinaiticum. I. Fotografski posnetek; II. Tekst s komentarjem*. Ljubljana, 1941–42.

L. Geitler (ed.). *Euchologium. Glagolski spomenik manastira Sinai brda*. Zagreb: Hartman, 1882.

### *Glagolita Clozianus* (Cloz)

A. Dostál (ed.). *Clozianus. Staroslověnský hlaholský sborník tridentský a innsbrucký*. Praha: Nakladatelství Ceskoslovenské Akademie Věd, 1959.

### *Hilandar folios*

S. M. Kul'bakin (ed.). *Xilandarskie listki. Otryvok kirillovskoj pis'mennosti XI v*. Sanktpeterburg: Izdanie Imperatorskoj Akademii Nauk, 1900.

### *Kiev Missal (*Kiev)

V. V. Nimčuk (ed.), *Kyjivs'ki hlaholični lystky*. Kyjiv: Naukova dumka, 1983. V. Jagić (ed.). *Glagolitica. Würdigung neuentdeckter Fragmente*. Wien: F. Tempsky, 1890.

### *Psalterium Sinaiticum (*Ps Sin)

S. N. Sever'janov (ed.). *Sinajskaja psaltyr'. Glagoličeskij pamjatnik XI v*. Petrograd: Rossijskaja Akademija Nauk, 1922.

### *Rila folios*

I. Gošev (ed.). *Rilski glagoličeski listove*. Sofija: Bălgarskata Akademija na Naukite, 1956.

### *Savva's Book* (Sav)

V. N. Ščepkin (ed.). *Savvina kniga*. Sanktpeterburg: Izdanie Imperatorskoj Akademii Nauk, 1903.

O. A. Knjazevskaja, L. A. Korobenko, and E. P. Dogramadžieva (eds.). *Savvina kniga. Drevneslavjanskaja rukopis' XI, XI–XII i konca XIII veka*. Vol. 1. *Rukopis'*; *Tekst*; *Kommentarii*; *Issledovanie*. Moskva: Indrik, 1999.

#### *Undolskij's fragments*

E. F. Karskij (ed.). *Listki Undol'skogo. Otryvok kirillovskogo evangelija XI v*. Sanktpeterburg: Izdanie Imperatorskoj Akademii Nauk, 1904.

# Dictionaries

### Meyer

K. H. Meyer. *Altkirchenslavisch-griechisches Wörterbuch des Codex Suprasliensis*. Glückstadt, Hamburg: Augustin, 1935.

### Penkova

P. Penkova. *Rečnik-indeks na Sinajskija evxologij*. Sofija: Marin Drinov, 2008.

### Sadnik

L. Sadnik and R. Aitzetmüller. *Handwörterbuch zu den altkirchenslavischen Texten*. S-Gravenhage: Mouton, 1955.

### Sreznevskij

I. I. Sreznevskij. *Materialy dlja Slovarja drevnerusskogo jazyka po pis'mennym pamjatnikam*. Sanktpeterburg: Izdanie Imperatorskoj Akademii Nauk, 1893.

### Trubačev

O. N. Trubačev et al (eds.). *Ėtimologičeskij slovar' slavjanskix jazykov*. *Praslavjanskij leksičeskij fond.* Volumes 1–41. Moskva: Nauka, 1974–2018.

#### Vasmer

M. Fasmer. *Ėtimologičeskij slovar' russkogo jazyka*. Moskva: Progress, 19862.

### Večerka

R. Večerka, R. M. Cejtlin, and Ė. Blagova. *Staroslavjanskij slovar'* (*Po rukopisjam X–XI vv.*). Moskva: Russkij jazyk, 1994.

# **Symbols and abbreviations**


FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

First name Last name, affiliation, affiliation country, email, ORCID

First name Last name, *Book title book title book title book title*, © 2021 Author(s), content CC BY 4.0 International, metadata CC0 1.0 Universal, published by Firenze University Press (www.fupress.com), ISSN 0000-0000 (online), ISBN 000-00-0000-000-0 (PDF), DOI 10.36253/000-00-0000-000-0 FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9



The meaning of any other auxiliary symbol is explained in the paragraph where that symbol is used.

# **Detailed contents**


PART I. SEGMENTAL GRAMMAR

#### **Chapter 1. Acquaintance 3**

Samples of segmental representations (§ 1) – 3. Wordforms (§ 2) – 4. Benchmark list of wordforms (§ 3) – 5. Graphics (§ 4) – 5. The phonological representation (§ 5) – 5. The morphophonological representation (§ 6) – 6. Sounding the text (§ 7) – 7. Segmental systems (§ 8) – 8. Setting up a segmental system (§ 9) – 8. Basic and derived units: segments, formatives, and wordforms (§ 10) – 8. Syntagmatics (§ 11) – 8. Mapping rules between systems (§ 12) – 9. Alloformy (§ 13) – 9. Aberrant spellouts and segmental aberrations (§ 14) – 9. The benchmark task of Part I of the book (§ 15) – 10.

#### **Chapter 2. Graphics and phonology 11**

**Alphabet.** Segment inventory (§ 16) – 11. Letter names (§ 17) – 11. Simple and iotated vowel letters (§ 18) – 12.

**Letter combinations.** Contextual conditions for vowel letters (§ 20) – 12. The distribution of simple and iotated letters (§ 21) – 13.

**Phoneme inventories.** Vowel membership (§ 22) – 14. Vowel invento-

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup\_best\_practice)

Anna Polivanova, *Old Church Slavic. Grammar and Dictionaries*, translated by Lev Blumenfeld, edited by Artemij Keidan, © 2023 Author(s), CC BY 4.0, published by Firenze University Press, ISBN 979-12-215- 0104-9, DOI 10.36253/979-12-215-0104-9

ry (§23) – 14. Vowel phoneme classes (§24) – 15. Phonological vowel pairings (§ 25) – 15. A terminological note: long and short vowels (§ 26) – 15. The set and inventory of consonants (§ 27) – 16. A note on dentopalatals (§ 28) – 16. A note on the phonemes *č* and *g* (§ 29) *–* 17*.* Classes of consonant phonemes (§30) – 17. Phonological consonant pairings (§31) – 17.

**Transformation from graphics to phonology and vice versa.** General considerations (§ 32) – 18. Graphics ⇔ phonology transformation rules (§33) – 18. From phonological spellout to normalization (§34) – 19. From normalization to phonological spellout (§ 35) – 20.

### **Chapter 3. Formatives and morphophonological representations of words 21**

**Positional classes of formatives.** Classes of formatives (§ 36) – 21. Segmental positions of formatives (§ 37) – 22. Nonstandard formatives (§ 38) – 23. Ambivalence (§ 39) – 24.

**Formative sequences in the word: morphological composition of stems.** Morphological composition of stems (§40) – 25. An overview of possible pRs schemata (§41) – 25. CVC norm for words (§42) – 26. Jakobson's law (§43) – 26. Jakobson's law and CVC norms (§44) – 27. Violations of Jakobson's law (§ 45) – 27. Hiatuses (§ 46) – 27. Clusters (§ 47) – 28.

### **Chapter 4. Phoneme syntagmatics and the mapping rules 31**

General (§ 48) – 31. Outline of the presentation (§ 49) – 31. Phoneme sequences in terms of the CVC schema (§ 50) – 32.

**Vowel syntagmatics.** *Sequences of the type xV.* General table (§51) – 33. Word-initial and postvocalic vowels (§ 52) – 34. A note on loose formative adjacency (§ 53) – 35. Vowel phonemes after morphophonologically soft consonants (§ 54) – 35. On the phonological independence of phonemes (§ 55) – 36. The law of the velars (§ 56) – 37. On the partial complementary distribution of vowels along the front-back dimension (§ 57) – 37. *Sequences of the type Vx.* Absolutely allowed and repairable sequences (§ 58) – 37.

**Consonant syntagmatics (clusters).** General table (§ 59) – 38. Data abbreviations in the general table (§ 60) – 38. On allowed and prohibited clusters (§ 61) – 39. Lists of syntagmatic anomalies (§ 62) – 40.

**The mapping from the morphophonological to the phonological representation.** General (§ 63) – 41. A note on internal and external expressions (§ 64) – 42. Cycles and blocks (§ 65) – 42. How to apply the rules (§66) – 43. Unchecked segment sequences and anomalies (§67) – 44. Rewrite rule format (§ 68) – 44. *Rules of the first cycle: the A.I and A.II blocks* (§ 69–71) – 45. *Rules of the main cycle: the B block* (§ 72– 76) – 47. *Rules of the final cycle: the C block* (§ 77) – 51.

### **Chapter 5. Polyformy of formatives 53**

**Alloformy of formatives and types of alloformy.** Monoformemic and polyformemic formatives. Families of formatives (§ 78) – 53. A note on intersections between formative families (§ 79) – 54. The internal organization of families (§ 80) – 54. A note on the goals of investigating varieties of formative families (§ 81) – 54. Standard types of alloformy (§ 82) – 55. A note on the distribution of alloforms (§ 83) – 56.

**Morphophonological variants.** A terminological note on morphophonological variants (§ 84) – 57. *Twofold rule*. Properties of the variants (§ 85) – 57. The twofold rule (§ 86) – 58. A technical note on the application of the twofold rule (§ 87) – 58. A note on the supine forms of velar-final verbs (§ 88) – 59. Special cases (§ 89) – 59. The list of cases (§ 90) – 59. A note on new twofold rule (§ 91) – 59. *CVC-ambivalence*. Properties of CVC-ambivalent variants (§ 92) – 60. The CVC agreement rule (§ 93) – 60. A note on the application of the CVC agreement rule (§94) – 60. The list of cases (§95) – 61. Special cases (§96) – 62. A note on new CVC ambivalence (§ 97) – 62.

**Segmental alloformy.** *Alternations as a tool for describing segmental alloformy*. General (§ 98) – 62. Alternations (§ 99) – 62. An order on the set of grades: the adjacency relation (§ 100) – 63. Segmental position tied to an alternation (§ 101) – 63. Alternations and alloformy (§ 102) – 63. Alternations and intraparadigmatic alloformy (§ 103) – 64. Alternations and interlexemic alloformy (§ 104) – 65. *Velar palatalization* (§ 105– 111) – 66. *Substitutive softening* (§ 112–118) – 68. *Fundamental vowel alternations* (§ 119–128) – 71.

#### **Chapter 6. Segmental peculiarities of sources 77**

Graphically alternative spellouts and aberrant forms (§ 129) – 77.

**Graphic systems of the sources.** Graphic system (§ 130) – 78. *Alphabets and transliteration.* The list of alphabets (§ 131) – 79. Transliteration (§ 132) – 79. Departures from sequential spellouts of letters (§ 133) – 81. *Basic writing systems in sources.* The basic correspondence and ways of characterizing it (§ 134) – 81. Basic writing system tables (§ 135) – 82. Phonological representations of alternative spellouts (§ 136) – 84. Phonological representations of alternative forms: special cases (§ 137) – 84. A note on the phonological representations of alternative spellouts (§ 138) – 87.

**Segmental aberrations.** General (§ 139) – 88. Atomic aberrations (§ 140) – 88. Modifying aberrations (§ 141) – 90.

#### **Chapter 7. Survey of the sources 91**

Treatment scheme for segmental idiosyncrasies of a source (§ 142) – 91.

*Codex Marianus* (§ 143–157) – 91. *Codex Zographensis* (§ 158–171) – 95. *Codex Assemanius* (§ 172–184) – 99. *Psalterium Sinaiticum* (§ 185–198) – 102. *Kiev Missal* (§ 199–211) – 107. *Sava's Book* (§ 212–226) – 109. *Codex Suprasliensis* (§ 227–241) – 114.

PART II. PARADIGMATICS

#### **Chapter 8. Acquaintance 123**

Lexemes and wordforms (§242) – 123. Paradigms and grammatical categories (§ 243) – 123. Grammatical description of a wordform (§ 244) – 124. The paradigmatic address and the paradigmatic call (§ 245) – 124. The paradigmatic construction (§ 246) – 125. A note on paradigmatic calls and paradigmatic addresses (§ 247) – 125. A note on the term *wordform* (§248) *–* 126*.* Grammatical classes of lexemes (§249) – 126. Morphological composition of wordforms (§ 250) – 126. Morphological skeleton and inflectional spellout (§ 251) – 127. Examples of morphological spellouts (§ 252) – 127. Examples of paradigmatic addresses and paradigmatic indices (§ 253) – 128. Paradigmatic synthesis (§ 254) – 130. Paradigmatic classes (§ 255) – 131. Profiles (§ 256) – 132. The paradigmatic dictionary and the benchmark list of lexemes and wordforms (§ 257) – 132. Workstems (§ 258) – 132. Selection of the terminals (§ 259) – 133. Boundary adjustment rules (§ 260) – 134. Morphologically anomalous forms (§ 261) – 135. Secondary forms (§ 262) – 135. Paradigmatic effects (§ 263) – 135. Aberrant forms and paradigmatic aberrations (§ 264) – 136. Benchmark task of paradigmatics (§ 265) – 136.

THE NOMINAL

#### **Chapter 9. Free nominal paradigms 141**

Nominal lexemes and their free paradigms (§266) – 141. Grammatical categories (§267) – 142. Substantives and adjectives (§268) – 142. *Long and short adjectives* (§269–273) – 143. *Participles* (§274–278) – 144. *Comparatives* (§ 279–281) – 146. *Secondary nominal forms* (§ 282) – 147.

### **Chapter 10. Formation of nominal forms. Terminals and paradigmatic classes 149**

**General.** Morphological composition of nominal forms (§ 283) – 149. Nominal form construction procedure (§ 284) – 149. Starting forms of nominal lexemes (§ 285) – 150. The first step: building the workstems (§ 286) – 151. Second step: selection of the terminal (§ 287) – 152. Third step: boundary adjustment rules (§ 288) – 152.

**Nominal terminals.** A catalog of terminals (§ 289) – 153. Explanation of notation (§ 290) – 153. The paradigmatic meaning of terminal sets (§ 291) – 155. An overview of terminal sets (§ 292) – 156. Overview of terminal types (§ 293) – 156. Disyllabic monocomponential terminals in standard sets (§ 294) – 157. Bicomponential terminals (§ 295) – 157. Simple and nontrivial compositions (§ 296) – 159. On the terminals оѭ/еѭ and ѫѭ (§ 297) – 159. On the etymology of 2-combi terminals (§ 298) – 160.

**Declension types.** Paradigmatic classes of nominals: declension types and inflectional classes (§ 299) – 160. A note on the notion of declension type (§ 300) – 161. Deformations (§ 301) – 162. List of inflectional classes (§ 302) – 163

#### **Chapter 11. An overview of nominal classes 165**

**Adjectives.** *Class 2/a:* новъ*/*нищь (§ 303–306) – 165. *Class 2/a\*:*  щ*-Part and* ш*-Part, and class 2/a\*\*: Compar* (§307–313) – 167. *Class 2/p:*  тъ*/*нашь (§314–318) – 172. *Class 2/p\*:* вьсь (§319–320) – 175. *Class 1/a:* триѥ (§ 321–322) – 176. *A note on adjectival forms of the* удобь *type* (§ 323) – 176. *Commentary on individual adjectival lexemes in the twofold declension* (§ 324–325) – 177.

**Morphologically masculine substantives.** *Class 2/m:* градъ*/*конь (§ 326–328) – 180. *Class 2/m\*:* дѣлател҄ь *and class 2/m\*\*:* граЖанинъ (§ 329–332) – 182. Class 1/m: пѫть (§ 333–335) – 183. *Commentary on individual morphologically masculine substаntive lexemes* (§ 336–337) – 184.

**Morphologically neuter substantives.** *Class 2/n:* село/полѥ (§ 338– 340) – 187. *Commentary on individual substаntive morphologically neuter lexemes* (§ 341) – 188.

**Morphologically feminine substantives.** *Class 2/f:* жена*/*землꙗ (§ 342–344) – 190. *Class 2/f\*:* рабыни (§ 345–348) – 191. *Class 1/f:*  кость (§ 349–351) – 192. *Commentary on individual substаntive morphologically feminine lexemes* (§ 352–353) – 193.

**Secondary forms** (§ 354–356) – 195.

#### Chapter 12. Unique nominal lexemes

A list of unique nominal lexemes (§ 357) - 199. Group หกล 0/n (\$358-359) - 200. Group отроча 0/n (\$360-361) - 200. Group слово 0/n (§ 362–363) – 201. Group око 0/n (§ 364–366) – 202. Group црькъ 0/f (\$367-368) - 203. Group мати O/f (\$369-370) -204. Lexeme господь 0/m (\$ 371-372) - 204. Lexeme десать 0/m (\$373-374) - 205. Pronouns кън 0/p and сь 0/p (\$375-376) - 206. Pronoun сь (\$377-378) - 207. Pronouns чьто 0/p and къто 0/р (\$379-380) - 208. Pronouns азъ 0/s, тъг 0/s, and см 0/s (\$ 381-382) - 209. Lexeme четъре 0/a (\$ 383-384) - 210.

#### Chapter 13. Aberrant nominal forms in sources

General (§ 385) - 211. Alien terminals (§ 386) - 211. Attestations of alien terminals in aberrant forms (\$ 387) - 212. Set intersections (§ 388) - 212. A note on 7-initial terminals (§ 389) - 213. A note on new twofold rule (§ 390) - 213. Notes on the diachronic interpretation of the alien terminals in the paradigm aberration (§391) - 214. Deformed terminals (§ 392) - 215. The order of examination of aberrant forms (§ 393) – 215. Aberrations in long adjectives: deformation of 2-combi set inflections (§ 394-395) - 216. Aberrations in 2/a\* adjectives (шт- and ш-Part) and 2/a\*\* (Compar) (\$ 396) - 218. Aberrations in substantives of classes 2/m\* (дълатель) and 2/m\* (гражданникъ) (\$ 397-398) - 220. Nonstandard terminals in lexemes of the main twofold declension (§ 399–400) – 221. Nonstandard and alien terminals in class 1/m substantives (\$ 401) - 224. Alien terminals in class 2/a and 2/p adjectives (\$ 402) - 225. Other aberrations in substantives (\$ 403) - 228.

#### THE VERB

#### Chapter 14. The free paradigm of the verb

The free paradigm (§ 404) - 233. Subparadigms: representations and systems (\$ 405) - 234.

Finite and nominal representations. Finite representations (\$406-407) - 234.

Secondary verb forms. An overview of secondary forms (\$ 408) - 235.

#### Chapter 15. Formation of verb forms

Morphological composition of verb forms (§ 409) - 237. Nominal forms in the paradigmatic synthesis of verbal paradigms (\$410) - 238. Verb form construction procedure (§ 411) - 238. Starting forms of verbal lexemes (§ 412) - 239. The first step: building verbstems (§ 413) - 239. The

211

second step: the selection of the appropriate terminal or suffix (§414) – 240. The third step: boundary adjustment (§ 415) – 240.

### **Chapter 16. Paradigmatic classes of the verb 241**

**Main paradigmatic classes.** Paradigmatic classes of the verb: acquaintance (§ 416) – 241. Unique verbs (§ 417) – 241. Main paradigmatic classes (§ 418) – 242. Splintering of the classification and irregular verbs (§ 419) – 243. Paradigmatic indices (§ 420) – 243. The principle of paradigmatic equivalence of members of a family (§ 421) – 243.

**Classification of the infinitives.** Morphological composition of the infinitive (§ 422–425) – 244.

**Basic verb stems.** Basic verb stems (§ 426) – 245. Distribution of basic stems by systems (§ 427) – 246. Basic stems and workstems and subparadigm systems (§ 428) – 247.

**Profiles of the type representatives for the main paradigmatic classes** (§ 429–430) – 247.

### **Chapter 17. Workstems of the verb 251**

**Workstems of regular verbs.** The segmental content of basic stems (§431) – 251. Workstem construction rules for regular verbs (§432) – 252. Paradigmatic effects and construction of truncated workstems of the main paradigmatic classes (§ 433) – 253.

**Workstems of irregular verbs.** The list of irregular verbs (§ 434) – 255. Classification of irregular verbs (§ 435) – 256. Assortment and distribution of workstems of irregular verbs (§ 436) – 256. Segmental content of irregular verb workstems (§ 437) – 256. Redistribution of workstems (§ 438) – 257. How to select the necessary workstem of irregular verbs (§ 439) – 257. Distribution of irregular verb workstems by subparadigms (§ 440) – 258. *Commentary of particular groups of irregular verbs.* метати 3° group (§ 441) – 259. пьсати 3\* group (§ 442) – 259. бьрати 3°\* group (§ 443) – 260. пл҄ьвати 3h\*⩨ group (§ 444) – 260. клѧти 4h group (§ 445) – 260. брати 4h• group (§ 446) – 260. плути 4h⤸ group (§ 447) – 261. бити 4h\*⤸ group (§ 448) – 261. крыти 4h\*⩨⤸ group (§ 449) – 262. пѣти 4h⤸ group (§ 450) – 262. мрѣти 4h\*⤹ group (§ 451) – 263. влѣщи 4c\*⤹ group (§ 452) – 263. чисти 4c\*⤹ group (§ 453) – 264. Profiles of type representatives of irregular verb groups (§ 454) – 265.

#### **Chapter 18. Verbal terminals and suffixes 267**

**Sets of terminals and suffixes.** A catalog of terminals and suffixes (§ 455) – 267. On notation (§ 456) – 269. Standard and nonstan-

dard sets in a subparadigm (§ 457) – 269. Terminals and suffixes in the present system (§ 458) – 269. Terminals and suffixes in the imperfect system (§ 459) – 269. Terminals and suffixes in the infinitive-aorist system (§ 460) – 270.

**Boundary adjustment rules** (§ 461–462) – 270.

### **Chapter 19. An overview of verb forms by system 273**

**The present system.** A general overview of forms (§ 463) – 273. Anomalous imperative forms (§ 464) – 274. A note on the so-called *j*-present (§ 465) – 274.

**The imperfect system.** A general overview of forms (§466) – 275. *Nonstandard imperfects (personal forms).* Contracted imperfect (§467) – 276. Illustrations and distribution of contracted imperfects (§ 468) – 276. Present imperfect (§ 469) – 276. Illustrations and distribution of present imperfects (§ 470) – 277. Iotated imperfect (§ 471) – 278. Illustrations and distribution of the iotated imperfect (§ 472) – 278. *New*  ш*-participle* (§ 473–474).

**The infinitive-aorist system.** A general overview of forms (§ 475) – 280. *Nonstandard aorists*. General (§476) – 280. Root aorist (идъ type) (§ 477) – 281. Old sigmatic 1 (нѣсъ type) (§ 478) – 281. Old sigmatic 2 (рѣхъ type) (§ 479) – 282. Secondary 2–3Sg forms (§ 480) – 282. Distribution of nonstandard aorist forms (§ 481) – 282. Nonstandard aorists: illustrations (§ 482) – 283. т*-participles* (§ 483–485) – 286.

#### **Chapter 20. An overview of verb classes 289**

любити *1 class* (§ 486–488) – 289. трьпѣти *2 class* (§ 489–491) – 290. *Commentary on individual verbal lexemes* (§ 492–493) – 290. плакати *3 class* (§ 494–498) – 293. нести *4 class* (§ 499–502) – 295. двигнѫти *5 class* (§ 503–505) – 296. миловати *6 class* (§ 506–507) – 297. *Commentary on individual verbal lexemes* (§ 508–511) – 297. дѣлати *7 class* (§ 512–514) – 300. *Commentary on individual verbal lexemes* (§ 515) – 301.

#### **Chapter 21. Unique verbs 303**

General table of unique verb profiles (§ 516) – 303. *The verb*  дати (§ 517–521) – 305. *The verb* ꙗсти (§ 522–525) – 307. *The verb*  вѣдѣти (§ 527–531) – 309. *The verb* имѣти (§ 532–537) – 310. *The verb*ѥсмь (§538–542) – 312. *The verb* быти (§543–549) – 314. *The verb*  хотѣти (§550–555) – 317. *The verb* довьлѣти (§556–560) – 319. *The verb* ити (§561–564) – 321. *The verb* ꙗти (§565–569) – 322. *The verb*  стати (§570–573) – 324. *The verb* съпати (§574–577) – 325. *The verb* 

въпити (§578–581) – 326. *The verb* сѣсти (§582–585) – 328. *The verb*  лещи (§ 586–589) – 329. *The verb* об.рѣсти (§ 590–593) – 330. *The verb* гънати (§594–597) – 332. *The verb* плѣти (§598–601) – 333. *The verb* дѣти (§ 602–605) – 334.

### **Chapter 22. Aberrant verbal forms in sources 337**

General (§ 606) – 337. The order of examination of aberrant verbal forms (§607) – 337. *Aberrations by paradigmatic class* (§ 608–610) – 338. *Aberrant personal terminals* (§ 611–612) – 341. Aberrant root vocalisms (§ 613) – 345. *Alien workstem expansion* (§ 614) – 346. *Other aberrations in the PRAE system* (§ 615–622) – 347.

PART III. ADDENDA

### **Chapter 23. Formative inventories: prefixes, roots, suffixes 357**

General (§ 623) – 357.

**On morphophonological representations.** Ty pes of data (§ 624) – 357. Two aspects of the problems (§ 625) – 358. Extraction of data from morphophonological representations (§ 626) – 358. Etymology and synchrony (§ 627) – 358. Parsing problems and roots: opaque stems (§ 628) – 359. Parsing problems and suffixes (§ 629) – 360. Parsing problems and prefixes (§ 630) – 360. Problems with establishing of the pRs schema (§ 631) – 361. Problems with the identification of prefixes (§ 632) – 361. Problems with the identification of suffixes (§ 633) – 361. Identification of roots (§ 634) – 361. The problem of establishing root alloforms (§ 635) – 362. The problem of borrowings (§ 636) – 363.

**Prefixes.** On the autonomous use of prefixal formatives (§ 637) – 364. Prefixes and opaque stems (§638) – 364. Prefixes and roots (§639) – 364. An inventory of prefixal formatives (§640) – 365. Commentary (§641) – 366.

**Roots.** General (§642) – 371. Root classes by main morphophonological features (§643) – 372. Order of exposition (§644) – 373. Order of illustration for alloformy (§645) – 373. A note on unattested alloforms (§646) – 373. A note on reconstructions of fictional alloforms and sources of alloformy (§647) – 374. A note on variant spellouts of roots (§648) – 374. *Roots with stable vocalism: standard* (§ 649–650) – 375. *Roots with stable vocalism: nonstandard* (§ 651–655) – 375. *Roots with unstable pure vocalism: standard and nonstandard* (§ 656–676) – 378. *Roots with unstable sonant vocalism: standard and nonstandard* (§677–682) – 383. *Sonant roots with H(n) vocalism* (§ 683–688) – 390. *Sonant roots with H(m) vocalism* (§ 689–691) – 391. *Sonant roots with H(n/m) vocalism* (§692–701) – 391. *Sonant roots with H(u) vocalism* (§702–729) – 392. *Sonant roots with H(j) vocalism* (§730–741) – 395. *Sonant roots with H(r) vocalism* (§ 742–770) – 396. *Sonant roots with H(l) vocalism* (§771–784) – 399. *Pronominal roots* (§785–805) – 400. *Anomalous roots* (§806–840) – 404.

**Suffixes.** General (§ 841) – 410. Alloformy of suffixes (§ 842) – 411. On the intersection of suffix families (§ 843) – 412. *Suffixes with a carrier consonant* (§ 844–859) – 412. *Suffixes without a carrier consonant* (§ 860–865) – 426.

### **Chapter 24. Supplement 431**

Trubeckoj's nonstandard phonology (§ 866) – 431. On word-initial vowels (§ 867) – 433. On wordform boundaries and loose and tight formative adjacencies (§ 868) – 435. On the law of the velars (§ 869) – 437. On partial neutralization of vowel advancement contrasts (§ 870) – 439. On the combinations бн, пн, and мн (§ 871) – 440. On combinations with з-final prefixes (§ 872) – 441. On the types of segmental pairings (§ 873) – 443. On verbs with unstable root vocalism (§ 874) – 446. On Jakobson's law (§ 875) – 448. On the morphological composition of stems and wordforms (§ 876) – 448.

**Excursus on the grammatical regularity of canonical wordforms** (§ 877–885) – 450.

**Excursus on aberrations** (§ 886–897) – 455.

**Excursus on** *yer* **aberrations and Havlík's rules** (§ 898–899) – 466.

**Excursus on the description of phoneme syntagmatics** (§900–909) – 474

**Excursus on the contamination and competition between paradigmatic classes** (§ 910–913) – 482.

**Excursus on the morphology of personal forms of the imperfect** (§ 914–915) – 492.

#### **Chapter 25. Summary 497**

Step-by-step construction of nominal wordforms (§ 916) – 495. Step-bystep construction of verbal wordforms (§ 917) – 497. An overview of participles (§ 918) – 498. *Comparative construction rules* (§ 919–924) – 499.

**Chapter 26. Chrestomathy 503**

**Normalized texts** – 504.

*Psalterium Sinaiticum –* 509.

*Codex Suprasliensis –* 511.

### PART IV. DICTIONARIES


# **Biblioteca di Studi Slavistici**


This book contains a synchronic grammar and grammatical dictionaries of Old Church Slavic. The framework is based on a substantially revised version of the classical descriptive methodology. The intent is to improve on the classical monographs by Vaillant, Diels, Lunt in the direction of utmost completeness, explicitness, and deliberate consistency between the grammatical structure, the corpus of texts (limited to the seven oldest OCS manuscripts), and the dictionaries. The grammar is intended as a set of rules that provide a complete characterization of any OCS wordform. Peculiarities in the language of each source are described as systematic departures from canonical OCS, a conventional constructed variety primarily described by the grammar. The book is addressed to linguists working in Slavic studies, as well as to specialists in the general theory of grammar, especially phonologists and morphologists. Data from the dictionaries are freely available as a searchable database online.

Anna Polivanova graduated in linguistics in 1967 at the Moscow State University, where she also obtained her PhD in 1970. Her research activity started at the Machine Translation Laboratory under the supervision of Igor Mel'čuk. Since the end of the 1960s, she taught several courses at her home university, and since the 1990s, also in other universities located in Moscow. She authored three monographs and numerous articles ranging from phonetics to morphology and syntax, from the study of ancient languages to the problems of mathematical methods in linguistics.

> ISSN 2612-7687 (print) ISSN 2612-7679 (online) ISBN 979-12-215-0103-2 (Print) ISBN 979-12-215-0104-9 (PDF) ISBN 979-12-215-0105-6 (XML) DOI 10.36253/979-12-215-0104-9 www.fupress.com

**FUP**

Old Church Slavic

Anna Polivanova