# **LATIN AND ARABIC**

Entangled Histories

Daniel G. König *Editor*

Latin and Arabic: Entangled Histories

#### **Heidelberg Studies on Transculturality – 5**

Series Editors:

Reuven Amitai, Jerusalem; David Armitage, Harvard; Christiane Brosius, Heidelberg; Beatrix Busse, Heidelberg; Prasenjit Duara, Durham; Christian Henriot, Lyon; Madeleine Herren, Basel; Joachim Kurtz, Heidelberg; Joseph Maran, Heidelberg; Axel Michaels, Heidelberg; Barbara Mittler, Heidelberg; Sumathi Ramaswamy, Durham; Roland Wenzlhuemer, Munich

# **Latin and Arabic**

Entangled Histories

Daniel G. König *Editor*

#### **Bibliographic information published by the Deutsche Nationalbibliothek**

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie. Detailed bibliographic data are available on the Internet at http://dnb.dnb.de.

This book is published under the Creative Commons Attribution 4.0 International License (CC BY-SA 4.0). The cover is subject to the Creative Commons License CC BY-ND 4.0.

The electronic, open access version of this work is permanently available on Heidelberg University Publishing's website: https://heiup.uni-heidelberg.de. urn: urn:nbn:de:bsz:16-heiup-book-448-6 doi: https://doi.org/10.17885/heiup.448

Text © 2019, by the authors.

Cover image: The first page of a seventeenth-century Arabic-Latin version of the Qurʾānic sūra 12 (sūrat Yūsuf), printed in Leiden by Thomas Erpenius (1584–1624) using his own typeset. Taken from Thomas Erpenius, *Historia Iosephi patriarchae, ex Alcorano, Arabicè. Cum triplici versione Latina, & scholijs Thomae Erpenii, cujus & alphabetum Arabicum praemittitur* (Leiden: Ex Typographia Erpeniana, 1617), D2.

ISSN 2365-7987 (Print) ISSN 2365-7995 (eISSN)

ISBN 978-3-947732-25-8 (Softcover) ISBN 978-3-947732-26-5 (Hardcover) ISBN 978-3-947732-24-1 (PDF)

# Table of Contents


# Preface

The collection of essays in *Latin and Arabic: Entangled Histories* is the fruit of a workshop held at the Heidelberg Cluster "Asia and Europe in a Global Context" on September 28 and 29, 2016. The idea of organizing a workshop on the entanglement of Latin and Arabic resulted from the desire to avoid and circumvent well-trodden paths of scholarship on the political, economic, social, cultural, and religious history of the Euromediterranean. Scholars interested in this history cannot avoid being confronted with wellknown binary oppositions—"Islam and the West," "Christianity and Islam," "tolerance and intolerance," "*convivencia* and the clash of civilizations," and so on. The focus on Latin-Arabic entanglement was deliberately chosen in the hope that a historical socio-linguistic approach to Euromediterranean history would open up the possibility of using different, maybe even more "neutral" categories and thus of providing a conceptual alternative to seemingly endless and ultimately pointless culturalist debates. In addition, a focus on Latin-Arabic entanglement also seemed to yield the prospect of highlighting the relevance of themes so far deemed secondary by historians of political, social, economic, and religious history, but highlighted in other fields of historical research.

Variants of Latin-Arabic entanglement currently play a role in various scholarly milieus. In the Arab world of the early twenty-first century, Latin-Arabic entanglement mainly results from scholarly engagement with the Latin language and its literature, as well as with Latin primary sources in various departments of classics and ancient and medieval history, most of them situated in Egypt.1 The foundation of a "Centre d'Études Latines" at the Lebanese Université Saint-Esprit de Kaslik in 2009 points to the fact that, due to historical relations between Rome and various Oriental churches since the medieval period, the study of Latin-Arabic entanglement is also of interest to Arabic scholars of ecclesiastical history.2

Before Western academia abandoned Latin as a language of scholarly endeavour over the course of the nineteenth century, Latin-Arabic entanglement in European and North American scholarly milieus was a by-product of scholarly engagement with the Arabic language. Today, however, Latin-Arabic entanglement has come to represent an object of analysis nurtured and advanced by a rather small number of specialists in niche areas of various academic disciplines. Forms of Latin-Arabic entanglement play a

<sup>1</sup> See the overview in Daniel G. König, "The Unkempt Heritage. On the Role of Latin in the Arabic-Islamic Sphere," *Arabica* 63, no. 5 (2016), 419–493, here 465–474.

<sup>2</sup> Centre d'Études Latines, Université Saint-Esprit de Kaslik, accessed November 29, 2017 http://www.usek.edu.lb/fr/centre-des-etudes-latines.

prominent role in several fields of research, such as the medieval history of philosophy and the sciences, Ibero-Romance philology and literature, and medieval and early modern European and Mediterranean history. They are of relevance to a much lesser extent in the fields of Islamic studies or Arabic and Latin philology and literature, excepting scholars whose areas of specialization include the Iberian Peninsula and southern Italy.

It is only in the fields of the history of philosophy and the sciences, however, that Latin-Arabic studies have achieved a certain degree of institutionalization in the form of research projects and research centres, often of a temporary nature. The project "*Speculum Arabicum*: Objectifying the contribution of the Arab-Muslim world to the history of sciences and ideas: the sources and resources of medieval encyclopaedism," was conducted by a group of researchers at the Université catholique de Louvain between 2012 and 2017.<sup>3</sup> The Centre for the History of Arabic Studies in Europe, hosted by the Warburg Institute in London4 and the "Digital Averroes Research Environment," hosted by the Thomas-Institute in Cologne,5 will hopefully prove more durable. This is also to be hoped for the "Forschungsstelle Philosophie- und Wissenschaftsgeschichte der griechisch-arabisch-lateinischen Tradition," hosted by the Department of Philosophy at Würzburg University,6 and the project "Ptolemaeus Arabus et Latinus," hosted by the Bavarian Academy of Sciences and Humanities.7

As academic fields of research, the history of philosophy and the history of the sciences indisputably boast a long and respectable tradition of Latin-Arabic studies, but there are still large corpora of sources awaiting thorough investigation.<sup>8</sup> This volume will prove beyond doubt, however, that it is impossible to reduce the history of Latin-Arabic entanglement to the field of intellectual history with its focus on processes of cultural transfer, the mobility of specific texts and ideas, the concomitant emergence of Arabic studies in early modern Europe, and related themes. Latin-Arabic entanglement also plays an important role in the fields of political,

<sup>3</sup> Speculum Arabicum, Université catholique de Louvain, accessed December 6, 2017, https://uclouvain.be/fr/instituts-recherche/incal/speculum-arabicum.html.

<sup>4</sup> Centre for the History of Arabic Studies in Europe, Warburg Institute, accessed November 29, 2017, https://warburg.sas.ac.uk/research/research-projects/cen tre-history-arabic-studies-europe-chase.

<sup>5</sup> Digital Averroes Research Environment, accessed December 6, 2017 http://dare. uni-koeln.de/.

<sup>6</sup> Julius-Maximilians-Universität Würzburg, accessed November 29, 2017, https:// www.philosophie.uni-wuerzburg.de/forschung/forschungsstellephilosophieun/.

<sup>7</sup> Ptolemaeus Arabus et Latinus, accessed November 29, 2017, http://ptolemaeus. badw.de/start.

<sup>8</sup> See the overview in Daniel G. König, "Übersetzungen und Wissenstransfer. Zu einem Aspekt der Beziehungen zwischen lateinisch-christlicher und arabischislamischer Welt, *Trivium* 8 (2011), § 10, accessed December 6, 2017, https://journals.openedition.org/trivium/3875, or Daniel G. König, "Traductions et transferts des savoirs. À propos des relations entre l'Occident latin et le monde arabomusulman," trans. Frédéric Vitoux, *Trivium* 8 (2011), § 10, accessed December 6, 2017, https://journals.openedition.org/trivium/3973.

economic, social, legal, and religious history. It forms an integral part of the history of the ancient Roman Middle East and plays a role in the regional histories of medieval North Africa, the Iberian Peninsula, and southern Italy. Trans-Mediterranean phenomena such as the Crusades or the late medieval Catalan and Italian trade networks cannot be fully understood if one fails to consider the interaction of various Mediterranean idioms, including the Romance languages and different forms of Arabic. Late medieval and early modern Roman-Catholic missionary policy in the Middle East produced many Latin–Arabic translations and milieus. The establishment of Latin studies in the Arab world of the twentieth century resulted in additional forms of Latin-Arabic entanglement, which feature clear links to the history of the ancient and medieval Mediterranean. Thus, to gain a fuller understanding of the macro-history of Latin-Arabic entanglement, it seemed necessary to move beyond the scope of intellectual history, with its focus on the transmission of philosophical and scientific texts.

Against this backdrop, the aim of the workshop was to bring together a large array of scholars from different fields of research on Latin-Arabic entanglement, and to foster communication and an exchange of ideas on how this topic relates to the wider history of the Euromediterranean. Various factors have prevented the production of a volume of collected essays that would provide a balanced, representative, and—as far as this is possible—a relatively exhaustive overview on the many existing forms of Latin-Arabic entanglement that played some part in approximately two thousand years of Euromediterranean history. Among these factors were limited resources and various impediments that prevented invited scholars from participating in the workshop or from contributing to this publication, but also the sheer mass of topics related to one form or other of Latin-Arabic entanglement. The bibliography at the end of the volume opens up further fields of enquiry and points to the achievements of many an important scholar in this area of research. These bibliographic references show clearly that this volume cannot claim to be more than a preliminary effort at understanding the various facets and ramifications of Latin-Arabic entanglement. It tries to draw together different research traditions, but cannot cover them all. Consequently, this volume only partly, often only implicitly, answers the research questions raised in the invitation to the workshop.

The invited scholars were asked to reflect upon the question of when, where, how, and why certain phenomena of Latin-Arabic entanglement often several at a time—appeared, while others did not. Moreover, they were asked to interpret the phenomena under scrutiny within the larger historical context of Latin-Arabic entanglement, e.g. by considering:


conditions that prepared the ground for the emergence of a specific form of entanglement;


All of these questions were raised with the aim of acquiring the material necessary to understand the interdependency between different forms of entanglement and their respective milieus of origin and to acquire an understanding of the different micro-histories of Latin-Arabic interaction and interpenetration. Such an understanding was deemed necessary in view of the self-evident observation that forms of Latin-Arabic entanglement emerged and existed in distinct social milieus. The latter formed part of a particular geopolitical framework and often stood at the crossroads of intersocietal relations, and/or resulted from a particular intrasocietal constellation marked by specific demarcations and boundaries. The respective geopolitical, intersocietal, and/or intrasocietal constellation supplied some of the resources necessary for the production of certain forms of Latin-Arabic entanglement, in particular linguistic mediators. It seems possible that specific forms of Latin-Arabic entanglement came into being thanks to unintended, "quasi-organic" processes of amalgamation involving different linguistic milieus and traditions. However, the respective constellation often required a particular reason, maybe even a social necessity, for creating the respective form in a given place and period of time. A systematic and comparative investigation of this complex of interdependencies might explain, for example, why the earliest documented translations of longer texts from Latin to Arabic were produced between the late ninth and the early tenth century, whereas the earliest substantial translations from Arabic to Latin only appeared in the eleventh century. It could also highlight shifting regional and chronological asymmetries, e.g. in the flows of texts and loanwords, or the occurrence of hybrid phenomena. An understanding of these shifting asymmetries is a prerequisite to writing a macrohistory of Latin-Arabic entanglement.

The six chapters that make up this volume do not even come close to fulfilling the above-mentioned research objectives. However, thanks to the intellectual enthusiasm, moral support of, and close cooperation with Benoît Grévin, the workshop and this volume have succeeded at making certain inroads into a preliminary analysis and systematization of the macro-history of Latin-Arabic entanglement. The volume is divided into two parts, the first featuring two different macro-historical perspectives, and the second consisting of four case studies from the late medieval, the early modern, the late modern, and the contemporary periods. Part One approaches the macro-history of Latin and Arabic first from a comparative structuralist view, then from a perspective that depicts the different phases of Latin-Arabic entanglement from Antiquity to the present. In its effort to cover the most relevant topics, this rather encyclopaedic overview exceeds the usual page limit of an article in a collected volume, but hopefully manages to give an overview that allows the reader to contextualize the case studies in the second part of the volume.

Although the authors of the first two chapters, Benoît Grévin and myself respectively, struggled to address the entire range of topics relevant to the comparative and to the entangled approach, it is clear that only micro-historical analysis can provide more detailed insight into the workings of individual Latin-Arabic milieus. The contributions of Part Two thus serve to elucidate how particular milieus of Latin-Arabic entanglement came into being and functioned. Chapter 3, by Daniel Potthast, focuses on bilingual or translated administrative and juridical documents produced in late medieval Iberia. In Chapter 4, Benoît Grévin offers a regional study that explains the roles played by Arabic in the different but interconnected milieus of late medieval and Renaissance Italy. This is followed, in Chapter 5, by Katarzyna K. Starczewska's analysis and contextualization of an Arabic-Latin Qurʾān manuscript produced by the Scottish Orientalist David Colville in the seventeenth century. In Chapter 6, Jan Scholz leads us into the nineteenth and twentieth centuries in the final case study dealing with the reception of what could be defined as the "Graeco-Roman" tradition in modern Arabic rhetorical manuals, many of them referred to for advice on preaching Islamic sermons.

Apart from the authors, whom I would like to thank very much for their contributions, many people participated in the production of this volume. The University of Heidelberg's "Field of Focus 3: Cultural Dynamics in Globalised Worlds" generously financed the initial workshop and parts of the publication process. I would like to express my sincere gratitude, not only for the possibility of bringing together scholars working on different facets of Latin-Arabic entanglement, but also for being able to offer a form of hospitality that facilitated an intensive intellectual exchange. Petra Kourschil and Patrick Zerner from the Cluster's finance office expertly handled concomitant financial matters and thus allowed me to focus on non-administrative matters. Many thanks are due to Rosanna Sirignano and Jan Scholz, both of whom ensured that the workshop itself ran smoothly and later lent a helping hand in the initial phase of the publication process. Dr. Andrea Hacker and, in particular, Russell Ó Ríagáin have accompanied the latter with great enthusiasm and helpful professional advice, thus giving the final touch to texts pre-edited by Thérèse Wassily Saba and subjected to conscientious final copy-editing by Amanda Gomez, and, in the last stage, by Joshua Elwer, Anna Larsson, and Chelsea Roden. In the stage of preparing the final manuscript, the comments of two anonymous external reviewers were very helpful in improving the quality of the book. The latter received its present form thanks to the efforts of several members of Heidelberg University Publishing, including Anja Konopka, Frank Krabbes, and Daniela Jakob. Finally, I would like to thank the directors, administrators, and researchers of the Cluster "Asia & Europe in a Global Context" for having provided an intellectually stimulating atmosphere encouraging forms of research that cross the boundaries of established fields of academic investigation. Working in this environment has been a highly enriching experience.

Last but not least, I would like to gratefully mention Jan Rüdiger, who first made me aware of the relevance of sociolinguistic issues for historical research, as well as Jocelyne Dakhlia, Bert Fragner, John Wansbrough, Dag Nikolaus Hasse, ʿAlī Fahmī Ḫušaym, and Benoît Grévin, whose research on mid- to long-term linguistic phenomena and developments encouraged me to approach the issue of Latin-Arabic entanglement from a macrohistorical perspective.9

Konstanz, July 2019, Daniel G. König

<sup>9</sup> Jan Rüdiger, *Aristokraten und Poeten. Die Grammatik einer Mentalität im tolosanischen Hochmittelalter* (Berlin: Akademie-Verlag, 2001); Jocelyne Dakhlia, *Lingua franca: Histoire d'une langue métisse en Méditerranée* (Arles: Actes Sud, 2008); Bert Fragner, *Die Persophonie. Regionalität, Identität und Sprachkontakt in der Geschichte Asiens* (Berlin: Das Arabische Buch, 1999); John E. Wansbrough, *Lingua Franca in the Mediterranean* (Richmond: Curzon, 1996); Dag Nikolaus Hasse, *Success and Suppression: Arabic Sciences and Philosophy in the Renaissance* (Cambridge, MA: Harvard University Press, 2016); *ʿAlī Fahmī Ḫušaym, Al-Lātīniyya al-ʿarabiyya. Dirāsa muqārana bayna luġatayn baʿīdayn qarībayn: muqaddima wa-muʿǧam* (Cairo: Markaz al-ḥaḍāra l-ʿarabiyya, 2002); Benoît Grévin, *Le parchemin des cieux: Essais sur le Moyen Âge du langage* (Paris: Éditions du Seuil, 2012).

# Part I Latin and Arabic: Macro-historical Perspectives

Benoît Grévin (CNRS, Paris)

# **1. Comparing Medieval "Latin" and "Arabic" Textual Cultures from a Structural Perspective**

The history of relations between medieval Latin and Arabic textual cultures is generally understood to be a multifaceted history of transmissions, contacts, and hybridizations. The study of these relations has become an entire subfield of medieval textual studies.1 The nature of the links between these two textual cultures raises many questions indeed, at different levels and in different fields. What forms of interaction were characteristic of the areas where Latin and Arabic coexisted over long periods, such as on the Iberian Peninsula, or in Sicily? More generally, what were the mechanisms that facilitated the transmission of Arabic knowledge or textual forms to the Latin West? Such questions have become the object of intense scientific investigations, as well as fierce first- and second-hand debates. In some Western academic milieus, we observed in recent years how representatives of right-wing political tendencies "denied," in a certain way, the influence of Arabic culture on the Latin West.2 At the same time, representatives of left-wing political tendencies managed to establish influential currents of thought such as the concept of "postcolonial medieval studies." Such currents, often having originated in the United States, propose to narrate the story of these Latin-Arabic entanglements on a new basis, thus implying that preceding investigations were conceptually invalid or at least ideologically biased.<sup>3</sup>

<sup>1</sup> For a bibliographical sketch, see Chapter 2 in this volume. For the now rapidly-developing sub-subgenre of studies on Latin translations of the Qurʾān, Thomas E. Burman, *Reading the Qur'ān in Latin Christendom, 1140–1560* (Philadelphia: University of Pennsylvania Press, 2007), provides one example of the general explosion of studies on Latin translations from Arabic. See also Chapter 5 in this volume.

<sup>2</sup> See the controversy that arose in France in 2008 around the book by Sylvain Gouguenheim, *Aristote au Mont Saint-Michel: Les racines grecques de l'Europe chrétienne* (Paris: Seuil, 2008). For a summary of reactions to this book, see Daniel D. König, "Traductions et transferts de savoirs: À propos des relations entre l'Occident latin et le monde arabo-musulman," *Trivium: Revue franco-allemande de sciences humaines et sociales* 8 (2011) https://trivium.revues.org. [Accessed October 31, 2017].

<sup>3</sup> For example, see Sharon Kinoshita, *Medieval Boundaries: Rethinking Difference in Old French Literature* (Philadelphia: University of Pennsylvania Press, 2006), which has a stronger focus on Romance languages and Arabic. See also Karla Malette, *The Kingdom of Sicily, 1100–1250: A Literary History* (Philadelphia: University of

## 1.1 A history of two (and more) languages: Can we deconstruct the "grand narratives" of Medieval Latin and Arabic?

There exists a different, complementary approach to the history of the relations between Arabic and Latin textual cultures. This approach consists in comparing the two linguistic cultures of the Islamic(ate) and Latin medieval spheres, thus treating them as two distinct, equivalent entities. It temporarily puts aside the problem of plausible or asserted relationships between the two spheres in order to examine possible structural similarities. This may evoke the somewhat old-fashioned structural and functionalist approach favoured by some researchers during the second half of the twentieth century, in the wake of the anthropology of Claude Lévi-Strauss (1908–2009). Many have pointed out that structural comparatism cannot be regarded as an adequate tool to study historical societies that experienced permanent changes, particularly with regard to their linguistic usages and cultures; for how can we model the similarities between two cultures in permanent evolution? Although the challenge seems overwhelming, our knowledge of the workings of language in medieval societies—in both the Latin and Islamic(ate) spheres—has advanced rapidly in the last thirty years.4 This includes, for example, our understanding of

Pennsylvania Press, 2005), on the interaction between Greek, Latin, and Arabic in Sicily, followed by Karla Malette, *European Modernity and the Arab Mediterranean: Toward a New Philology and a Counter-Orientalism* (Philadelphia: University of Pennsylvania Press, 2010). This is perhaps the most elaborate theorization of the doctrine of "post-colonial medievalism."

<sup>4</sup> On the sociolinguistic evolution of Latin in the Christian world of Late Antiquity, see e.g. Michel Banniard, *Viva Voce: Communication écrite et communication orale du IVe au IXe siècle en Occident latin* (Paris: Institut des Études augustiniennes, 1992). For the period from Late Antiquity to the late Middle Ages, see Pascale Bourgain and Marie-Clotilde Hubert, *Le latin médiéval* (Turnhout: Brepols, 2005). On the birth of humanism, see Ronald Witt, *"In the Footsteps of the Ancients": The Origins of Humanism from Lovato to Bruni* (Leiden: Brill, 2000). For the early modern period, see Françoise Waquet, *Le latin ou l'empire d'un signe*, *XVIe –XX<sup>e</sup> siècle* (Paris: Albin Michel, 1998). For a contrastive examination of the dawn of the Western vernaculars, see Michèle Goyens, Werner Verbeke, eds, *The Dawn of the Western Vernacular in Western Europe* (Louvain: Presses de l'Université de Louvain, 2003). For an accurate study of the interactions between the vernacular and Latin in a teaching context during this crucial period, see Anna A. Grotans*, Reading in Medieval St. Gall* (Cambridge: Cambridge University Press, 2012). On the history of Arabic, see the seminal but now dated Johann Fück, ʿ*Arabīya: Untersuchungen zur arabischen Sprach- und Stilgeschichte* (Berlin: Akademie-Verlag, 1955). Kees Versteegh, *The Arabic Language* (Edinburgh: Edinburgh University Press, 2001) provides a more recent synthesis. The *Encyclopaedia of Arabic Language and Linguistics*, ed. Kees Versteegh, 5 vols. (Leiden: Brill, 2006–2009), largely reflects the current state of the art. On various aspects of Arabic sociolinguistics in the medieval period, neglected or unknown until quite recently, see Li Guo's exploration of popular Egyptian poetry of the Mamlūk era, *The Performing Arts in Medieval Islam: Shadow Play and Popular Poetry in Ibn Daniyal's Mamluk Cairo* (Leiden: Brill, 2012). For medieval reading practices, see Konrad Hirschler, *The Written Word in the Medieval Arabic Lands: A Social and Cultural History of Reading Practices* (Edinburgh: Edinburgh University Press, 2012). For the shifting

memorizing and mnemonic processes,<sup>5</sup> of the oral dimension of traditional literature, of pragmatic writing and techniques of writing, and of medieval multilingualism. Consequently, it seems plausible that a relatively old tool such as "structural comparatism" can be reused with some effectiveness, provided that it is correctly adapted to the present needs.

It is worth asking whether it is actually possible to establish a valid frame for such an experiment.6 Researchers from the two fields of textual studies—of the Latin Middle Ages and of classical Islam—might deny the validity of such a comparison on a broader scale right from the start, for a number of reasons. The histories of Arabic and Latin—understood here as cultural tools and linguistic mediums—differ enormously, from a chronological as well as from other points of view. The assumption that the histories of Latin and Arabic are ultimately incomparable necessarily contains some truth. No history of a great, culturally influential language, that is, a language used as the ultimate reference language in a large number of cultural sectors, can be identical to other, grossly similar histories. However, such an assumption misses the mark to a certain extent. We should consider that we cannot reduce the history of a highly complex sociolinguistic field to a "grand narrative" that explains the emergence or decline of a language in teleological terms. Such a reduction is equally impossible if the task is to compare two highly complex sociolinguistic fields and their evolution.

A good starting point to approach the method of structuralist comparatism from a new angle consists in cross-examining the traditional ways in which the broad histories of Latin and Arabic during the Middle Ages are put into perspective. There are naturally some basic, unavoidable, and apparently considerable differences between the sociolinguistic histories of the two languages as employed by their speakers between ca. 550 and 1500. During Late Antiquity (ca. 300–650), Latin was already a wildly diffused idiom, a language of culture used at different levels of communication in the western Mediterranean as well as in the Romano-Germanic kingdoms that had emerged within the space formerly held by the Western Roman

relations between Arabic and non-Arabic languages in the teaching of Arabic in non-Arabophone areas, see Travis Zadeh, *The Vernacular Qurʾān: Translation and the Rise of Persian Exegesis* (Oxford: Oxford University Press, 2012).

<sup>5</sup> For these questions in connection with the Latin sphere, see the now classical work of Mary Carruthers, *The Book of Memory*: *A Study of Memory in Medieval Culture* (Cambridge: Pilgrim Books, 1990); see also Mary Carruthers, *The Craft of Thought: Meditation, Rhetoric, and the Making of Images, 400–1200* (Cambridge: Cambridge University Press, 1998).

<sup>6</sup> A tentative approach has been tested in Benoît Grévin, *Le parchemin des cieux: Essai sur le Moyen Âge du langage* (Paris: Seuil, 2012), an essay in comparative sociolinguistic history between the Latin Christian West and classical Islam, focusing on the period 565–1500. For another, still broader perspective, comparing the medieval and early modern Latin West, Islam, and the Orthodox world, see Siegfried Tornow, *Abendland und Morgenland im Spiegel ihrer Sprachen: Ein kulturhistorischer Vergleich* (Wiesbaden: Harrassowitz, 2009), where the author focuses on the differing developments of the three sociolinguistic spheres from an evolutionist perspective.

Empire. The common representation of the history of this language during the thousand years stretching from 500 to 1500 is that some late Latin varieties were still spoken until the Carolingian period in the linguistic space usually known as *Romania*, 7 but that they shifted gradually to forms that became more and more alienated from classical Latin. At the end of this process, which took place between 650 and 950 depending on the region, Latin remained the written tool of the entire "Latin sphere," whereas the population spoke not only Romance, but also German and Celtic vernaculars. Then, in a third phase, the so-called modern languages—to which various Slavic and Finno-Ugric languages must be added in line with the pace of the Christianization of Central and Northern Europe<sup>8</sup>—entered a process of literarization that resulted in the progressive marginalization of Latin as a written tool. This process was still incomplete in Central Eastern Europe by around 1500,9 but the tendency was relatively clear. Thus, the entire story seems to be one of a gradual process of the birth and rise of modern languages and of the progressive sclerosis and death of Latin.

When we try to map out the history of Arabic during the Middle Ages, our first impression is that of a linguistic history diametrically opposed to that of Latin. Apparently not an important language of culture outside the Arabian peninsula and its peripheries before the beginning of Islam, pre-classical and classical Arabic was rapidly diffused into the expanding Islamic(ate) area during the first centuries of Islam. In the centre and in the


west, from Syria and Iraq to the Maghreb, it thus progressively (and more or less radically) marginalized the pre-existing languages, such as Aramaic in the Fertile Crescent, Coptic in Egypt, or Berber in the Maghreb. In the East, on the contrary, it deeply influenced the formation of classical (neo-) Persian, but did not prevail as a spoken language in a linguistic landscape characterized by a multiplicity of Iranian and Turkic idioms. In various parts of the Mashreq, its prestige was thus counterbalanced in some important areas of communication. According to some proponents of Arabic literary studies, Arabic suffered a sort of literary and linguistic crisis after 1100<sup>10</sup> (we will not enter into the problem of Middle Arabic here).11 Even considering this, one must acknowledge that Arabic was still spoken from Morocco to Oman at the end of the Middle Ages, and was even progressing as a vernacular language in Africa.

In view of this proposed dichotomy between a supposedly "Latin" history of extinction and an "Arabic" history of successful propagation, the prerequisites for a comparison between the two linguistic cultures seem to be non-existent—at least at first sight. However, a more detailed analysis helps to downplay some of these differences, especially if it questions the scale of the frame that constituted the basis of comparison so far.

First, classical Arabic was certainly introduced into all societies of the medieval Islamic(ate) sphere. However, even in the early centuries, the majority of the population never learnt to express themselves in this language as native speakers. From the start, they began to create as many Arabic dialects as there were local societies, and these dialects or varieties were *in some aspects* as different from classical Arabic as early medieval Romance languages were from classical Latin. This situation resulted in what, in the 1950s, the linguist Charles Ferguson defined as structural

<sup>10</sup> For a standard view of the stylistic, literary, and linguistic decadence of Arabic at the time of the Turcization of political power in the Mashreq (to be followed after 1500 by an analogous Turcization in a large part of the Maghreb), see Djamel Kouloughli, *L'arabe* (Paris: Presses universitaires de France, 2007), 93–94, 100. There is a lot to be said against the tendency to analyse the period 1200–1800 as a time of general linguistic, literary, and stylistic decadence, an idea that is too heavily influenced by literary criteria and by the legacy of the *nahḍa* to be of much use from a sociolinguistic and sociohistorical point of view. The persisting sociolinguistic importance of Arabic, even in a context of partial Turcization, is illustrated by the linguistic acculturation of the Turkic elites in Mamlūk Egypt. On this, see Ulrich Haarmann, "Arabic in Speech, Turkish in Lineage: Mamluks and their Sons in the Intellectual Life of Fourteenth-Century Egypt and Syria," *Journal of Semitic Studies* 33 (1988), 81–114. However, on the importance given in Egypt to Turkic and Turkish studies, to the point of starting a "grammatization" of the language, see Robert Ermers, *Arabic Grammars of Turkic: The Arabic Linguistic Model Applied to Foreign Languages and Translation of Abū Ḥayyān al-Andalusī's Kitāb al-Idrāk li-Lisān al-Atrāk* (Leiden: Brill, 1999).

<sup>11</sup> On Middle Arabic as a conceptual tool for examining the intermediate levels between theoretically "pure" classical Arabic and "pure" dialect, and on the confusions that result from different uses of the concept, see Pierre Larcher, "Moyen arabe et arabe moyen," *Arabica: Revue d'études arabes et islamiques* 48 (2001), 578–609; Jacques Grand'Henry and Jérôme Lentin, eds, *Moyen arabe et variétés mixtes de l'arabe à travers l'histoire* (Leuven: Peeters, 2008).

diglossia.12 Not only people in the "Latin," but also in the "Arabic" sphere used highly different varieties of language to write and to speak, with an entire scale of intermediate levels mingling on different occasions.

Second, the eastern half of the Islamic sphere never switched to Arabic at the level of daily speech: Iran and Central Asia were integral parts of the classical Islamic sphere, and they consequently developed a relationship with Arabic that was more akin to the interaction with Latin characteristic of German- and Slavic-speaking areas in the Latin-Christian sphere. In these regions, Arabic persisted as a prestigious written (and in some contexts oral) language, but Persian progressively acquired some pre-eminence in the fields of poetry and even administration. Other languages, in turn, such as the Iranian vernacular languages of Afghanistan or Khwarezm, or the Turkic languages, remained confined to a predominantly oral dimension until very late in the medieval period.13 Symbolically and conceptually, Arabic remained at the pinnacle of the hierarchy of languages. At the written level, however, it interacted with other idioms in increasingly complex forms of triangulation. We can thus find the same kind of linguistic complexity, and the same kind of linguistic and sociolinguistic latent or open tensions, in thirteenth-century Seljuk Anatolia or in fifteenth-century Transoxiana as in, for example, fourteenth- and fifteenth-century Bohemia or England, with a Persian–Turkic–Arabic triangle versus a German–Czech– Latin triangle, or a French–English–Latin triangle (see Fig. 1.1). In the last


Figure 1.1, a–d: Four linguistic triangles with Latin and Arabic in a semi-symbolic, semi-effective domineering position at the end of Middle Ages. The arrows symbolize lexical and stylistic influence; the lightning arrows symbolize tensions.

centuries of the Middle Ages, be it in the Latin or in the Islamic(ate) sphere, symbolic competition for linguistic pre-eminence was no longer reduced to a dialectical rivalry between the theoretically most prestigious language (Arabic, Latin) and an idiom with an inferior status (e.g. Persian, French, German). In late Seljuk Anatolia or fifteenth-century Transoxiana, tensions arose between Persian—a prestigious intermediate language now firmly established as a court and administrative medium, and early Ottoman or Chagatai—the latter being a language widely spoken by the average population, which lacked the prestige of its courtly rival.14 In this process, Arabic

<sup>14</sup> In the case of fifteenth-century Central Asia, this tension comes to the fore in ʿAlī Šīr Nawāʾī's pamphlet on the pre-eminence of Chagatai over Persian, see Robert

still retained religious pre-eminence and even exerted lexical and stylistic influence on Persian and Turkic, even if the latter was already being deeply influenced by the strongly Arabicized language of Persian. Similarly, in late Plantagenet England, the theoretically uncontested status of Latin as the most prestigious ecclesiastical and royal language did not preclude the progressive reinforcement of French as a courtly and administrative language from the twelfth century onwards. Thus, rather than there being an issue with the use of Latin, the actual linguistic battle occurred between French—still a very prestigious medium even after the progressive Anglicization of the nobility, and Middle English—a language that styled itself as an outsider courtly medium under the double influence of Latin and French.15 The establishment and assertion of a courtly form of Czech presents a somewhat similar case. Czech emerged as a written tool under the influence of German in a Bohemian linguistic landscape still largely dominated by Latin, the latter still serving intellectual and ecclesiastical purposes throughout the fourteenth century.16

As soon as we turn to an analysis of a more dialectical relationship, between either Arabic or Latin and a less prestigious vicinal language, a detailed history of the medieval textual cultures of Arabic and Latin offers many counter-narratives that seriously question the idea of a dialectically opposed evolution in the two linguistic spheres. Recent works, for example, have re-evaluated the role of specific forms of Berber in the construction of the Almohad ideology in the Maghreb of the twelfth and early thirteenth centuries. An attempt to promote Berber as an alternative sacred language, symbolically opposed to Arabic as "the Maghrebian/Occidental language" (*al-lisān al-maġribī*) was accompanied by its temporary promotion as a courtly language. Admittedly, the traces of this experiment are relatively scarce, since the collapse of the Almohad Empire and the subsequent *damnatio memoriae* of its cultural and religious programme

Devereux, "Judgment of Two Languages: Muḥākamāt al-Lughatain by Mīr ʿAlī Shir Nawāʾī: Introduction, Translation and Notes," *Muslim World* 54 (1964), 270– 287, and 55 (1965), 28–45. See Ernst Werner, *Die Geburt einer Großmacht: Die Osmanen (1407–1480)* (Berlin: Akademie Verlag, 1966), on the occasional tension between Persian, considered the elite language, and Anatolian Turkic at the end of the Seljuk era in western Anatolia. Werner's study contains a sketchy presentation of the initial attempts to substitute Turkic for Persian and Arabic in the political and administrative areas of late Seljuk Anatolia in a period of political turmoil.

<sup>15</sup> On the relations between Latin, French, and English in late medieval England, see Serge Lusignan, "Communication in the Later Plantagenet Empire: Latin and Anglo-Norman as Regal Languages," in *The Plantagenet Empire, 1259–1453*, ed. Peter Crooks et al. (Donington: Shaun Tyas, 2016), 273–289.

<sup>16</sup> See Éloïse Adde-Vomácka, *La chronique de Dalimil: Les débuts de l'historiographie nationale tchèque en langue vulgaire au XIVe siècle* (Paris: Publications de la Sorbonne, 2015), on the rise of Czech as a subsidiary courtly and written language in the shadow of German and under the general umbrella of Latin. This process took place amid rising tensions between the German-speaking minority and the Czech-speaking majority in Bohemia.

deprived this linguistic initiative of a future.17 However, while we cannot deny that Arabic relegated the Berber language to the margins of the written tradition in the long-term, this process was neither unilinear nor total during the Middle Ages. Nor was it the case of a simple battle of Arabic against Berber. According to what can be deduced from its few surviving testimonies, Almohad courtly Berber was elaborated as a new medium under the strong lexical and stylistic influence of Arabic. Comparable to the case of Persian or later Turkic, the creation of a sophisticated literary medium occurred through a process of hybridization, in which Arabic served as a stylistic and lexical matrix, in accordance with its status as a reference language.

At the same time, the conventional idea that Latin's evolution from a dominant to a marginalized language in the political communication of the Christian West was essentially linear must be somewhat tempered. The Canadian researcher Serge Lusignan, for example, has demonstrated quite successfully that the progressive emergence of the king's (Parisian) French as the political language of the French royal administration during the fourteenth century was not a smooth process.18 Indeed, after an initial period during which the proportion of written French documentation had steadily increased, the royal chancery suffered a total process of "re-Latinization." King John II (r. 1361–1365) even considered the (at that time) comparatively recent use of French for royal communication one of the factors of decay that led to the early French defeats in the Hundred Years' War. Consequently, he ordered the royal chancery to re-establish the Latin monopoly for writing royal letters and mandates. It was only after John II's death that the process of "Francization" began anew, resulting in the total elimination of Latin, but only at the beginning of the sixteenth century. The eastern parts of the classical Islamic sphere offer parallels to this notion that the use of the traditional idiom would have politically stabilizing effects, which, in this case also, resulted in a process of transitory linguistic restoration.19

However, my point here is more general. While the histories of the attempted promotion of Berber as an alternate, sacred, and courtly language in the Almohad Empire and of the ephemeral effort to re-Latinize the French chancery during the reign of John II diverge strongly, both suggest one thing. In the broader, almost infinite variety of sociolinguistic

<sup>17</sup> On this point, see Mehdi Ghouirgate, *L'ordre almohade (1120–1269): Une nouvelle lecture anthropologique* (Toulouse: Presses universitaires du Mirail, 2014), 215– 251; and Mehdi Ghouirgate, "Le berbère au Moyen Âge: Une culture linguistique en cours de reconstitution," *Annales HSS* 70 (2015), 577–605.

<sup>18</sup> See Serge Lusignan, *La langue des rois au Moyen Âge: Le français en France et en Angleterre* (Paris: Presses universitaires de France, 2004), particularly 107–116.

<sup>19</sup> On this point, see Richard N. Frye, "The Sāmānids," in *The Cambridge History of Iran*, vol. 4. *The Period from the Arab Invasion to the Saljuqs*, ed. Richard N. Frye (Cambridge: Cambridge University Press, 1975), 136–161, particularly 144–145, which addresses hesitations to switch from Arabic to Persian as the chancery language in the period of the Samanid and Ghaznavid dynasties.

constellations that characterized the two enormous spheres during the medieval period, there is no lack of examples to challenge an overly strict teleological vision of the histories of either Arabic or Latin. The former represented far more than a language that was bound to prevent the emergence of other courtly languages west and south of the Turco-Iranian-speaking world. The latter, in turn, cannot be reduced to an idiom doomed to become obsolete as a consequence of the increasing use of the European vernacular languages for administrative and political purposes. Both Arabic and Latin could temporarily assume a defensive or offensive role, for example, when the Almohads promoted Berber to the detriment of Arabic, or when the French royal administration restored Latin as a chancery language in 1351. Even when both languages were confined to a prestigious but not pervasive role linked to the sacred sphere and learned controversies rather than to courtly and direct political use, they played a somewhat analogous role as media of supra-regional and even global linguistic communication. This we have seen in connection with late Seljuk Anatolia or fifteenth-century Transoxiana in the case of Arabic, fourteenth-century England and Bohemia in the case of Latin.

This suggests that we cannot reduce the comparative history of Latin and Arabic to the times and spaces in which they were most intensively spoken and/or held an almost total monopoly in the sphere of writing. On the contrary, their status as reference languages, in societies where other languages competed for intermediate positions as courtly prestigious languages, bears striking similarities, in particular because they remained the ultimate source of inspiration for the stylistic and semantic improvement of idioms emerging more or less rapidly as prestigious tools of communication. Elizabethan English, after all, is a combination of the triple heritage of Medieval Latin, Old French and Anglo-Norman French, and Middle English, just like Ottoman Turkish depends in almost equal parts on classical and post-classical Arabic, semantically and stylistically Arabized Persian, and Anatolian Turkic. Consequently, there is a lot to be said in favour of a new comparative history of Medieval Latin and Arabic that fully considers the role of both languages as symbolically and conceptually central in multilingual areas. A new comparative approach should neither exclude those areas during the classical Islamic period where Arabic ultimately disappeared or was marginalized as a spoken language, nor should it consider as secondary the history of late Medieval Latin as a communication tool that naturally competed in many fields with the written vernacular languages. A comparison of the two linguistic spheres must consider the various stages and different spaces of their respective histories, including a variety of sociolinguistic combinations between the two reference languages and other idioms.

#### 1.2 Thinking the world through two languages: Limits and perspectives of a comparative study of Medieval Latin and Arabic as conceptual tools

We have seen that it is possible to establish a number of structural similarities between two linguistic histories that mainstream scholarship has tended to regard as essentially different. These similarities have more than just general implications, as they apply to a variety of sociolinguistic levels as well. In the classical Islamic sphere, for example, a tradition of grammatical excellence existed among scholars who were not native Arabic speakers but of Iranian origin. To some degree, this tradition echoes the relevance of non-Romance speakers among some important schools of grammar in the Medieval Latin sphere, such as the Danish grammarians of the Parisian "modist" school of the thirteenth century. The reason for this perhaps is that these literati were not native speakers of language varieties that displayed a strong genetic relation to either Arabic or Latin. This would have allowed them to analyse either Arabic or Latin from an external, more analytical perspective.20

At a broader level, processes of literarization in the shadow of the respective reference language also display interesting parallels. The literarization of modern languages in the Latin West did not begin with the Romance languages such as French, Occitan, or Italian, but rather with

<sup>20</sup> On this grammatical current, see Martinus de Dacia, *Opera*, ed. Heinrich Roos (Copenhagen: G. E. C. Gad, 1961); Irène Rosier, *La grammaire spéculative des Modistes* (Lille: Presses Universitaires de Lille, 1983); Costantino Marmo, *Semiotica e linguaggio nella scolastica: Parigi, Bologna, Erfurt 1270–1330. La semiotica dei modisti* (Rome: Istituto Storico Italiano per il Medioevo, 1994). On the Arabic grammarians of Persian origin in early and classical Islam, see Victor Danner, "Arabic Literature in Iran," in *The Cambridge History of Iran*, vol. 4. *The Period from the Arab Invasion to the Saljuqs*, ed. Richard N. Frye (Cambridge: Cambridge University Press, 1975), 566–594. For Zamaḫšarī, see Djamel Kouloughli, *Le résumé de la grammaire arabe par Zamaḵšarī* (Paris: ENS Éditions, 2007). Naturally, there were also numerous Arabic grammarians with some sort of Arabic as their native language, as well as Latin grammarians issuing from Romance-speaking milieus. We must assume, however, that the interference between dialects and classical Arabic, or between Romance languages (or, at an early stage, colloquial late Latin) and classical Latin, was bound to create some difficulties. In fact, the necessity to learn Latin or Arabic as a foreign language could actually prove useful from a conceptual point of view. See what seems a strange confirmation of this hypothesis *per absurdum* in a Franciscan correspondence edited by Michael Bihl and Arthur C. Moule, "De duabus epistolis Fratrum Minorum Tartariae Aquilonaris," *Archivum franciscanum historicum* 16 (1923), 89–112. While trying to convert Qipchak-speaking populations in the first half of the fourteenth century, a number of friars based in a Crimean convent complain that the Italian and French brothers are unable to learn the language correctly, whereas the Anglo-Saxons, Germans, and Hungarians have no particular problem. Could this be a testimony to the formation of good linguistic learning habits, facilitated by the initial effort made by speakers of non-Romance languages to master Latin and one or two Romance *linguae francae*? Such learning habits would stand in contrast to a more instinctive and less grammaticalized approach to Latin on the part of Romance speakers that later obstructed their systematic learning of foreign languages.

the Germanic or Celtic languages. This is comparable to the Arabic sphere, where—apart from certain sociolinguistic exceptions, such as Judaeo-Arabic,21 and certain forms of mixed poetry—we do not find many early attempts to write dialectal Arabic regularly, but an early tradition of writing neo-Persian.22 With regard to dialectal Arabic or the Romance languages, the strong etymological and linguistic interferences with classical Arabic or classical Latin certainly contributed to obstructing a clear distinction between the two varieties of language, thus delaying the emergence of written forms. This is why Italian speakers refrained from theorizing about the existence of Italian, as distinct from the Latin, until the twelfth century. When they finally created such a theoretical framework, they did not use the term "Italian," but used the term "volgare," meaning a vulgar variety of speech, just as an average Arab literate would speak of his Egyptian in the fourteenth century as his dialect or "loose language" (using the term *ʿāmmiyya*, or a similar one), not as a separate tongue.23 This situation persisted in the Arabic-speaking sphere after the medieval period, leading to the actual imbalance between the neoclassical *ʿarabiyya mutawassiṭa* used for literary or press purposes and for highly formal levels of communication,


and the pervasiveness of dialectalized forms in everyday speech. Today, a blatant difference exists between the linguistic regimes of western and central European countries, where Latin has disappeared even as a stylistic reference and where almost every country elaborates and defends its national language, and the Arabic-speaking sphere, where the transformation of dialect-based languages into national tools of communication is met with fierce resistance at different levels. However, this observation of contemporary phenomena does not invalidate the fact that many similarities existed between the two linguistic spheres in earlier periods. Rather, it explains why scholarship has downplayed these similarities in accordance with a general tendency to separate the two histories.

A reasonable guess is that the sociolinguistic similarities characteristic of earlier periods mirror structural similarities between two hierarchically organized clusters of societies that were marked by restricted literacy. In these societies, the "high" written varieties of language were mastered by a caste of linguistic specialists who had no intention whatsoever of imposing their linguistic tools on the rest of the society. In these societies, it was an accepted fact that commoners spoke all sorts of local, vernacular languages. The ruling and intermediate elites, in turn, developed a variety of intermediate levels, for example, using a form of linguistic *koinē* for their poetic production such as the poetical Langue d'Oc in southern France and, temporarily, in large parts of the Iberian Peninsula and Italy, or standard poetical neo-Persian in the Turco–Iranian world. We would thus have had at the very least a three-level linguistic organization in these societies, with a permanent interaction between those three broad levels, the mechanisms of which could vary according to time and place. During the Middle Ages, there would have been symbolic and concrete rivalries between languages of the "low" and the intermediate levels. Latin and classical Arabic, however, would have occupied the most prestigious level. This would have gone uncontested, because these languages were so strongly associated with the respective holy books, as well as with the religious and social orders that went with them (Table 1.1).

This structural isomorphism leads us to another set of questions. Can we retrace some similarities between the linguistic ideas prevalent in these societies, despite their obvious differences, determined by the original histories of the two languages? We certainly cannot deny that certain conceptual differences existed. The Qurʾān, for example, was created or revealed only through Arabic, while part of the Bible was originally written in Hebrew, another in Greek, whereas Latin was but the third "official" vehicle of the text (only for the Roman Catholic Church, and leaving Aramaic aside). However, looking beyond such a fundamental difference, we can find some common tendencies to analyse the respective sacred texts. Latin theologians took advantage of the Bible's multilingualism to define a form of "biblical Latin," a linguistic variant interspersed with Hebrew and Greek, a sort of super-language that potentially included the three "languages of the Cross." Some Islamic theologians, in turn, asserted that, due to the



use of terms of Coptic, Greek, Persian, Ethiopian, and Syriac origin in the Qurʾān, the holy text, in fact, contained every language of the world.<sup>24</sup>

More generally, perhaps it is possible to say that—both in the medieval Latin and the Islamic(ate) spheres—linguistics were characterized by two complex and interfering trends: on the one hand, a trend towards scientific, logical approaches to the language; on the other hand, a mystical tendency which saw language as a magical tool providing access to the supernatural. Both trends resulted from the exceptional status given to classical Arabic and classical Latin in their respective spheres. As the languages of holy texts and the original written linguistic norms, they came to define what language as such actually is, and did so for many centuries to come.

A good example of the practical consequences resulting from these similarities is the common asymmetrical development of grammatical analysis of other languages that used Arabic and Latin grammars as their starting point. As a science, grammar had experienced a rigorous development in

<sup>24</sup> On the statements and disputes among early scholars about the Qurʾān as a sacred text written in Arabic but containing mundane language, see Claude Gilliot and Pierre Larcher, "Language and Style of the Qurʾān," in *Encyclopaedia of the Qur'ān*, vol. 3, ed. Jane Dammen McAuliffe (Leiden: Brill, 2003), 109–135, particularly 117–118.

the Arabic sphere since the eighth and above all in the ninth century. The medieval Latin West, in turn, inherited and developed a Latin grammatical science, which had been forged during Late Antiquity and further developed during the sixth century by Priscian.25 In the late Middle Ages, new grammatical techniques and schools flourished in northern France, then in Italy and Germany. In spite of this, 99 per cent of the grammatical thought preserved in the Latin sphere is devoted to classical Latin until the end of the medieval period. The remaining one per cent, naturally, has attracted a lot of attention, since the first tentative descriptions of the grammar of Middle French, Occitan, Icelandic, or Old English represent precious monuments of the linguistic and literary history of Europe.26 These texts have a common quality: whatever the language concerned, it is always strictly analysed through the grammatical categories of Latin. Late Middle French, for example, did not have declensions. Notwithstanding this, fifteenth-century grammars still present this language with reference to the six cases of Latin.

Robert Ermers made the (unfortunately still relatively neglected) observation that Muslim scholars also produced some tentative but brilliant grammatical analyses of non-Arabic languages during the late Middle Ages. In particular, Egyptian (or Egypt-based) Arabic-writing scholars of the fourteenth and fifteenth centuries developed a tradition of analysing Turkic—a consequence of the prominent role of Mamlūk elites in Egyptian and Syrian societies at the time. Some treatises of these masters have survived.27 We also know that similar analyses have been attempted for Persian, even if the greater part of the manuscripts concerned seem to have been lost.28 A closer look at such material reveals that the relationship between Arabic and Turkic as manifest in such treatises displays striking parallels to the relationship between Latin and late Middle French. In terms of percentages, the number of grammars of Turkic as opposed to the masses of Arabic grammars is as small as the number of late Middle French grammars as opposed to the masses of Latin grammars. Moreover,

<sup>25</sup> On Priscian, see Petrus Helias, *Summa super Priscianum*, ed. Leo Reilly, 2 vols. (Toronto: Pontifical Institute of Medieval Studies, 1993); *Priscien: Transmission et refondation de la grammaire, de l'antiquité aux modernes*, ed. Marc Baratin et al. (Turnhout: Brepols, 2009).

<sup>26</sup> For Occitan, see e.g.: John Henry Marshall, *The "Donatz proensals" of Uc faidit* (London: Oxford University Press, 1969), and John Henry Marshall, *The Razos de trobar of Raimon Vidal and Associated Texts* (London: Oxford University Press, 1972); for (Middle) French see Pierre Swiggers, "Le Donait françois: la plus ancienne grammaire du français," *Revue des langues romanes* 89 (1985), 235–251; and Pierre Swiggers, "Les premières grammaires de vernaculaires gallo-romans face à la tradition latine: stratégies d'adaptation et de transformation," in *L'héritage des grammairiens latins de l'Antiquité aux Lumières*, ed. Irène Rosier (Paris: Société pour l'information grammaticale, 1988), 259–269.

<sup>27</sup> Ermers, *Arabic Grammars of Turkic*, principally deals with the works of Abū Ḥayyān al-Andalusī on Turkic.

<sup>28</sup> See Ermers, *Arabic Grammars of Turkic*, 25, for the (apparently lost) works of Abū Ḥayyān on Persian, Coptic, and Ethiopian. When one thinks of the quality of this grammar on Turkic, one cannot but regret bitterly the apparent loss of these other treatises.

the grammatical categories used to analyse Turkic are the same categories that are used to describe classical Arabic in traditional Arabic grammars as well. Thus, Arabic grammarians analysed Turkic according to the grammatical categories invented for classical Arabic, just as Latin grammarians subjected late Medieval French to the system of Latin declension. These parallels suggest that we cannot oppose a Latin sphere radically open to other languages to an Arabic sphere radically closed to other languages at a conceptual level. Instead, we are confronted with two linguistic spheres, which combined the widespread use of a variety of languages with a conceptual predominance of a high-standard language, with permanent consequences for the process of linguistic conceptualization.

In the wake of the Mongol conquests, parallel attempts to create impressive polyglot dictionaries and linguistic tools in both cultural spheres offer yet another example of similar potentialities, produced this time by similar geopolitical and sociolinguistic impulses. In both cases, the conceptual predominance of Latin and Arabic overshadowed attempts to develop full linguistic programmes. The *Codex Cumanicus*, a complex working tool for "Latin" travellers and preachers venturing into Mongol Inner Asia, contains a trilingual Latin–Persian–Qipchak Turkic lexicon, as well as other (generally Latin, but also German) tools for learning Qipchak Turkic.<sup>29</sup> In some ways, it mirrors the almost contemporary Rasūlid Hexaglott, a six-columned lexicon created under the auspices of the Yemenite sultan al-Malik al-Afḍal (r. 764–778/1363–1377) in the second half of the fourteenth century. This complex artefact contains a copious list of terms in six languages: Arabic, Persian, Turkic, Greek, Mongol, and Armenian.30 It was part of a set of linguistic books created under the patronage of al-Malik al-Afḍal, most of which concerned merely Arabic.31 In the two cases, the linguistic aperture vis-à-vis other spheres had been accelerated by geopolitical circumstances, that is, the *pax Mongolica* and its aftermath. Notwithstanding these new Eurasian linguistic horizons, the linguistic conceptualizations that form the basis of both works continue to reflect the unavoidable predominance of the two reference languages of Latin and Arabic.

<sup>29</sup> On the *Codex Cumanicus*, see the pioneering edition *Codex Cumanicus bibliothecae ad templum Divi Marci Venetiarum*, ed. Geza Kuun (Budapest: Editio scientifica Academiae Hungariae, 1880), with an interesting loop-effect: as his name suggests, Geza Kuun was a Hungarian of possible Cuman descent. Also see *Codex Cumanicus: Édition diplomatique avec Fac-Similés*, ed. Vladimir Drimba (Bucharest: Editura Enciclopedica, 2000); and *Il codice cumanico e il suo mondo*, ed*.* Felicitas Schmieder and Peter Schreiner (Rome: Edizioni di Storia e di letteratura, 2005).

<sup>30</sup> On the Hexaglott, see the edition *The King's Dictionary: The Rasūlid Hexaglot*, ed. Peter B. Golden (Leiden: Brill, 2000), with new details provided in Éric Vallet, "La grammaire du monde: Langues et pouvoir en Arabie occidentale à l'âge mongol," *Annales HSS* 70 (2015), 637–664.

<sup>31</sup> For an assessment of his production, see *The King's Dictionary*. For a court treatise with some examples of linguistic/stylistic anecdotes concerning Arabic, see Renato Traini, *Uno "specchio per principi" yemenita: la nuzhat aẓ-ẓurafāʾ wa-tuḥfat al-ḫulafāʾ del sultano Rasūlide al-Malik al-Afḍal (m. 778/1377)* (Rome: Accademia nazionale dei lincei, 2005).

Such an effort to register similar intellectual tendencies linked to comparable sociolinguistic contexts can yet lead to other discoveries in a broad range of fields. This is the case, for example, with the more classical lexicography, that is, the lexicography of the most famous unilingual lexicons of the two spheres.

One of the most important Arabic medieval lexicons is undoubtedly the *Qāmūs al-Muḥīt* of al-Fīrūzābādī (d. 817/1415).32 This lexicon, like many others, is organized around roots, that is, the consonant stemmata of individual lexemes. Each entry contains dozens to hundreds of words, based on the (predominantly) triliteral root system characteristic of the Semitic languages. Medieval lexicographers thus classified entries of semantically very different words, such as *fāris* (the Persian), *fāris* (the horseman), and *fāris* (the lion), among many other words using the root *fāʾ*–*rāʾ*–*sīn*, all under the same section. Moreover, they tried to explain the derivation of these terms from the same root with etymological creativity.33 At first glance, this seems to provide evidence for the relativistic idea that the structure of a language predetermines the ideas, rhetoric, and to some extent, the uses that people make of it—for the triliteral root system uncontestably encourages this trend. What kind of results will we obtain, however, if we compare the gigantic entries characteristic of Medieval Arabic lexicons to the organization of a classical lexicon of Medieval Latin, such as the *Derivationes* of Uguccione da Pisa (d. 1210)? In the second half of the twelfth century, this Italian scholar created a lexicon that was to become one the most popular tools of the late Middle Ages. The bulk of its entries are strangely akin in their dimensions to the multiple-word entries of the *Qāmūs*. In this lexicon, words are not analysed separately, but are regrouped according to their supposed etymological affinities, for example, *augere* (augment), *augustus*  (emperor), *augur* (augur), *avis* (bird), and so on. Just like al-Fīrūzābādī, Uguccione da Pisa makes an effort to justify these semantic constellations with the help of sophisticated etymological reasoning.<sup>34</sup> We could conclude that traditional societies displayed a general tendency to understand the links between words, their forms, and their definitions in terms of broad semantic clouds, rather than from strictly analytical perspectives. The comparison between the *Qāmūs* and the *Derivationes* also seems useful to relativize the idea that the basic structure of the language represents the only determinant factor that conditioned linguistic thought. To be sure, the internal

<sup>32</sup> On the place of the *Qāmūs* in classical and post-classical Arabic lexicography, see John A. Haywood, "Arabic Lexicography," in *Wörterbücher—Dictionaries—Dictionnaires: Ein internationales Handbuch zur Lexikographie—An International Encyclopediae of Lexicography—Encyclopédie internationale de lexicographie*, ed. Franz Joseph Hausmann et al. (Berlin: Walter de Gruyter, 1991), vol. 3, 2438–2447.

<sup>33</sup> In connection with the root *fāʾ–rāʾ–sīn*, see the complex reasoning in al-Fīrūzābādī, *Al-Qāmūs al-muḥīṭ*, ed. Muḥammad Naʿīm al-ʿArqasūsī (Damascus: Maktabat al-risāla, 1998), 562–563, to justify the homonymy between *fāris* ("lion") and *fāris* ("gentleman").

<sup>34</sup> Uguccione da Pisa, *Derivationes*, ed. Enzo Cecchini, 2 vols. (Florence: Galluzzo, 2004).

structure of classical Arabic and its modes of derivations incited scholars to arrange words with different meanings but the same triliteral root into groups—a tendency that has been prolonged in numerous modern dictionaries. But the detailed logic of construction that is characteristic of these traditional lexicons certainly also had to do with traditional modes of thinking—the same modes of thinking that were more or less reflected in the "etymological lexicons" of the late Latin Middle Ages. Thus, I believe that the linguistic structure of the languages examined here constantly interfered with the needs of a traditional society boasting a caste of linguistic specialists who controlled linguistic thought. The latter were not only highly educated in the art of writing, but also—both in the Arabic as in the Latin sphere—strongly dependent on very similar processes of learning and memorizing.

## 1.3 Language in society (I): On some mnemonic, metrical, and rhythmical tools and the logics of medieval teaching

Indeed, one of the apparently original facts of certain medieval Arabic lexicons is that they are constructed according to the last consonants of the roots of the words, rather than to the first one. What, from a modern point of view, would seem to be an aberration must have been a very useful feature in a traditional society, which used the poetic medium in a considerable amount of its textual production. Since classical Arabic poetry combines a metrical system broadly akin to the quantitative system of Latin metres<sup>35</sup> with the systematic presence of rhymes, scholars, whether just starting out or established, had some need for a lexicon organized according to the endings of the words to be retrieved.<sup>36</sup>

The poetic medium was in actual fact essential to the learning processes in both cultural spheres. This led to the creation of numerous versified teaching tools. The study of these didactic versified manuals, as they were used in the Latin sphere, has long been neglected in medievalist scholarship, because positivist researchers despised them as pedagogical aberrations. The situation has improved notably in recent years,<sup>37</sup> but there is

<sup>35</sup> On this system see e.g. Sandro Boldrini, *La prosodia e la metrica dei Romani* (Rome: Carocci, 1992).

<sup>36</sup> On Arabic metrics in early and classical times, see Bruno Paoli, *De la théorie à l'usage: Essai de reconstitution du système de la métrique arabe ancienne* (Damascus: Institut français du Proche Orient, 2008).

<sup>37</sup> On this matter, see the important article by Vivien Law, "Why Write a Verse Grammar," *The Journal of Medieval Latin* 9 (1999), 46–76. For a very short selection of various metrical Latin pedagogical treatises of the thirteenth century recently edited or studied, see Elsa Marguin-Hamon, *L'Ars lectoria Ecclesie de Jean de Garlande: Une grammaire versifiée du XIII<sup>e</sup> siècle et ses gloses* (Turnhout: Brepols, 2003). This is a study on a metrical treatise of liturgical reading, to be compared with the Arabic arts of psalmody. Moreover, see Rüdiger Lorenz, *Summa Iovis: Studien zu Text und Textgebrauch eines mittelalterlichen Lehrgedichts* (Cologne: Böhlau, 2013), the study of a metrical treatise on the writing of prose letters.

still much to do in this area. As every specialist in Arabic traditional literacy knows, an entire set of counterparts—functional and structural—of these Latin instruments exists in the medieval and early modern literary culture of the Arabic sphere. The hypothesis that one culture borrowed them from the other to enhance the significance of this similarity is not required.

In the thirteenth century, famous Latin versified grammars such as the *Graecismus* of Eberhard of Béthune (d. ca*.* 1212)38 find a parallel in the *Alfiyya* of the Andalusī scholar Ibn Malik (d. 672/1274).39 On both shores of the Mediterranean, we encounter versified medical treatises, or rather introductory lessons to university medicine. The poem *Urǧūza fī l-ṭibb* by Ibn Sīnā/Avicenna (d. 428/1037), composed in the *rağaz* metre and later translated into Latin under the poetic title of the *Canticle of Avicenna*, provides one example,40 the *Regimen sanitatis* of the Salerno school, composed in Latin hexameters, another.41 There are some apparent limits to these parallel manifestations of comparable linguistic phenomena. In the Latin West, for example, a metrical abstract of the Bible existed, the so-called *Summarium Bibliae*, <sup>42</sup> whereas I do not know of a versified abstract of the Qurʾān. However, this apparent asymmetry has a functional cause. The Latin Bible, being far longer than the Qurʾān, was never entirely committed to memory. Nevertheless, memorizing sections of it was part of the clerk's average curriculum. Against this backdrop, it is not surprising that the Book of Psalms, i.e. the section of the Bible that the majority of clerks would have already memorized during their schooling, is the only portion of the Bible that is not contained in the verses of the *Summarium Bibliae*.

We should not classify such similarities as superficial. The same writing and reading processes often form part of an entire textual cycle. Thus, the versified grammars did not only use the most common metrical form the Latin hexameter in one case, the Arabic *raǧaz* metre in the other.<sup>43</sup> We also notice the same tendency to use these texts as mnemonic support for

See also Alexander de Villa-Dei, *Das Doctrinale des Alexander de Villa-Dei*, ed. Dietrich Reichling (Berlin: A. Hoffmann, 1893, reprint Aalen, 1974), which represents a classical grammatical and lexical manual.

<sup>38</sup> On the *Graecismus*, see Anne Grondeux, *Le Graecismus d'Évrard de Béthune à travers ses gloses: Entre grammaire positive et grammaire spéculative du XIIIe au XVe siècle* (Turnhout: Brepols, 2000).

<sup>39</sup> On the *Alfiyya*, see Antoine Sylvestre de Sacy's pioneering work in Ibn Mālik, *Alfiyya* [= *Alfiyya ou la quintessence de la grammaire arabe: ouvrage de Djémal eddin Mohammed connu sous le nom d'Ebn Malec*], ed. Antoine Sylvestre de Sacy (Paris: Imprimerie Nationale, 1833).

<sup>40</sup> See the Arabic-Latin edition in al-Ḥusayn Ibn ʿAbd Allāh Ibn Sīnā, *Urjūza fī ṭ-Ṭibb— Cantica Avicennae. Texte arabe, traduction française, traduction latine du XIIIe siecle avec introductions, notes et index*, ed. Henri Jahier and Abdelkader Noureddine (Paris: Les Belles Lettres, 1956).

<sup>41</sup> On the *Regimen sanitatis*, see *Flos medicine (regimen sanitatis salernitanum)*, ed. Virginia de Frutos González (Valladolid: Universidad de Valladolid, 2010).

<sup>42</sup> On this central Latin mnemonic tool, understudied despite its expansive diffusion, see Lucie Doležalová, "*Biblia quasi in saculo: Sumarium Biblie* and Other Medieval Bible Mnemonics," *Medium Aevum Quotidianum* 56 (2007), 5–35.

<sup>43</sup> On the *rağaz*, see Jaakko Hämeen-Antilla, "Rajaz," in *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh, vol. 4 (Leiden: Brill, 2009), 32–37.

broader, ever-growing commentaries. Just as the *Graecismus* served as an anchor for a textual web of glosses and notes that proliferated during the fourteenth century, the *Alfiyya* saw the development of a thicket of commentaries and sub-commentaries. These similarities and parallels form an integral part of so many types of knowledge characteristic of the Arabic and Latin Middle Ages that they deserve further analysis.

# 1.4 Language in society (II): *ʿilm al-inšāʾ* and *ars dictaminis*

I would now like to discuss briefly the existence of another "structural" similarity between the medieval Latin and Arabic literary cultures, which so far has received only partial recognition in medievalist scholarship; I allude to the importance of the two arts of creating musicalized, ornate prose. They are part of the global set of writing techniques mastered in both literary cultures, and are known respectively as *ars dictaminis* and *ʿilm al-inšāʾ*.

Numerous forms of rhythmical prose existed during the Latin Middle Ages. The system of political, solemn, epistolary Latin communication of the late Middle Ages, for example, was dominated by a set of writing techniques known as *ars dictaminis*. This roughly translates as the "art of composition," with connotations quite similar to the Arabic term *ʿilm al-inšāʾ*, which roughly translates as "the science of [literary] production."<sup>44</sup> The principal characteristic of this Latin writing style is the use of a set of rhythmical ornamentations called *cursus rhythmicus*. The presence of these rhythmical embellishments was mandatory before every minor or major pause of the phrase.45 These ornaments are found in an enormous num-

<sup>44</sup> On *Ars dictaminis*, see Martin Camargo, *Ars dictaminis ars dictandi* (Turnhout: Brepols, 1991); and Martin Camargo, ed., *Medieval Rhetorics of Prose Composition: Five English* "*Artes dictandi" and their Tradition* (Binghamton, NY: Binghamton Press, 1995); Anne-Marie Turcan-Verkerk, "Répertoire chronologique des théories de l'art d'écrire en prose (milieu du XI<sup>e</sup> s.–années 1230). Auteur, œuvre(s), inc., édition(s) ou manuscrit(s)," *Archivum latinitatis medii aevi* 64 (2006), 193–239. See also Benoît Grévin and Anne-Marie Turcan-Verkerk, eds, *Le dictamen dans tous ses états: Perspectives de recherche sur la théorie et la pratique de l'ars dictaminis (XIe –XV<sup>e</sup> siècles)* (Turnhout: Brepols, 2015), which contains an extensive and updated bibliography.

<sup>45</sup> The theory and the use of the *cursus* was reactivated at the papal chancery after a long period of relative neglect. From there it was diffused in a semi-standardized form to every important European laical or ecclesiastical chancery from the twelfth century onwards. See Gudrun Lindholm, *Studien zum mittellateinischen Prosarhythmus: Seine Entwicklung und sein Abklingen in der Briefliteratur Italiens* (Stockholm: Almqvist & Wiksell, 1963); Tore Janson, *Prose Rhythm in Medieval Latin from the 9th to the 13th Century* (Stockholm: Almqvist & Wiksell, 1975); Benoît Grévin, "L'empire d'une forme: Réflexions sur la place du cursus rythmique dans les pratiques d'écriture européennes à l'automne du Moyen Âge (XIII<sup>e</sup> –XV<sup>e</sup> siècle)," in *Parva pro magnis munera: Études de littérature tardo-antique et médiévale offertes à François Dolbeau par ses élèves*, ed. Monique Goullet (Turnhout: Brepols, 2009), 857–881; Anne-Marie Turcan-Verkerk, "La théorisation progressive du cursus et sa terminologie entre le XI<sup>e</sup> et la fin du XIV<sup>e</sup> siècle," *Archivum latinitatis medii aevi* 73 (2015), 179–259.

ber of average or sophisticated Latin prose texts from the twelfth to the fourteenth century. Since solemn political writings were composed according to these techniques, these rhythmical effects were not limited strictly to the literary field, but also affected a vast array of texts, from law via epistolary communication to diplomas and official historiography.<sup>46</sup>

The example presented below is taken from a banal epistle written in the papal chancery during the first half of the thirteenth century. The passages in italics show the sequences that were constructed according to rhythmical schemes; the accents indicate the stressed syllables that musicalized these segment-endings; *velox*, *tardus*, and *planus* are the respective names of the rhythmical combinations thus created. The translation gives insight into the topic of the text, but is not really relevant for the discussion, since every possible theme could be the object of similar "musicalizing" processes.

Cur Florentie *tráxeris tantam móram*velox gravans *ecclésias sine cáusa*velox, *scíre non póssumus*tardus nec id *grátum habémus*planus. Quare tibi presentium *tenóre mandámus*planus, quatenus ad nos visis *preséntibus revertáris*velox. 47

Why do you have to stay so long in Florence and why do you burden the churches without good reasons, that is what we can neither know, nor approve! Consequently, according to the tenor of the present letters, we order you, as soon as you will have read them, to come back to us!

<sup>46</sup> For the most important letter collections, which were formalized according to this technique and served as a model for papal and royal propaganda during the thirteenth and fourteenth centuries, see Matthias Thumser, "Les grandes collections de lettres de la curie pontificale au XIII<sup>e</sup> siècle: Naissance, structure, édition," in *Le dictamen dans tous ses états: Perspectives de recherche sur la théorie et la pratique de l'ars dictaminis (XIe –XV<sup>e</sup> siècles)*, ed. Benoît Grévin and Anne-Marie Turcan-Verkerk (Bibliothèque d'histoire culturelle du Moyen Âge 16) (Turnhout: Brepols, 2015), 209–241. For an application of the same technique to ordinary administrative royal correspondence, see *Il registro della cancelleria di Federico II del 1239–1240*, ed. Cristina Carbonetti Venditelli, 2 vols. (Rome: Istituto Storico Italiano, 2002), the edition of a chancery register that covers six months of administrative mandates in the Kingdom of Sicily under Frederick II. For its literary use see e.g. the famous essay on the love of books *Philobiblon*, written in rhythmical prose by Richard of Bury: *The Philobiblon of Richard de Bury*, ed. Ernest C. Thomas (London: Paul Kegan, Trench and Co, 1888).

<sup>47</sup> Thomas of Capua, "*Summa dictaminis*," in *Die Briefsammlung des Thomas von Capua: Aus den nachgelassenen Unterlagen von Emmy Heller und Hans Martin Schaller*, ed. Matthias Thumser and Jakob Frohmann (Monumenta Germaniae Historica, 2011) http://www.mgh.de/fileadmin/Downloads/pdf/Thomas\_von\_Capua. pdf. [Accessed October 24, 2017], Book I, chapter 62, 46. This collection regroups letters written in the name of various popes from the thirteenth century, as well as the personal correspondence of some members of the papal chancery. It has been selected as an example here because, from the 1270s onwards, it became one of the major formularies for political correspondence in the Latin West.

However, despite its ubiquity, the mechanisms of composing the rhythmical prose of *ars dictaminis* are familiar only to a handful of specialists of Medieval Latin. Their study represents a very small part of the actual scholarship on medieval textual history. To the imperial, royal, or papal notaries, to the clerks and administrators of the thirteenth and fourteenth centuries, however, these small ornaments signified a lot more than just simple rhetoric. Some of them even regarded these techniques of formalization as indispensable tools that reflected the harmony of the universe. Consequently, they credited these rhythms with a number of highly emotional, almost magical powers.<sup>48</sup>

Comparing this Latin art of composition with the sociolinguistic and stylistic cultures of the classical Islamic sphere might enhance our understanding of both the uses and social implications of these styles. The Arabic counterpart of *ars dictaminis*, the so-called ʿ*ilm al-inšāʾ*, is far better known and studied. This "science of composition" was taught to the Arabic-writing scribe (*kātib*) enabling him to write every sort of political or personal prose texts in a lavish, rhythmical, and rhymed prose.49 Obviously there are significant differences between *ars dictaminis* and ʿ*ilm al-inšāʾ*, mainly because the ornamentations of *ʿilm al-inšāʾ* require not only the use of rhythm, but above all of rhyme (in the form of *sağʿ*). Rhyme is a pervasive feature of classical Arabic poetry and clearly plays an important role in the Qurʾān.50 In classical Latin poetry, by contrast, it was generally absent, as it was from the Latin version of the Bible. Consequently, rhyme had not conditioned the development of sophisticated Latin prose to the same degree, even though rhyme had acquired some importance in Medieval Latin, in a complex process of interaction between vernacular and Latin poetry. An extract from a letter of submission (*bayʿa*), which was addressed to the caliph of Baghdad and composed for the ephemeral ruler of a province of the Spanish Levante by an Andalusī stylist during the first half of the thirteenth century, provides an example of the classical rhyme (and rhythm) effects of a solemn document composed according to the science of *ʿinšāʾ*. <sup>51</sup> While

<sup>48</sup> On this question, see the first pages of one classical *ars dictandi* (theoretical treaty of *ars dictaminis*), the *Candelabrum* of Bene of Florence (1225), that is, Bene Florentini, *Candelabrum*, ed. Gian Carlo Alessio (Padua: Antenore, 1983), and, above all, the short treatise of Giovanni del Virgilio, edited in Paul Oskar Kristeller, "Un '*ars dictaminis*' di Giovanni del Virgilio," *Italia medioevale e umanistica* 4 (1961), 181–200. According to this treatise, the choice of appropriate rhythmical ornaments "moves the soul" of the auditor.

<sup>49</sup> For an introduction that remains valid and contains a substantial but dated bibliography on *ʿilm al-inšaʾ*, see H. R. Roemer, "Inshāʾ," in *Encyclopédie de l'Islam*, vol. 3 (Leiden: Brill, 1971), 1273–1276. On epistolography, one of the major fields of application, see the bibliography in Werner Diem, "Arabic Letters in Pre-Modern Times: A Survey with Commented Selected Bibliographies," *Asiatische Studien/ Études Asiatiques* 62, no. 3 (2008), 843–883.

<sup>50</sup> On the *saǧʿ* in the Qurʾān, see Devin J. Stewart, "Sajʿ in the Qurʾān: Prosody and Structure," *Journal of Arabic Literature* 21 (1990), 101–139.

<sup>51</sup> Quoted from Werner Diem, *Ehrendes Kleid und ehrendes Wort: Studien zu tašrīf in mamlūkischer und vormamlūkischer Zeit* (Würzburg: Ergon Verlag, 2002), 120; extracted from Aḥmad b. ʿAlī al-Qalqašandī, *Kitāb Ṣubḥ al-aʿšā fī ṣināʿat al-inšāʾ*,

the passages directly conditioned by the rhythmic and rhyming effects are proportionally less extensive than the rhythmic sections of the papal missive, the redoubling effect of the two-, three- or even four-syllable rhyming sequences is striking in musical terms. The passages in bold characters show those parts of the text that seek to create rhymed and rhythmical parallelisms):

Wa-ʿadda ilā s-sulṭāni fulānini l-mušāri ilayhi min tašrīfi d-dīwāni l-ʿazīzi n-nabawīyi mā wasamahu mina l-faḫāri bi-ağalli *wasmih* / wa-qalladahu s-sayfa ṣ-ṣārima wa-sammāhu *bi-smih* // fa-talāqā s-sayfāni l-maḍrūbu *wa-l-ḍārib* / wa-štabaha l-wasfāni l-māḍī *wa-l-qāḍib*. Wa-barazat tilka l-ḫilaʿu fa-byaḍḍa wağhu l-islāmi *min sawādihā* / wa-wuḍiʿa l-kitābu fa-kādati l-manābiru tasʿā ilayhi šawqan *min aʿwādihā.* 52

He [the caliph] has transmitted to the Sultan "NN," the elected, the honour/cloth-gift of the powerful and prophetical dīwān, an honour that has impressed on him the pride of the most exalted mark, and he has vested him with the edged sword when he named him with a name, so as to bind the two sabres, the one that is struck and the one that strikes, thus combining the two qualities of penetration and edge. And when the gift-clothes appeared, the face of Islam whitened under their darkness, and when the writing was bestowed, one could have said that the minbar-chairs were running toward it, such was the impulse of their wood.<sup>53</sup>

Apart from the fact that they accorded a different role to rhyme, we can postulate that the two rhetorical techniques broadly served the same purposes. They were employed to magnify the linguistic liturgies of medieval power and enhance its communication. They were supposed to enable

<sup>14</sup> vols., ed. Yūsuf ʿAlī Ṭawīl (Beirut: Dār al-kutub al-ʿilmiyya, 1987), vol. 9, 310. Diem's selection of ceremonious texts—a genre of the *tašrīf*, letters that accompany cloth presents—are good examples of an official, courtly *inšāʾ*. They are generally of *Mašriqī* origin, although the text quoted here is an exception. For *Maġribī* usages of political *inšāʾ*, see the *taqdīm* letters in Almohad formularies, such as those edited in Pascal Buresi and Hicham el Aallaoui, *Gouverner l'empire: La nomination des fonctionnaires provinciaux dans l'empire almohade, Maghreb, 1224–1269: manuscrit 4752 de la Ḥasaniyya de Rabat contenant 77 taqdīm-s "nominations"* (Madrid: Casa de Velázquez, 2013); English version: *Governing the Empire: Provincial Administration in the Almohad Caliphate (1224–1269): Critical Edition, Translation, and Study of Manuscript 4752 of the Ḥasaniyya Library in Rabat Containing 77 Taqādīm ("Appointments")* (Leiden: Brill, 2012). One can observe the same tendency that prevails in Western Latin culture: the *inšāʾ* is not restricted to a literary (or to a reputedly non-literary) genre. It is used in the literature of the *maqāmāt*, as well as in the official correspondences of sovereigns, or in epistolary exchanges among literati; it transcends textual boundaries.

<sup>52</sup> Diem, *Ehrendes Kleid und ehrendes Wort*, 120.

<sup>53</sup> This translation, not really necessary since every kind of rhythmic prose could serve as an example, is only tentative, given that the text is full of word play.

scribes and professional writers to make full use of their memorized poetic knowledge, even as they wrote prose documents. Finally, they were intended to satisfy a general aesthetic of writing which could not conceive of a text written without some musical effect.

In the case of this writing technique, as with many others, comparing the two literary spheres helps us recall more effectively that the textual cultures of pre-modern times were as different from ours now as they were similar to each other then. Today, we no longer possess an intermediate level between poetry and purely administrative, political, or epistolary prose. In the Islamic or Latin Middle Ages, however, such an intermediate level not only existed but was ubiquitous: rhymed or rhythmical prose texts were a prevalent feature of literary and also political and administrative communications, and the rhetorical techniques invented to compose such texts inside the linguistic frame of Latin and Arab were progressively transposed into vernacular languages such as, for example, German, Italian, Persian, or Turkic.<sup>54</sup>

# 1.5 Conclusion

The ornate prose systems of the Latin and Arabic Middle Ages are certainly not a popular research theme—outside of a handful of specialists obsessed with stylistic questions, few scholars care to explore their intricacies.<sup>55</sup> However, more than one researcher has noticed the similarities

<sup>54</sup> Taking a comparative perspective, an additional aspect of the comparison between *inšāʾ* and *ars dictaminis* seems relevant: the two stylistic concepts originated in Arabic and Latin writing cultures, but they were tentatively adapted to other languages. For a good example of Arabo-Persian *inšāʾ* used in princely political communication, see David Durand-Guédy, "Diplomatic Practice in Salğūq Iran: A Preliminary Study Based on Nine Letters about Saladin's Campaign in Mesopotamia," *Oriente Moderno*, new series 88, no. 2 (2008), 271–296. For the first attempts at Italianization of *ars dictaminis*, see e.g. Matteo dei Libri, *Arringhe*, ed. Eleonora Vincenti (Milan: Riccardo Ricciardi, 1974). The work contains models of political speeches formulated in the northern Italian communes by the notary Matteo dei Libri in around 1260. Partly due to the differences of accentuation between Latin and many vernacular languages, Western Christian scholars encountered various difficulties when they tried to replicate the schemes of *ars dictaminis* and to adapt them to the vernacular languages. Such difficulties are a testimony to the limits of a structural, isomorphic comparison, when it collides with an evolutionistic approach.

<sup>55</sup> There certainly is an imbalance between Islamic (particularly, but not only Arabic) and Western Latin studies. In Islamic studies, *inšāʾ* is perhaps not always admired—for its association with the decadence of Arabic, see Kouloughli, *L'arabe*, 93–94. However, it is relatively well-known as an important device of traditional writing practices, even among historians who are not particularly interested in the theme. In Western Latin studies, in turn, the study of *ars dictaminis* and the rhythmical ornamentation of medieval Latin prose is the reserve of a handful of specialists. An interest in conceptualizing political or annalistic prose practice from this angle is lacking almost everywhere, a discrepancy that can be explained by the continuity of this writing practice in Arabic letters into the nineteenth century, as opposed to the collapse and the oblivion of the

between *dictamen* and *inšāʾ*. In one case at least, it has led the scholar George Makdisi to postulate that the development of these writing techniques in the Latin world depended on the importation of Arabic writing techniques via Sicily and Italy.56 Makdisi was one of the forerunners of the "post-colonial" variant of medieval studies to which I briefly alluded in the introduction to this chapter. Makdisi's hypotheses are probably founded on an error: he equated the very striking structural similarities of numerous textual forms and cultural processes in the medieval Latin and Arabic spheres with a direct causal relationship between the two phenomena. It is undeniable that the Latin West was influenced in some important ways by Arabic knowledge, particularly in the spheres of philosophy and the sciences. However, on a mere formal linguistic level, we do not have to postulate as such that Latin versified grammars or Latin rhythmical prose texts were formally dependent on their Arabic equivalents.<sup>57</sup> The history of these linguistic and literary universes, rather, implies that pre-existing tools were adapted to common needs, with the result that similar causes led to similar effects. For example, the two metres most commonly used


*dictamen* in the Western world after 1500. More broadly, this raises the question of what impact the apparent continuity between classical and written modern Arabic, on the one hand, and the almost total discontinuity between Medieval Latin and the vernacular modern languages, on the other, has had on the form assumed by medievalist textual studies during the twentieth and twenty-first centuries. Scholars interested in the older Latin written culture do not seem to have developed the same empathy with their sources as scholars interested in the older Arabic written culture, with the benefits, but also the disadvantages of a greater estrangement.

to create the versified didactic manuals, the dactylic hexameter and the *rağaz*, have a very long, independent history in the two literary cultures, which relate to remote antiquity.<sup>58</sup> Rather than a direct cultural transfer from one area to the other, the importance of poetry as a teaching tool of the classical language, be it Latin or Arabic, explains why similar matters, such as grammar or medicine, were taught in a similar way.<sup>59</sup>

I am aware that a structural comparison can go too far and that there is the risk of over-stressing the similarities. In order to strike the correct balance between structural equivalences and concrete differences in the textual cultures of Latin Europe and medieval Islam, we must be able to underscore the differences too. The same sociolinguistic potentialities were certainly not bound to develop in an identical way at every level of textual production, even if many similarities are worth investigating. I believe, however, that a fresh start in the investigation of the structural affinities between the two textual cultures would contribute not only to anthropologists' efforts to define what traditional literacy and traditional writing are, but also help to highlight aspects of these ancient cultures

59 In specific cases, the comparative history based on structural affinities and the history of entanglements (translations, etc.) between Arabic and Latin can intersect. See e.g. the existence of numerous medical poetic Latin treatises composed in Italy (school of Salerno) or in France (*studium* of Montpellier) in the twelfth to thirteenth centuries, and the concomitant translation into Latin of the metrical *Urğuza fī l-ṭibb* of Ibn Sīnā. Even in such a case, it seems that we must resist the temptation to attribute the surge of the Latin genre to the attention accorded to putative Arabic models. The Latin translation of Ibn Sīnā's text (eds. Jahier and Noureddine) was composed during the late thirteenth century, after a variety of independent poetic Latin treatises on medicine had already been written. Gilles de Corbeil provides an example at the beginning of the thirteenth century. See Camille Viellard, *Gilles de Corbeil, médecin de Philippe Auguste et chanoine de Notre-Dame, 1140–1224* (Paris: Champion, 1909); and now Gilles de Corbeil, *Liber de virtutibus et laudibus compositorum medicaminum*, ed. Mireille Ausécache (Florence: Sismel, 2017). The idea of using the metrical Latin form was probably more efficiently stimulated by the surge of all sorts of pedagogical Latin metrical treaties at this time, rather than by the specific influence of the *Urğuza fī l-ṭibb*, which, to my knowledge, was the only Islamic metrical text translated into Latin during the Middle Ages, for obvious medical reasons. However, such connections must be examined accurately in order to gauge the possibility that structural similarities could, in some cases, have resulted from more or less punctual formal influences. For my part, I remain persuaded that the complexity of the cultural foundations of the prosody and stylistics of classical Arabic and classical Latin exclude the possibility that a profound formal influence (such as the imitation of metrical or specific rhythmical ornaments) could have taken place between the two spheres at this level (classical Arabic towards sophisticated Latin). A more direct influence, at lower stylistic levels (Arabic popular or intermediate poetry towards Romance popular poetry) remains more probable, or at least plausible. However, the question remains eminently complex, since we must also take other forms of influence between the two languages into account. The semantic impact of Arabic on the languages of the Iberian Peninsula, for example, necessarily had some influence on a part of the Latin vocabulary there, even if it was relatively limited, if compared with the impact of Arabic on Castilian or Portuguese. However, this problem leads us away from the issue of a structural comparison.

<sup>58</sup> On *rağaz*, see Hämeen-Antilla, "Rajaz." On the dactylic hexameter, see Boldrini, *La prosodia e la metrica dei Romani*.

that—for cultural, modern reasons—are not always clearly perceived by current research. We could thus help to re-valorize ancient ideas, such as the multilingual complexity of the Qurʾān, highlighted by some early Muslim scholars, by comparing them with the Latin analysis of the multilingual lexicon of the Bible. Reciprocally, we could assist scholars of Latin philology to grasp the importance of the Latin *ars dictaminis* and musicalized prose with the help of its structural, better studied Arabic equivalent, *ʿilm al-inšāʾ*. Without devaluing the interest of other, more entangled narrations, a comparative analysis of Arabic and Latin textual cultures could thus serve as a sort of corrective lens that helps to sharpen our understanding of both the Latin and Arabic textual histories.

Daniel G. König (University of Konstanz) **2. Latin-Arabic Entanglement: A Short History**

As linguistic systems comprising a large variety of written and oral registers including derivate languages and dialects, Latin and Arabic have been of paramount importance for the history of the Euromediterranean since Antiquity. Due to their long-term function as languages of administration, intellectual endeavours, and religion, Latin and Arabic are often regarded as cultural markers of Europe and the (Arabic-)Islamic sphere respectively. With regard to Latin, this conviction was already formulated by the humanist scholars Lorenzo Valla (d. 1457) and Juan Luis Vives (d. 1540).<sup>1</sup> It also lay at the basis of the Finnish government's proposal, during its EU-presidency in 2006, to reintroduce Latin as a pan-European medium of communication.2 In Islamic(ate) societies, in turn, the Qurʾānic message of Islam is intrinsically tied to the Arabic language.3 The latter is still used as a medium of communication in a region stretching from Morocco to Iraq. Adherents to the idea of pan-Arabism have highlighted repeatedly that Arabic has to be regarded as an essential cultural feature of this area. In the 1950s, in particular, they underscored the necessity of upholding a standardized form of Arabic in the various national Arab schooling systems and media, rather than adapting the latter to the colloquial realities of an Arab world marked by diverging dialects.4 It is questionable, however, whether Latin and Arabic really constitute cultural identity markers of a clearly defined European and an (Arabic-)Islamic sphere. Latin boasts an important extra-European history in North Africa and has long ceased

<sup>1</sup> Jozef Ijsewijn, "Latin as lingua franca: Renaissance Humanism and Translation," in *Translation. An International Encyclopedia of Translation Studies*, ed. Harald Kittel, vol. 2 (Berlin: de Gruyter, 2007), 1429–1430.

<sup>2</sup> Government Communications Newsletter, "Finland's EU Presidency: Conspectus rerum Latinus 19/2006," December 27, 2006, accessed February 6, 2018, http://www.archive-fi-2013.com/fi/e/2013-08-29\_2754308\_40/Finland-8217 s-EU-Presidency-About-this-site/.

<sup>3</sup> See Qurʾān 2:2; 13:37; 46:12, which mention Arabic as the linguistic medium of Qurʾānic revelation.

<sup>4</sup> This debate is very well documented in the almost contemporary book by Anwar G. Chejne, *The Arabic Language: Its Role in History* (Minneapolis: University of Minnesota Press, 1969), 13, 123, 158–159, 165–166.

to be a defining feature of cultural activity in European societies.5 Arabic, in turn, was never used exclusively by Muslims.6

In Chapter 1 of this volume, Benoît Grévin offered a structural comparison of Latin and Arabic as linguistic systems. The present contribution, in turn, recounts the entangled history of Latin and Arabic from a macro-historical perspective. This entangled history can be divided into three phases. In Phase One, both linguistic systems came into contact in the ancient Roman Near East, in a time in which Arabic as a standardized supraregional language had not yet fully emerged. In Phase Two, the Arabic-Islamic expansion into the western Mediterranean, dominated linguistically by Latin and Romance, ushered in a period of intensive Latin-Arabic entanglement. Lasting approximately from the seventh to the fifteenth century, this period was particularly dynamic: the expansion of so-called "Latin-Christian societies" into Mediterranean regions hitherto under Muslim rule considerably transformed the geopolitical equilibrium of linguistic interaction. In Phase Three, i.e. from the late medieval and early modern periods onwards, the interaction of Latin and Arabic progressively receded into the sphere of academic endeavours. This was mainly due to the rise and fully fledged development of Romance languages. The latter successively replaced Latin in its various fields of interaction with different forms of Arabic.7 Today, at the (preliminary?) "end" of this long history of entanglement, Latin and Arabic mainly seem to meet within an academic framework, both in Europe and the Arab world.

Tracing this macro-history of Latin-Arabic entanglement is interesting for several reasons. On the one hand, its reconstruction proves that the interaction of so-called "Latin-Christian" and "Arabic-Islamic" societies produced a large number of diverse Latin-Arabic milieus in which Latin and Arabic merged to some degree. Although research on many of these milieus boasts a long tradition, this historical-linguistic evidence has rarely been used to counter the dichotomizing master narrative juxtaposing and often opposing "Islam" and "the West." Such a macro-history of what we could define as "Latin-Arabic transculturation"8 traces how different Latin-Arabic

<sup>5</sup> Jürgen Leonhardt, *Latein. Geschichte einer Weltsprache* (Munich: C. H. Beck 2009).

<sup>6</sup> See e.g. *Judaeo-Arabic Studies*, ed. Norman Golb (Hoboken: Taylor and Francis, 2013); Georg Graf, *Geschichte der christlich-arabischen Literatur*, 4 vols. (Città del Vaticano: Biblioteca Apostolica Vaticana, 1944–1953).

<sup>7</sup> See the comprehensive overviews in Reinhold Kontzi, "Das Zusammentreffen der arabischen Welt mit der romanischen und seine sprachlichen Folgen," in *Substrate und Superstrate in den romanischen Sprachen*, ed. Reinhold Kontzi (Darmstadt: WBG, 1982), 387–450; Reinhold Kontzi, "Arabisch und Romanisch," in *Lexikon der romanistischen Linguistik, vol. VII: Kontakt, Migration und Kunstsprachen. Kontrastivität, Klassifikation und Typologie*, ed. Günter Holtus et al. (Tübingen: Niemeyer, 1998), 328–347; Gustav Ineichen, *Arabisch-orientalische Sprachkontakte in der Romania: ein Beitrag zur Kulturgeschichte des Mittelalters* (Tübingen: Niemeyer, 1997).

<sup>8</sup> The conceptual term *transculturación* was coined by Fernando Ortiz in his study on processes of cultural reconfiguration in early modern Cuba. Ortiz used the term to highlight that processes of interaction between groups of different cultural origin not only result in processes of transmission, reception, adaptation,

forms of entanglement came into being over the centuries, thus creating shared spaces between spheres framed by a different religious system and reference language, that are generally considered culturally apart. Since the chronological scope of this chapter covers the period from the Roman intrusion into the ancient Middle East up to the twentieth century, it is possible to provide an overview on very different forms of entanglement and to acquire a typological understanding of the subject. Forms of Latin-Arabic entanglement range from descriptive and analytical comments on the other language to linguistic policy statements, and from different forms of transformation and appropriation—e.g. oral and written translations, bilingual word-lists, glossaries, loans, and calques—to graphic, literary, or systemic forms of linguistic hybridity. The identification and specification of different forms of entanglement gives rise to the question—not systematically answered here—why particular forms of entanglement appeared in a specific historical setting, and not in another. Studying processes of Latin-Arabic transculturation thus gives us the possibility of explaining the link between specific entangled forms and their respective milieu of origin, and consequently allows for defining factors that encouraged such processes.

On the other hand, the history of Latin-Arabic entanglement can also be read as a history of cultural segregation. Many forms and milieus of Latin-Arabic entanglement that resulted from the above-mentioned processes of transculturation succumbed to destructive pressures or retreated into the recesses of academia. We can often explain the disintegration or academic isolation of Latin-Arabic milieus as an effect of different manifestations of religious and cultural othering, intolerance, and even violence. An analysis of the driving forces of Latin-Arabic segregation shows, however, that concomitant processes are too manifold and too intertwined as to support dichotomizing master narratives which draw clear boundaries between Christians and Muslims or between "Islam" and "the West," not least because the disintegration of many Latin-Arabic milieus resulted from the rise of the Romance vernaculars.

Simultaneously tracing the macro-histories of Latin-Arabic transculturation and cultural segregation thus allows us to highlight the complexity of cultural processes in the Euromediterranean. This two-pronged approach deconstructs culturalistic narratives of Euromediterranean history formulated, for example, by proponents of Orientalist and Occidentalist ideologies.9 At the same time it explains, rather than substantiates, why certain voices feel justified in assigning Latin to a "European," Arabic to an "Islamic" cultural heritage, rather than regarding both as intrinsic elements of a shared Euromediterranean history.

and assimilation, but also lead to the transformation and amalgamation of previously distinct cultural elements within a new cultural synthesis. See Fernando Ortiz, *Contrapunteo cubano del tabaco y el azúcar* (Havana: Jesús Montero, 1940); Fernando Ortiz, *Cuban Counterpoint. Tobacco and Sugar* (New York: Knopf, 1947).

<sup>9</sup> Daniel G. König, "Islamic Studies: A Field of Research under Transcultural Crossfire," *Transcultural Studies* 2 (2016), 101–135.

# 2.1 Early entanglement: Rome in the Middle East

Cultural ascriptions that assign Latin to the European, and Arabic to the Islamic sphere play no role, as soon as we turn to the beginnings of Latin-Arabic entanglement in Antiquity. In this period of Roman imperialism spanning three continents, the term Europe did not yet represent a cultural category, while the interfaith rivalries of Jewish, Christian, and Muslim interpretations of monotheism were still inexistent and thus irrelevant for the interaction of Latin and Arabic.

#### 2.1.1 THE ROMAN TAKEOVER

The earliest forms of Latin-Arabic entanglement came into being thanks to the Roman military intrusion into the ancient Middle East of the first century bce. The political constellation of this period was marked by the disintegration of the Seleucid Empire, the expansionist ambitions of the Armenian ruler Tigranes, and the rivalry between Hyrcanus and Aristobulus, two pretenders to the throne in Judaea. All this provided the pretext for Pompeius's interference in Syrian affairs and led to the creation of the Roman province of Syria in 64 bce, to be followed by the military expedition of the proconsul Aemilius Scaurus against the Nabataeans in 62 bce.

The political unrest that followed the murder of Caesar in 44 bce strengthened Rome's foothold in the eastern Mediterranean and thus produced various forms of interaction between Romans and various Middle Eastern populations. Octavian's victory over the joint forces of Marcus Antonius and Cleopatra in the Battle of Actium (31 bce), for example, was supported by the Nabataean ruler Malichus and brought Egypt under full Roman control. Cooperation between Romans and Nabataeans seems to have been a precondition for the earliest Roman expedition into the Arabian Peninsula, led by Aelius Gallus in 25 bce. Roman-Nabataean relations oscillated between cooperation and confrontation in the next century: Nabataean troops under Aretas IV seem to have unilaterally taken control of Damascus in 37–40 ce, but supported the emperor Titus during his conquest of Jerusalem in 70 ce. Under Trajan, the Nabataean realm was integrated into the Roman Empire in 106 ce. From around 111 ce onwards, a new Roman road, the *Via Nova Traiana*, connected the Syrian city of Bostra with the city of Ayla at the Gulf of ʿAqaba.<sup>10</sup>

Under the Severan dynasty, the Roman East began to play an important role in imperial politics. In 198 ce, Septimius Severus re-established the province of Mesopotamia, a territory already held under Trajan for a short time. He gave support to the city of Palmyra, which, in the early third century, brought forth a Roman senator named Septimius Odaenathus. In addition, Septimius Severus married Julia Domna, an aristocratic woman

<sup>10</sup> Glen W. Bowersock, *Roman Arabia* (Cambridge: Cambridge University Press, 1994).

from the Syrian town of Emesa, whose great-nephews Elagabalus (r. 218– 222) and Severus Alexander (r. 222–235) were soon to rule the empire. Backed by Irfan Shahîd, Glen Bowersock claimed that, thanks to the influence of the house of Septimius Severus, "Arabs reached the pinnacle of Roman government."11 This seems corroborated by the epithet "Arabs" ascribed to the emperor Philip I (r. 244–249) in Latin and Greek sources of the fourth and fifth centuries ce. It seems dangerous, however, to attribute too much significance to this ethnic category. Considering the divergent living conditions of Arab groups in the ancient Middle East, many of which remained independent of Rome, the rise to power of a number of Romanized Syrian families cannot be regarded as an "Arabization" of the imperial elite in the third century ce. Later Arabic-Islamic historiography never defines Severan emperors or Philip I as "Arabs."<sup>12</sup> From a historical socio-linguistic point of view, however, the Roman intrusion into the Middle East produced a large variety of relations and thus set the stage for various forms of linguistic interaction, including forms relevant to the early and later history of Latin-Arabic entanglement.

#### 2.1.2 ROME'S LINGUISTIC IMPACT

Rome exerted cultural and linguistic influence on the Middle East and a number of its Arab inhabitants from the first, and especially from the second century ce onwards. Latin inscriptions could now be seen in many parts of the region, ranging from the city of Palmyra,13 via the forty-two milestones of the *Via Nova Traiana* from Bostra to Ayla,14 the Nabataean city of Hegra in the north of the Arabian Peninsula,15 to the Farasan islands in the southern part of the Red Sea.16 Roman army units used a form of Latin as a means of internal communication, with Latin ranks and certain Latin orders surviving far into Byzantine times.17 The settlement of recruits

<sup>11</sup> Bowersock, *Roman Arabia*, 118.

<sup>12</sup> Daniel G. König, *Arabic-Islamic Views of the Latin West*. *Tracing the Emergence of Medieval Europe* (Oxford: Oxford University Press, 2015), 29.

<sup>13</sup> Michał Gawlikowski, "Deux inscriptions latines de Palmyre," *Studia Palmyreńskie* 3 (1969), 77–83 [dating from ca. 206–207 CE]; Khaled Asʿad, "Inscriptions latines de Palmyre," *Revue des Etudes Anciennes* 104 (2002), 363–400 [ca. 52 CE].

<sup>14</sup> David F. Graf, "The *Via Nova Traiana* in Arabia Petraea," in *Rome and the Arabian Frontier: From the Nabaeans to the Saracens,* ed. David F. Graf (Farnham: Ashgate, 1997), art. VI, 1–33.

<sup>15</sup> Dhaifallah al-Talhi and Mohammad al-Daire, "Roman Presence in the Desert: A New Inscription from Hegra," *Chiron. Mitteilungen der Kommission für alte Geschichte und Epigraphik des Deutschen Archäologischen Instituts* 35 (2005), 205–217.

<sup>16</sup> François Villeneuve, Carl Philipps, and William Facey, "Une inscription latine de l'archipel Farasân (sud de la mer Rouge) et son contexte archéologique et historique," *Arabia* 2 (2004), 143–192 [143–144 CE].

<sup>17</sup> Ladislav Zgusta, "Die Rolle des Griechischen im römischen Kaiserreich," in *Die Sprachen im Römischen Reich der Kaiserzeit*, ed. Günter Neumann and Jürgen Untermann (Bonn: Habelt, 1980), 121–145, 132–133.

and veterans created a number of Latin-speaking communities in some parts of the region.<sup>18</sup> Most notable in this context is the city of Beirut. At the end of the second century, it became a centre of Roman legal culture, giving a home to such important Roman jurists as Aemilius Papinianus (d. 212 ce), Domitius Ulpianus (d. ca. 223 ce), and Modestinus Herrenius (d. after 228 ce), whose legal reasonings contributed to establishing Roman precedence in matters of jurisdiction throughout the ancient and late antique Euromediterranean.19 However, Rome's linguistic influence on the Middle East should not be exaggerated. In the eastern Mediterranean, Greek remained the dominant elite language up to the Byzantine era. In the Greater Syrian region, Aramaic held the ground well into the Islamic era in its Judaeo-Aramaic and Syriac variants, the latter increasingly put into writing from the second century onwards.20 In Egypt, Greek retained comparable dominance alongside Coptic, increasingly put into writing from the second century ce onwards.21

Defining the regional Latin influence on Arabic in the period after the first century bce is made difficult by the fact that Arabic as a language of regional importance only began to emerge in this same period. A very rough chronology places the fusion of several northern Arabian dialects into a form of Old Arabic in the period between the seventh and the third century bce. Old Arabic seems to have acquired regional significance because it was used as an oral vehicular language within the Nabataean trade network. Until the fourth century ce, the latter used a particular form of Aramaic in its written documentation that features various Old Arabic terms. Eventually, the stele of Imru l-Qays, dated 328 ce, presents us with a clearly readable Arabic text, represented graphically in the Nabataean variant of Aramaic. Its claim that Imru l-Qays was ruler of all Arabs suggests that this form of Arabic meanwhile facilitated communication between various tribal confederations of Late Antiquity. It then developed to become a poetic prestige language used at the courts of Arab confederations that served as a buffer between the Byzantine and Sassanid empires of the sixth century, i.e. the Ǧafnīds/Ghassanids and the Naṣrīds/ Lakhmids. With the rise of Islam in the early seventh century, this form of

<sup>18</sup> Benjamin Isaac, "Latin in Cities of the Roman Near East," in *From Hellenism to Islam. Cultural and Linguistic Change in the Roman Near East*, ed. Hannah M. Cotton et al. (Cambridge: Cambridge University Press, 2009), 43–72.

<sup>19</sup> See Linda Jones Hall, *Roman Berytus: Beirut in Late Antiquity* (Abingdon: Routledge, 2004), 195–220, esp. 200–201.

<sup>20</sup> Sebastian P. Brock, *A Brief Outline of Syriac Literature* (Kerala: S. Ephrem Ecumenical Research Centre, 1997), 7–8, 18.

<sup>21</sup> Tonio Sebastian Richter, "Greek, Coptic and the 'Language of the Hijra.' The Rise and Decline of the Coptic Language in Late Antique and Medieval Egypt," in *From Hellenism to Islam. Cultural and Linguistic Change in the Roman Near East*, ed. Hannah M. Cotton et al. (Cambridge: Cambridge University Press, 2009), 401–446.

Arabic additionally acquired the status of a sacred language and slowly but surely received the graphic form known today.22

Latin influence on this emerging Arabic language becomes manifest in around forty to fifty loanwords identified by scholarship of the last century. Among these loanwords, listed in Table 2.1, we find ethnonyms and toponyms, ranks and titles, military and administrative terms, terms from the sectors of logistics, transport, finance, commerce, and measurement, and finally a number of terms pointing to forms of higher culture and defining objects of daily life.

The list raises a number of problems. First and foremost, the etymologies are often based on speculation and cannot be substantiated. In most cases, scholars established equivalents between Latin, Greek, Aramaic, and Arabic terms to show how a Latin word moved into the Arabic orbit via Greek and Aramaic. It is often doubtful, however, if the respective Latin term is really of Latin and not, for example, of Greek origin. Christoph Luxenberg went as far as to invert some of the etymologies after pointing out that, in a Mediterranean landscape characterized by several centuries of linguistic interaction between Greek, Latin, Phoenician, and other so-called Semitic languages, it is often impossible to identify the origin of a word.<sup>23</sup>

Even if we assume that the Arabic terms listed in Table 2.1 are of Latin origin, it is generally impossible to pinpoint where and when they were borrowed. Some terms, such as the ethnonym *Romani* > *al-Rūm* or the title *caesar* > *qayṣar* must have entered the one or the other form of Arabic relatively early, as is confirmed in the case of the term *centurio*, which is already attested in a Nabataean inscription.24 Other ranks and titles may have only entered the Arabic language later, e.g. during exchanges between Byzantine Constantinople on the one hand and the Ǧafnīds/Ghassanids, Umayyads, or even Abbasids on the other hand.

<sup>22</sup> Christian Robin, "Les plus anciens monuments de la langue arabe," *Revue des mondes musulmans et de la Méditerranée* 61 (1991), 113–125; Michael C. A. Macdonald, "ARNA Nab 17 and the Transition from the Nabataean to the Arabic Script," in *Philologisches und Historisches zwischen Anatolien und Sokotra. Analecta Semitica in Memoriam Alexander Sima*, ed. Werner Arnold et al. (Wiesbaden: Harrassowitz, 2009), 207–240; Ernst-Axel Knauf, "Arabo-Aramaic and ʿArabiyya: From Ancient Arabic to Early Standard Arabic, 200 CE–600 CE," in *The Qurʾān in Context: Historical and Literary Investigations into the Qurʾānic Milieu*, ed. Angelika Neuwirth et al. (Leiden: Brill, 2011), 197–254; Thomas Bauer, "The Relevance of Early Arabic Poetry for Qurʾanic Studies Including Observations on Kull and on Q 22:27, 26:225, and 52:31," in *Qurʾān in Context*, ed. Neuwirth et al., 699–732; Zbigniew T. Fiema, Ahmad Al-Jallad, Michael C. A. Macdonald, and Laïla Nehmé, "Provincia Arabia: Nabataea, the Emergence of Arabic as a Written Language, and Graeco-Arabica," in *Arabs and Empires before Islam*, ed. Greg Fisher (Oxford: Oxford University Press, 2015), 373–433.

<sup>23</sup> Christoph Luxenberg, *The Syro-Aramaic Reading of the Koran: A Contribution to the Decoding of the Language of the Koran* (Berlin: H. Schiler, 2007), 226–229.

<sup>24</sup> David F. Graf, "The Nabataean Army and the *Cohortes Ulpiae Petraeorum*," in David F. Graf, *Rome and the Arabian Frontier. From the Nabataeans to the Saracens*  (Farnham: Ashgate, 1997), art. V, 289.



In the case of Latin military, administrative, and logistic terminology, it seems quite plausible that Roman imperial presence in the Middle East manifested itself in the linguistic surroundings of the emerging Arabic language. A case in point may be the word *ṣirāṭ*. Probably deriving from the Latin word *strata*, it could attest to the impression Roman roads such as the *Via Nova Traiana* made on Arabic speakers of the Middle

<sup>25</sup> Based on Siegmund Fraenkel, *De vocabulis in antiquis Arabum carminibus et in Corano peregrinis* (Leiden: Brill, 1880); Siegmund Fraenkel, *Die aramäischen Fremdwörter im Arabischen* (Leiden: Brill, 1886); Arthur Jefferey, *Foreign Vocabulary in the Qurʾān* (Baroda: Oriental Institute, 1938); Irfan Shahîd, "Latin Loanwords," in *Encyclopedia of the Arabic Language and Linguistics*, ed. Kees Versteegh, 5 vols. (Leiden: Brill, 2008), vol. 3, 6–8, with further references.

East, before becoming a theological keyword in the first Qurʾānic sura, i.e. *sūrat al-fātiḥa*, one of the most frequently recited Muslim prayers worldwide.26 The Roman presence also had an economic dimension, in that it involved the paying of taxes as well as connections to producers, markets, and consumers farther west. Consequently, the borrowing of financial and commercial terminology also seems plausible. Much more difficult to explain is how certain terms connected to daily life may have entered the Arabic language: it does not seem credible that Middle Eastern populations had to wait for the Romans to become acquainted with such basic objects as buckets, soap, napkins, cloaks, mirrors, and given the long history of literacy in the Middle East—writing materials!

In retrospective, it is impossible to prove, trace, and explain every instance of terminological borrowing implied by the apparent relationship between approximately forty to fifty Latin and Arabic words. It is clear, however, that between the Roman intrusion into the Middle East in the first century bce and the rise of Islam in the early seventh century—Latin and Arabic became entangled in such a way that it becomes possible to imagine complex forms of linguistic interaction, the details of which are lost to us today.

One should note that early forms of Arabic were not only at the receiving end. Various terms of Middle Eastern origin already became part of the Latin language in Antiquity, as is attested in a number of ancient Latin texts. Certain parts of the Arabian flora and fauna were assimilated into the Latin vocabulary, a prime example being Middle Eastern terms for "camel," predating the Arabic form *ǧamal*, which form the basis for the Greek and Latin terms κάμηλος and *camelus*. 27 This also applies to a number of ethnonyms such as *ʿArab* > *Arabes*, *Ṯamūd* > *Thamudeni*. More disputed is the hypothesis that the (pre-)Arabic root *š-r-k* stands at the basis of the Latin term *Saracenus* which, in the course of the early medieval period, became the most frequently used term for Muslims in Latin texts.28 In addition, Table 2.2 shows that both Latin and Arabic adopted a number of Hebrew, Aramaic, and Greek terms, many of which originated in the orbit of Judaism and early Greek Christianity, others designating miscellaneous objects.

<sup>26</sup> Fraenkel, *De vocabulis*, 24; Jefferey, *Foreign Vocabulary*, 195; Guy Monnot, "Ṣirāṭ," *Encyclopaedia of Islam*, second ed., vol. 9 (Leiden: Brill, 1997), 670–671; Irfan Shahîd, *Byzantium and the Arabs in the Sixth Century*, vol. II, part 1 (Washington, DC: Dumbarton Oaks Research Library, 2002), 63. On the ensuing theological interpretation of the term, see Walid A. Saleh, "The Etymological Fallacy and Qurʿanic Studies: Muhammad, Paradise, and Late Antiquity," in *Qurʾān in Context*, ed. Neuwirth et al., 665–666. Cf. Luxenberg, *Syro-Aramaic Reading*, 226–229, who questions the etymological link.

<sup>27</sup> Karl Ernst Georges, *Ausführliches lateinisch-deutsches Handwörterbuch* (Darmstadt: WBG, 1998), vol. 1, 943–944; Cornelia Becker and Philip de Souza, "Camel," *New Pauly* http://dx.doi.org/10.1163/1574-9347\_bnp\_e607690.

<sup>28</sup> David F. Graf, "The Saracens and the Defense of the Arabian Frontier," in David F. Graf, *Rome and the Arabian Frontier: from the Nabataeans to the Saracens* (Farnham: Ashgate, 1997), art. IX, 14–15; Irfan Shahîd and Clifford E. Bosworth, "Saracens," *Encyclopaedia of Islam*, 2nd ed., vol. 9 (Leiden: Brill, 1997), 27; Robert Hoyland, *Arabia and the Arabs. From the Bronze Age to the Coming of Islam* (Abingdon: Routledge, 2001), 229–247, esp. 235.



It is important to highlight that this early period of Latin-Arabic entanglement cannot only be reconstructed by drawing on loanwords. Two bilingual Latin-Nabataean inscriptions from the first century ce found in Rome and Puteoli not only attest to the presence of Nabataeans in Italy;30 seen in conjunction with the aforementioned Nabataean inscription bearing a transcription of the term *centurio*, <sup>31</sup> they point to the existence of a certain degree of Latin-Nabataean, if not Latin-Old Arabic bilingualism.

All this evidence, however, cannot obscure the fact that the majority of speakers of Latin and of various early forms of Arabic were separated from each other geographically. This geographical distance was overcome only temporarily thanks to the creation of a trans-Mediterranean Roman imperial sphere in the first century bce. It began to increase again when the western and the eastern part of the empire began to drift apart irrevocably, at the latest at the end of the fourth century. Latin still played a certain role in the emerging Byzantine Empire of the fifth and sixth centuries: a number of historiographical texts, legal compendia such as the *Codex Iustinianus*, and Latin military commands in the *Strategikon* of Maurikios attest to its lingering impact.<sup>32</sup> Despite this, the political separation into a western and eastern Roman Empire, both subject to very different

<sup>29</sup> Again going by Fraenkel, *De vocabulis*; Fraenkel, *Fremdwörter*.

<sup>30</sup> Knauf, "Arabo-Aramaic," 230 fn. 104, with reference to CIS II 159, an inscription with one line of Nabataean and three lines of Latin. See George A. Cooke, *A Text-Book of North-Semitic Inscriptions. Moabite, Hebrew, Phoenician, Aramaic, Nabataean, Palmyrene, Jewish* (Oxford: Clarendon Press, 1903), 256–257, with the Nabataean inscription from Puteoli, Italy.

<sup>31</sup> Graf, "Nabataean Army," 289.

<sup>32</sup> Hubert Petersmann, "Vulgärlateinisches aus Byzanz," in *Zum Umgang mit fremden Sprachen in der griechisch-römischen Antike*, ed. Carl Werner Müller et al. (Stuttgart: Franz Steiner Verlag, 1992), 226–227.

experiences between the late fourth and the seventh century, radically diminished the influence of Latin on the societies of the eastern Mediterranean. In this period, Latin was again relegated to the Roman and post-Roman West, extending its influence to the north and northeast, whereas Greek retained its status and gradually affirmed its influence as the language of political power in the Roman-Byzantine East.33

#### 2.2 The creation of a linguistic contact zone in the western Mediterranean

The geographic distance between the majority of Latin and Arabic speakers was again overcome in the seventh and the early eighth century with the Arabic-Islamic expansion into the western Mediterranean. The establishment of Arabic-speaking Muslim elites in North Africa and on the Iberian Peninsula created a linguistic contact zone that, from then on, facilitated the intensive interpenetration of the linguistic systems of Latin and Arabic.

The Arab groups that had been called together under the banner of Islam, either by Muḥammad (d. 10/632) or by his successors Abū Bakr (r. 10‒12/632‒634) and ʿUmar b. al-Ḫaṭṭāb (r. 13‒23/634‒644), had been in touch with different languages before they moved westwards to Egypt, North Africa, and the Iberian Peninsula.34 This is attested frequently in Arabic-Islamic works describing this early period. Arab groups in the border zone of the Byzantine Empire, some of them Christianized, were certainly acquainted with Syriac and Greek.<sup>35</sup> However, Arabs from Mecca and Medina were also in touch with other linguistic groups, either thanks to their exposure to Jewish and Christian scriptural traditions,<sup>36</sup> or because of relations connecting the Ḥiǧāz to Syria and Egypt.37

<sup>33</sup> For further reading, see Fergus Millar, *A Greek Roman Empire: Power and Belief under Theodosius II (408*–*450)* (Berkeley: University of California Press, 2006), 84–116; Scott Fitzgerald Johnson, ed., *Languages and Cultures of Eastern Christianity: Greek* (Abingdon: Routledge, 2016).

<sup>34</sup> Daniel G. König, "Herrschaftsübernahme durch Multilingualismus. Die Sprachen der arabisch-islamischen Expansion nach Westen," *Historische Zeitschrift* 308, no. 3 (2019), 637–674, esp. 642–646.

<sup>35</sup> Hoyland, *Arabia*, 204; Fiema et al., "Provincia Arabia," 395–399.

<sup>36</sup> Sidney H. Griffith, *The Bible in Arabic. The Scriptures of the "People of the Book" in the Language of Islam* (Princeton: Princeton University Press, 2013), 7–53.

<sup>37</sup> Documented, for example, in Ibn ʿAbd al-Ḥakam, *Futūḥ Miṣr wa-aḫbāruhā*, ed. Charles Torrey (Cairo: Madbūlī, 1999), 46–48 (exchange between Muḥammad and the Coptic patriarch resulting in the prophet's marriage with the Coptic woman Māriya), 53 (pre-Islamic visit of ʿAmr b. al-ʿĀṣ to Syria and Egypt).

#### 2.2.1 LATIN IN THE MULTILINGUAL SITUATION OF POST-CONQUEST EGYPT

Coptic and Greek skills may have been lacking occasionally among the Arab conquerors who ventured into Egypt in the 640s. According to Ibn ʿAbd al-Ḥakam (d. 257/871), a group of invading Arabs drew on a Byzantine interpreter whom "they ordered to speak to them in Arabic."38 However, the fact that the earliest Greek-Arabic papyri date from the 640s, i.e. the years immediately after the conquest, proves that linguistic mediators must have been available.<sup>39</sup> Ensuing bilingualism is attested by many bilingual Greek-Arabic<sup>40</sup> and Coptic-Arabic papyri41 of the seventh and the eighth century. While we could assume that, aside from Greek, some form of Aramaic may have played a role in facilitating exchanges between Egypt and the Syrian Levant, Latin seems to have been almost of no relevance in Egypt at the time of the Arab conquest.42

Papyrological evidence shows that Latin had still been studied in some parts of Egypt in Late Antiquity, e.g. in the oasis of Daḫla in western Egypt.43 Monastic literature of the late fourth and early fifth century suggests that some Egyptian monks were capable of receiving and dealing with Latin-speaking pilgrims from the western parts of the Roman Empire.44 Up to the fifth century, Latin seems to have played a certain role in the military as well as in the legal sphere, its influence diminishing continuously, however.45 Even if we believe Joseph Karabacek, who claimed that post-conquest Egypt produced a number of Latin-Arabic coins featuring Latin versions of the Islamic creed (*al-šahāda*), the latter mainly feature Greek characters open to different interpretations and do not indicate a flourishing Latin

<sup>38</sup> Ibn ʿAbd al-Ḥakam, *Futūḥ Miṣr*, ed. Torrey, 78: "fa-amarū Rūmiyyan an yukallimahum bi-l-ʿarabiyya."

<sup>39</sup> Adolf Grohmann, "Aperçu de papyrologie arabe," *Études de papyrologie* 1 (1932), 41–43, plate IX.

<sup>40</sup> Roger S. Bagnall, *Everyday Writing in the Graeco-Roman East* (Berkeley: University of California Press, 2010), 105; Lennart Sundelin, "Introduction," in *Papyrology and the Study of Early Islamic Egypt*, ed. Petra M. Sijpesteijn and Lennart Sundelin (Leiden: Brill, 2004), 7 fn. 20; Petra M. Sijpesteijn, *Shaping a Muslim State: The World of a Mid-Eighth-Century Egyptian Official* (Oxford: Oxford University Press, 2013), 235–236.

<sup>41</sup> Petra M. Sijpesteijn, "Multilingual Archives and Documents in Post-Conquest Egypt," in *The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids*, ed. Arietta Papaconstantinou (Abingdon: Routledge, 2016) 113, 121.

<sup>42</sup> Arietta Papaconstantinou, "Introduction," in *The Multilingual Experience in Egypt*, ed. Papaconstantinou, 4–5; James N. Adams, *Bilingualism and the Latin Language* (Cambridge: Cambridge University Press, 2003), 528.

<sup>43</sup> Sarah Joanne Clackson, "Papyrology and the Utilization of Coptic Sources," in *Papyrology and the Study of Early Islamic Egypt*, ed. Petra M. Sijpesteijn and Lennart Sundelin (Leiden: Brill, 2004), 37.

<sup>44</sup> Sofia Torallas Tovar, "Linguistic Identity in Graeco-Roman Egypt," in *The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids*, ed. Arietta Papaconstantinou (Abingdon: Routledge, 2016), 36–37, 41.

<sup>45</sup> Sijpesteijn, *Shaping a Muslim State*, 34; Papaconstantinou, "Introduction," 4–5.

culture.46 Petra Sijpesteijn, in turn, claims that the Arab conquerors of Egypt, rather than being confronted with the remnants of a partly Latinized administration, actually—and probably unconsciously—reintroduced the originally Latin terms *siǧill* (< *sigillum*, i.e. "seal," "document") and *barīd* (< *veredus*, i.e. "postal service") to the multilingual Egyptian bureaucracy.<sup>47</sup>

A recent find by Dario Internullo, the British Library Papyrus 3124, may still slightly modify our picture of Latin-Arabic entanglement in Egypt in and after the period of the Muslim conquest. The papyrus is unique in that it contains five lines of vulgar Latin, followed by eighteen lines of Arabic text in Latin letters. According to Internullo's paleographical analysis, the document dates from the seventh or eighth century. In the Latin portion of the text, the sender asks about the addressee's health in a rather clumsy and repetitive manner. Given the difficulties of reconstructing an early medieval Arabic text written in Latin letters, the Arabic portion has only been partially deciphered so far by Arianna D'Ottone Rambach.48 It is obvious, however, that this part of the letter deals with business matters. Phrases such as "uctubuli bihabar elbida" could be interpreted as "uktubū lī bi-ḫabar al-biḍāʿa," i.e. "write to me about the merchandise."49 Written on an Egyptian writing material and also found in Egypt, the letter seems to have been produced in that country. In addition, the internal reference to something coming "from Jerusalem" indicates that its sender operated in the eastern Mediterranean.50 The phrasing "min Ierusalem" represents a curiosity in itself in that it uses an Arabic preposition, but not the usual Arabic(-Islamic) term for Jerusalem, i.e. *Īliyāʾ* and *Bayt al-Maqdis* in the seventh and eighth centuries, *al-Quds* or *al-Bayt al-Muqaddas* in later periods.<sup>51</sup>


<sup>46</sup> Joseph Karabacek, "Kritische Beiträge zur lateinisch-arabischen Numismatik," *Numismatische Zeitschrift* 2 (1870), 457–462. I would like to thank Michael Marx for pointing out this article to me. Karabacek's hypothesis was severely attacked by Carl Heinrich Becker, "Das Lateinische in den arabischen Papyrusprotokollen," *Zeitschrift für Assyriologie und verwandte Gebiete* 1 (1886), 166–193, and is not confirmed by more recent research on the numismatic evidence from Egypt, see Michael L. Bates, "Coins and Money in the Arabic Papyri," in *Documents de l'Islam médiéval: nouvelles perspectives de recherche*, ed. Yūsuf Rāġib (Cairo: IFAO, 1991), 43–64.

<sup>47</sup> Frank R. Trombley, "Sawīrus ibn al-Muqaffaʿ and the Christians of Umayyad Egypt. War and Society in Documentary Context," in *Papyrology and the Study of Early Islamic Egypt*, ed. Petra M. Sijpesteijn and Lennart Sundelin (Leiden: Brill, 2004), 219; Sijpesteijn, *Shaping a Muslim State*, 91 fn. 299.

<sup>48</sup> Dario Internullo, "Un *unicum* per la storia della cultura. Su un papiro latinoarabo della British Library (P. Lond. inv. 3124)," *Mélanges de l'École française de Rome—Moyen Âge* 128, no. 2 (2016), 128–122, http://mefrm.revues.org/3233. A full but tentative transcription is published in Arianna D'Ottone Rambach and Dario Internullo, "Arabic in Latin Letters: The Case of the Papyrus British Library 3124," in *Palaeography between East and West. Proceedings of the Seminars on Arabic Palaeography at Sapienza University of Rome* (Pisa: Fabrizio Serra Editore, 2018), 53–72. I would like to thank both authors for allowing me to read a draft version of the paper.

Another interesting but ambiguous feature of the letter is that the sender uses the term "Allāh" six times, three times in the form "insalla," i.e. "in šāʾ Allāh," "if God wills,"52 once in the form "sellimu alla biramati," possibly to be transcribed as "sallimu llāh bi-raḥmati[hi]," i.e. "may God bless through his mercy."53 In spite of these indicators, one cannot be completely sure if this document has to be assigned to a Muslim environment, given that the term "Allāh" was not only used by Muslims, but also by the Meccans of pre-Islamic times.54

As interesting as this unique document may be, it neither shows that Egypt of the early seventh century constituted a primary location for Latin-Arabic encounters, nor does it imply that a form of Latin used in Egypt may have facilitated acts of communication during the ensuing Muslim conquest of North Africa.

#### 2.2.2 THE EXPANDING MUSLIMS AND THE LINGUISTIC CONSTELLATION OF NORTH AFRICA

In pre-Islamic North Africa of the late sixth and early seventh century, we can distinguish three major linguistic groups, i.e. the Greek-speaking Byzantine authorities, the Latin-speaking Romanized African population, and the Berber-speaking autochthonous non-Romanized populations,55 which may have even retained some remnants of the Punic language.56 For the eve of the Arab conquest, i.e. the first half of the seventh century, we may add a sizeable number of Armenians<sup>57</sup> as well as various Middle Eastern groups from Syria, Palestine, and Egypt, fleeing first from the Persian invasion of Egypt, then from the Byzantine religious policies vis-à-vis Judaism and certain interpretations of Christianity, and finally from the Arab invasion of Egypt.58 Thus, in the Romanized milieus of North Africa, the invading Arabs must have encountered a rather large group of people able to speak one of the languages also current in the Middle East.

<sup>52</sup> British Library P. 3124, lines 8, 14, 15, 19.

<sup>53</sup> British Library P. 3124, lines 6–7.

<sup>54</sup> William Montomery Watt, "The 'High God' in pre-Islamic Mecca," in *Actes du Ve congrès international d'Arabisants et d'Islamisants* (Brussels: Publications du Centre pour l'étude des problèmes du monde musulman contemporain, 1971), 449–505; Christian Robin, "Du paganisme au monothéisme," *Revue des mondes musulmans et de la Méditerranée* 61 (1991), 139–155.

<sup>55</sup> Walter E. Kaegi, *Muslim Expansion and Byzantine Collapse in North Africa* (Cambridge: Cambridge University Press, 2010), 66.

<sup>56</sup> Wolfgang Röllig, "Das Punische im Römischen Reich," in *Die Sprachen im Römischen Reich der Kaiserzeit*, ed. Günter Neumann and Jürgen Untermann (Bonn: Habelt, 1980), 298.

<sup>57</sup> Kaegi, *Muslim Expansion*, 102.

<sup>58</sup> Averil Cameron, "The Byzantine Reconquest of North Africa and the Impact of Greek Culture," in *Changing Cultures in Early Byzantium*, ed. Averil Cameron (Aldershot: Ashgate, 1996), art. X, 157; Kaegi, *Muslim Expansion*, 74, 84.

Considering that late antique North Africa had been a flourishing centre of Latin rhetorics and literature,<sup>59</sup> and Latin continued to be used under Byzantine rule,60 one would surmise that Latin was still widely used at the time of the conquest, thus making some kind of Latin-Arabic encounter inevitable. This is certainly valid for the urbanized coastal areas. It is quite difficult, however, to reconstruct how far the Latin language had penetrated into the coastal hinterland, which, in the seventh century, constituted an immense dialect continuum of so-called Berber languages reaching from the Western desert of Egypt to the Atlantic.61 Linguists such as Hugo Schuchardt and Otto Rössler pointed to the large number of Latin or Romance loanwords in modern Berber dialects.62 Maarten Kossmann, however, has recently emphasized the difficulties of distinguishing between "Latin loans" from the Roman period, "African Romance loans" from the post-Roman period, "precolonial non-African Romance loans" dating from the medieval and early modern period, and "colonial and post-colonial Romance loans."63 From the viewpoint of the historian highlighting the intensive interaction between Romans and non-Romans in ancient and late antique North Africa, Yves Modéran argued for a sizable population of Romanized Africans in the cities and the connected countryside who used a vulgar form of Latin as their dominant language on the eve of the Arab conquest.64 Against this backdrop, we can probably assume that (partly) Romanized Berber groups displayed some knowledge of a vulgar form of Latin, a fact not lacking relevance, considering that such Berber groups may have participated in the invasion and settlement of the Iberian Peninsula after 711.

Much of the evidence for Latin-Arabic entanglement in North Africa of the conquest period is difficult to interpret. The supposition that the Arabic term *al-Barbar* derives from the Latinized *barbari* or the Greek βάρβαρoι, says little more than that the Arab conquerors adopted an ethnocultural distinction prevalent among the Romanized elites of North Africa, regardless of their speaking Greek or Latin.65 An Arabic merchant letter, probably written in al-Qayrawān in the seventh century on a fifth-century parchment containing a fragment of the Latinized Book of Exodus, only tells us that the Arab conquerors reused local writing

<sup>59</sup> Leonhardt, *Latein*, 88–89.

<sup>60</sup> Cameron, "Byzantine Reconquest," 153–154, 157–158.

<sup>61</sup> Otto Rössler, "Der semitische Charakter der libyschen Sprache," *Zeitschrift für Assyriologie und Vorderasiatische Archäologie* 50 (1952), 121; Maarten Kossmann, *The Arabic Influence on Northern Berber* (Leiden: Brill, 2013), 16–25, esp. 16–17.

<sup>62</sup> Hugo Schuchardt, *Die romanischen Lehnworte im Berberischen* (Vienna: Alfred Hölder, 1918); Otto Rössler, "Libyen von der Cyrenaica bis zur Mauretania Tingitana," in *Die Sprachen im Römischen Reich der Kaiserzeit*, ed. Günter Neumann and Jürgen Untermann (Bonn: Habelt, 1980), 282.

<sup>63</sup> Kossmann, *Arabic Influence*, 63–64.

<sup>64</sup> Yves Modéran, *Les Maures et l'Afrique romaine (IVe* –*VIIe siècle)* (Rome: École française de Rome, 2003), 685–709.

<sup>65</sup> Moderan, *Les Maures*, 685–709.

materials.66 We do not know, however, if and to what extent the Muslim takeover entailed the systematic destruction of Latin writings. Leo Africanus, a North African Muslim convert to Christianity who wrote an Italian history of Africa at the papal court of the sixteenth century, claims that the Arabs only encountered Latin texts during their conquest of North Africa. This statement is of no help to reconstruct concrete linguistic encounters and seems highly doubtful, considering the role of Greek in seventh-century North Africa. Leo's ensuing claim that the majority of "African books" were burnt with the aim of squelching any ideas that might question Islamic doctrine, when North Africa seceded from the authority of Abbasid Baghdad, smacks of Christian anti-Islamic polemics and finds no corroboration in other evidence.<sup>67</sup> Finally, al-Idrīsī's (d. 560/1165) remarkable description of the twelfth-century inhabitants of Gafsa (in modern-day Tunisia) as "Berberized, the majority of them speaking the African Latin language," strongly suggests that—in spite of the penury of evidence for continuous Latin literacy in Muslim North Africa—pockets of Latin speakers continued to exist long after the Arab conquest.<sup>68</sup> However, al-Idrīsī's statement is too late to enable a reconstruction of Latin-Arabic encounters in the period of conquest. Only numismatic evidence from the late seventh and the early eighth century confirms clearly that the Arab conquerors of North Africa consciously began to engage with the Latin language in a linguistic landscape also marked by Greek. The coins can be divided into four series.69

Series 1 adapted the local Byzantine iconography. It bears neither date, nor the name of the mint, but was probably produced after the final Muslim seizure of Carthage in 79/698–699. As in the North African Byzantine model, the legends are in Latin, but formulate either variations of the

<sup>66</sup> Yūsuf Rāġib, "La plus ancienne lettre arabe de marchand," in *Documents de l'islam médiéval. Nouvelles perspectives de recherche*, ed. Yūsuf Rāġib (Cairo: Institut français d'archéologie orientale, 1991), 1–9.

<sup>67</sup> Johannes Leo, *A Geographical Historie of Africa*, trans. and commented by Iohn Pory [London 1600] (Amsterdam: Da Capo Press, 1969), Book I, 28–29.

<sup>68</sup> Al-Idrīsī, *Nuzhat al-muštāq / Opus geographicum sive Liber ad eorum delectationem qui terras peragrare studeant*, ed. Alesio Bombaci et al., 2 vols. (Cairo: Maktabat al-ṯaqāfa l-dīniyya, 2002), vol. 1, 278: "wa-ahluhā mutabarbarūn wa-akṯaruhum yatakallam bi-l-lisān al-laṭīnī l-ifrīqī."; Tadeusz Lewicki, "Une langue romane oubliée de l'Afrique du Nord. Observations d'un arabisant," *Rocznik Orientalistyczny* 17 (1951–1952), 415–480. Mark A. Handley, "Disputing the End of African Christianity," in: *Vandals, Romans and Berbers. New Perspectives on Late Antique North Africa*, ed. A. H. Merrills (Aldershot: Ashgate, 2004), 302–309; and John Conant, *Staying Roman. Conquest and Identity in Africa and the Mediterranean, 439–700* (Cambridge: Cambridge University Press, 2012), 363, with references to the continuous use of Latin as an epigraphic language.

<sup>69</sup> Trent Jonson, *A Numismatic History of the Early Islamic Precious Metal Coinage of North Africa and the Iberian Peninsula*, 2 vols. (unpublished PhD thesis, University of Oxford, 2014), vol. 1, 32. Abd El Hamid Fenina, "L'arabisation du monnayage d'Ifrīqiya. Étapes et signification," in *Civilisations en transition (II): Sociétés multilingues à travers l'histoire du Proche-Orient*, ed. Jean-Luc Fournet et al. (Byblos: Centre International des Sciences de l'Homme, n.d.), 115–168.

Islamic creed (*al-šahāda*) or monotheistic invocations to God.70 One of the earliest specimens, for example, features the inscription "DeUS TuUS DeUS ET AliUS NON Est," i.e. "God is your God, and there is no other," in a mixture of Latin and Greek letters.71

Series 2 represents a Latin epigraphic type without images that dates the coin according to the current year of the Roman tax-cycle (*indictio*), making it possible to assign these coins to the governorship of Mūsā b. Nuṣayr.<sup>72</sup> It was minted in three phases, from 84–87/703–706, then from 89–92/707–711, and finally, from 95–96/713–715. The coins bear abbreviated monotheist inscriptions in Latin, e.g. "eternal God" (*deus eternus*), "great God" (*deus magnus*), or "God, the creator of all" (*deus omnium creator*) in the first phase. From the second phase onwards, they contain a Latin variation of the *basmala*, e.g. the abbreviated inscription "in the name of the merciful Lord, this *solidus* was made in Africa" (*in nomine dominis misericordis solidus feritus in Africa*), as well as abbreviated Latin versions of the *šahāda*, e.g. "there is no God but one, who does not have a similar associate" (*non est deus nisi unus cui non socius alius similis*), i.e. a rough Latin translation of the Arabic "lā ilāha illā llāh waḥdahu lā šarīk lahu."73

Series 3 represents an Arabic-Latin bilingual type that was struck between 97/715–716 and 99/717–718. Coins now feature the Arabic version of the Islamic creed, an often highly corrupted Latin version of the *šahāda*, as well as a Latin indication of the place and date of minting, exchanging the dating by tax-cycle with a date in *hiǧrī*-years. That empty spaces in the Latin legend are often filled up with meaningless repetitions of groups of letters from the first part of the legend, suggests that the responsible mints reproduced Latin formulae rather than retranslating them.74 Finally, series 4, not relevant for this study, abolished Latin epigraphy and merged with the post-reform coinage minted in the eastern parts of the Umayyad caliphate.

In view of this evidence, Northwest Africa constitutes the earliest area in which a respectable number of direct Latin-Arabic encounters took place after the initial contact of both linguistic systems in the ancient Middle East. The late seventh and early eighth century witnessed intensive Arab-Berber engagement and the Muslim establishment in the former Roman provinces of North Africa. Here, Latin, Greek, Berber, Arabic, and maybe even other linguistic elements mingled in such a way as to bridge the linguistic divide between Arabic-speakers on the one side, and Romanized groups using a late antique or early medieval form of Latin on the other side. While it is

<sup>70</sup> Jonson, *Numismatic History*, vol. 1, 33–34, 90–110.

<sup>71</sup> Michael L. Bates, "Roman and Early Muslim Coinage in North Africa," in *North Africa from Antiquity to Islam*, ed. Mark Horton, Thomas Wiedemann (Bristol: University of Bristol, 1995), 12.

<sup>72</sup> Bates, "Roman and Early Muslim Coinage," 13; Jonson, *Numismatic History*, vol. 1, 85.

<sup>73</sup> Jonson, *Numismatic History*, vol. 1, 34–37, 111–140, 188–193.

<sup>74</sup> Jonson, *Numismatic History*, vol. 1, 37–38, 194–200, esp. 200.

difficult to gauge the extent to which Berber groups had been Romanized in linguistic terms, it seems plausible that a number of Latin or Romance speakers came from their ranks and not only from the more strongly Romanized urban population of the coastal regions. The numismatic evidence proves that Latin and Arabic occasionally became highly entangled in what seem to be the earliest Latin versions of Islamic formulae of faith. Although intermediate languages may have played a role here and elsewhere, it seems possible that some forms of Latin-Arabic bilingualism emerged before Arab and Berber groups crossed the Straits of Gibraltar in around 711. Much speaks in favour of the hypothesis that people from North Africa constituted the mediators who facilitated this crossing in linguistic terms.

#### 2.2.3 LANGUAGES OF THE IBERIAN PENINSULA

Contemporary Latin and later Arabic-Islamic sources describing the Muslim invasion of the Iberian Peninsula affirm that the Berber- and Arabic-speaking groups crossing the straits, fighting their way north, and establishing themselves in the former Visigothic kingdom were assisted by various people and groups. Among those who helped them to enter and find their way around the Iberian Peninsula in and after 711, we find a North African noble (*uir illustris*) either called "Urbanus" or defined as a city-dweller (*urbanus*), as well as the highly stylized count Julian, allegedly Visigothic governor of Ceuta, both of them Christians.<sup>75</sup> North African and Iberian Jews may have also supported the invasion. The Seventeenth Council of Toledo, held in 694, accuses the Jews of the kingdom of collaborating with their brethren overseas, whereas later Arabic-Islamic sources claim that Jewish groups in Granada and Elvira were immediately accorded special treatment by the Muslim invaders.76 Finally, the sources also mention various inhabitants of Visigothic Iberia who supported the invasion, ranging from undefined locals to the sons of the former Visigothic king Witiza (r. 701–710).<sup>77</sup>

<sup>75</sup> E.g. "Continuatio Hispana a. DCCLIV," in *Chronica Minora* vol. 3, ed. Theodor Mommsen, Monumenta Germaniae Historica, Auctores antiquissimi 11 (Berlin: Weidmann, 1894), 322–369, at § 77, 355; cf. Ibn ʿAbd al-Ḥakam, *Futūḥ Miṣr*, ed. Torrey, 206; *Aḫbār maǧmū*ʿ*a*, ed. and trans. Don Emilio Lafuente y Alcántara (Madrid: Rivadeneyra, 1867), 5 (AR), 19–20 (ES), on Ṭāriq b. Ziyād's cooperation with a certain Julian, defined as Visigothic governor of Ceuta.

<sup>76</sup> Concilium Toletanum [Toledo] XVII (694), ed. and trans. José Vives, *Concilios visigóticos e hispano-romanos* (Madrid: Consejo Superior de Investigaciones Científicas, 1963), 524; *Aḫbār maǧmūʿa,* ed. Lafuente y Alcántara, 12, 16 (AR), 25, 29 (ES). On the disputed role of Jews during the invasion, see Daniel König, *Bekehrungsmotive. Untersuchungen zum Christianisierungsprozess im römischen Westreich und seinen romanisch-germanischen Nachfolgern* (Husum: Matthiesen, 2008), 412; Kaegi, *Muslim Expansion*, 84; Norman Roth, "The Jews and the Muslim Invasion of Spain," *Jewish Social Studies* 38, no. 2 (1976), 145.

<sup>77</sup> *Aḫbār maǧmū*ʿ*a*, ed. Lafuente y Alcántara, 7 (AR), 21 (ES); Ibn al-Qūṭiyya, *Tārīḫ iftitāḥ al-Andalus*, ed. Ibrāhīm al-Abyārī (Beirut: Dār al-kitāb al-lubnānī, 1989),

Unfortunately, all of these sources fail to mention in which languages the Berber and Arabic conquerors communicated with the inhabitants of the Iberian Peninsula, be they shepherds or nobles. One wonders in which language the wife or daughter of the former Visigothic king spoke to the early governor ʿAbd al-ʿAzīz b. Mūsā b. Nuṣayr, allegedly convincing him as Latin and Arabic sources relate—to crown himself king of Spain.78 In view of the evidence collected in the preceding section, however, it seems fairly probable that the advancing Muslim troops could draw on linguistic mediators from North Africa. These would have included Latin-speaking North African Romanized city-dwellers such as the aforementioned "Urbanus," who, depending on their respective region of origin, had had the possibility of adjusting to Muslim rule for several decades. In view of the partial Latinization of North Africa's coastal hinterland, we can assume that some of the Berbers fighting on the side of the Muslims spoke a form of African Latin and thus encountered no unsurmountable communication barrier when facing the inhabitants of the Iberian Peninsula. The contemporary Hispano-Latin *Chronica muzarabica*, for example, mentions no communication problems when it reports that the Frankish *dux* Eudo of Aquitaine concluded a marital alliance by giving his daughter to a Berber leader named Munnuz in around 731.<sup>79</sup>

Concrete evidence for early encounters between Latin and Arabic is again numismatic. Muslim coins with Latin inscriptions were produced in 93/711–712, giving rise to the idea that Mūsā b. Nuṣayr had brought one or several mobile mints along from North Africa, which struck several coins along the lines of the aforementioned second series of North African coins produced in the second phase. They bear the mint name "Africa" and are dated to 91–92/710–711.80 A second series of Iberian coins was then produced between 93–95/712–714. It differed from the North African model in that it introduced a star and added the *hiǧrī* date to the indiction date, but contains Latin words and phrases already known from the North African models.81 A third series of bilingual Latin-Arabic dinars was then produced in al-Andalus in the sole year of 98/716–717. Scholarship believes that these bilingual coins were issued in an independent Andalusian mint in Cordoba installed by the governor al-Ḥurr (r. ca. 97–100/716–719).82 Their Latin

<sup>29–32;</sup> Ibn al-Qūṭīya, *The History of Ibn al-Qūṭīya*, trans. David James (Abingdon: Routledge, 2009), 49–51.

<sup>78</sup> "Continuatio Hispana," ed. Mommsen, § 79, 356; Ibn ʿAbd al-Ḥakam, *Futūh*̣ *Mis*̣*r*, ed. Torrey, 211–213; *Aḫbār maǧmū*ʿ*a*, ed. Lafuente y Alcántara, 20 (AR), 31 (ES).

<sup>79</sup> "Continuatio Hispana," ed. Mommsen, § 102, 361; *Gesta episcoporum Autissiodorensium*, ed. Georg Waitz, Monumenta Germaniae Historica, Scriptores in folio 13 (Hanover: Hahn, 1881), 394.

<sup>80</sup> Almudena Ariza Armada, "Los dinares bilingües de al-Andalus y el Magreb," *Revista Numismática Hécate* 3 (2016), 146; Bates, "Roman and Early Muslim Coins," 15; Jonson, *Numismatic History*, vol. 1, 86–87, 142–144.

<sup>81</sup> Jonson, *Numismatic History*, vol. 1, 144–187.

<sup>82</sup> Alejandro García Sanjuán, *La conquista islámica de la península Ibérica y la tergiversación del pasado. Del catastrofismo al negacionismo* (Madrid: Marcial Pons, 2013), 159–168; Armada, "Los dinares bilingües," 153–154.

inscription "this *solidus* was made in Spain in the year 91" (*FERITOS SOLIdus IN SPANia Anno XCI*) is accompanied by the Arabic text "Muḥammad, messenger of God" (*Muḥammad rasūl Allāh*) as well as an Arabic legend bearing a different date, i.e. "this dīnār was coined in al-Andalus in the year 98" (*ḍuriba hāḏā l-dīnār bi-l-Andalus sana ṯamān wa-tisʿīn*).83 In spite of these new elements, the coins do not differ from their North African predecessors enough to force us to believe that the minting of them necessarily involved renewed efforts of translating from Arabic to Latin.

Although we cannot assume that communication between the conquerors and the conquered posed no problems in the immediate wake of the Muslim invasion of the Iberian Peninsula, it seems very plausible that the issue of bridging the linguistic divide between Latin and Arabic speakers was less novel on the Iberian Peninsula than it had been in North Africa. In certain cases, Greek may have been used as an intermediate language. Greek skills are attested for parts of the Jewish population of the Iberian Peninsula, evidence being provided by various inscriptions in and beyond the region, some of which also feature Hebrew and even Aramaic.<sup>84</sup> Greek skills may also have played a role in those southern and southeastern coastal regions that had been under Byzantine rule until the reign of Suinthila (r. 621‒631),85 and which seem to have maintained sporadic contact with Greek-speaking communities until the first half of the eighth century.86 It seems very plausible, however, that forms of vulgar Latin would have been used rather frequently by mediators, initially probably of North African origin, to facilitate interaction between conquerors and the conquered until new generations of bilingual mediators had come into being.

<sup>83</sup> Jonson, *Numismatic History*, vol. 1, 201–206.

<sup>84</sup> David Noy, *Jewish Inscriptions of Western Europe, vol. I: Italy (excluding the city of Rome), Spain and Gaul* (Cambridge: Cambridge University Press, 1993), 310–311. Norman Roth, *Jews, Visigoths, and Muslims in Medieval Spain: Cooperation and Conflict* (Leiden: Brill, 1994), 25.

<sup>85</sup> Kaegi, *Muslim Expansion*, 261; Jamie Wood, "Defending Byzantine Spain: Frontiers and Diplomacy," *Early Medieval Europe* 18, no. 3 (2010), 292–319.

<sup>86</sup> The "Continuatio hispana," ed. Mommsen, § 74–75, 354, claims that the Visigothic noble Theodemir defeated Greek sailors and communicated with eastern Christians before he died in 744. The Hispano-Latin *Chronica Byzantia-Arabica* or *Chronicle of 741* contains information on the Muslims' Middle Eastern origins that allegedly derive from a Greek source of information, see Theodor Nöldeke, "Epimetrum," in *Chronica Minora* vol. 3, ed. Theodor Mommsen, Monumenta Germaniae Historica, Auctores antiquissimi 11 (Berlin: Weidmann, 1894), 368–369.

#### 2.3 Exchange and hybridization in the linguistic contact zone

The Arabic-Islamic expansion to the west thus created a linguistic contact zone that would, from then on, continue to connect Latin and Arabic, understood here—this should be emphasized again—as linguistic systems comprising a large variety of written and oral registers including derivate languages and dialects. All Latin-Christian territories affected directly by the Arabic-Islamic expansion—i.e. North Africa in the seventh, the Iberian Peninsula and southern France in the eighth, the Mezzogiorno in the ninth century—were not (yet) subjected to the Carolingian reform of Latin liturgy that, according to Roger Wright, created a clear and conscious distinction between written and spoken Latin on the one side, and spoken Romance on the other, in all territories under Carolingian control from around the late eighth century onwards.87 It is necessary, however, to distinguish between the different parts of the linguistic contact zone: North Africa, Sicily with the southern parts of mainland Italy, southern France, and the Iberian Peninsula differ clearly with regard to the documentation and intensity of Latin-Arabic entanglement.

#### 2.3.1 THE SUPPOSED DEMISE OF LATIN AND LATIN-ARABIC ENTANGLEMENT IN NORTH AFRICA

As we have seen in section 2.2.2, North Africa of the seventh century must have still featured a sizable population capable of speaking and writing a form of Latin.<sup>88</sup> However, much evidence speaks in favour of a certain destruction and then demise of autochthonous Latin culture in North Africa in the period following the Muslim conquest. Neither the fifth-century Latin parchment re-employed as an Arabic business letter in al-Qayrawān in the seventh century, nor Leo the African's claim that religious fanaticism led to the burning of non-Islamic books provide sufficient evidence for this process.<sup>89</sup> Decline, rather, is suggested by the fact that there exists no definite proof for the continued production of Latin or Afro-Romance texts in North Africa up to the high Middle Ages. Al-Idrīsī (d. 560/1165) claimed that the Berber population of Gafsa still spoke an African form of Latin in the twelfth century.90 However, apart from this, there exists only scant concrete evidence that a productive Latinized culture continued to exist. Among this evidence we find a limited number of

<sup>87</sup> Roger Wright, *Late Latin and Early Romance in Spain and Carolingian France* (Liverpool: Liverpool University Press, 1982), 261.

<sup>88</sup> Cameron, "Byzantine Reconquest," 157–158.

<sup>89</sup> Rāġib, "La plus ancienne lettre," 1–9; Leo, *A Geographical Historie of Africa*, trans. Pory, Book I, 28–29.

<sup>90</sup> Al-Idrīsī, *Nuzhat al-muštāq / Opus geographicum*, ed. Bombaci et al., 278.

inscriptions.91 Moreover, Dominique Valérian lists several Latin sources, all of them written north of the Mediterranean, which mention African Christians and even African bishops for the period between the eighth and the early thirteenth century. Their relations with Rome imply that they shared some language of communication with the Roman church, presumably Latin. Some of the references, especially the earlier ones, imply that we are really dealing with autochthonous African Christians, and not with Latin-Christian merchants from the northern Mediterranean. The fact remains, however, that we possess no substantial Latin document produced in North Africa in the early medieval period.92 Consequently, scholarship tends to support the theory of the demise of a written Latin culture in North Africa, without denying the existence of individual pockets of, so to speak, mainly oral "linguistic resilience." <sup>93</sup> References to early medieval acts of communication between Christians from Salerno and Amalfi on the one hand, and North African Muslims on the other hand, imply that ongoing exchange in the western Mediterranean contributed to the mingling of Latin and Arabic in North Africa in the early medieval period.94 In addition, there is agreement on the matter that Latin-Christian culture would be revived in North Africa from around the late twelfth century onwards, when large numbers of expatriate Latin Christians from the northern shores of the Mediterranean sought a living in and around North African *fondachi*, i.e. legally protected urban enclaves of European-Christian merchants.95

#### 2.3.2 ORAL ENTANGLEMENT IN SICILY, SOUTHERN ITALY, AND SOUTHERN FRANCE

Given the lack of sources, reconstructing Latin-Arabic entanglement in early medieval Sicily, southern Italy, and southern France—three regions that only witnessed temporary Muslim rule—is fraught with difficulties.

<sup>91</sup> Handley, "Disputing the End," 302–309; and Conant, *Staying Roman*, 363, with references.

<sup>92</sup> Dominique Valérian, "La permanence du christianisme au Maghreb: l'apport problématique des sources latines," in *Islamisation et arabisation de l'Occident musulman médiéval (VIIe* –*XIIe siècle)*, ed. Dominique Valérian et al. (Paris: PUPS, 2011), 136–138.

<sup>93</sup> Lewicki, "Une langue romane oubliée," 415–480; Serge Lancel, "La fin et la survie de la latinité en Afrique du Nord. État des questions," *Revue des études latines* 59 (1981), 269–297; Christian Schmitt, "Die verlorene Romanität in Afrika: Afrolatein / Afroromanisch," in *Romanische Sprachgeschichte, 1. Teilband*, ed. Gerhard Ernst et al. (Berlin: de Gruyter, 2003), 668–675; Conant, *Staying Roman*, 363.

<sup>94</sup> The tenth-century *Chronicon Salernitanum. A Critical Edition with Studies on Literary and Historical Sources and on Language*, ed. Ulla Westerbergh (Stockholm: Almqvist and Wiksell, 1956), § 110–111, 122–123, for example, quotes conversations between a North African Muslim (*Agarenus*) and the duke Guaifar (r. 861–880) in Salerno, as well as between the same Muslim and Amalfitan merchants in an unnamed place in North Africa.

<sup>95</sup> See section 2.4.3.

As opposed to North Africa, these regions retained a strong Latin- or Romance-speaking population throughout the period of Muslim dominance. We must thus assume that speakers of Arabic and Latin/Romance found some kind of *modus vivendi*, but can mainly infer this from contemporary narrative sources, which refer to acts of communication without mentioning the language(s) employed.

When Sicily gradually came under Muslim rule in the course of the ninth century, the Muslim conquerors of Berber and Arab stock encountered groups of Greek- and Latin-speakers. In his linguistic study of Siculo-Arabic, Dionisius Agius supposes that many of the island's inhabitants were already bi- or multilingual before the Muslim takeover, but assigns a preponderant position to Latin. To him, "it seems clear that Latin or some early Latin type was essentially the spoken and official language of the island, if not on the coast line [. . .]." Muslim Sicily produced no sources that allow us to reconstruct specific milieus of Latin/Romance-Arabic entanglement. However, the reports of foreign visitors such as Ibn Ḥawqal (d. after 378/988), who harshly criticized the practice of rearing Christian girls and Muslim boys among Muslim-Christian couples of the Sicilian countryside, describe an atmosphere of intensive social interaction that must have had linguistic implications.96 Agius confirms that Arabic had a lasting impact on the island's linguistic landscape, also to the detriment of Berber.97 Nevertheless, he assumes that, as soon as relations in the field of artisanship, trade, and administration had been established and a certain degree of intermarriage had taken place, "a portion of the Muslim colonizers were Arabic and Romance speakers."<sup>98</sup>

The Italian mainland, in contrast, never witnessed the same degree and intensity of Muslim settlement as Sicily. However, contemporary Greek, Latin, and Arabic sources attest to intensive communication between Christian and Muslim inhabitants of the region. Arabic-Islamic authors such as al-Masʿūdī (d. 345/956) confirm that Muslims from al-Andalus temporarily took control of and lived in cities such as Bari (*Bārī*), Tarent (*Ṭārniyyū*), and Salerno (*Šabarāma*), until they were reconquered by what he defines as Langobards (*al*-*Nawkubard*).<sup>99</sup> In his travel account, the monk Bernard reports how he sought out the emir of Bari around 867 to demand travel documents to Egypt and the Holy Land.100 Referring to the same emir, Constantine VII Porphyrogennetos (r. ca*.* 913–959) even claims "that the nobles of Capua and Benevent used to go to the sultan [of Bari] and ask

<sup>96</sup> Ibn Ḥawqal, *Kitāb Ṣūrat al-arḍ*, ed. J. H. Kramers (Leiden: Brill, 1938), 129.

<sup>97</sup> On the role of Berber languages in Sicily, see Alex Metcalfe, *Muslims and Christians in Norman Sicily: Arabic Speakers and the End of Islam* (Abingdon: Routledge, 2003), 67.

<sup>98</sup> Dionisius A. Agius, *Siculo Arabic* (Abdingdon: Routledge, 1996), 96–98.

<sup>99</sup> Al-Masʿūdī, *Murūǧ al-ḏahab wa-maʿādin al-ǧawhar*, ed. Charles Barbier de Meynard, Abel Pavet de Courteille, and Charles Pellat, 7 vols. (Beirut: al-Ǧāmiʿa l-lubnāniyya, 1965–1979), § 921, 151.

<sup>100</sup> Bernardus monachus, "Itinerarium," in *Patrologiae cursus completus*, ed. Jean-Paul Migne, Patrologia Latina 121 (Paris: Migne, 1852), cap. 2, 569.

him questions about the treatment and care of cattle and other matters because of his age and experience."101 In several letters, Pope John VIII criticizes the bishop of Naples and Christian rulers in southern Italy for cooperating with the "Saracens," only to negotiate the payment of tribute with Saracen raiders around 878.102 The tenth-century *Chronicon Salernitanum* mentions Muslims spending time in various cities of southern Italy such as Salerno and Naples.103 In one case, the Latin author quotes a Muslim (*Agarenus*) swearing "by the son of Maria whom you venerate as God," clearly a Latin translation of the Arabic-Islamic epithet "Ibn Maryam" that emphasizes Jesus's human nature.104 According to a twelfth- or thirteenthcentury medical treatise from Salerno, Constantine the African decided to translate medical books from Arabic to Latin and to settle in Italy when, during a commercial visit to Rome around the middle of the eleventh century, Saracen slaves translated the diagnosis of a local physician to him.<sup>105</sup> Even if we do not believe such anecdotes in detail, they do convey an impression of the many and highly varied instances of communication in early medieval southern Italy between Christians and Muslims, some of whom must have been bilingual and proficient in the language usually not associated with their respective religious allegiance.

With regard to southern France, the extant sources provide much fewer details. It is not likely, however, that Muslim troops held Narbonne for almost four decades, i.e. approximately between 719–720 and 759, without communicating with the local population at least occasionally.106 At the end of the ninth century, Muslims established the so-called raider colony of Fraxinetum and subjected the surrounding countryside to frequent raids, until the colony was destroyed at the end of the tenth century. Latin sources credit the Saracens of Fraxinetum with speaking to their

<sup>101</sup> Constantinus Porphyrogenitus, *De administrando imperio*, new rev. ed., ed. Gyula Moravcsik, trans. R. H. J. Jenkins (Washington, DC: Dumbarton Oaks Center for Byzantine Studies, 1967), cap. 29, 131. See a variation of related anecdotes in Symeon magister, "Annales," ed. Immanuel Bekker, in *Theophanes continuatus: Ioannes Cameniata, Symeon magister, Georgius monachus*, ed. Immanuel Bekker (Bonn: Weber, 1838), cap. 20, 694–697.

<sup>102</sup> Iohannis VIII papa, "ep. 273 (March 14, 881)," in *Epistolae Carolini aevi* vol. 5, ed. Erich Caspar, Monumenta Germaniae Historica, Epistolae in Quart 7 (Berlin: Weidmann, 1928), 241.

<sup>103</sup> See e.g. *Chronicon Salernitanum*, ed. Westerbergh, § 99, 100.

<sup>104</sup> *Chronicon Salernitanum*, ed. Westerbergh, § 110, 122: "Idem ipse Agarenus: Per filium Marie te obtestor, quem ut Deum colitis, ut fideliter meis dictis illius innuas [. . .]."

<sup>105</sup> See the "glosula a magistro Mathaeo Ferrario (Salerno, ca. 1160) super diaetas universales Isaaci composita" (MS 13. Jh. südital. Hand), ed. Rudolf Kreutz, "Die Ehrenrettung Konstantins von Afrika," *Studien und Mitteilungen zur Geschichte des Benediktinerordens* 49, no. 18 (1931), 40–41. See also Thomas Ricklin, "Der Fall Gougenheim," *Historische Zeitschrift* 290 (2010), 128.

<sup>106</sup> Philippe Sénac, *Les Carolingiens et al-Andalus* (Paris: Maisonneuve et Larose, 2002), 16–17, 37–40.

kidnapped prey,107 claim that they intermarried with local women,<sup>108</sup> and accuse them of sheltering political dissidents and of collaborating with king Hugh of Italy.<sup>109</sup> Here again, we must assume, but cannot prove that Latin-/ Romance-speaking and Arabic-speaking groups mingled to such a degree that certain oral forms of Latin-Arabic entanglement were inevitable.

#### 2.3.3 THE EFFECTS OF PARALLEL LINGUISTIC ARABIZATION AND ROMANIZATION ON THE IBERIAN PENINSULA

In comparison to North Africa, Sicily, southern Italy, and southern France, the Iberian Peninsula provides the best documentation of early medieval Latin-Arabic entanglement by far.<sup>110</sup> As has been shown in section 2.2.3, the Arab-Berber invasion of the Iberian Peninsula around 711 involved various acts of communication between North Africans and Iberian locals. Sources of the eighth and early ninth centuries do not contain explicit references to linguistic developments, but allow us to reconstruct an ever-increasing number of acts of communication that accompanied the processes of transculturation involving the conquerors and the conquered. Contemporary Latin as well as later Arabic sources, for example, provide a detailed account of the administrative measures taken by the early Arab governors between 713 and 756 and the first Umayyad emir ʿAbd al-Raḥmān I (r. 138‒172/756‒788) to ensure that fiscal revenues were extracted from the Romance-speaking Christian population.111 Marital alliances between Arab conquerors and Romance-speaking Christian women are attested from the early eighth century onwards, thus implying that the initial separation between conquerors and conquered was slowly thawing.112 A letter written by Pope Hadrian I to the bishops of Spain between 785 and 791 complains about many Christians who interacted with Jews and non-baptized

110 For an extensive analysis, see Cyrille Aillet, *Les Mozarabes. Christianisme, islamisation et arabisation en péninsule Ibérique (IXe* –*XIIe siècle)* (Madrid: CSIC, 2010), 131–244, under the title "Latinité et arabisation."

<sup>107</sup> Rodulfus Glaber, *Historiarum libri quinque* / *The Five Books of the Histories*, ed. and trans. John France (Oxford: Oxford University Press, 1989), lib. I, cap. iv, § 9, 19–23.

<sup>108</sup> Ekkehardus IV, *Casuum Sancti Galli continuatio*, ed. D. Ildephonsus ab Arx, Monumenta Germaniae Historica, Scriptores in folio 2 (Hanover: Hahn, 1829), 110.

<sup>109</sup> Liutprandus Cremonensis, *Antapodosis*, ed. Joseph Becker, Monumenta Germaniae Historica, Scriptores rerum Germanicarum in usum scholarum 41 (Hanover: Hahn, 1915), lib. V, cap. 16–17, 139; Liutprandus Cremonensis, *Liber de rebus gestis Ottonis magni imperatoris*, ed. Joseph Becker, Monumenta Germaniae Historica, Scriptores rerum Germanicarum in usum scholarum 41 (Hanover: Hahn, 1915), cap. 4, 160–161.

<sup>111</sup> See Daniel G. König, "Charlemagne's 'Jihād' Revisited. Debating the Islamic Contribution to an Epochal Change in the History of Christianization," *Medieval Worlds* 3 (2016), 13–14. See also Pedro Chalmeta Gendrón, *Invasión e islamización. La sumisión de Hispania y la formación de al-Andalus* (Jaén: Universidad de Jaén, 2003), 121, 206–209.

<sup>112</sup> König, *Arabic-Islamic Views*, 48–49.

heathens and even chose to give their daughter(s) in marriage "to the infidels" (*cum infidelibus*), the latter "thus being delivered to the pagan people" (*sic populo gentili traderetur*).113

Approximately from the middle of the ninth century onwards, the sources begin to provide information on linguistic issues. The phenomenon of the so-called "martyrs of Cordoba," a comparatively large number of Christians executed by the Muslim authorities either for denigrating Islam or for apostasy from Islam in the 850s, sheds light not only on the religious, but also on the linguistic effects of intermarriage. The highly stylized lives of the martyrs, written by extremely biased members of the martyr movement, mention young people, such as a certain Aurelius, who were obviously bilingual. Son of a Christian mother and a "pagan father," he was educated in what the hagiographer Eulogius (d. 859) calls "Arabica litteratura." Aurelius was strongly influenced by his Christian aunt and eventually joined the martyr movement, where he intensively engaged with the biblical scriptures, presumably in Latin, under the tutelage of a certain Albarus of Cordoba (d. ca. 861).<sup>114</sup>

Conveniently compiled in the *Corpus Scriptorum Muzarabicorum*, the bulk of Latin literature produced on the Iberian Peninsula under Muslim rule is clearly confined to texts written in the eighth and ninth centuries.<sup>115</sup> Since the middle of the ninth century, various sources suggest that Arabic was beginning to supplant Latin/Romance in those parts of the Iberian Peninsula under Muslim control. Engaging in fierce polemic against Islam, Albarus of Cordoba harshly criticized the Christian youth of his time for abandoning the Latin literature of the Church fathers (*uolumina* [. . .] *doctorum Latine*), for not regarding themselves as Latin-speakers anymore (*linguam propriam non aduertunt Latini*), and for not being able to write a respectable letter (*litteras*). Instead, the Christian youth collectively received recognition for its Arabic skills (*harabico eloquio sublimati*), boasted of its knowledge of what he polemically defines as "Chaldean words" (*Caldaicas uerborum*), and engaged enthusiastically in what he calls "pagan erudition" (*gentilici[a] eruditjoni*), "the works of the Chaldeans" (*uolumina Caldeorum*), and the complex rhythms of "the poetry of these peoples" (*ab ipsis gentibus carmine*).<sup>116</sup> In approximately the same social environment and period, i.e.

<sup>113</sup> *Codex Carolinus* (ep. 95: Hadrianus papa ad episcopos Hispaniae), ed. Wilhelm Gundlach, Monumenta Germaniae Historica, Epistolae 3, Epistolae Merowingici et Karolini Aevi 1 (Berlin: Weidmann, 1892), 643.

<sup>114</sup> Eulogius Cordubensis, *Memoriale Sanctorum*, ed. Ioannes Gil, Corpus Scriptorum Muzarabicorum 2 (Madrid: CSIC, 1973), lib. II, cap. X, 1, 416; lib. II, cap. X, 18, 423. On the connection between the movement of the so-called "Martyrs of Cordoba" and a revival of Latin erudition, see Roger Wright, "The End of Written Ladino," in *The Formation of al-Andalus, Part 2: Language, Religion, Culture and the Sciences*, ed. Maribel Fierro and Julio Samsó (Farnham: Ashgate, 1998), 25–28.

<sup>115</sup> David Wasserstein, "The Language Situation in al-Andalus," in *Studies on the Muwaššaḥ and the Kharja*, ed. Alan Jones and Richard Hitchcock (Reading: Ithaca, 1991), 7.

<sup>116</sup> Albarus Cordubensis, *Indiculus luminosus*, ed. Ioannes Gil, Corpus Scriptorum Muzarabicorum 1 (Madrid: CSIC, 1973), cap. 35, 43–64, 314–315.

the urban Iberian milieu under Muslim rule around the end of the ninth century, the Psalter was translated from Latin into Arabic. The translator, a certain Ḥafṣ b. Albar, explains in the prologue to his translation that he regarded this task, supported by some and criticized by others, as necessary. Criticizing the shortcomings of an earlier translation, he takes great pains to explain why he used the Arabic metre *raǧaz* to translate the Latin *iambus* and describes his problems with translating certain terms and expressions.117

The evidence presented so far implies that, by the late ninth century, Arabic had successfully supplanted a hitherto flourishing Latin-Romance linguistic landscape in those parts of the Iberian Peninsula under Muslim rule, even in the sphere of Christian liturgy.118 However, this process of linguistic Arabization was complemented by a parallel process of linguistic Romanization. The unequal distribution of native Arabic-speakers in al-Andalus,119 outweighed numerically by Berbers and local Romance-speakers, did not lead to an establishment of Berber languages, but certainly guaranteed the survival of some forms of Latin or Romance, not only in the lower echelons of society. Al-Ḫušanī's (d. 371/981) *History of the Judges of al-Andalus* contains references to Muslim judges understanding and speaking "the non-Arabic language" (*al-ʿaǧamiyya*).120 According to the Middle Eastern geographer al-Muqaddasī (d. after 380/990), two languages were spoken in al-Andalus—a form of Arabic as well as "another language similar to the Roman" (*wa-lahum lisān āḫar yuqārib al-rūmī*).121 One may even gain the impression that, by the eleventh century, most Arab and Berber settlers on the Iberian Peninsula had learned to express themselves in the local Romance idiom. This would explain why Ibn Ḥazm (d. 456/1064) regarded it as noteworthy that neither the women nor the men of the Arab group of Balī north of Cordoba were able to speak Latin, but only Arabic (*lā yuḥsinūn al-kalām bi-l-laṭīniyya lākin bi-l-ʿarabiyya faqaṭ*).<sup>122</sup>

In the scholarly discussion about Latin/Romance-Arabic bilingualism on the Iberian Peninsula, a specific form of Andalusī literature known as *muwaššaḥ*-poetry constitutes an additional important corpus of sources.

<sup>117</sup> Ḥafṣ bin Albar, "Urjūza," ed. and trans. Marie-Thérèse Urvoy, v. 28, 47–48, 50, 75, 98, 110, in Marie-Thérèse Urvoy, *Le Psautier mozarabe de Ḥafṣ le Goth* (Toulouse: Presses universitaires, 1994), 15–19. See the English translation in Douglas M. Dunlop, "Ḥafṣ b. Albar: The Last of the Goths?," *Journal of the Royal Asiatic Society of Great Britain and Ireland* 3, no. 4 (1954), 140–144.

<sup>118</sup> Ángeles Vicente, *El proceso de arabización de Alandalús: Un caso medieval de interacción de lenguas* (Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, 2007), 45–82.

<sup>119</sup> See the settlement patterns reconstructed by Chalmeta, *Invasión e islamización*, 159.

<sup>120</sup> Al-Ḫušanī, *Tārīḫ quḍāt al-Andalus / Historia de los jueces de Córdoba por Aljoxani*, ed. and trans. Julián Ribera (Madrid: Imprenta Ibérica, 1914), 111–112, 139 (AR).

<sup>121</sup> Al-Muqaddasī, *Aḥsan al-taqāsīm fī maʿrifat al-aqālīm*, ed. Michael de Goeje (Leiden: Brill, 1877), 243.

<sup>122</sup> Ibn Ḥazm, *Ǧamharat ansāb al-ʿArab*, ed. ʿAbd al-Salām Muḥammad Hārūn, 2 vols., fifth ed. (Cairo: Dār al-maʿārif, n.d.), vol. 2, 443. See also Wasserstein, "Language Situation," 9.

Approximately between the tenth and the fourteenth century, this genre was used to express love, desire, and praise. Around three hundred and seventy specimens are known to have been written in Arabic; around seventy additional poems feature a final stanza called *ḫarǧa* which is written in Arabic letters, but either in Romance or a garbled mixture of Arabic and Romance.123 This corpus, however, raises several problems of interpretation. In twentieth-century Spain, the origins of this poetical bilingualism were fiercely debated in a general discourse revolving around the effects of the Muslim invasion of 711 on Spanish national identity.124 In this context, it made a great difference if the so-called Romance *ḫarǧa*s represented a kind of "autochthonous" native lyric poetry, a manifestation of "Spanishness" (*hispanidad*) that had survived the establishment of Muslim elites,125 or rather an expression of "domestic loyalty" and "folkloric interest" on the part of Muslim and Jewish poets.126 Obscuring further research on the social context and significance of the Romance *ḫarǧa*, the debate on origins was not able to explain why the fifth stanza containing the Romance *ḫarǧa* is generally attributed to a female using "the non-Arab language" (*mantiqin ʿaǧamiyyi*),127 who responds to the four preceding stanzas sung by a male in Arabic. At first sight, the relationship between Arabic and Romance characteristic of this genre seems to express a power-asymmetry between Muslim males and Christian females, especially considering that enslaved non-Muslim singing-girls formed an integral part of Arabic-Islamic courtly culture in and beyond al-Andalus.128 This interpretation cannot go uncontested: in fact, *muwaššaḥ*-poetry featuring Romance *ḫarǧa*s came into being in the tenth century, the heyday of Umayyad power on the Iberian Peninsula. However, the greater number of poems stem from a period when Muslim rule was increasingly threatened by the takeover of more and more Muslim territories by Romance-speaking Christians. Moreover, the same linguistic and gender characteristics that seem to imply clear hierarchies between Islam and Christianity, also apply to *muwaššaḥ*-poetry written by Jews. This should caution us against exaggerating the cultural and ideological implications of this poetic genre. The latter's existence

<sup>123</sup> Numbers taken from Consuelo López-Morillas, "Review of 'Poesía dialectal árabe y romance en Alandalús' by F. Corriente," *Journal of Arabic Literature* 29, no. 3–4 (1998), 218–221.

<sup>124</sup> On this discourse, see García Sanjuán, *La conquista islámica*.

<sup>125</sup> Emilio García Gómez, *Las jarchas romances de la serie árabe en su marco* (Madrid: Sociedad de Estudios y Publicaciones, 1990), 19–75; James T. Monroe, *Hispano-Arabic Poetry: A Student Anthology* (Piscataway: Gorgias Press, 2004), 1–71.

<sup>126</sup> Federico Corriente, *Poesía dialectal árabe y romance en Alandalús* (Madrid: Gredos, 1997), 14; cf. Otto Zwartjes, *Love Songs from al-Andalus. History, Structure and Meaning of the Kharja* (Leiden: Brill, 1997).

<sup>127</sup> García Gómez, *Las jarchas romances*, muwaššaḥa XXXVIII, 401.

<sup>128</sup> See Otto Zwartjes, "Thematical Correspondences between the Romance and Hispano-Arabic *xarja*-s," in *Proceedings of the 17th Congress of UEAI*, ed. The Society of Friends of the St. Petersburg Branch of the Institute of Oriental Studies (Saint Petersburg: Thesa, 1997), 295–315, 299–306; Joseph Fees, *New Directions in Kharja Studies. Gender, Sexuality and Religion* (unpublished PhD thesis: University of Texas, 2013), 153, 196–197.

proves, in any case, that Latin and Arabic—understood again as linguistic macro-systems—mingled creatively in the linguistic and literary landscape of Muslim al-Andalus. The latter certainly reflected prevailing socio-cultural hierarchies. Seen in a wider context, however, the relationship between literary landscape and religious hierarchies cannot be framed in terms of clear-cut socio-cultural and socio-linguistic dichotomies.

Scholarship has proposed different hypotheses with regard to the balance between Latin/Romance and Arabic in al-Andalus. Some scholars believe that an Iberian form of Romance remained the primary language of communication throughout the period of Muslim rule, with Arabic in the position of a superficial superstratum.129 Others opt for Romance-Arabic bilingualism, whereas a third group proposes that a Romance-influenced form of Andalusian Arabic slowly but surely replaced Romance as the majority language of communication.130 Considering the uneven distribution of Arabic-, Berber-, and Latin- or Romance-speakers in the different parts of the Iberian Peninsula in the post-invasion period, differing linguistic conditions in urban and rural settings, and an ever-changing geopolitical situation that clearly affected processes of linguistic Arabization or Romanization, one has to account for great local and regional differences as well as an evolution of the linguistic landscape. In linguistic terms, the early expansionist period of Arab-Berber settlement must have differed considerably from the period of establishment in the Umayyad emirate and caliphate, or the time when the growing impetus of the so-called *Reconquista* subjected great numbers of Muslims to Christian Romance-speakers entering and settling territories hitherto under Muslim rule.

It is undeniable, in any case, that the regional Andalusian form of Arabic was considerably influenced by Romance elements. This is already attested by the Middle Eastern geographer al-Muqaddasī, who described Andalusī Arabic as "difficult to comprehend" (*munġaliqa*) and "different from what we have mentioned concerning [other] regions" (*muḫālifa limā ḏakarnā fī l-aqālīm*).131 In the fourteenth century, Ibn Ḫaldūn (d. 808/1406) explained that, in al-Andalus, interaction between Arabs, Galicians, and Franks had led to the emergence of a "mixed language. The non-Arab element in it was dominant and it had diverged considerably from the original language," which he defined as "the language of the Muḍar," a tribal group from the Arabian Peninsula associated with the prophet Muḥammad.132 According to Federico Corriente, such evidence proves that "Andalusī Arabic soon became the main linguistic link between all the

<sup>129</sup> Agius, *Siculo-Arabic*, 97: "Proportionately their number [of Romance speakers among the colonizing Muslims] is smaller than the Andalusi Muslims who were primary speakers of a Romance dialect."

<sup>130</sup> On these positions, see Otto Zwartjes, "Al-Andalus," in *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh, vol. 1 (Leiden: Brill, 2008), 96–101.

<sup>131</sup> Al-Muqaddasī, *Aḥsan al-taqāsīm*, ed. de Goeje, 243.

<sup>132</sup> Ibn Ḫaldūn, *Tārīḫ*, ed. Suhayl Zakkār and Ḫalīl Šaḥāda, 8 vols (Beirut: Dār al-fikr, 2000–2001), vol. 1, 771: "wa-ṣārat luġatan uḫrā mumtaziǧatan. wa-l-ʿuǧmatu fīhā aġlaba limā ḏakarnāhu fa-hiya ʿan al-lisāni l-awwali abʿadu."

inhabitants of al-Andalus, if due allowance is made for the initial stages of that historical entity and for remote areas where Romance monolingualism might have lasted longer."<sup>133</sup> Whereas the logic of contact linguistics permits the formulation of plausible theories about how Andalusī Arabic came into being,134 the constellation of sources makes it extremely difficult to reconstruct this particular dialect. Andalusī Arabic was rarely written down by native speakers and generally documented rather late. Some terms feature in Arabic-Islamic treatises on botany recording local plant names, the aforementioned Arabic-Romance *ḫarǧa*s, and in bilingual Latin-Arabic treaties of the late medieval period. Often, however, Andalusī Arabic is documented in the latest phase of the Reconquista by Christians intent on conveying oral Arabic skills to facilitate the conversion of Muslims to Christianity.135 Consequently, it is very difficult to trace if, why, when, and to which extent a Latin or Romance word became part of this regional dialect.<sup>136</sup>

Vice versa, it is also very difficult to trace the early medieval impact of Arabic on the emerging Romance languages of the Iberian Peninsula. Again, the problem lies in the lack of documentation, which, in the case of the many Iberian variants of Romance, is only available in respectable quantities from the thirteenth century onwards.<sup>137</sup> Linguistic scholarship claims that Romance languages borrowed more nouns than other wordtypes from Arabic, that—in spite of centuries of daily interaction between

<sup>133</sup> Federico Corriente, "Andalusi-Arabic," in *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh, vol. 1 (Leiden: Brill, 2008), 102.

<sup>134</sup> Kees Versteegh, "The Origin of the Romance Languages and the Arabic Dialects," in *Islão e arabismo na península ibérica*, ed. Adel Sidarus (Évora: Universidade de Évora, 1986), 344.

<sup>135</sup> See the list of sources listed in Federico Corriente, *Dictionary of Andalusi Arabic*  (Leiden: Brill, 1997), xiii-xvii. Also see Chapter 3.1.

<sup>136</sup> See Corriente, *Dictionary of Andalusi Arabic*, 2, 42, 44. Although rich in material, it is difficult to gauge the relevance of many entries in this dictionary for the development of Andalusī Arabic. A term such as {ʾBRKN}, said to mean "musket" and to be a late derivation from Castilian "robadoquín," the latter deriving from French "ribaudequin," obviously only entered the Arabic language in the late medieval period in connection with the introduction of firearms. The term {ʾPŠṬLY}, with the meaning "formal letter," obviously derives from the Latin "epistola." Given that Arabic possesses an equivalent term, i.e. "risāla," one wonders if this Andalusī Arabic term was ever used outside a Christian or missionary milieu, dealing, for example, with the Pauline epistles. Given the precedence of Arabic over Latin in medicine before the modern period, it is doubtful that Andalusī Arabic would have used a Latin-derived term for "navel," i.e. {ʾMLQ}, unless it was in a Christian context, e.g. when describing Jerusalem as "navel of the world" (*umbilicus mundi*). Other terms, such as {BRBR}, meaning "to speak Berber" cannot really be classified as characteristic of Andalusī Arabic. Speaking Berber was described in the same terms in North Africa. The term {BRD}, i.e. "barīd," for postal service was already part of the Arabic lexicon before the expansion to the west, as has been shown in section 2.1.2.

<sup>137</sup> See the scarcity of early medieval primary texts as documented in Ramón Menéndez Pidal, *Documentos lingüísticos de España I: Reino de Castilia* (Madrid: Sucesores de Hernando, 1919), as well as the importance accorded to a "cheeselist" written in the late tenth-century Leonese form of Romance in Wright, *Late Latin and Early Romance*, 173–175.

Romance- and Arabic-speakers!—many loanwords were appropriated via medieval Latin rather than through direct contact, and that the attested loanwords mainly include toponyms, names of plants, and terms for food, clothes, artisanship, and social organization, as well as weights and measures, but few terms from the semantic field of emotion.138 However, such estimates of how many and which types of words of Arabic origin form part of Castilian, Catalan, Portuguese, etc. are based on texts written after the initial period of intensive contact between the eighth and the twelfth century. This makes it difficult, not only to trace the chronology of the Arabic impact on Ibero-Romance languages, but also to judge how strong this impact was in the early centuries of linguistic encounter. Considering that we lack approximately four hundred years of documentation, such results can obviously only provide a fleeting impression of linguistic contact between Romance and Arabic on the Iberian Peninsula before the period of intensified Latin-Christian expansionism.139

#### 2.3.4 ARABIC ENGAGEMENT WITH LATIN TEXTS ON THE IBERIAN PENINSULA

In this early medieval Iberian environment of flourishing, but mainly oral Latin/Romance-Arabic entanglement, Latin texts also soon entered the fray. Already mentioned above, Ḥafṣ b. Albar's Arabic translation of a Latin version of the psalms mainly catered to the needs of an Arabicized Christian community under Muslim rule. Although this is occasionally also said of the second important Latin-Arabic translation produced in al-Andalus, the "Book of Orosius" or *Kitāb Hurūšiyūš,*140 one must acknowledge that the latter received a much larger Muslim audience. The *Kitāb Hurūšiyūš* is a restructured version of the late antique Euromediterranean history by the Hispano-Latin historiographer Orosius of Braga (d. ca*.* 417). Working in the late ninth or early tenth century, the unknown compiler-editor expanded the book's chronology, covering the period from Orosius's death to the Muslim invasion in 711, and added more sources, including excerpts from

<sup>138</sup> Reinhard Kiesler, "Ibero-Romance," *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh, vol. 2 (Leiden: Brill, 2008), 281–286. For further lexical material see Reinhart Dozy and W. H. Engelmann, *Glossaire des mots espagnols et portugais dérivés de l'arabe: avec une introduction linguistique, un index de mots européens et un index arabe*, second rev. ed. (Leiden: Brill, 1869); Federico Corriente, *Dictionary of Arabic and Allied Loanwords* (Leiden: Brill, 2008).

<sup>139</sup> Nevertheless, note the impressive efforts at synthesis by Kontzi, "Zusammentreffen," 387–450; Kontzi, "Arabisch und Romanisch," 328–347; Ineichen, *Arabisch-orientalische Sprachkontakte*.

<sup>140</sup> Hans Daiber, "Weltgeschichte als Unheilsgeschichte. Die arabische Übersetzung von Orosius' Historiae adversus paganos als Warnung an die Muslime Spaniens," in *Christlicher Norden—Muslimischer Süden. Ansprüche und Wirklichkeiten*, ed. Matthias Tischler and Alexander Fidora (Münster: Aschendorff, 2011), 191–200.

a variety of other Latin texts, such as works by Isidore of Seville (d. 636).<sup>141</sup> Although the conditions that facilitated the production of this work are highly disputed, there is no doubt that it introduced masses of information on the Roman and post-Roman West to Arabic-Islamic textual culture and opened the way for a more intensive engagement with the contents, if not the language of Latin literature.142

After the translation, several Arabic-Islamic scholars from al-Andalus began to mention and even comment on the Latin language: Ibn Ǧulǧul (d. after 384/994), who gives one of several explanations of how the *Kitāb Hurūšiyūš* became available to Andalusian Muslim scholars, mentions "Latins able to speak the Latin language" (*al-Laṭīniyyīn man yaqraʾuhu bi-llisān al-laṭīnī*) in al-Andalus, capable of translating the book "from Latin to Arabic" (*min al-laṭīnī ilā l-lisān al-ʿarabī*).143 After several centuries, during which Arabic-Islamic scholars had regularly mixed up Greek and Latin and used the terms "rūmī," i.e. "Roman," or even "ifranǧī," i.e. "Frankish," to describe Latin,144 Ṣāʿid al-Andalusī (d. 462/1070), *qāḍī* of Toledo, made a clear distinction between an ancient language of the Greeks (*al-Yūnāniyyīn*) called "al-iġrīqiyya," and a language of the Romans (*al-Rūm*) called "al-laṭīniyya."145 His contemporary al-Bakrī (d. 487/1094) quoted "experts of the Latin language" (*ahl al-ʿilm bi-l-lisān al-laṭīnī*) and even tried to transcribe the correct Latin pronunciation of the toponym "Toledo" (*maʿnā Ṭulayṭula bi-l-laṭīnī Tūlāẓū*).<sup>146</sup> In probably the most detailed Arabic description of the Latin language in pre-modern times, Ibn Ḥazm (d. 456/1064) commented thrice on certain particularities of the Latin language, which he compared to their Arabic equivalents in a treatise on Aristotelian logic.147 Thus, the textual entanglement of Latin and Arabic as written languages also reached its first apogee in al-Andalus of the ninth to the eleventh century.

<sup>141</sup> *Kitāb Hurūšiyūš (traducción árabe de las historiae adversus paganos de Orosio)*, ed. Mayte Penelas (Madrid: CSIC, 2001).

<sup>142</sup> König, *Arabic-Islamic Views*, 84–86, 134–135.

<sup>143</sup> Quoted in Ibn Abī Uṣaybiʿa (d. 668/1270), *ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ*, ed. August Müller, 2 vols. (Cairo: al-Maktaba l-wahbiyya, 1881–1882), vol. 2, 47.

<sup>144</sup> Daniel G. König, "The Unkempt Heritage: On the Role of Latin in the Arabic-Islamic Sphere," *Arabica* 63, no. 5 (2016), 428–432.

<sup>145</sup> Ṣāʿid al-Andalusī, *Kitāb Ṭabaqāt al-umam*, ed. Ḥayāt Bū ʿAlwān (Beirut: Dār al-ṭalīʿa, 1985), 96.

<sup>146</sup> Al-Bakrī, *Kitāb al-Masālik wa-l-mamālik*, ed. Adrian P. van Leeuwen and André Ferré (Tunis: al-Dār al-ʿarabiyya li-l-kitāb, 1992), § 1521, 907; ibid., § 1513, 902.

<sup>147</sup> Ibn Ḥazm, *Al-Taqrīb li-ḥadd al-manṭiq wa-l-madḫal ilayhi bi-l-alfāẓ al-ʿāmmiyya wa-l-amṯila l-fiqhiyya*, ed. Iḥsān ʿAbbās, in Ibn Ḥazm, *Rasāʾil Ibn Ḥazm*, vol. 4 (Beirut: al-Muʾassasa l-ʿarabiyya li-l-dirāsa wa-l-našr, 1983), 109–110, 153, 155–156; König, "Unkempt Heritage," 433–435.

#### 2.3.5 LANGUAGE CONTACT IN THE RELATIONS BETWEEN ARABIC-ISLAMIC AND LATIN-CHRISTIAN SOCIETIES

Latin-Arabic entanglement cannot only be regarded as an intrasocietal phenomenon characteristic of multi-religious and multi-ethnic societies under Muslim rule. It also formed part of economic, diplomatic, and other relations between Latin-Christian and Arabic-Islamic societies that were established in the wake of the Arabic-Islamic expansion of the seventh and eighth centuries. The emerging Christian realms on the Iberian Peninsula maintained regular exchanges with Umayyad al-Andalus and the early *ṭāʾifa*-principalities. This is also valid for the Carolingian Empire, whose rulers additionally engaged with Muslims in Italy and even established short-lived relations with the Abbasids. The eastern Frankish ruler Otto I exchanged embassies with al-Andalus and, by taking control of parts of Italy in the 950s, involved his successors in dealings with the Muslims of southern Italy. Last but not least, Italian cities such as Venice, Amalfi, Naples, and Salerno established and maintained commercial and other relations with the Muslims of southern Italy and North Africa, with Genoa and Pisa entering the scene in the early eleventh century.<sup>148</sup>

In view of these relations, Arabic and Latin anthroponyms, ethnonyms, and toponyms inevitably became part of the respective other textual culture, challenging authors and scribes with the problem of transcription. Mentioning a Muslim delegation from the northern parts of al-Andalus to the court of Charlemagne in 777, the *Annales regni Francorum* Latinize Arabic anthroponyms, defining the envoys as "Ibn al-Aʿrābī and Ibn Yūsuf, who is called Joseph in Latin" (*Ibin al Arabi et filius Deiuzefi, qui et latine Ioseph nominatur*).149 Aside from acknowledging that interaction led to such rather simple forms of Latin-Arabic entanglement, one must consider the linguistic implications of these intersocietal relations.

Older scholarship propounded the hypothesis that intersocietal relations were facilitated linguistically by Jews and Christians under Muslim rule. It acknowledged a few Muslim merchants, sailors, and harbour officials with a "working colloquial knowledge" of Romance languages, but placed them at "a low level of society with little or no cultural influence." This was explained by asserting that, among Muslims, "knowledge of foreign languages was not an esteemed qualification," but rather "a specialized craft belonging to the non-Muslim communities and, like some other occupations, marked with a stigma of social inferiority."<sup>150</sup> A more nuanced approach shows, however, that the linguistic facets of intersocietal

<sup>148</sup> König, *Arabic-Islamic Views*, 52–54.

<sup>149</sup> *Annales Regni Francorum*, ed. Georg Heinrich Pertz and Friedrich Kurze, Monumenta Germaniae Historica, Scriptores rerum Germanicarum in usum scholarum 6 (Hanover: Hahn, 1895), a. 777, 48. The source probably refers to the son of the Muslim governor of Barcelona, Sulaymān al-Aʿrābī, who is mentioned in *Aḫbār maǧmū*ʿ*a*, ed. Lafuente y Alcántara, 112–113 (AR), 103 (ES).

<sup>150</sup> Bernard Lewis, *The Muslim Discovery of Europe* (New York: Norton, 2001), 81.

exchange between Latin-Christian and Arabic-Islamic societies cannot be reduced to a situation characterized by the alleged linguistic inertness of adherents to Islamic religion and culture, the latter buttressed and enhanced by the general availability of non-Muslim linguistic resources in societies under Muslim rule.

Taking on the perspective of those scholars who believe that Romance idioms were an integral feature of Jewish, Christian, and Muslim lifeworlds under Muslim rule in the western Mediterranean may explain better why early medieval sources either imply or describe acts of communication between Muslims and Christians pertaining to different societies without mentioning any form of linguistic mediation. Such acts of communication are frequently attested, e.g. when a certain Solinoan, Muslim governor of Barcelona and Geronda, submitted to the Carolingian King Pippin III in 752,151 or when the Umayyad dissident ʿAbd Allāh and his two sons sought refuge at the court of Charlemagne in 797.<sup>152</sup> A panegyric to Louis the Pious dating from 829–830 features a "Moorish" raider (*maurus*) who effortlessly speaks with a Frankish man from Aquitaine named Datus, whose mother he has taken captive.153 One wonders if Ǧamīla, the sister of the renegade rebel Maḥmūd b. ʿAbd al-Ǧabbār had to acquire Romance skills when her brother was killed in 226/841 and she was allotted to a Christian noble and baptized, later to become ancestor of an archbishop of Santiago de Compostela.154 When a certain Frankish Christian named Bodo converted to Judaism and moved to al-Andalus in 847, he allegedly incited the Muslim authorities to forcibly convert their Christian subjects, thus triggering Hispano-Christian complaints to the Frankish court.155 After his defeat around 871, the emir of Bari is said to have communicated intensively with Louis II of Italy as well as with the local rulers, e.g. of Benevent.156 The duke Guaifar of Salerno (r. 861–880) is reported to have spoken to an "Agarenus" on the marketplace of Salerno, who—upon his return to North Africa—asked an Amalfitan merchant to warn the prince of an upcoming Aghlabid attack.157

<sup>151</sup> *Annales Mettenses priores*, ed. Bernhard von Simson, Monumenta Germaniae Historica, Scriptores rerum Germanicarum in usum scholarum 10 (Hanover: Hahn, 1905), a. 752, 43–44.

<sup>152</sup> Ibn Ḥayyān, *Al-Sifr al-ṯānī min kitāb al-muqtabis [al-Muqtabis II-1]*, ed. Maḥmūd ʿAlī Makkī (Riyad: Markaz al-malik Fayṣal li-l-buḥūṯ wa-l-dirāsāt al-islāmiyya, 2003), fol. 90a, AH 181, 97; Ibn Ḥayyān, *Crónica de los emires Alḥakam I y ʿAbdarraḥmān II*, trans. Mah*̣*mūd ʿAlī Makkī and Federico Corriente (Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, 2001), 20.

<sup>153</sup> Ermoldus Nigellus, *Carmen in honorem Hludowici*, ed. Ernst Dümmler, Monumenta Germaniae Historica, Poetae 2 (Berlin: Weidmann, 1884), lib. 1, v. 207– 260, 11–13.

<sup>154</sup> Ibn Ḥayyān, *Al-Muqtabis II-1*, ed. Makkī, fol. 184r, AH 225, 445 (AR), 306 (ES).

<sup>155</sup> *Annales Bertiniani auctore Prudentio*, ed. Georg Waitz, Monumenta Germaniae Historica, Scriptores rerum Germanicarum in usum scholarum 5 (Hanover: Hahn, 1883), a. 847, 34–35.

<sup>156</sup> Constantinus Porphyrogenitus, *De administrando imperio*, ed. Moravcsik, trans. Jenkins, cap. 29, 131. See a variation of related anecdotes in Symeon magister, *Annales*, ed. Bekker, cap. 20, 694–697.

<sup>157</sup> *Chronicon Salernitanum*, ed. Westerbergh, § 110, 122–123.

In the period between 899 and 906 a North African Muslim captive named ʿAlī interacted with Bertha, the Carolingian wife of the margrave of Tuscany, providing her with information about the Abbasid caliph.<sup>158</sup> The Leonese King Ordoño does not seem to have needed an interpreter to instruct renegade Muslim scouts in 303/915.159 Merchants from Verdun, known for their profitable export of completely castrated eunuchs to the Iberian Peninsula,160 were chosen as guides for the Ottonian envoy to the Umayyad court in 953 because they were known to have experience in the country.161 In 1076, Pope Gregory VII wrote to the Ḥammādid prince al-Nāṣir of Bejaïa in Latin, maybe surmising that local Christians such as the bishop of Carthage, addressed in another letter from the same period, might inform the prince of the letter's content.162

Not all of these sources can be taken at face value. They show, however, which acts of communication between Christians and Muslims pertaining to different societies were deemed possible by contemporary authors. Consequently, they support the supposition that the western Mediterranean of the early Middle Ages constituted a linguistic sphere in which oral forms of Latin, i.e. various Romance idioms, had maintained a certain currency shared by Christians, Christian converts to Islam and their descendants, some Berbers, and even families of Arab stock. We can also assume that several centuries of commercial, political, and military interaction with Arabic-Islamic societies gave Christians from Romance-speaking societies a certain degree of knowledge about Arabic. This is implied in the description of a raid on Narbonne at the beginning of the eleventh century executed by "Moors from Cordoba" (*Cordubensis Mauri*). The author, Adhémar of Chabannes (d. 1034), seems to have expected that these Moors would speak Arabic: he is astonished that the captives taken during the raid do not speak "the Saracen language" (*loquela Sarracenisca*), but "speak like puppies and seem to bark" (*sed more catulorum loquentes, glatire videbantur*).163

160 Liutprandus, *Antapodosis*, ed. Becker, lib. VI, cap. 6, 155–156.

<sup>158</sup> Al-Rašīd b. al-Zubayr, *Kitāb al-Ḏaḫāʾir wa-l-tuḥaf*, ed. Muḥammad Ḥamīdullāh (Kuwait: Dāʾirat al-maṭbūʿāt wa-l-našr, 1959), 50–51; *Book of Gifts and Rarities— Kitāb al-Hadāyā wa al-Tuḥaf*, trans. Ghāda al Ḥijjāwī al-Qaddūmī (Princeton: Princeton University Press, 1996), § 69, 92. König, *Arabic-Islamic Views*, 200–202.

<sup>159</sup> Ibn Ḥayyān, *al-Muqtabas li-Ibn Ḥayyān al-Qurṭubī (al-ǧuzʾ al-ḫāmis*) *[al-Muqtabis V]*, ed. Pedro Chalmeta and Federico Corriente (Madrid: Instituto hispano-árabe de cultura: 1978), AH 303, 120; *Cronica del Califa ʿAbdarraḥmān III an-Nāṣir entre los años 912 y 942*, trans. María Jesús Viguera and Federico Corriente (Zaragoza: Anubar ediciónes, 1981), § 81, 101.

<sup>161</sup> Iohannis abbas s. Arnulfi, *Vita Iohannis abbatis Gorziensis*, ed. Georg Heinrich Pertz, Monumenta Germaniae Historica, Scriptores in folio 4 (Hanover: Hahn, 1841), cap. 117, 370; Jean de Saint-Arnoul, *La Vie de Jean, abbé de Gorze*, ed. and trans. Michel Parisse (Paris: Picard, 1999), § 117, 144–145.

<sup>162</sup> Gregorius VII, *Registrum*, ed. Erich Caspar, Monumenta Germaniae Historica, Epistolae selectae in usum scholarum 2,1 (Berlin: Weidmann, 1920), lib. III, ep. 21, 287–288.

<sup>163</sup> Ademarus Cabannensis, *Chronicon (recensiones beta et gamma)*, ed. Pascale Bourgain, Richard Landes, and Georges Pon, (Turnhout: Brepols, 1999), lib. III, cap. 52, 171.

If it is accepted that bi- and multilingualism were an integral feature of Christian-Muslim interaction in the early medieval Mediterranean, then it becomes necessary to explain why linguistic obstacles and mediators are mentioned at all.164 It seems obvious, for example, that it was much more difficult to find human resources bridging the linguistic divide between Latin/Romance and Arabic in the eastern than in the western Mediterranean. Thus, instances of linguistic mediation are occasionally addressed: an Anglo-Saxon pilgrim, imprisoned in Syria around 724 by the Umayyad authorities, needed the help of a "homo hispanus," i.e. a person from the western Mediterranean, who acted as mediator.<sup>165</sup> Reports on Pippin III's and Charlemagne's diplomatic exchanges with the Abbasid caliphs al-Manṣūr (around 765) and Hārūn al-Rašīd (around 801), in turn, do not mention interpreters.166 Then again, a letter sent by Bertha, wife of the Margrave of Tuscany to the Abbasid caliph al-Muktafī bi-llāh in 906, is said to have been translated first into Greek by a "Frank" working in the caliph's wardrobe, then into Arabic by the Syrian Christian Isḥāq b. Ḥunayn.<sup>167</sup> Obviously impressed by their linguistic skills, the Abbasid geographer Ibn Ḫurdāḏbah (d. *c*.300/911) explains that the so-called Radhanite Jews were able to move goods from the western Mediterranean via the eastern Mediterranean to China and back again because they were able to speak "Frankish" (*al-ifranǧiyya*) and "Andalusian" (*al-andalusiyya*), presumably Frankish and Iberian forms of Romance, in addition to Slavonic, Greek, Arabic, and Persian.<sup>168</sup>

While it seems plausible that exchanges between the Christian societies of Western Europe and the Middle East needed to be mediated linguistically, this should not imply that linguistic obstacles and mediators were inexistent in the western Mediterranean. When they are mentioned, however, this is generally done because linguistic issues played a preponderant role for the narrative. As opposed to the Carolingians of the western Frankish realm, the Ottonian court of the 950s, for example, had practically no experience in dealing with the Muslim sphere. After receiving a letter from the Umayyad caliph, ʿAbd al-Raḥmān III (r. 300–350/912–961), the Ottonian court was in the dark about how to deal with this unexpected "pagan" effort to establish contact. Consequently, it handled its first embassy to a Muslim ruler rather ineptly. Considering how strange Muslim al-Andalus

<sup>164</sup> See König, *Arabic-Islamic Views*, 66.

<sup>165</sup> Hugeburc, *Vita Willibaldi*, ed. Oswald Holder-Egger, Monumenta Germaniae Historica, Scriptores in folio 15 (Hanover: Hahn, 1887), cap. 4, 95.

<sup>166</sup> See Michael McCormick, "Pippin III, the Embassy of Caliph al Mansur and the Mediterranean World," in *Der Dynastiewechsel von 751: Vorgeschichte, Legitimationsstrategien und Erinnerung*, ed. Matthias Becher and Jörg Jarnut (Münster: Scriptorium, 2004), 221–241; Michael Borgolte, *Der Gesandtenaustausch der Karolinger mit den Abbasiden und mit den Patriarchen von Jerusalem* (Munich: Arbeo, 1976), for the relevant sources.

<sup>167</sup> Al-Rašīd b. al-Zubayr, *Kitāb al-Ḏaḫāʾir*, ed. Ḥamīdullāh, 48–54; *Book of Gifts*, trans. al Ḥijjāwī al-Qaddūmī § 69, 91–98.

<sup>168</sup> Ibn Ḫurdāḏbah, *Kitāb al-Masālik wa-l-mamālik*, ed. Michael de Goeje (Leiden: Brill, 1896), 153.

seemed to the protagonist of the *Life of John of Gorze*, the Ottonian envoy to ʿAbd al-Raḥmān III in Cordoba, it comes as no surprise that both John and the author of his *Life* seemed impressed by the fact that the influential Umayyad courtier Recemundus was a good Catholic and possessed firm knowledge of both Latin and Arabic literary cultures.169

The most important Arabic-Islamic chronicler of early medieval al-Andalus, Ibn Ḥayyān (d. 469/1076) mentions several occasions in which the Umayyad caliph al-Ḥakam II (r. 350–366/961–976) received delegations from Iberian Christian rulers, but only refers to the services of interpreters twice. In connection with a delegation in 360/971, the interpreters seem noteworthy because linguistic mediation is facilitated by important Christian authorities from Cordoba and Seville, including a certain Aṣbaġ b. Nabīl, judge (*qāḍī*) of the Christians in Cordoba.<sup>170</sup> In connection with a Leonese delegation that took place in 363/973–974, Ibn Ḥayyān describes how the caliph punished this individual for having translated the delegation's vituperative attacks against the caliph verbatim.171 In the case of ʿAlī, the son of Muǧāhid, the eleventh-century Muslim ruler of the *ṭāʾifa*-principality of Denia, linguistic issues were of relevance because the future ruler of Denia may not have been capable of speaking Arabic before his accession to power. According to Ibn al-Ḫaṭīb (d. 776/1375), ʿAlī and his Christian mother had been taken captive during a raid of Muǧāhid's troops on Sardinia. When he was released around 423/1031, after seventeen years of captivity, he was only able to speak the language of his (probably Pisan) captors.172 One of the sources on the biography of Constantine the African, the translator and author of various medical treatises in Salerno of the second half of the eleventh century, proposes to explain why and how the latter became a translator. It claims that Constantine was a Saracen merchant who, during a visit to Rome, allegedly conversed with a local physician through the mediation of the latter's Saracen slaves. Thanks to this conversation, Constantine became aware of the Latins' penury in medical books, moved to Italy, converted to Christianity, and began translating. It seems evident that his successful efforts to study "the Roman

<sup>169</sup> Iohannis abbas, *Vita Iohannis*, ed. Pertz, § 128, 374: "Recemundus quidam, adprimae catholicus, et litteris optime tam nostrorum quam ipsius inter quos versabatur linguae Arabicae institutus"; Jean de Saint-Arnoul, *La Vie de Jean*, § 128, 154–155.

<sup>170</sup> Ibn Ḥayyān, *Al-Muqtabis fī aḫbār balad al-Andalus [al-Muqtabis VII]*, ed. ʿAbd al-Raḥmān ʿAlī l-Ḥaǧǧī (Beirut: Dār al-ṯaqāfa, 1965), AH 360, 64.

<sup>171</sup> Ibn Ḥayyān, *Al-Muqtabis VII*, ed. al-Ḥaǧǧī, AH 363, 146: "wa-tawaṣṣala ilayhi baʿdahum rusul Ḥalwīra ʿammat al-ṭāġiya amīr Ǧillīqiyya wa-kāfilatihi fa-takallamū ʿan mursalatihim bi-kalām badā fīhi baʿḍ al-ǧifāʾ, tarǧamahu naṣṣan ʿanhum Aṣbaġ bin ʿAbd Allāh Ibn Nabīl qāḍī l-naṣārā bi-Qurṭuba al-mutawallī ḏālika ʿan al-aʿāǧim, ankarahu l-ḫalīfa li-waqtihi, fa-izwarra li-l-mutarǧim wanaharahu, wa-amara bi-taʾḫīr al-rusul ʿanhu wa-nālahum bi-baʿḍ al-tawbīḫ, waalzama Aṣbaġ al-mutarǧim ḏanbahu, wa-amara bi-iqṣāʾihi wa-ʿazlihi ʿan quḍāʾ al-naṣārā wa-ihānatihi."

<sup>172</sup> Ibn al-Ḫaṭīb, *Kitāb Aʿmāl al-aʿlām*, ed. Évariste Lévi-Provençal (Beirut: Dār al-makšūf, 1956), 221.

and Latin language" (*romana [et] latina lingua*) would be highlighted in this context.<sup>173</sup>

All this evidence for communication between members of societies under Latin-Christian or Arabic-Islamic rule respectively shows not only that communication was frequent, but also that it was not regarded as problematic. The small number of references to interpreters in narrative sources suggests that either bilingualism or linguistic mediation were considered normal, and the activity of interpreters only deemed noteworthy in particular circumstances.

## 2.4 Linguistic effects of Latin-Christian expansionism

The evidence compiled in section 2.3 of this chapter served to provide an overview of the different variants of Latin-Arabic entanglement that resulted from the Arabic-Islamic expansion into the western Mediterranean. From the late eleventh century onwards, new milieus of linguistic entanglement emerged as the Mediterranean power balance between Latin-Christian and Arabic-Islamic societies was gradually reversed. Latin-Christian expansion into the Mediterranean sphere as represented by the Norman conquest of Sicily, the so-called *Reconquista*, the Crusades, and the commercial enterprise of European-Christian maritime powers considerably enforced the spread of Latinate languages into Mediterranean regions hitherto under Muslim and Byzantine control.

Although Latin played a role in this process of linguistic expansion, Romance languages, increasingly put into writing from the thirteenth century onwards, became more dominant than before, at least in the written documentation. Two corpora of sources allow us to trace the rising impact of Romance languages.

Bilingual commercial and political treaties concluded between European-Christian maritime powers such as the Crown of Aragon and the Italian republics with North African Muslim polities constitute the first corpus.<sup>174</sup> Treaties of the late twelfth and early thirteenth centuries were generally formulated in Latin and Arabic. Although such Latin-Arabic treaties were produced until the fifteenth century, Catalan and—slightly later—Italo-Romance versions begin to appear and to increase in number from around the middle of the thirteenth century onwards.<sup>175</sup>

Arabic transcriptions of European-Christian titles and anthroponyms in Arabic-Islamic historiography of the thirteenth to the fifteenth century constitute the second corpus. The transcriptions "ray dā Farans" or "rawā Farans" obviously derive from the French "roi de France." The transcription

<sup>173</sup> See the edition in Kreutz, "Ehrenrettung," 40–41. See also Ricklin, "Fall Gougenheim," 128.

<sup>174</sup> On such documents, see the contribution by Daniel Potthast in Chapter 3 of this volume.

<sup>175</sup> König, "Unkempt Heritage," 441 fn. 83.

"rayd Araġūn" possibly leads back to the Catalan "Reis d'Aragó" or the French "Roi d'Aragon," whereas the toponym "Inkaltīra" for England, ruled by a king called "al-Inkitār" possibly derives from the Anglo-Norman variant of the French "Angleterre" or the Castilian "Inglaterra." The title "malik al-Almān," i.e. "king of the Germans," clearly originates in the French denomination "Allemands." The technical term "ḥukm kumūn," i.e. "communal government," obviously derives from an early Italian variant of the term "comune." We even find distinctions between a formal and a colloquial pronunciation of the imperial title, contrasting "al-inbaraḏūr" (*imperator*, *imperador*) to "al-anbarūr" (*empereur*).<sup>176</sup>

Apart from these generalities, European-Christian expansionism took on different forms, and thus affected the respective target regions differently. For this reason, it is necessary to approach the associated linguistic effects of expansionism in regional order, thus dealing separately with Sicily, the Iberian Peninsula, North Africa, and the eastern Mediterranean.

#### 2.4.1 SICILY

The Norman conquest of Sicily in the last third of the eleventh century was facilitated by linguistic mediators working to the advantage of the Normans. These were recruited from among Christians hitherto under Muslim rule as well as from among the occasional Muslim renegade. A fourteenth-century French version of the Latin history of the Normans by Amatus de Montecassino (d. after 1061 or 1078) mentions a deacon called Pierre who, because "he understood and talked very well like the Saracens" (*entendoit et parloit molt bien coment li Sarrazin*), was sent out as a spy by Robert Guiscard (r. 1059–1085). In addition he refers to (presumably Arabic-speaking) Christians "who did not wish to live subjected to the pagans" and consequently joined Robert's forces.177 That the Normans employed Christian Arabic-speakers is confirmed by Gaufredus de Malaterra (d. after 1101). He mentions that a certain Philipp, son of the patrician Gregorius, was sent out to reconnoitre the Saracen fleet "since he and all the sailors who set out with him were fluent in their language as well as in Greek."178 In some cases, even Muslims seem to have defected to the Norman camp.179

<sup>176</sup> König, *Arabic-Islamic Views*, 88 (imperial title), 104–105 (Aragon), 106 (imperial title), 221–225 (France), 277–278 (England), 281–284 (Germans), 286–288 (imperial title), 298–290 (communal rule).

<sup>177</sup> Amatus de Montecassino, *Ystoire de li Normant*, ed. Vincenzo De Bartholomaeis (Rome: Tip. del Senato, 1935), lib. V, cap. 24–25, 244: "qué desirroient de non estre subjette à li Paien."

<sup>178</sup> Gaufredus de Malaterra, *De rebus gestis Rogerii Calabriae et Siciliae comitis et Roberti Guiscardi ducis fratris eius*, ed. Ernesto Pontieri (Torino: Bottega d'Erasmo, 1972), lib. IV, cap. 2, a. 1085, 86: "nam et lingua eorum, sicut et graeca, ipse et nautae omnes, qui cum ipso processerant, peritissimi erant [. . .]"

<sup>179</sup> Gaufredus de Malaterra, *De rebus gestis*, ed. Pontieri, lib. III, cap. 30, a. 1082, 75: "Elias Cartomensis—qui ex Saracenis ad fidem Christi conversus."

As soon as Sicilian territories had come firmly under Norman control, the island seems to have received new "Latin" settlers. This is implied by a diploma issued by Roger II in 1133 in response to a legal dispute between the inhabitants of Patti and Bishop John of Lipari-Patti. The diploma refers to an earlier *memoratorium* and states that Ambrosius of Lipari, "the first abbot under the consulship of Roger [I, d. 1101], the land's conqueror, ordered men of Latin language into the fort Patti." Ironically, the descendants of these "men of Latin language" (*ho[m]i[n]es q[u]ic[um]q[ue] sint Latinę linguę*) were not able to understand the Latin *memoratorium:* its contents were explained to them "in the vulgar language (*vulgariter*)."180 At the end of the twelfth century, the influential monastery Monreale took the decision to translate its Arabic land registers "de saracenico in latinum" or, as the Arabic version states, "min al-ʿarabī ilā l-laṭīnī." In this context, the Greek word for official registers, i.e. διϕθέρα, appropriated in Arabic as *daftar* pl. *dafātir*, was Latinized as *deftarii / deptarii.*<sup>181</sup>

Against this backdrop, neither is it surprising that Stephen of Antioch (second quarter of the twelfth century) referred the readers of his trilingual glossary of medicaments as documented in Dioscorides's pharmacological treatise to Sicily and Salerno, where they would find Greek- and Arabic-speakers able to define those plants and herbs which he had been unable to identify.<sup>182</sup> Nor does it seem exaggerated to speak with Petrus de Ebulo (d. before 1220) of Palermo as a "happy city endowed with a trilingual people" (*urbs felix populo dotata trilingui*).183 When Pope Innocent III wrote to "all the Saracens established in Sicily" in 1199, and to various judges (*quḍā*, pl. of *qāḍī*, transcribed *Archadio*) and leaders (*quwwād*, pl. of *qāʾid*, transcribed *Gaietis*) in 1206, he probably did not need to worry that his call for supporting the young Frederick II would not be understood.184


Frederick II is described by Ibn Wāṣil (d. 697/1298) as having conversed—without reference to an interpreter—with the *qāḍī* of Nablus during his sojourn in Jerusalem in 1229.185 Ibn Naẓīf al-Ḥamawī (d. after 631/1233) cites the emperor's Arabic letter sent to the Ayyubid sultan al-Malik al-Kāmil and Faḫr al-Dīn b. Šayḫ al-Šuyūḫ around 627/1230. The letter is perfectly structured and phrased, partly in prose, partly in verse.<sup>186</sup> However, in spite of these strong indicators of Frederick II's interest in Arabic-Islamic culture and even linguistic skills, he seemingly failed to interact, in cultural terms, with the Arabic-speaking population of Sicily and the Mezzogiorno. Although Frederick sponsored a number of Arabic-Latin translations, including works of Aristotle, Avicenna / Ibn Sīnā, and Averroes / Ibn Rušd, he did not seek for the necessary human resources among the Muslims of Sicily, but drew on Christian translators from other regions: Michael Scot (d. ca. 1235) came from Toledo, Theodore of Antioch (d. before 1250) from the Syrian Levant.187

Frederick's choice to import Arabic-speaking Christian intellectuals may have been due to a constant decrease in Arabic skills among the population of Sicily, already noticeable during the emperor's reign. In the transitory period of Muslim, Norman, and Staufen rule, population movements and processes of acculturation transformed the linguistic landscape of Sicily. Many Muslim intellectuals had emigrated to North Africa during and in the wake of the Norman conquest.188 In addition, Alex Metcalfe has detected onomastic shifts from Arabic to Graeco-Latin names among Christians in Collesano, which suggest that Arabic-speaking Christians were slowly succumbing to a linguistic and cultural process of "Latinization."189 Finally, Frederick II's deportation of thousands of Muslims from the island to Lucera in Apulia from the 1220s onwards contributed further to leaving the island devoid of Arabic speakers.190 In 659/1252, Frederick II's son, Manfred, still seems to have been surrounded by a certain number of Muslims and, according to the report of Ibn Wāṣil, had

Plon, 1852–1861), vol. 1,1, 37–40: "Innocentius [. . .] universis Saracenis in Sicilia constitutis [. . .]." [a. 1199]; ibid., 118–120, 118: "Archadio [*al-qāḍī*] et universis Gaietis [*al-qāʾid*] Antelle, Platane, Jaci, Celsi et omnibus Gaietis et Sarracenis per Siciliam constitutis [. . .]." [a. 1206].

<sup>185</sup> Ibn Wāṣil, *Mufarriǧ al-kurūb fī aḫbār Banī Ayyūb*, ed. Ḥasanayn Muḥammad Rabīʿ and Saʿīd ʿAbd al-Fattāḥ ʿĀšūr, vol. 4 (Cairo: Dār al-kutub wa-l-waṯāʾiq al-qawmiyya, 1957), 244–245.

<sup>186</sup> Ibn Naẓīf, *al-Tārīḫ al-manṣūrī*, ed. Abū l-ʿĪd Dūdū and ʿAdnān Darwīsh (Damascus: Maṭbaʿat al-Ḥiǧāz, 1982), 190. See Chapter 4.2.1 in this volume.

<sup>187</sup> *Kulturtransfer und Hofgesellschaft im Mittelalter. Wissenskultur am sizilianischen und kastilischen Hof im 13. Jh.*, ed. Gundula Grebner and Johannes Fried (Berlin: Akademie-Verlag, 2008). See Chapter 4.3.1 in this volume.

<sup>188</sup> Julie Anne Taylor, *Muslims in Medieval Italy: The Colony at Lucera* (Lanham: Lexington Books, 2005), 11–14, 72; Metcalfe, *Muslims and Christians*, 101.

<sup>189</sup> Metcalfe, *Muslims and Christians*, 85–86; Hubert Houben, "Möglichkeiten und Grenzen religiöser Toleranz im normannisch-staufischen Königreich Sizilien," *Deutsches Archiv für die Erforschung des Mittelalters* 50 (1994), 159–198.

<sup>190</sup> Al-Ḥimyarī, *Kitāb al-Rawḍ al-miʿṭār fī ḫabar al-aqṭār*, ed. Iḥsān ʿAbbās (Beirut: Maktabat Lubnān, 1975), 514.

even ordered the construction of a building dedicated to the theoretical sciences.191 Arabic-Latin translations effected under Manfred, however, were also produced by Christians, not by Muslims, in this case William of Luna.192 After Manfred's death and the Angevin takeover in Sicily and southern Italy, documented translators from Arabic to Latin were all Jews.193 Some of them, such as Faraǧ b. Sālim of Agrigento, seem to have been sufficiently fluent in Arabic and Latin to translate al-Rāzī's treatise on medicine, procured by Charles I of Anjou from Tunis.194 Another translator, Moses of Palermo, seems to have been capable of reading Arabic, but in need of a teacher "to teach and inform him about Latin literature."<sup>195</sup> As late as the fifteenth century, Jews in Sicily and the Mezzogiorno seem to have constituted the prime mediators between Latin and Arabic, within Sicily as well as in relations with North Africa, as is attested repeatedly in archival documents collected by Henri Bresc and Shelomo D. Goitein.196 According to Henri Bresc and Alex Metcalfe, this evidence suggests that, by the late thirteenth century, Muslim converts to Christianity and the Christian-Arabic population of Sicily had adapted so thoroughly to the Latinized culture of Staufen and Angevin Sicily that they were no longer able to fill the linguistic void left by the departed and deported Muslims


of Sicily.197 Consequently, Arabic-speaking Sicilian Jews, enriched numerically by Arabic-speaking immigrants from Almohad al-Andalus and North Africa,198 fulfilled the task of linguistic mediation and translation in Angevin and Aragonese Sicily199 and even influenced the early humanist study of Arabic in Italy.200 At the same time, they also succumbed to the process of Latinization as is proven by onomastic shifts in the form of calques, e.g. in the substitution of the Judaeo-Arabic name "Ḥabīb" (i.e. beloved) with the name "Amatus" (i.e. beloved),201 a process either enforced or interrupted, when the Aragonese authorities forcibly converted or expelled the Jewish population from Sicily in 1492.202 Thus, at the end of the fifteenth century, Sicily was probably not completely devoid of Arabic speakers, but was certainly not able to compete with the new, humanist-influenced centres of Arabic studies that had meanwhile emerged in mainland Italy.203

#### 2.4.2 IBERIAN PENINSULA

When Christian conquerors took over territories hitherto held by Muslims in the course of the so-called *Reconquista*, this usually entailed an administrative arrangement. The latter was often spelt out in Latin or Romance documents that allow us to identify specific forms of Latin-Arabic or Romance-Arabic entanglement. After his conquest of Menorca in 1231, for example, the Aragonese King James I granted certain rights to the Muslim inhabitants of the island. Acknowledging the internal organization of the Muslim community, he was obliged to employ certain Arabic terms in the Latin document he issued, not only anthroponyms such as *Aboabdille Abenixem* (probably Abū ʿAbd Allāh b. Hišām), but also administrative titles such as *alfaqui* (*al-faqīh*, i.e. jurist), *alcayd* (*al-qāʾid*, leader), and *almoxariff* (*al-mušrif*, i.e. overseer).<sup>204</sup> The Castilian King Alfonso X (r. 1252–1284), in turn, obviously deemed it necessary to create a legal framework for the commercial interaction of Romance- and Arabic-speakers not able to understand each other. This may have become a rather

<sup>197</sup> Henri Bresc, *Arabes de langue, juifs de religion: l'évolution du judaïsme sicilien dans l'environnement latin, XIIe –XV<sup>e</sup> siècle* (Saint-Denis: Éditions Bouchène, 2001); Metcalfe, *Muslims and Christians*, 23–117.

<sup>198</sup> Giuseppe Mandalà, "La migrazione degli ebrei del *Garbum* in Sicilia (1239)," *Materia giudaica* 11, no. 1–2 (2006), 179–198.

<sup>199</sup> Bresc, *Arabes de langue*, 39–47; Cohn, "Jüdische Übersetzer," 240–260.

<sup>200</sup> Bénoît Grévin, "Connaissance et enseignement de l'arabe dans l'Italie du XV<sup>e</sup> siècle: quelques jalons," in *Maghreb-Italie: des passeurs médiévaux à l'orientalisme moderne, XIIIe –milieu XX<sup>e</sup> siècle*, ed. Bénoît Grévin (Rome: École française de Rome, 2010), 103–138. See Chapters 4.2.1 and 4.2.3 in this volume.

<sup>201</sup> Bresc, *Arabes de langue*, 43–44.

<sup>202</sup> Bresc, *Arabes de langue*, 42.

<sup>203</sup> See Chapter 4 in this volume.

<sup>204</sup> Louis de Mas Latrie, *Traités de paix et de commerce et documents divers concernant les relations des chrétiens avec les Arabes de l'Afrique septentrionale au moyen âge*, 2 vols. (Paris: Henri Plon, 1866, 1872), vol. 1, 183.

frequent problem as soon as Romance-speaking settlers from the Iberian north, without previous contact with Muslim al-Andalus or Arabic, moved into former Muslim territory in what Spanish scholarship usually defines as *repoblación*. In his legal compilation known as the *Siete Partidas*, the Castilian king obliged such people to use an interpreter accepted by both sides.<sup>205</sup>

The *Siete Partidas* form part of a large body of Castilian texts, many of which had been commissioned by Alfonso X. According to L. P. Harvey, Alfonso X's commitment transformed Castilian, formerly "one relatively uncultivated vernacular Romance dialect among many" into "an established medium of expression in which was available a large body of writings both original and in translation."206 The development of Castilian to a language fully operative in all fields of knowledge was achieved partly by translation, mainly from Latin, but also from Arabic.207 According to Don Juan Manuel's (d. 1348) *El libro dela caza*, the king "commissioned the translation of the entire sect of the Moors [*sic*] so that, in this way, the errors into which their false prophet Muḥammad pushed them and which they still adhere to today, would appear."<sup>208</sup> The universalist Christian approach to Islam which becomes apparent in this project of translating Islamic religious texts also led to the production of other documents of linguistic interest, most notably linguistic manuals serving the aim of proselytizing among the newly subjected Muslim populations. Ramón Martí's *Vocabulista in Arabico*, written around 1275, provides an example of a Latin-Arabic and Arabic-Latin dictionary that could be used to spread the Christian faith among Muslims. Given its lexical breadth, however, it was certainly not confined to this function.209 A more pertinent example is Pedro de Alcalá's *Arte para ligera mente saber la lengua arauiga*, an introduction to the Arabic dialect of Granada published in 1505, i.e. three years after the Castilian monarch had obliged the Muslims of the kingdom to either depart or to convert to Christianity. Pedro de Alcalá states in the prologue that "the time of fulfilment or the fulfilment of time has come in which it pleased the Sovereign Piety to extricate this newly converted people from darkness."

<sup>205</sup> *Siete Partidas del rey Don Alfonso el Sabio*, ed. La Real Academia de Historia, 3 vols. (Madrid: Real Academia, 1807), vol. 3, partida quinta, capitulo XI, ley I, 255.

<sup>206</sup> L. P. Harvey, "The Alfonsine School of Translators: Translations from Arabic into Castilian Produced under the Patronage of Alfonso the Wise of Castile (1221– 1252–1284)," *Journal of the Royal Asiatic Society of Great Britain and Ireland* 1 (1977), 111.

<sup>207</sup> Lloyd Kasten, "Alfonso el Sabio and the Thirteenth-Century Spanish Language," in *Emperor of Culture*, ed. Robert I. Burns (Philadelphia: University of Pennsylvania Press, 1990), 33–45; María Ángeles Gallego García, "The Languages of Medieval Iberia and their Religious Dimension," *Medieval Encounters* 9, no. 1 (2003), 111–113.

<sup>208</sup> Don Juan Manuel, *El libro dela caza*, ed. Georg Baist (Halle: Max Niemeyer, 1880), 1–2: "Otrosi fizo trasladar toda la secta delos moros, porque paresçiesse por ella los errores enque mahomad el su falso propheta les puso e enque ellos estan oy en dia."

<sup>209</sup> *Vocabulista in Arabico*, ed. Celestino Schiaparelli (Florence: Tipografia dei successori Le Monnier, 1871).

His manual contains an introduction to the Arabic alphabet and tables of Arabic verbs, as well as a dictionary restricted to nouns.210

While such documents seem to imply that studying Arabic was on the rise among speakers of Latin and Romance, other documents show that the *Reconquista* also destroyed infrastructures of Arabic learning,211 paradoxically creating further variants of Romance-Arabic entanglement. In a Castilian text interspersed with Arabic religious terms, the "wise and honoured *muftī* and *faqīh* of the *aljama* of the Muslims of the noble and loyal city of Segovia," a man known under the various names ʿĪsā b. Ǧābir, Iça Jedih, Yça Gidelli, etc., produced a Castilian summary of Islamic norms around 1462. The reason he gives is as follows:

"Because the Muslims of Castile, given that they suffer from great subjection, heavy tribute as well as many toils and labours, have declined in their wealth and lost their schools of Arabic [. . .] very many of my friends [. . .] begged me to compile in Romance a short text on our Holy Law and Sunna, of all that which every good Muslim ought to know and to follow [. . .]."<sup>212</sup>

Considering that he produced a Castilian translation of the Qurʾān in Burgundy under the supervision of Juan de Segovia, to whom he even wrote a Latin letter,213 ʿĪsā b. Ǧābir certainly occupies a special place in the history of Latin-Arabic entanglement. However, ʿĪsā b. Ǧābir was not the only one to use mixed forms of Arabic and Romance to ensure the survival of basic Islamic ideas among the (crypto-)Muslim population of the late medieval and early modern Iberian Peninsula. Luís F. Bernabé Pons cites texts which explain that they contain extracts of the Qurʾān and the *šarīʿa* "in the letters of the Christians" (*en letra de cristianos*) to enable those who "cannot read the letters of the Muslims" (*no [saben] la letra de los muslimes*),

<sup>210</sup> Petrus Hispanus, *De lingua arabica libri duo*, ed. Paul de Lagarde (Göttingen: Diederich, 1883), 1 (prologue): "Venido el tiempo del complimiento o el complimiento del tiempo, enel qual plugo ala soberana piedad sacar a esta gente nueuamente conuertida delas tiniebras"; also see p. 3 (alphabet), 75 (table of verbs), 89 (glossary of nouns).

<sup>211</sup> For regional variants see Brian A. Catlos, *Muslims of Medieval Latin Christendom, c.1050–1614* (Cambridge: Cambridge University Press, 2014), 249–251, 448–451.

<sup>212</sup> Içe de Gebir, *Suma de los principales mandamientos y devedamientos de la Lez y Çunna*, ed. Pascual de Gayangos, *Tratados de legislación musulmana* (Madrid: Real Academia, 1873), 247–249; re-edited in Gerard Wiegers, *Islamic Literature in Spanish and Aljamiado. Yça of Segovia (fl. 1450), His Antecedents and Successors*  (Leiden: Brill, 1994), 236–237: "Y porque los moros de Castilla con gran sujeçion y muchos tributos y grandes fatigas y trabachos an descaeçido de sus riquezas y an perdido las escuelas y del arabigo, [. . .] muy muchos amigos míos [. . .] me rrogaron que en rromance yo quisiese copilar vna tan breue escriptura de nuestra sancta ley y açuna de todo aquello que todo buen moro deuia saver y vsar [. . .]."; Catlos, *Muslims of Latin Christendom*, 198.

<sup>213</sup> José Martínez Gázquez, "El Prólogo de Juan de Segobia al Corán (Qurʾān) trilingüe (1456)," *Mittellateinisches Jahrbuch* 38, no. 2 (2003), 394–410. See ʿĪsā b. Ǧābir's Latin letter, edited in Wiegers, *Islamic Literature*, 230–235.

to understand them.214 Other texts transmit their contents in the regional Ibero-Romance idiom, but in Arabic letters. This fusion of a Romance idiom with the Arabic alphabet, already known from the earlier Romance *ḫarǧa*s of Andalusī *muwaššaḥ*-poetry,215 came to stand at the basis of an entire corpus of literature known as Aljamiado. Produced until around the seventeenth century, it covers such diverse genres as Qurʾānic exegesis, Islamic juridical texts, and Islamic hagiography, but also notarial forms, contracts, and profane literature.216

Accessing Arabic and Aljamiado literature constituted a challenge for the Spanish Inquisition, intent on eradicating all traces of Islam after the forced conversion of the Muslims of Castile in 1502 and of Aragon in 1526, which seems to have been accompanied by the burning of Arabic books.217 Ana Labarta's study on the translators recruited by the inquisition in Valencia between 1565 and 1609 distinguishes between translators and interpreters with active and/or passive skills in written and/or oral Arabic. The case of Hiéronymo de Mur, a Jesuit involved in thirty-two trials between 1575 and 1601, presents us with an "interpreter of the Arabic language and a qualifier thereof" (*intérprete de la lengua arábiga y calificador della*). Among other things, his function was to distinguish between religious and profane texts. Although classified as "able to read, write, and speak the Arabic language and letters" (*por saber leer y screvir y ablar la lengua y letra arábiga*), an analysis of his translations shows that he was more competent in the Valencian dialect than in the written form of Arabic.<sup>218</sup>

The anti-Christian polemic and travel account of Aḥmad b. Qāsim al-Ḥaǧarī (d. after 1640), a Morisco who had left the Iberian Peninsula around 1599, i.e. around ten years before the official expulsion of the Moriscos in 1609, provides a good example of the linguistic make-up of an Iberian crypto-Muslim who successfully managed to flee and to establish himself in North Africa. Still in Spain, Aḥmad b. Qāsim hesitated to reveal his Arabic skills for fear of being burned, but also mentions that the reading of profane Arabic books was permitted to the inhabitants of Valencia and encountered two "Andalusian interpreters" who "had the excuse that they had learned Arabic in their youth, close to the Islamic period."219 Sent as an envoy to France by the Moroccan sultan Mawlāy Zīdān

<sup>214</sup> Luís F. Bernabé Pons, "Los manuscritos aljamiados como textos islámicos," in *Memoria de los moriscos. Escritos y relatos de una diáspora cultural* (Madrid: Sociedad Estatal de Conmemoraciones Culturales, 2010), 29.

<sup>215</sup> See section 2.3.3.

<sup>216</sup> Bernabé Pons, "Manuscritos aljamiados," 33.

<sup>217</sup> See José Martínez Gázquez, *The Attitude of the Medieval Latin Translators Towards the Arabic Sciences* (Florence: SISMEL, 2016), 177–180.

<sup>218</sup> Ana Labarta, "Notas sobre algunos traductores de árabe en la inquisición valenciana (1565–1609)," *Revista del Instituto Egipcio de Estudios Islámicos en Madrid* 21 (1981–1982), 106–108.

<sup>219</sup> Aḥmad b. Qāsim al-Ḥaǧarī, *Kitāb Nāṣir al-dīn ʿalā l-qawm al-kāfirīn / The Supporter of Religion Against the Infidels*, ed. and trans. Pieter S. van Koningsveld, Qasim al-Samarrai, and Gerard A. Wiegers, Fuentes Arábico-Hispanas 35, 2nd ed. (Madrid: CSIC, 2015), 23–24 (AR) / 87–89 (EN): "al-mutarǧimīn al-Andalus [*sic*] fa-kānū

(r. ca*.* 1012–1039/1603–1628), he sojourned in Holland around 1615, where he confided to Prince Maurice that he was able to speak Arabic, Castilian, and Portuguese, and to understand French.220 Aḥmad b. Qāsim's precarious youth in Catholic Spain, in combination with his settlement in North Africa and his travels to France, had certainly enhanced his linguistic versatility in Arabic and several Romance languages.221

#### 2.4.3 NORTH AFRICA

Iberian Muslims fleeing the *Reconquista* arrived in a North African linguistic landscape that still seems to have featured some linguistic remnants of the Roman past apart from Latin inscriptions on architectural remains from the Roman period. Ibn Ḫaldūn (d. 808/1406) claims that the inhabitants of Ifrīqiya believed that the "Franks" of the pre-Islamic period had buried treasures, to the effect that, in his time,

"Berber petitioners approach well-to-do people with papers that have torn margins and contain either non-Arab writing or what they claim to be the translation of a document written by the owner of buried treasures containing indications on them with regard to their location, hoping by this to receive their sustenance from them by means of what they spend on excavating."222

In addition, high and late medieval North Africa was home to large numbers of Romance-speakers from the northern shores of the Mediterranean. Throughout the medieval and early modern periods, Christian-European captives were brought to North Africa.223 In the twelfth century, the Almoravids deported entire Christian communities from the Iberian Peninsula

šuyūḫan wa-yastaʿḏirun bi-annahum taʿallamū l-qirāʾata l-ʿarabiyya fī ṣuġarihim bi-qurbi ʿahd islām [*sic*]."

<sup>220</sup> Aḥmad b. Qāsim, *Kitāb Nāṣir al-dīn*, ed. and trans. van Koningsveld et al., 226 (AR), 225 (EN).

<sup>221</sup> Gerard A. Wiegers, "A Life Between Europe and the Maghrib. The Writings and Travels of Aḥmad b. Qâsim ibn Aḥmad ibn al-faqîh Qāsim ibn al-shaykh al-Ḥajarî al-Andalusî (born ca. 977/1569–70)," in *The Middle East and Europe: Encounters and Exchanges*, ed. Geert J. van Gelder and Ed de Moor (Amsterdam: Atlanta, 1993), vol. 1, 87–115.

<sup>222</sup> Ibn Ḫaldūn, *Tārīḫ*, ed. Zakkār and Šaḥāda, vol. 1, 481: "wa-naǧid kaṯīran min ṭalabat al-Barbar [. . .] yataqarrabūn ilā ahl al-dunyā bi-l-awrāq al-mutaḫarrima l-ḥawāšī immā bi-ḫuṭūṭ ʿaǧamiyya aw bi-mā turǧima bi-zaʿmihim minhā min ḫuṭūt ahl al-dafāʾin bi-iʿṭāʾ al-amārāt ʿalayhā fī amākinihā yabtaġūn bi-ḏālik alrizq minhum bi-mā yabʿaṯūnahu ʿalā l-ḥafr [. . .]." Translation adapted from Ibn Khaldûn, *The Muqaddimah: An Introduction to History*, trans. Franz Rosenthal, ed. and abridged. N. J. Dawood (Princeton: Princeton University Press, 1989), chapter V,4, 302.

<sup>223</sup> See e.g. König, *Arabic-Islamic Views*, 60–61, 292; Wolfgang Kaiser, *Le commerce des captifs: les intermédiaires dans l'échange et le rachat des prisonniers en Méditerranée, XVe –XVIII<sup>e</sup> siècle* (Rome: École française de Rome, 2008).

to North Africa.224 From the twelfth century onwards, various expatriate professionals from the northern shores of the Mediterranean made a living in or at the margins of the Muslim societies of North Africa. These included European-Christian mercenaries working for Muslim overlords,225 and, of course, large merchant communities of mainly Aragonese and Italian origin. Soon headed by a consul, the latter organized their sojourn in North African countries within the physical and legal space of the urban *fondaco*. <sup>226</sup> They employed European-Christian notaries as well as Jewish, Christian, and Muslim interpreters to facilitate their communication with the local Muslim authorities,227 while mainly Franciscan and Dominican chaplains administered to their religious needs and ensured relations with the papacy.<sup>228</sup> Vice versa, the Latin documentation also attests to the presence of "Saracens," i.e. Muslims, in the economic hubs of the northern Mediterranean, e.g. in Venice.229

Commercial diplomacy gave rise to milieus of Latin-Arabic and Romance-Arabic interaction that flourished between the late twelfth and the beginning of the sixteenth century when the Ottoman expansion and the increase in piracy considerably modified the conditions of trade and diplomatic interaction in the western Mediterranean.230 Late medieval Arabic-Islamic texts from North Africa provide only a few glimpses into the linguistic effects of this European-Christian presence in North Africa, e.g. Ibn Ḫaldūn citing a late twelfth-century magic spell that transcribes Romance


words such as "Alfonso," "Barcelona," and "French king," subsuming this under the category "Rūm."231 Fortunately, however, the archives of the Crown of Aragon and the Italian maritime republics of Genoa, Pisa, and Venice have preserved dozens of letters and treaties in Arabic, Latin, Catalan, and several variants of Italo-Romance that allow the linguistic dimensions of this intensive exchange to be understood.

A large number of bilingual Latin-Arabic and Romance-Arabic letters and treaties provide the most important evidence. They attest to the flow of Latin and Romance loanwords into Arabic, most of them administrative terms serving the function of identifying forms of authority particular to the republican systems of Italy's maritime cities, as can be shown in Table 2.3.


Table 2.3: Examples of Latin and Romance Loanwords in Arabic Documents of Commercial Diplomacy (twelfth–fifteenth centuries).232

<sup>231</sup> Ibn Ḫaldūn, *Tārīḫ*, ed. Zakkār and Šaḥāda, vol. 1, 669: "wa-qsim ʿalā l-qaṭri wa-kun mutafaqqidan / fa-in šiʾta li-l-Rūmi fa-bi-l-ḥarri šakkilan / fa-Fanšun wa-Baršanūnu l-rāʾu ḥarfuhum / wa-Ifransuhum dālun wa-bi-l-ṭāʾi kammilan."

<sup>232</sup> Bibliographical references in the table refer to Roman document numbers and Arabic page numbers in Michele Amari, *I Diplomi Arabi del R. Archivio Fiorentino* [abbr. "DA"] (Florence: Felice Le Monnier, 1863); Michele Amari, *Nuovi ricordi arabici su la storia di Genova* [abbr. "NR"] (Genoa: Tipografia del R. Istituto sordomuti, 1873); as well as to John Wansbrough, "A Mamluk Letter of 877/1473," *Bulletin of the School of Oriental and African Studies* 24, no. 2 (1961), 200–213.

Moreover, these documents often provide the earliest testimony for the introduction of Arabic loanwords into medieval Latin and Romance languages. Table 2.4 shows that these are often terms for hitherto unknown goods and products, as well as North African forms of administration facilitating trans-Mediterranean commerce.


Table 2.4: Examples of Arabic Loanwords in Latin and Romance Documents of Commercial Diplomacy (twelfth–fifteenth centuries).233

In addition, these documents often mention interpreters and describe their procedure and techniques of translating the respective texts in detail. Thus, a treaty of 1264 was translated "de lingua arabicha [*sic*] in latina,"234 whereas the Arabic text of a Genoese-Mamluk treaty of 1290 was furnished with a "Frankish" interlineary translation reviewed by two Mamluk interpreters.235 The text of a treaty of 1313 was read aloud in Arabic, then translated word for word (*de verbo ad verbum*) to Catalan, and then written out in Latin.236 An Arabic letter written in 1452 was translated "de moriscu in latinu" in Genoa, the translation reviewed by a second translator.<sup>237</sup>

<sup>233</sup> Dates based on Raja Tazi, *Arabismen im Deutschen. Lexikalische Transferenzen vom Arabischen ins Deutsche* (Berlin: de Gruyter, 1998). For further lexical material see John Derek Latham, "Arabic into Medieval Latin," *Journal for Semitic Studies* 17 (1972), 21, no. 1–2 (1976), 24 (1989), 30–67, 120–137, 459–469.

<sup>234</sup> Mas Latrie, *Traités de paix*, vol. 1, 47.

<sup>235</sup> Amari, *Nuovi ricordi*, 16–17 (AR), 63–65 (IT).

<sup>236</sup> Mas Latrie, *Traités de paix*, vol. 1, 189–192.

<sup>237</sup> Mas Latrie, *Traités de paix*, vol. 1, 147.

Taken together, the many bilingual Latin/Romance-Arabic documents produced between North Africa and the northern shores of the Mediterranean prove that professional linguistic mediation involving Latin, various Romance languages, and Arabic became a regular feature of commercial diplomacy in the western Mediterranean of the late twelfth to fifteenth centuries. Many bilateral treaties condition the legal validity of commercial transactions on the use of officially accredited interpreters, impose various rules on the interpreters themselves, and regulate the latter's payment.<sup>238</sup> In official letters, interpreters feature as agents who try to de-escalate tensions,<sup>239</sup> or even advertise their skills and loyalty to potential protectors and employers.240 In one case, we even find a letter that is written in a rather unusual form of "Aljamiado." In the letter, dated 30 Ramaḍān 767 or June 10, 1366, the Hafsid ruler Abū l-ʿAbbās Aḥmad b. Abī ʿAbd Allāh Muḥammad from Bejaïa addresses Giovanni dell'Agnello de'Conti, the doge of Pisa, *capitano generale*, and governor of Lucca, in a letter written in Arabic script, but in the Pisan variant of Italo-Romance.<sup>241</sup>

Juxtaposing the extant Latin, Romance, and Arabic versions of specific treaties and letters allows us to evaluate the quality of the respective translations. This includes observing procedures and techniques of translation as well as the liberties occasionally taken or not taken by the translators.

In letters, transcriptions of honorary titles allow us to understand the extent to which the political symbolism of the respective addressee was understood by the sender and vice versa. A letter by the ruler of Tunis ʿAbd Allāh b. ʿAbd al-ʿAzīz b. ʿAbd al-Ḥaqq b. Abī Ḫurasān addressed "to the exalted and most noble archbishop, the archbishop of Pisa" (*ilā l-ark al-ǧalīl al-akram ark Bīsha*), dating from July 10, 1157, extensively describes the former's victory over the Almohads in strong religious language full of typical Islamic idiomatic expressions. The Latin version enlarges and thus corrects the archbishop's title, thus addressing "the archbishop of the Pisans, the primate and vicar of Corsica and Sardinia" (*Archiepiscopo Pisanorum [. . .], Corsice et Sardinie primati atque vicario*). It eliminates the rhythmic rhyme prose of the Arabic original as well as all Islamic idiomatic expressions,

<sup>238</sup> König, "Übersetzungskontrolle," 480.

<sup>239</sup> Amari, *I Diplomi*, 39: In this Arabic letter, dated 597/1201, ʿAbd al-Raḥmān b. Abī l-Ṭāhir, overseer of the dīwān in Tunis (*al-nāẓir bi-dīwān Tūnis*), reports to the archbishop and the commune of Pisa that Pisan scribes residing in Tunis harshly criticized the captains and crews of two Pisan ships which had attacked a Muslim ship in the harbour of Tunis.

<sup>240</sup> Amari, *I Diplomi*, 75–76: In this Arabic letter, dated 604/1207, Aḥmad b. Tamīm from Bejaïa asks the Pisan Lamberto del Vernaccio for a letter of reference to the dīwān.

<sup>241</sup> Amari, *I Diplomi*, doc. XXXIII, 119–122, here: 119: [transliteration of the Arabic text] "Inm ddī kī bīṭūs mzrkrdiyyūs daura bir suwā msāǧ Mḥmd lkrsiyyūs wa-brṭūṭ lsuwā knbāni al-suwā skāš ǧinirār salūṭām / da ʿAbd-Allāh lunbaradūr subra lasuwā sarayin riyā [. . .]."; transcribed in Italian by Amari as: "In nome di Dio che, pietoso, misericordiosi, darà pel suo messagio Maometto il grazioso, e per tutti i suoi compagni e i suoi seguaci, general salute. Da Abd-Allah, l'imperatore sopra [la sua] saracineria [. . .]." This Arabic-Pisan Aljamiado version is preceded by the Arabic original, doc. XXXII, 115–118.

but retains the idea that the ruler of Tunis overcame the Almohads "by the grace of the Creator" (*gratia Creatoris*). As opposed to the Arabic original, the Latin version additionally defines the defeated enemy as belonging to the tribal confederation of the Maṣmūda. Moreover, it legitimizes the report on this victory by stating that the author of the letter deemed it proper to inform his "true friends, whom I treasure more than anyone else in the Christian sphere" (*veris amicis meis quos pre ceteris mundi christiani diligo*) about his current state of affairs.<sup>242</sup>

A Pisan letter to the Almohad ruler Abū Yaʿqūb Yūsuf, dated April 23, 1181, features a Latin transcription that reflects an understanding of Almohad political ideology. The addressee, Abū Yaʿqūb Yūsuf (r. 558– 580/1163–1184), had succeeded ʿAbd al-Muʿmin (d. 558/1163), who had been designated as caliph by the Almohads' founding figure, the *mahdī* Ibn Tūmart (d. 524/1130). Respecting these political-theological claims, the Pisan letter addresses Abū Yaʿqūb Yūsuf as "Commander of the faithful by the grace of God" (*Dei gratia, elmire Elmomini*), as "son of the Commander of the faithful" (*filio domini elmire Elmomini*), and as "venerable priest of the *Mahdī*" (*venerabili sacerdoti Elmachadin*). Both the Latin original and the Arabic version pursue the objective of demanding the liberation of Pisan citizens held captive in Almohad territory, but differ decidedly in tone. In the Latin version, the archbishop of Pisa "humbly pleads with your Majesty" (*majestam vestram humiliter exoramus*) for the captives' liberation. The Arabic version is considerably enriched with Islamic formulae describing God and expresses the archbishop's humility within an honorary address to "our kind lord, the Commander of the faithful, may God support him to the effect that his high command be executed" (*faḍl sayyidinā amīr al-muʾminīn ayyadahu Allāh li-yunfaḏ amruhu l-ʿalī*). As opposed to the Latin original, however, the Arabic version does not merely plead, but firmly insists on the Almohads' legal obligation to guarantee the security of Pisan citizens and their property, according to "the binding command and the concluded treaty to be implemented" (*al-amr al-multazim wa-l-ʿahd al-nāfiḏ al-muḥkam*).243

Such examples of considerable divergences show that translators obviously adapted the language, style, and even the tone of a letter to the addressee, probably with the aim of making the letter more effective. One would expect that the different versions of legally binding bilateral treaties would not feature comparable distortions—especially considering that many treaties describe mechanisms of control that aim at ensuring a precise rendering of the treaty's wording in both languages.244 A general survey of existing bilingual treaties conveys the impression that simple and uncontroversial issues were formulated as closely as possible in both languages. In some cases, however, slight variations in the Latin and

<sup>242</sup> Amari, *I Diplomi*, doc. VI, 255 (LAT) = doc. I, 1 (AR).

<sup>243</sup> Amari, *I Diplomi*, doc. XIII, 269 (LAT) = doc. II, 7–9 (AR).

<sup>244</sup> See König, "Übersetzungskontrolle," 480–481.

Arabic text seem to reflect disagreement on certain rights and obligations. A treaty concluded June 1, 1181 between the lord of Mallorca, Abū Ibrāhīm Isḥāq b. Muḥammad, and the Genoese ambassador, Rodoanus de Moro, for example, deals with the rights and obligations of both parties in the case of a Genoese shipwreck off the Mallorcan coast. The Arabic version of the treaty guarantees the Genoese possession of all flotsam and jetsam, i.e. goods floating to the coast, and allows the Genoese to recover their sunken goods, i.e. lagan and derelict, with the help of rented hands.<sup>245</sup> The Latin version of the treaty also guarantees the Genoese possession of flotsam and jetsam, and also discusses the issue of paid help to recover lagan or derelict. However, its formulations oblige the Mallorcans not only to respect Genoese property rights, but also to actively save shipwrecked persons and their goods.246 The divergent translations thus show that the Muslims of Mallorca were not prepared to render the exact same services demanded by the Genoese.

In view of this evidence, it is clear that different degrees of Latin-Arabic and Romance-Arabic entanglement were a recurring feature of communication in the western Mediterranean, in spite of the fact that North African Arabic-Islamic literature rarely comments upon these phenomena.<sup>247</sup> North African exposure to Romance languages was probably reinforced by the emigration of Muslims and Moriscos from the Iberian Peninsula to North Africa in the late medieval and early modern period. According to Ibn Ḫaldūn, this emigration also had linguistic effects in that Andalusian emigrants imposed their (Arabic) writing style on North African chanceries, e.g. in al-Qayrawān, al-Mahdiyya, and Tūnis.248 The linguistic skills displayed by Aḥmad b. Qāsim al-Ḥaǧarī, the Morisco emigrant and later Moroccan envoy to France mentioned at the end of section 2.4.2, implies that the influx of Andalusian Muslims also increased the number of Muslim Romance-speakers in North Africa of the same period.

Against this backdrop, it is not surprising that the so-called "Lingua franca" became an important means of communication in the western Mediterranean. Attested among Christians and Muslims until the nineteenth century, it represents a pidgin or vehicular language that is made

<sup>245</sup> Antoine-Isaac Silvestre de Sacy, "Pièces diplomatiques tirées des archives de la république de Gênes," in *Notices et extraits des manuscrits de la bibliothèque du roi et autres bibliothèques*, vol. 11 (Paris: Imprimerie royale, 1827), 9 (AR): "an lā yuʿtaraḍū fī šayʾ mimmā aḫraǧahu l-baḥr wa-ramiya bihi ilā l-barr wa-matā ḏahabū ilā l-kirāʾ ʿalā iḫrāǧ mā fī l-baḥr fa-ḏālika mubāḥ lahum in šāʾ Allāh taʿālā."

<sup>246</sup> Mas Latrie, *Traités de paix*, vol. 1, 111–112 (LAT): "quod homines sui eos salvare naufragos et eorum bona nec inde auferre vel minuere, sed quicquid inde habere possent restituere; excepto si de pecunia qui jacet in fundo recuperanda ullam inde conventionem cum Sarracenis [et] Christianis fact[a esset] que conventio, si intercederet, firma sicut equum est servetur."

<sup>247</sup> Mohamed Tahar Mansouri, "Les milieux marchands européens et la langue arabe au Maghreb médiéval," in *Trames de langues: usages et métissages linguistiques dans l'histoire du Maghreb*, ed. Jocelyne Dakhlia (Paris, Tunis: Maisonneuve et Larose, Institut de recherche sur le Maghreb contemporain, 2004), 283.

<sup>248</sup> Ibn Ḫaldūn, *Tārīḫ*, ed. Zakkār and Šaḥāda, vol. 1, 528–529.

up of various linguistic elements joined together in a flexible mixture. Scholarship on the Lingua franca disagrees on many points—e.g. if the Lingua franca constitutes a proper language or only a form of "broken" Italian, French, Catalan, or Castilian; if it dates from the high and late medieval or from the early modern period; and if its origins and strongest diffusion are to be sought in the western or in the eastern Mediterranean. Regardless of their respective stance on these issues, scholars agree that exchange in the early modern Mediterranean was facilitated by a flexible linguistic medium made up—in its majority—of Romance, but also of Arabic, Turkish, Greek, and other elements. <sup>249</sup>

#### 2.4.4 THE EASTERN MEDITERRANEAN

The scholarly dispute presented in the preceding section about the origins and diffusion of the Lingua franca in the western or eastern Mediterranean shows that it is also necessary to consider the eastern Mediterranean as a region relevant to the history of Latin/Romance-Arabic entanglement. European-Christian expansion of the high and late medieval period had a strong linguistic impact in that it led to a diffusion of Romance languages as well as to the production of Latin and Romance documents in that region. Different variants of French, Italian, and Catalan were imported into the eastern Mediterranean by crusaders, various Italian merchants, and mercenary groups such as the so-called Catalan Company, as well as the military orders—all of them taking control temporarily over territories, trade, and shipping routes in the Syrian Levant, the Peloponnese, various eastern Mediterranean islands, Byzantium, and the Black Sea from the First and, particularly, from the Fourth Crusade onwards.250 The *Chronicle of Morea*, which recounts the establishment of the duchy of Achaia (1204–1432) on the Peloponnese, provides an excellent example in that the existence of Middle Greek, French, Italian, and Aragonese versions clearly attests to the linguistic Romanization brought about by European-Christian expansion.251 It is against this backdrop that some scholars also regard

<sup>249</sup> Compare the positions of Jocelyne Dakhlia, *Lingua franca: Histoire d'une langue métisse en Méditerranée* (Arles: Actes Sud, 2008), 15–16, 21, 25, 42–46, 85–91; with Cyril Aslanov, "Débat sur l'ouvrage de Jocelyne Dakhlia, Lingua franca: histoire d'une langue métisse en Méditerranée," *Langage et société* 134/4 (2010), 103–113. Both continue discussions summarized a few years earlier by Laura Minervini, "La Lingua franca mediterranea. Plurilinguismo, mistilinguismo, pidginizzazione sulle coste del Mediterraneo tra tardo medioevo e prima età moderna," *Medioevo romanzo* 20, no. 2 (1996), 231–301. On scholars dealing with the same issue in the eastern Mediterranean, see the following section 2.4.4.

<sup>250</sup> Cyril Aslanov, *Le français au Levant jadis et naguère. À la recherche d'une langue perdue* (Paris: Honoré Champion, 2006), 75–76.

<sup>251</sup> W. J. Aerts, "The Chronicle of Morea as a Mirror of a Crusader State at Work," in *East and West in the Crusader States. Context—Contacts—Confrontations*, vol. 2, ed. Krijnie Cigaar and Herman Teule (Leuven: Peeters, 1999), 153–162.

the eastern Mediterranean as an alternative breeding ground for the early modern Lingua franca.<sup>252</sup>

Multilingual lifeworlds in the eastern Mediterranean involved various forms of Romance-Arabic entanglement. One may surmise that much linguistic mediation was effected by local Christian groups who were invited to settle in Jerusalem after 1101,<sup>253</sup> intermarried with crusader families,<sup>254</sup> and seemingly fulfilled various administrative functions within crusader administrations, e.g. in the commercial tribunals (*cour de la Fonde*),255 or as scribes in tax-stations.256 Although communication problems are attested, e.g. between the Syrian-Arab Muslim noble Usāma b. Munqiḏ (d. 584/1188) and a Frankish woman,257 or between Dominican monks and *ʿulamāʾ* at the court of the Ayyubid governor al-Manṣūr Ibrāhīm in Ḥimṣ in 1245,258 we also find miscellaneous evidence for language learning and multilingualism connecting Muslims and European Christians. Intermarriage, shorter or longer bouts of captivity, and defections to the other camp, as well as occasional conversions, seem to have opened up possibilities for native speakers of Arabic or Romance to learn the other language.259 The attested crusaders' employment of Oriental Christian, Jewish, and Muslim physicians may serve as an


indicator of various forms of cross-lingual communication in the crusader milieu.260 We even possess anecdotal evidence describing a form of tandem language learning in which the son of a "pagan" noble is sent to the court of the king of Jerusalem to learn "the Gallic language" (*Gallicum*), while the king's son is sent to the noble's residence "to learn the Saracen language" (*ad discendum idioma Sarracenicum*).<sup>261</sup> In addition, we find references to several noble crusaders able to speak Arabic,262 to "Saracen scribes" in the service of the Kingdom of Jerusalem or the Order of the Temple,<sup>263</sup> to interpreters ensuring communication between Muslim rulers such as Saladin and foreign kings such as Richard the Lionheart,264 as well as to the linguistic challenges of setting up a treaty in Arabic that would fulfil the stylistic requirements of trained Arabic scribes and simultaneously conform to a previous oral, and thus linguistically less-sophisticated agreement with "the Franks."265 In view of the relevance of Romance languages in the late medieval eastern Mediterranean, autochthonous groups seem to have made efforts to acquire some working knowledge of these languages. Evidence is provided by two documents from Egypt—an Arabic-Old French glossary in Coptic letters that seems to have been produced for Coptic travellers visiting Acre in the course


<sup>260</sup> "Constitutiones Nicosiensis," in *Sacrorum conciliorum nova et amplissima collectio*, ed. Joannes Dominicus Mansi, vol. 26 (Venice: Antonius Zatta, 1784), 328–329, prohibit the employment of Jewish and Muslim physicians. See also Johannes Pahlitzsch, "Ärzte ohne Grenzen. Melkitische, jüdische und samaritanische Ärzte im Ägypten und Syrien der Kreuzzüge," in *Gesundheit—Krankheit. Kulturtransfer medizinischen Wissens von der Spätantike bis in die Frühe Neuzeit*, ed. Florian Steger and Kay Peter Jankrift (Cologne: Böhlau, 2004), 101–119.

of the thirteenth century,266 and the Geniza-fragment of an Arabic-Castilian glossary in Hebrew letters, compiled between 1424 and 1430.<sup>267</sup>

The evidence suggests that, in the eastern Mediterranean of crusader times, Romance languages played a much more preponderant role than Latin, even in the written sphere: the legal code known as the *Assises de Jérusalem*, 268 or the account of the fall of Acre by the so-called Templar of Tyre that is included in the so-called *Gestes des Chiprois*, were written in French.269 Latin text production in the eastern Mediterranean seems to have been limited to the ecclesiastical sphere, to the acts of Church councils such as the councils of Nablus (1120) or Nicosia (mid-thirteenth century), or to such authors as William of Tyre (d. 1184) and William of Tripolis (mid-thirteenth century). In the latter's treatises on Islam one finds several Arabic words, including a Latin transcription and translation of the *šahāda*, the Muslim creed.270 In addition, some translation activity also seems to have taken place in crusader principalities such as Antioch, attested by the trilingual glossary of medicaments taken from the pharmacological treatise of Dioscorides, produced by Stephen of Antioch (or Pisa) in the second quarter of the twelfth century.271

All in all, it is difficult to evaluate the extent to which Latin/Romance-Arabic entanglement in the eastern Mediterranean had a long-term impact on the Romance languages used in Europe and Arabic as used in the Middle East. Older scholars such as Arnold Steiger supposed that the crusading movement contributed greatly to enriching European languages with Arabic loanwords, but admitted that is was difficult to distinguish between Arabic loanwords of western and of eastern Mediterranean origin.272 Cyril Aslanov's judgement that some Arabic words were adopted in what he calls "le français d'Outremer," and, via this variant of French, survived in modern French, seems more modest and, ultimately, more realistic. Vice versa, the Arabic-Islamic documentation of the crusading period only contains a few

<sup>266</sup> Cyril Aslanov, "Languages in Contact in the Latin East: Acre and Cyprus," *Crusades* 1 (2002), 157–158.

<sup>267</sup> Hayim Y. Sheynin, "Genizah Fragments of an Unknown Arabic-Castilian Glossary," *The Jewish Quarterly Review, New Series* 71, no. 3 (1981), 151–166.

<sup>268</sup> *Assises de Jérusalem ou Recueil des ouvrages de jurisprudence composés pendant le XIIIe siècle dans les royaumes de Jérusalem et de Chypre*, ed. Auguste-Arthur Beugnot, 2 vols. (Paris: Imprimerie royale, 1843).

<sup>269</sup> "Chronique du Templier de Tyre (1242–1309)," in *Gestes des Chiprois*, ed. Gaston Raynaud (Geneva: Jules-Guillaume Fick, 1887), 139–334.

<sup>270</sup> Wilhelm von Tripolis [Guilelmus Tripolitanus], *Notitia de Machometo et de libro legis qui dicitur alcoran et de continentia eus et quid dicat de fide domini nostri Iesu Christi*, ed. and trans. Peter Engels (Würzburg: Echter, 1992), cap. 3, 204–205: "The way, by which somebody becomes and is acknowledged a Saracen, is by declaring in any kind of manner: 'lā ilāha illā llāh, Muḥammad rasūl Allāh', which means: 'there is no God but God and Muḥammad is the messenger of God.'" (*Forma vero talis est, per quam quis iudicatur et efficitur Sarracenus, quocumque modo eam proferat: Le Ellech ella Alla Machomet resol Alla, quod est: Non est deus nisi Deus et Machometus Dei nuntius.*)

<sup>271</sup> Burnett, "Antioch as a Link," 38–39.

<sup>272</sup> Arnold Steiger, "Aufmarschstraßen des morgenländischen Sprachgutes," *Vox Romanica* 10 (1948–1949), 42.

loanwords of Latin or Romance origin. Apart from the transcribed titles of European-Christian rulers mentioned at the beginning of this section, the unparalleled Arabic transcription of the French term "bourgeois," i.e. "burǧāsī," as found in the memoirs of Usāma b. Munqiḏ, has to be counted among the most spectacular finds.273 In spite of the fact that the eastern Mediterranean certainly became a sphere of Latin-Arabic and, in particular, Romance-Arabic entanglement between the twelfth and fifteenth centuries, the area's contribution as a mediator of words and texts seems less important than that of the western Mediterranean.

#### 2.4.5 ANNEXATION OF A SCIENTIFIC HERITAGE? ARABIC–LATIN TRANSLATIONS (TWELFTH–SIXTEENTH CENTURIES)

The age of European-Christian expansionism into the Muslim Mediterranean closely correlates with the age of Arabic-Latin translation. It is true that the earliest attested efforts to appropriate knowledge available in the Arabic-Islamic sphere lead back to a period preceding the Norman conquest of Sicily, the first intensive phase of the so-called *Reconquista*, and the Crusades. Gerbert of Aurillac, usually the first Latin-Christian scholar associated with "Arabic science," acquired his knowledge in tenth-century Catalonia—itself a Carolingian zone of expansion into territory previously held by Muslims.274 Constantine the African, said to have been a North African who decided to translate Arabic books to Latin when he noticed the dearth of such material in Italy of the mid-eleventh century, can certainly not be regarded as an exponent of European-Christian expansionism.275 It is undeniable, however, that the greatest quantity of Arabic-Latin translations was produced at the height of European-Christian expansion into regions that had previously been under Muslim control—i.e. the Iberian Peninsula, Sicily, and southern Italy, as well as, to a much more limited degree, the crusader states.

In the very early stage, medical issues seem to have aroused most interest, as is attested by the translations of Constantine the African in Salerno of the late eleventh and Stephen of Pisa in Antioch in the second quarter of the twelfth century. From the twelfth century onwards, translators on the Iberian Peninsula, particularly in Toledo, began to take the lead and to enlarge the thematic scope of translated texts.276 At the beginning of the twelfth

<sup>273</sup> Usāma b. Munqiḏ, *Kitāb al-Iʿtibār*, ed. Hitti, 141; Aslanov, *Français*, 42–43.

<sup>274</sup> Marco Zuccato, "Gerbert of Aurillac and a Tenth-Century Jewish Channel for the Transmission of Arabic Science to the West," *Speculum* 80 (2005), 742–763.

<sup>275</sup> Kreutz, "Ehrenrettung," 40–41; Raphaela Veit, "Quellenkundliches zu Leben und Werk von Constantinus Africanus," *Deutsches Archiv für Erforschung des Mittelalters* 59 (2003), 121–152.

<sup>276</sup> Charles Burnett, "Translation from Arabic to Latin in the Middle Ages," in *Übersetzung: Ein internationales Handbuch zur Übersetzungsforschung*, ed. Harald Kittel (Berlin: de Gruyter, 2007), vol. II/2, 1220–1231. Also see the old, but still highly valuable overview by Charles Homer Haskins, *Studies in the History of Mediaeval Science* (Cambridge: Harvard University Press, 1924), 3–140, 242–271;

century, Ibn ʿAbdūn of Seville advised the readers of his manual for market inspectors not to sell any scientific books to Jews and Christians, since they would translate them and then ascribe the authorship to one of their co-religionists.<sup>277</sup> Independent scholars such as Adelard of Bath, translator of Euclid from Arabic to Latin,<sup>278</sup> certainly valued the possibility of citing new Arab authorities in an intellectual landscape that only accepted knowledge invested with the prestige of past authorities.279 While rather independent translators such as Hermann of Carinthia (d. ca. 1155) and Robert of Ketton were enthusiastically extracting knowledge "from the depths of the treasuries of the Arabs" (*ex intimis Arabum thesauris*),280 the influential abbot of the monastery of Cluny, Peter the Venerable (d. 1156), grasped the chance of drawing on these two translators to "transfer the writings on the origins, life and teaching of this damnable person [i.e. Muḥammad] as well as his legal book, the Qurʾān, from Arabic to Latin," a task completed around 1143.281 The lasting impact of this first Latin translation of Arabic texts defining and describing Islamic dogma should not obscure the fact that most translators did not pursue a religious or polemic agenda. According to the obituary composed by his students, Gerard of Cremona (d. 1187), probably the most prolific translator from Arabic to Latin, allegedly went to the Iberian Peninsula because he was searching for Ptolemy's "Μαθηματικὴ Σύνταξις," generally known under the Latin version of its Arabic title "Almagest." This book "he could not find at all among the Latins." In Toledo, however,

"seeing the abundance of books in Arabic on every subject, which he could not find at all among the Latins, and regretting the poverty of the Latins in these things, which he knew well, he learned the Arabic language in his desire to translate [. . .]."282

Dag Nikolaus Hasse, "The Social Conditions of the Arabic–(Hebrew–)Latin Translation Movements in Medieval Spain and in the Renaissance," in: *Wissen über Grenzen. Arabisches Wissen und lateinisches Mittelalter,* ed. Andreas Speer and Lydia Wegener. (Berlin: de Gruyter, 2006), 3–21.


In the thirteenth century, royal courts in Castile and southern Italy became important centres of translation activity—be it from Arabic to Castilian under the patronage of Alfonso X (r. 1252–1284),283 or from Arabic to Latin under the patronage of Frederick II, Manfred, and later Charles I of Anjou.<sup>284</sup> However, royal courts did not monopolize translation activities. We find a large number of translators outside the royal sphere, e.g. in southern France, Italian cities, or—in the late fifteenth and early sixteenth centuries—in the orbit of the Venetian consulate of Damascus.<sup>285</sup>

Muslims only rarely participated in these translation activities, even though they are occasionally attested, e.g. in connection with Peter the Venerable's translation of the Qurʾān in around 1143,286 the Arabic-Portuguese translation of al-Rāzī's tenth-century history *Aḫbār mulūk al-Andalus*  in the thirteenth century,287 and a later Castilian translation of the Qurʾān commissioned by Juan de Segovia around 1454.288 Many Latin-Christian translators seem to have drawn on Arabic- and Romance-speaking Christians and Jews, and, especially in the earlier period, to have produced the final Latin text with their help by translating word for word from Arabic via Romance to Latin.<sup>289</sup> This does not preclude that many translators also demonstrated independence in their approach to the text and language of the original—abbreviating the original and publishing works based on available translations such as Michael Scot,290 being native speakers of


Toletum perrexit, vbi librorum cuisque facultatis habundantiam in arabico cernens et latinorum penurie de ipsis, quam nouerat, miserat, amore transferendi linguam edidicit arabicam [. . .]." Translation adapted from Michael McVaughan, in Edward Grant, *A Source Book in Medieval Science*, vol. 1 (Cambridge, MA: Harvard University Press, 1974), 35–39.

Arabic themselves, such as Theodore of Antioch,291 or engaging explicitly and didactically with Arabic grammar, such as William of Luna.292 As we have seen in sections 2.4.1 and 2.4.2, the Arabic-speaking population of southern Italy and the Iberian Peninsula continuously decreased from the thirteenth century onwards. Emigration, deportation, or linguistic assimilation made it increasingly necessary to draw on the linguistic skills of recent immigrants, such as Jews from the Almohad realm in Sicily,<sup>293</sup> or on people who had acquired Arabic as a second language, such as Jesuits working for the Spanish Inquisition.<sup>294</sup> In consequence, it became expedient to produce Latin and Romance works that would serve as aids to study Arabic, a necessity that contributed to the gradual academic institutionalization of Arabic studies in Christian Europe that will be addressed in more detail in section 2.5.

The Arabic-Latin translation movement of the twelfth to the sixteenth centuries thus featured various intellectual milieus of Latin-Arabic entanglement which differed with regard to the environment of patronage, the topics and works chosen for translation, and the concrete techniques of transferring texts from Arabic to Latin. The debate on the effects of the Arabic-Latin translation movement on European cultural and intellectual history has elicited highly polemic contributions ranging from statements that attribute a large part of developments in European cultural, intellectual, technical, and scientific history to Arabic or Islamic influence,295 to statements that explicitly negate and discredit any impact of Arabic texts on the Latin-Christian intellectual landscape of the later medieval and early modern period.296 The extant prefaces to Latin translations of Arabic texts leave no room for doubt that the translators themselves were fascinated by the range of new literature that had become available to them in Arabic.297 From the historical sociolinguistic point of view chosen in this chapter, it is clear that the intellectual repercussions of the Arabic-Latin translation movement often had little to do with concrete forms of Latin-Arabic entanglement because they drew on the existing Latin translations. Thomas Aquinas's (d. 1274) Latin refutation of Averroist thought contains few traces

Officina di studi medievali, 2012), 149–178; Charles Burnett, "Michael Scot and the Transmission of Scientific Culture from Toledo to Bologna via the Court of Frederick II Hohenstaufen," *Micrologus* 2 (1994), 101–111.

<sup>291</sup> Benjamin Z. Kedar and Etan Kohlberg, "The Intercultural Career of Theodore of Antioch," *Mediterranean Historical Review* 10, no. 1–2 (1995), 164–176.

<sup>292</sup> Delle Donne, "Un'inedita epistola," 232–236.

<sup>293</sup> Bresc, *Arabes de langue*, 39; Mandalà, "La migrazione degli ebrei," 179–198.

<sup>294</sup> Labarta, "Notas sobre algunos traductores," 102–132.

<sup>295</sup> Sigrid Hunke, *Allahs Sonne über dem Abendland. Unser arabisches Erbe* (Stuttgart: Deutsche Verlags-Anstalt, 1960); Juan Vernet, *Lo que Europa debe al Islam de España* (Barcelona: Ariel, 1978); Jonathan Lyons, *The House of Wisdom: How the Arabs Transformed Western Civilization* (New York: Bloomsbury Press, 2009); Jim Al-Khalili, *The House of Wisdom. How Arabic Science Saved Ancient Knowledge and Gave Us the Renaissance* (New York: Penguin Press, 2011).

<sup>296</sup> Sylvain Gouguenheim, *Aristote au Mont Saint-Michel. Les racines grecques de l'Europe chrétienne* (Paris: Seuil, 2008).

<sup>297</sup> Martínez Gázquez, *Attitude of the Medieval Latin Translators*.

of Arabic,<sup>298</sup> as does the condemnation of "pagan books" by the bishop of Paris in 1277.299 Roger Bacon's (d. after 1292) exhortation to study Arabic for scholarly purposes does little more than mention important exponents such as al-Fārābī, Ibn Sīnā, Ibn Rušd, and others,300 as does Guy de Chauliac's (d. 1368) *Chirurgia magna*, which lists the works of various Muslim physicians.301 Starting with Petrarca (d. 1374) and continuing up to the sixteenth century, Graecophile humanist polemics were directed against the corpus of Latin translations from Arabic, but rarely engaged with the Arabic texts themselves.<sup>302</sup> Among the "Arabists," meaning either propagators of the Arabic scientific heritage lacking Arabic skills, or the emerging group of Arabic philologists, only the latter continued to deal with Arabic texts. As opposed to the Graecophile humanists, however, the "Arabists" referred positively to the contributions of the Arabic-Latin translation movement to the advancement of science in Christian Europe.<sup>303</sup>

Although dictionaries of Arabic loanwords in European languages list a number of scientific terms adopted from the Arabic language in the fields of mathematics, chemistry, astronomy, pharmacology, etc.<sup>304</sup>—frequent


examples being algorithm (< *al-Ḫwārizmī*), cipher/zero (< *ṣifr*), alcohol (< *al-kuḥūl*), zenith (< *samt*), syrup (< *šarāb*)—the long-term impact of the Arabic-Latin translation movement on European scholarly vocabularies is smaller than one would expect. With regard to mathematics, for example, André Allard has shown that mathematical innovations produced in or transmitted via the Arabic-speaking sphere could easily be expressed in Latin, which also featured a wide lexical range of expression in this field.305 Danielle Jacquart, in turn, not only emphasized the point that the Arabization of scholarly vocabulary varied from field to field,306 but analysed a number of medical texts to see when individual translators opted for the transliteration of an Arabic term and when they replaced it with a Greek or Latin equivalent.307 Her analysis of Renaissance corrections of earlier medieval Arabic-Latin translations shows that terms of Arabic origin were often systematically replaced, either with calques, i.e. loan translations whose Arabic origin is difficult to recognize, or, considering the increasing impact of humanist thought, with Greek and Latin terms.<sup>308</sup> These processes of substitution affected scientific language in early modern Europe to such a degree that, in the nineteenth century, the Egyptian traveller and intellectual Rifāʿat al-Ṭahṭawī (d. 1290/1873) observed during his sojourn in France between 1826 and 1831 that,

"when the French became proficient in the sciences, they took their scientific terms from the languages of the respective people, most of the specialized terms being derived from Greek."<sup>309</sup>

Thus, Arabic was only granted a short period between the twelfth and sixteenth centuries during which it was able to temporarily influence the scientific language(s) of European-Christian societies.

Latin," *Journal for Semitic Studies* 17 (1972), 21, no. 1–2 (1976), 24 (1989), 30–67, 120–137, 459–469; Tazi, *Arabismen*; Corriente, *Dictionary of Arabic and Allied Loanwords*.


# 2.5 Retreat into academia (I): Studying Arabic in Europe

The rise of Romance languages as written languages considerably influenced the fate of Latin-Arabic entanglement from the late medieval period onwards. Latin remained an important means of supraregional communication and intellectual endeavours in early modern Europe and was even commended for its potential in unifying Christian Europe by contemporaries.310 It is clear, however, that the European vernaculars slowly but surely replaced Latin in all fields of communication. Latin managed to retain a foothold in ecclesiastical and academic circles where most encounters of Latin and Arabic have to be situated in the early modern period. In the protected zone of European academia, Latin-Arabic entanglement flourished in various forms, until, in the course of the nineteenth and twentieth centuries, even the most conservative sectors of society abandoned its use in favour of what were now defined as "national" languages.311

In direct relations with the Arabic-speaking sphere, Romance languages and the Lingua franca—rather than Latin—became the prime means of communication aside from Arabic itself. The prestigious but secondary status of Latin is exemplified by the following text, written by the royal Spanish interpreter Diego de Urrea when he signed his adscription in Arabic and slightly corrupt Latin to become a member of the Accademia dei Lincei of Naples in 1612. The translation of the Arabic text, roughly equivalent to the Latin version, reads:

Diego of Urrea Conca, private secretary to the great sultan of Spain concerning the issues of North Africa and Asia in the languages Arabic, Turkish, Persian, a *Linceus* at the age of 50 in the year of the incarnation 1612 on 2 February in Naples, with my own hand.

ّ للسلطان االعظم سلطان / اسبانية ىف مسايل ديقوا ذى اوريه قونقه كاتب السر ّة ابللسان العريب، والرتكي، / والفارسى لنجيوس عمرى مخسني سنة من افريقية وآسي عام التلحيم الف / وستماية واثىن عشر اثىن يوم شهر فربارس ىف نبلس / خبط يدي

Ego Didacus de Urrea Conca Joannis Aloy / sij filius lynçeus neapolitanus a se / cretis epistolarum Africae et Asiae in lingua arabica turcica et persica Regnis / ispaniarum aetatis maee [*sic*] anno 50 salutis / 1612 die 2 februariy neapoli manu mea scripsi.<sup>312</sup>

<sup>310</sup> Ijsewijn, "Latin as lingua franca," 1429–1435.

<sup>311</sup> Leonhardt, *Latein*, 1–6, depicts this development much more positively than is usually expected from a Latin philologist.

<sup>312</sup> Mercedes García Arenal and Fernando Rodriguez Mediano, "De Diego de Urrea a Marcos Dobelio, intérpretes y traductores de los plomos," in *Los plomos del Sacromonte: Invención y tesoro*, ed. Manuel Barrios Aguilera and Mercedes García-Arenal (València: Universitat de València, 2006), 317. On Diego de Urrea, see Mercedes García-Arenal and Fernando Rodrigo Mediano, *The Orient in Spain*: *Converted Muslims, the Forged Lead Books of Granada, and the Rise of Orientalism* (Leiden: Brill, 2013), 225–241.

This is one of the few extant texts in which the truly multilingual Diego de Urrea had recourse to Latin, a language he had only learned at the rather late age of thirty after growing up in North Africa.313 In the adscription, Diego de Urrea seems to have employed Latin to demonstrate intellectual prowess in a distinguished academic milieu. In his actual working environment as an interpreter and translator of Arabic, Turkish, and Persian, Castilian and Italian played a much more important role than Latin.314

#### 2.5.1 PROSELYTISM AND INTELLECTUAL ENGAGEMENT: THE EMERGENCE OF ARABIC STUDIES IN LATIN

Outside the direct sphere of personal, commercial, and diplomatic interaction, Latin retained an important role as an intellectual means to access the Arabic language and its literature. Intellectuals who were engaged in institutionalizing the study of Arabic in the academic centres of European-Christian societies of the early modern period employed and cherished Latin as an established language of intellectual endeavours. Their engagement continued a medieval tradition of engaging with Islam and Arabic literature. It is interesting to note, however, that early modern centres of Arabic studies in Europe were no longer situated in the former linguistic contact zones of the Iberian Peninsula and southern Italy, but were located farther north in places such as Rome, Milan, Paris, Leiden, and Oxford, i.e. places that had never witnessed Muslim rule.<sup>315</sup>

The early beginnings of Arabic studies in Latin are usually associated with the first Latin translation of the Qurʾān commissioned by the Cluniac abbot Peter the Venerable around 1141, as well as with the Arabic-Latin translation movement of the same period. Both involved and resulted in the intensive engagement of Latin-trained intellectuals with Arabic texts from the twelfth century onwards. Ideas of institutionalizing Arabic studies in Christian Europe first came up in the late thirteenth or early fourteenth century in an ecclesiastical milieu keen on spreading the Roman-Catholic faith among Muslims, Oriental Christians, and peoples farther east. Already in 1248, Pope Innocent IV informed the chancellor of the University of Paris that he would send ten boys raised in Arabic or another Oriental language who were to be taught sound doctrine to the effect that they "would then be able to teach others in the territories

<sup>313</sup> See José M. Floristán, "Diego de Urrea (c. 1559–octubre de 1616), traductor de árabe, turco y persa en la corte de España: nuevas noticias biográficas," *Boletín de la Real Academia de la Historia* CCX, no. 2 (2013), 228–229.

<sup>314</sup> Also compare the biographies of other Spanish interpreters in José Manuel Floristán Imízcoz, "Intérpretes de lenguas orientales en la Corte de los Austrias: tres notas prosopográficas," *Silva: Estudios de humanismo y tradición clásica* 2 (2003), 41–60.

<sup>315</sup> See the overview in Gerald J. Toomer, *Eastern Wisedome and Learning: The Study of Arabic in Seventeenth Century England* (Oxford: Oxford University Press, 1996), 7–56.

beyond the sea the way to salvation."316 In another letter, Humbert of Romans, Master General of the Order of Preachers, informed the Dominican friars at the University of Paris in 1256 of the progress achieved by certain brothers in the study of Oriental languages. This had facilitated their preaching among the Arabic-speaking Oriental-Christian Maronites as well as among the Muslims of Spain, where certain friars "now for many years have studied Arabic among the Saracens."317 Between 1285 and 1314, masters from the University of Paris argued vis-à-vis the Pope that the establishment at Paris of a *studium* in the Greek, Arabic, and Tartar languages, taught by six regent-masters and attended by twenty students, seemed extremely profitable to them, given that such students could either become missionaries or translate hitherto unknown Arabic and Greek texts.<sup>318</sup> Another appeal, directed by Ramón Llull (d. 1315– 1316) to the staff of the University in 1298/1299, urging them to win the king of France's support for the aforementioned project,319 was followed by several full-fledged plans to prepare missionaries linguistically for preaching outside Latin Christendom.

In 1305, Ramón Llull proposed setting up four monasteries, one each dedicated to the study of Arabic, Syriac, Hebrew, and what he defined as "Tatar." Staffed with native speakers, to be lured to the respective monastery with the promise of receiving a salary, each monastery should train twelve future missionaries to be sent to the relevant missionary ground after completing their linguistic education.<sup>320</sup> One year later, Pierre Dubois (d. ca*.* 1321) proposed an alternative way to expand the influence of the Roman Catholic Church in a treatise discussing strategies to regain the Holy Land, lost to the Mamluks in 1291. In this treatise, Pierre Dubois explained the procedure of recruiting Oriental-Christian children in the Middle East, of providing them with a thorough dogmatic and linguistic education, and of sending them back to the Middle East as missionaries.<sup>321</sup> Both plans were


not implemented immediately, but bore fruit to a certain extent and, seen in connection with previous efforts of promoting the study of Arabic, thus contributed to the establishment of Arabic studies in various European-Christian centres of learning, many of which already boasted universities.

Ramón Llull's idea was modified and developed by the ecclesiastics present at the council of Vienne (1311–1312). They decreed the establishment of two chairs each for the study of Arabic, Aramaic, and Hebrew in all residences of the Roman curia as well as at the universities of Paris, Oxford, Bologna, and Salamanca. The chair-holders were to translate foreign books and to impart the linguistic skills necessary to "teach the infidels in the holy commandments and to join them to the community of Christians through the teaching of the faith and the reception of holy baptism." In this way, they would create human resources, who, "sufficiently taught and educated in these languages, can bring forth fruit with the grace of God and spread the faith among and for the salvation of infidel peoples."322

Outside the sphere of the early universities, various actors made additional contributions to the promotion of Arabic studies. Ecclesiastics such as Riccoldo da Monte di Croce (d. 1320) and Juan de Segovia (d. 1458) engaged with the Arabic language either directly or with the help of a translator. Their aim was to arrive at a better understanding of Islam in order to refute it more efficiently.323 Jewish converts to Christianity such as Guglielmo Raimondo Moncada used their skills in Hebrew and Arabic to impress papal and humanist circles in Italy of the late fifteenth century.324 Interest in the Arabic language became great enough to produce the first book printed in Arabic, a Christian devotional "Book of Hours," printed at the behest of Pope Leo X (Fano 1514), several years before the latter became acquainted with Leo Africanus, alias al-Ḥasan al-Wazzān al-Fāsī.325

ed. Bridges, vol. 1, pars tertia, 95, on the necessity to study Oriental languages to be able to communicate with Oriental churches subject to Rome.

<sup>322</sup> Concilium Viennense (a. 1311–1312), decretum 24, ed. and trans. Giuseppe Alberigo and Joseph Wohlmuth, in *Konzilien des Mittelalters, vol. 2: Vom ersten Laterankonzil (1123) bis zum fünften Laterankonzil (1512–1517)*, ed. Joseph Wohlmuth (Paderborn: Schöningh, 2000), 378–379: "qui infideles ipsos sciant et valeant sacris institutis instruere christicolarum que collegio per doctrinam christianae fidei ac susceptionem sacri baptismatis aggregare. [. . .] ut instructi et edocti sufficienter in linguis huiusmodi fructum speratum possint deo auctore producere fidem propugnaturi salubriter in ipsos populos infideles."

<sup>323</sup> Jean-Marie Mérigoux, "L'ouvrage d'un frère prêcheur florentin en Orient à la fin du XIII<sup>e</sup> siècle. Le 'Contra legem Sarracenorum' de Riccoldo da Monte di Croce," *Memorie domenicane* 17 (1986), 62; Gázquez, "El Prólogo de Juan de Segobia al Corán," 399, lines 160–183.

<sup>324</sup> As related in *Il diario romano di Jacopo Gherardi da Volterra dal 7 settembre 1479 al 12 agosto 1484*, ed. Enrico Carusi, Rerum italicarum scriptores 23.3 (Città di Castello: Scipione Lapi, 1904), 49. See section 4.2.3 in this volume.

<sup>325</sup> Philip Kh. Hitti, "The First Book Printed in Arabic," *Princeton University Library Chronicle* IV/1 (1942), 5–9; Miroslav Krek, "The Enigma of the First Arabic Book Printed from Movable Type," *Journal of Near Eastern Studies* 38, no. 3 (1979), 203– 212. On Pope Leo X and Leo Africanus, see Natalie Zemon Davis, *Trickster Travels. A Sixteenth-Century Muslim Between Worlds* (London: Faber and Faber, 2007).

DANIEL G. KÖNIG

The printing of the Arabic version of the Lord's Prayer,<sup>326</sup> as well as the repeated production of polyglot bibles containing parallel Greek, Latin, Syriac, Arabic, and other versions of the New Testament,<sup>327</sup> testify to an early modern urge of engaging linguistically with eastern forms of Christianity. The idea, formulated by Pierre Dubois in 1306, of spreading Roman-Catholic influence among Arabic-speaking Christians had not been lost. The establishment of the *Collegium maronitum* in Rome as well as several efforts to recruit missionaries among and for Christian groups of the Middle East may not have been successful in every sense.328 Nonetheless, such efforts created various links between European and Oriental Christians that are of utmost importance for the history of Latin-Arabic entanglement. Not only the aforementioned polyglot bibles, but a large corpus of Latin-Arabic translations of Christian devotional, liturgical, historiographical, and other explicitly Christian texts, produced between the fifteenth and nineteenth centuries by Oriental Christians and European missionaries, provide definite proof for the emergence of numerous, specifically Christian Latin-Arabic milieus.329 By the middle of the seventeenth century, Roman-Catholic influence on Oriental Christians had become so great that the Anglican protestant Arabist Edward Pococke (d. 1691) felt the need to translate Hugo Grotius's (d. 1645) protestant treatise *On the Truth of the Christian Religion* (*De veritate religionis Christianae*) into Arabic, published in 1660.<sup>330</sup>

#### 2.5.2 ACADEMIC ENGAGEMENT WITH ARABIC IN EARLY MODERN UNIVERSITIES

By the middle of the seventeenth century, the study of Arabic boasted a number of chairs at various European universities. However, the latter had not been established in the wake of the council of Vienne (1311–1312), but seem to have emerged gradually in the course of the sixteenth and seventeenth centuries. These chairs came into being in an academic environment that treated Arabic as one among several Oriental languages worth studying—in Latin, of course.<sup>331</sup>

<sup>326</sup> Angelo Michele Piemontese, "Venezia e la diffusione dell'alfabeto arabo nell'Italia del cinquecento," *Quaderni di Studi Arabi* 5, no. 6 (1987–1988), 641–660.

<sup>327</sup> Graf, *Geschichte*, vol. 1, 93–96.

<sup>328</sup> Bernard Heyberger, *Les Chrétiens du Proche-Orient au temps de la réforme catholique (Syrie, Liban, Palestine, XVIIe* –*XVIIIe siècles* (Rome: École française de Rome, 1994).

<sup>329</sup> These translations are listed in König, "Unkempt Heritage," 449–455.

<sup>330</sup> Hugo Grotius, *De veritate religionis Christianae*, translated as *Kitāb fī ṣiḥḥat al-šarīʿa l-masīḥiyya nuqila min al-lāṭīnī ilā l-ʿarabī*, trans. Edward Pococke (Oxford: no ed., 1660). See Graf, *Geschichte*, vol. 1, 98–99, for details of Arabic bible translations based on Protestant versions.

<sup>331</sup> On the diversity of teaching environments related to Arabic, see *The Teaching and Learning of Arabic in Early Modern Europe*, ed. Jan Loop, Alastair Hamilton, and Charles Burnett (Leiden: Brill, 2017).

Guillaume Postel (1510–1581), for example, was appointed a royal lecturer for Greek, Arabic, and Hebrew in 1538 by the French King Francis I in the orbit of what was to become the future *Collège de France*. <sup>332</sup> His famous *Grammatica Arabica*, a short introduction to the Arabic language, actually constitutes an extended version of a chapter within a larger oeuvre dedicated to the alphabets and grammars of twelve languages of North Africa and the Middle East.<sup>333</sup>

Joseph Scaliger (1540–1609) held a chair for Oriental languages at the University of Leiden. His major contribution to the discussion about the calendar reform of 1582, *On the Emendation of the Times* (*De emendatione temporum*), draws on dating systems in various Oriental languages, including Arabic.334 In his funeral oration, held in 1609 by Daniel Heinsius, Scaliger was commended for his extensive knowledge of several languages, initially of Greek and Hebrew, to which "he added Chaldean, Syriac, Arabic, soon also Ethiopic, Persian, and Punic." Heinsius excels in hyperbole:

There was, there was that time, when in a single house of this city one man was master of more languages than any one in Europe. There was, there was that time, when the house of one man in this city was the museum of the whole world: distant Maronites and Arabs, Syrians and Ethiopians, Persians, and some of the Indians had in this city the man to whom they could unfold their thoughts through the interpretation of language.<sup>335</sup>

Scaliger's successor, Thomas Erpenius (1584–1624), could build on the collection of manuscripts that Joseph Scaliger had bequeathed to the university's library.<sup>336</sup> Author of another *Grammatica Arabica*, Erpenius held

<sup>332</sup> Marion L. Kuntz, *Guillaume Postel. Prophet of the Restitution of All Things, His Life and Thought* (The Hague: Springer, 1981), 28–29.

<sup>333</sup> Postel, *Linguarum duodecim*, 28–31.

<sup>334</sup> Joseph Scaliger, *De emendatione temporum* (Cologny: Typis Roverianis, 1629 [enlarged repr. of Paris 1583]). See also Anthony Grafton, *Joseph Scaliger. A Study in the History of Classical Scholarship*, vol. 2: *Historical Chronology* (Oxford: Clarendon Press, 1993).

<sup>335</sup> Daniel Heinsius, "In Iosephi Scaligeri exequiis habita," in Daniel Heinsius, *Orationum editio nova* (Leiden: Ex officina Elseviriana, 1642), 50, 59: "Chaldaeam, Syriacam, Arabicam, mox & Aethiopicam, Persicam ac Punicam adjunxit. [. . .] Fuit fuit [*sic*] illud tempus, cum in vna hujus ciuitatis domo, vnus homo plures linguas quam Europaeorum quisquam teneret. Fuit fuit [*sic*] illud tempus, cum unius in hac vrbe viri domus, orbis vniuersi museum esset. Cum remoti Maronitae & Arabes, Syria & Aethiopes, Persae & ex Indis nonnulli, eum in hac vrbe hominem haberent, cui mentem animi, lingua interprete explicare possent." Translation adapted from Daniel Heinsius, *Funeral Oration on the Death of Joseph Scaliger*, trans. George W. Robinson (Cambridge, MA: no ed., 1915), 6, 15.

<sup>336</sup> On Erpenius, see Johann W. Fück, *Die arabischen Studien in Europa bis in den Anfang des 20. Jahrhunderts* (Leipzig: Otto Harrassowitz, 1955), 59–73. On his relationship to Scaliger, see Arnoud Vrolijk, "The Prince of Arabists and his Many Errors. Thomas Erpenius's Image of Joseph Scaliger and the Edition of the *Proverbia Arabica* (1614)," *Journal of the Warburg and Courtauld Institutes* 73 (2010), 297–325.

a chair for Hebrew and Arabic, and legitimized the study of the latter by pointing to the uses the study of Arabic could bring to the understanding of Hebrew in his *Three Orations on the Dignity of the Hebrew and Arabic Languages* (*Orationes tres de linguarum Ebraeae atque Arabicae Dignitate*).337

In England, Edward Pococke first studied Hebrew before he received his initial training in Arabic around 1624–1626 at the hands of Matthias Pasor, an exiled mathematician from Heidelberg. As a minister of the Anglican church, he then spent several years in Aleppo, where he catered to the spiritual needs of English merchants, improved his Arabic skills, began learning Syriac, and engaged in buying manuscripts in various Oriental languages. He returned to England when the archbishop of Canterbury, William Laud (1573–1645), instituted a chair for Arabic studies at the University of Oxford and offered him this post in 1636.338

These examples show that, in the sixteenth and seventeenth centuries, the study of Arabic was carried out in an increasingly academic environment that engaged with Arabic as one of several Oriental languages. However, all above-mentioned scholars highlighted that Arabic was worth studying in its own right. Guillaume Postel, for example, explained in an extensive Latin passage why knowledge of Arabic was needed, underscoring how much knowledge had been acquired thanks to the Arabic-Latin translation movement, and presenting Arabic as a means to various ends:

With its help, we can transmit the most excellent authors and disciplines to our men, we can destroy all the enemies of the Christian faith with the sword of the scriptures, we can participate in the commerce of the entire world through the knowledge of one single language.<sup>339</sup>

In his *De emendatione temporum*, Joseph Scaliger accorded a special relevance to Arabic texts which had played a major role in allowing him to correlate Graeco-Latin and extra-European chronologies.340 In the

<sup>337</sup> Thomas Erpenius, *Orationes tres de linguarum Ebraeae atque Arabicae Dignitate* (Leiden: ex typographia auctoris, 1621), 67.

<sup>338</sup> *The Theological Works of the Learned Dr. Pocock, Sometime Professor of the Hebrew and Arabick Tongues, in the University of Oxford, and Canon of Christ-Church, containing His Porta Mosis, And English Commentaries on Hosea, Joel, Micah, and Malachi. To which is prefixed An Account of his Life and writings, never before printed*, ed. Leonard Twells, 2 vols. (London: R. Gosling, 1740), vol. 1, 1–84, esp. 7, 9. On Pococke and his Arabist connections, see Toomer, *Eastern Wisedome*, 93–146.

<sup>339</sup> Postel, *Linguarum duodecim*, cap. "De lingua Punica, Arabica'ue," 30–31: "Qua optimos authores & disciplinas possimus nostris hominibus tradere, omnes Christianae fidei hostes scripturarum gladio confodere, eos ipsorum dogmatibus refellere, totius orbis commercio unius linguae cognitione uti."

<sup>340</sup> Scaliger, *De emendatione temporum*, vii: "All these things have come to light for the first time from the writings of the Arabs, and to such a degree that this entire treatise is new to our men [meaning European-Christian scholars]." (*Omnia nunc primum ex Arabum scriptis prodeunt: atque adeo omnis tractatio nostris hominibus noua est.*)

introduction to his *Grammatica Arabica*, Thomas Erpenius lists the uses of Arabic, a language he describes as "extensive, extremely old, and very elegant" (*longè & antiquissima & elegantissima*), knowledge of which seemed "useful and necessary to the highest degree" (*summe utilem & necessariam*) to him, because it procured the advantages already mentioned by Postel.<sup>341</sup>

#### 2.5.3 LATIN-ARABIC ENTANGLEMENT AND THE DEMISE OF ACADEMIC LATIN

These early studies of the Arabic language and its literature were all written in Latin, which, in the early modern period, still represented the most important language of academic endeavours. One tends to forget that, in terms of quantity, the great mass of Latin texts was produced, not in Roman times, but in the medieval and early modern periods,<sup>342</sup> when Latin still fulfilled the function of a lingua franca—occasionally even outside of Europe.<sup>343</sup> The modern period, however, also witnessed the successive replacement of Latin by the European vernaculars in all sectors of textual production. In this context, Latin was also ousted from the field of Oriental and Arabic studies. Although it is not possible to trace a simple linear development,<sup>344</sup> bibliometrical analysis of Orientalist and Arabist works produced in Western Europe between the sixteenth and twentieth centuries shows that the study of Arabic and Arabic texts—the primary field of Latin-Arabic entanglement in early modern Europe—was increasingly executed in the vernaculars.

Between the sixteenth and early eighteenth centuries, renowned European scholars interested in Arabic texts and the Arabic language generally used Latin as their only, or at least their favoured language of scholarly activity. Leonhard Fuchs (1501–1566), Thomas Erpenius (1584–1624), and Johann Elichmann (ca. 1601–1639) published their entire scholarly oeuvre in Latin,345 whereas Guillaume Postel (1510–1581), Edward Pococke

<sup>341</sup> Thomas Erpenius, *Grammatica Arabica* (Leiden: In Officina Raphelengiana, 1613), i.

<sup>342</sup> Leonhardt, *Latein*, 4–6; Martin Korenjak, *Geschichte der neulateinischen Literatur* (Munich: C. H. Beck, 2016).

<sup>343</sup> Ijsewijn, "Latin as lingua franca," 1429–1430.

<sup>344</sup> Leonhardt, *Latein*, 144–146.

<sup>345</sup> See e.g. Leonhart Fuchs, *Paradoxorum Medicinae libri tres, in quibus sane multa a nemine hactenus prodita, Arabum aetatisque nostrae medicorum errata non tantum indicantur sed & probatissimorum autorum scriptis [. . .] confutantur* [. . .] (Basel: Bebelius, 1535); Thomas Erpenius, *Historia Saracenica. Qva res gestae Muslimorum, inde a Muhammede arabe, Vsque ad initium Imperij Atabacaei per XLIX Imperatorum successionem fidelissimè explicantur. [. . .] Arabicè olim exarata a Georgio Elmacino* (Leiden: Ex typographia Erpeniana linguarum orientalium, 1625); Johann Elichmann, *Literae Exoticae Scriptae Arabice* (Jena: Philipp Lippoldt, 1636); Johann Elichmann, *Tabula Cebetis graece, arabice, latine: Item aurea carmina Pythagorae, cum paraphrasi arabica* (Leiden: Typis Ioannis Maire, 1640).

(1604–1691), and Johann Jakob Reiske (1716–1774) chose Latin for their more scholarly works and translations,<sup>346</sup> the respective vernacular either for more popular and didactic writings in the case of Postel,347 non-Arabist commentaries on Holy Scripture in the case of Pococke,<sup>348</sup> or scholarly correspondence in the case of Reiske.349 In this early period, the only Orientalist and Arabist of any name who eschewed publishing in Latin was Barthélemy d'Herbelot (1625–1695), whose encyclopaedia *Bibliothèque Orientale* was directed at a larger public.<sup>350</sup>

Between the eighteenth and twentieth centuries, one notices an increasing tendency to publish in the vernacular. If Latin was used, it either served to demonstrate a scholar's ability to move within a Latin tradition of academic endeavours—such as early dissertations—or to guarantee a durable international reception of a scientific achievement considered fundamental—e.g. in the case of manuscript catalogues. Although he still seems to have used Latin as a working language,351 Antoine-Isaac Silvestre de Sacy (1758–1838) wrote important works in French, and only seems to have published in Latin in a German context.352 The scientific production of his German pupil, Heinrich Orthobius Fleischer (1801–1888), in turn, is still entirely in Latin.353 This seems to confirm the Latinist observation that German-speaking academia retained Latin slightly longer than other Western European academic circles.354 From the second half of the nineteenth century onwards, it is possible to note a functional difference between Latin and vernacular publications: in France, Ernest Renan (1823–1892)


published his PhD thesis in Latin in the same year in which his well-known French work *Averroës et l'Averroïsme* became available on the market in French.355 In Germanophone regions, Moritz Steinschneider (1816–1907) only published manuscript catalogues in Latin, his other scientific production being in German.<sup>356</sup> Later generations exclusively used the vernacular, Ignaz Goldziher (1850–1920) publishing in Hungarian and in German,357 Carl Heinrich Becker (1876–1933) and Gotthelf Bergsträsser (1886–1933) only in German.<sup>358</sup>

In the nineteenth century, it became increasingly necessary to justify writing in Latin. Published in 1831, Fleischer's introduction to his critical edition and Latin translation of the pre-Islamic section of the historiographical work *al-Muḫtaṣar fī aḫbār al-bašar* by Abū l-Fidāʾ (d. 732/1331) provides an example. Fleischer explains that he had started out with the aim of producing a French translation after having studied with Antoine-Isaac Silvestre de Sacy in Paris.359 However, his publisher counselled him to produce a Latin translation to ensure a wider diffusion of the work.360 Fleischer received a partial Latin translation by Silvestre de Sacy with the latter's notes, which he took as the basis for his work.361 Here, he mentions his difficulties in finding the right Latin style to render the not very elegant, but nonetheless sophisticated Arabic style of Abū l-Fidāʾ. He describes his Latin style as impure, full of non-Latin words, and lacking coherence, but emphasizes that it was necessary to use neologisms given the inadequacy of Latin as opposed to Greek or the modern European vernaculars. After claiming that Cicero—had he been in Fleischer's position—would have also used

<sup>355</sup> Ernest Renan, *De Philosophia Peripatetica, apud Syros* (Paris: A. Durand, 1852); Ernest Renan, *Averroës et l'Averroïsme* (Paris: Calmann Lévy, 1852).

<sup>356</sup> Moritz Steinschneider, *Catalogus codicum Hebraeorum Bibliothecae Academiae Lugduno-Batavorum* (Leiden: Brill, 1858); Moritz Steinschneider, *Alfarabi des Arabischen Philosophen Leben und Schriften* (St. Petersburg: Commissionnaires de l'Académie Impériale des sciences, 1869).

<sup>357</sup> Ignaz Goldziher, *Beiträge zur Geschichte der Sprachgelehrsamkeit bei den Arabern*, 3 vols. (Vienna: Karl Gerolds Sohn, 1871–1873); Ignaz Goldziher, *A nemzetiségi kérdés az araboknál* [The Question of Nationality among the Arabs] (Budapest: Eggenberger Ferdinánd, 1873).

<sup>358</sup> Carl Heinrich Becker, *Beiträge zur Geschichte Ägyptens unter dem Islam*, 2 vols. (Strasbourg: Trübner, 1902–1903); Carl Heinrich Becker, *Islamstudien. Vom Werden und Wesen der islamischen Welt*, 2 vols. (Leipzig: Quelle and Meyer, 1924/1932); Gotthelf Bergsträsser, *Die Negationen im Kurʾān* (Leipzig: Pries, 1911); Gotthelf Bergsträsser, *Nichtkanonische Koranlesarten im Muḥtasab des Ibn Ğinnī* (Munich: Verlag der Bayerischen Akademie der Wissenschaften, 1933).

<sup>359</sup> Fleischer, *Abulfedae historia*, iv: "Igitur hanc Abulfedae historiam, quam quondam per horas subsecivas, consilio parum eorum, e Cod. Biblioth. Reg. Paris. 615 transcripseram, [. . .] e latebris protraxi et gallice vertere institui."

<sup>360</sup> Fleischer, *Abulfedae historia*, v: "Confecto igitur utrinque negotio, quum bibliopola versionem gallicam minus e re sua fore judicaret, de latina facienda cogitandum fuit."

<sup>361</sup> Fleischer, *Abulfedae historia*, v–vi.

neologisms,<sup>362</sup> he ends with an explanation of how he Latinized certain Arabic letters and patronyms.<sup>363</sup>

As late as 1890, Theodor Nöldeke (1836–1930) published a critical edition of pre-Islamic Arabic poetry with the help of August Müller (1848– 1892). This edition is framed by an introduction, a critical apparatus, and comments in Latin. In the Latin introduction, Nöldeke justifies his use of this language in the following way:

We have been asked to use the Latin language. For since many of those growing up, of whom we hope that they find this book useful in the future, will, without doubt, not be proficient enough in our languages, it seemed necessary, either to produce two or better three editions, or to content ourselves with the Latin language. We have made a great effort, however, to be understood by everyone as well as to please the imitators of Cicero. Some things I have written in the foreign way (*barbare*): I have generally not declined Arabic names, but have occasionally added the case endings *i*, *o*, *um* to avoid ambiguities as in Zaid, Zaidi (gen.); Abû Bekr, Abû Bekrum (acc.).364

Fleischer's and Nöldeke's introductions are among the latest texts that employ Latin in Arabist scholarship. In the course of the nineteenth century, Latin as a language of scientific analysis was given up in European


societies, with the exception of the Latin-based fields of ancient and medieval history.<sup>365</sup> This development was deplored by no less a person than Arthur Schopenhauer, who wrote in 1851:

The abolition of Latin as the common language of scholars and the petty-bourgeoisie of national literatures that has been introduced instead, has been a real disaster for the sciences in Europe, first and foremost because a common European scholarly public only existed because of the Latin language. [. . .] Barbarism is coming again, in spite of trains, electric wires, and hot-air-balloons.366

For the earliest Arabists working in a universitarian framework in the sixteenth and seventeenth centuries, there had been no alternative to using Latin in their scholarly output: it was the established language of scholarship that opened up centuries of rich literary culture and investigative efforts. Slowly but surely, Latin was substituted by the increasingly powerful and expressive national vernaculars, earlier in France and England than in Germany, first in popular works, then in individual studies, finally in theses necessary to acquire academic degrees as well as in the most durable and fundamental works of text-based scholarship, i.e. manuscript catalogues and critical editions.

## 2.6 Retreat into academia (II): Studying Latin in the Arab world

The academic situation in the Arab world of the same period was completely different: Latin was not studied systematically in the Arabic-speaking sphere until the early twentieth century.367 In spite of the fact that Arabic-speakers had established rule over populations with a Latin literary tradition when they took over regions in the western Mediterranean in the late seventh and early eighth centuries, it took a long time until a clear concept of a Latin language emerged in Arabic writings. Middle Eastern Arabic texts of the ninth to eleventh centuries do not feature an Arabic transcription of the term "Latin" and refer to the primary language of the

<sup>365</sup> The *Monumenta Germaniae Historica*, for example, the multi-volume edition of late antique and medieval primary sources regarded as being connected, in one way or another, with the history of the so-called Germanic groups of Late Antiquity and their medieval legacy, produced Latin paratexts until the beginning of the twentieth century.

<sup>366</sup> Arthur Schopenhauer, *Parerga und Paralipomena. Kleine Philosophische Schriften*, 2 vols. (Berlin: A.W. Hahn, 1851), vol. 2, § 255, 407: "Die Abschaffung des Lateinischen als allgemeiner Gelehrtensprache und die dagegen eingeführte Kleinbürgerei der Nationalliteraturen ist für die Wissenschaften in Europa ein wahres Unglück gewesen. Zunächst weil es nur mittelst der lateinischen Sprache ein allgemeines Europäisches Gelehrtenpublikum gab. [. . .] Die Barbarei kommt wieder, trotz Eisenbahnen, elektrischen Drähten und Luftballons."

<sup>367</sup> König, "Unkempt Heritage," 419–493, for references.

Roman and post-Roman West as either "Roman" (*al-rūmiyya*) or "Frankish" (*al-ifranǧiyya*). The term "Latin" (*al-luġa l-laṭīniyya*) was first transcribed in Andalusian texts produced between the late ninth and the tenth centuries, and then more frequently used in various texts of the eleventh century written by such scholars as Ṣāʿid al-Andalusī (d. 462/1070) and al-Bakrī (d. 487/1094). Among the Andalusian scholars of this period who mention the Latin language—e.g. explaining that it was the language of the Romans, distinguishing it from Greek, referring to the pronunciation of Andalusian toponyms as provided by experts on the Latin language, etc.—Ibn Ḥazm of Cordoba (d. 456/1064) left the most detailed and sophisticated comments. In a treatise on logic, he compares Latin and Arabic with regard to their respective ability of giving expression to certain Aristotelian categories.<sup>368</sup>

In al-Andalus in the following four centuries, slowly succumbing to the pressures of the so-called *Reconquista*, this basic knowledge of Latin was not cultivated.369 Some of it was diffused to North Africa and the Middle East from the twelfth century onwards, thus leaving traces in the Arabic works of such scholars as Ibn Abī Uṣaybiʿa (d. 668/1270), Ibn Ḫallikān (d. 681/1281), Ibn Ḫaldūn (d. 808/1406), al-Qalqašandī (d. 821/1418), al-Maqrīzī (d. 845/1442), and even the Ottoman polymath Ḥaǧǧī Ḫalīfa or Kātip Çelebī (d. 1068/1657).370 However, such traces do not amount to a systematic study of Latin, let alone an establishment of Latin studies that could in any way be compared to the emergence of Arabic studies in European universities of the same period.<sup>371</sup> Apart from a few intellectuals at the Ottoman court involved in translating European scientific writings in the seventeenth and eighteenth centuries,<sup>372</sup> the only group of people within the Islamic(ate) sphere that intensively engaged with Latin texts between the fifteenth and nineteenth centuries were Oriental Christians such as the Maronites. Maintaining intensive relations with the Roman-Catholic church, they translated large numbers of explicitly Christian writings including conciliar decrees, devotional texts, and theological tracts—from Latin into Arabic.<sup>373</sup>

<sup>368</sup> König, "Unkempt Heritage," 428–436.

<sup>369</sup> In reaction to the ideological claims of the so-called *Reconquista*, Arabic-Islamic scholars from the Muslim West seem to have lost interest in the Iberian Peninsula's pre-Islamic heritage. See König, *Arabic-Islamic Views*, 182–185.

<sup>370</sup> König, "Unkempt Heritage," 436–448.

<sup>371</sup> Fück, *Die arabischen Studien*, 25–129.

<sup>372</sup> Pinar Emiralioğlu, *Geographical Knowledge and Imperial Culture in the Early Modern Ottoman Empire* (Farnham: Ashgate, 2014), 149–151.

<sup>373</sup> König, "Unkempt Heritage," 448–455; Mireille Isa, ed., *Le latin des Maronites* (Paris: Geuthner, 2017).

#### 2.6.1 LATIN AS AN ELEMENT OF ARAB ENGAGEMENT WITH EUROPEAN MODERNITY

In the course of the seventeenth to the nineteenth century, the occasional traveller to Europe observed the use of Latin in the educational establishments in Spain, the Kingdom of Naples, and France. The writings of Muḥammad b. ʿAbd al-Wahhāb al-Ġassānī (d. 1119/1707), Muḥammad b. ʿUṯmān al-Miknasī (d. 1213/1799) and, of course, Rifāʿat al-Ṭahṭāwī (d. 1290/1873) provide evidence for an increasing interest in the role and function of the Latin language in European systems of education.<sup>374</sup> However, the systematic study of Latin in the Arab world only began in the twentieth century.

In the French-dominated Maghreb, an intensive engagement with French intellectual culture—including French scholarship on Roman history—had already taken place before the actual colonial takeover in the course of the nineteenth century. An example is provided by the Ottoman-Tunisian reformer Ḫayr al-Dīn al-Tūnisī (d. 1307/1890) who, in his political treatise on the "modernized" countries of the world published in 1867, quoted French historical literature by the pageful. In the treatise, he also commented on the Roman institution of the dictatorship.375 The French colonial powers then seem to have introduced Latin as part of a multifaceted colonial education system. One wonders how the Algerian reformer Aḥmad al-Tiǧānī, alias Abū l-ʿAbbās Aḥmad b. al-Hāšimī, was inspired by knowledge about the Latin language—diffused as part of French-Maghrebian relations within a colonial framework—when he justified the use of the Arabic popular dialect for the purpose of religious teaching in 1936. In an article entitled "After the Alienation of the Arabic Language, We Have Begun Fearing for the Dialect" the author compares the difference between Latin and the Romance languages with that of standard Arabic and the local Arabic dialects, thus displaying some superficial knowledge about the history of Latin and the evolution of Romance languages.<sup>376</sup>

Whereas manifestations of Latin-Arabic entanglement in the Maghreb seem to have been closely associated with the French colonial project, the search for "autochthonous" efforts to promote the study of Latin leads to Egypt of the 1920s. Egypt boasts the first secular university of the Arab world. Founded in 1908, it was known as "The Egyptian University" (*al-ǧāmiʿa l-miṣriyya*) between 1908 and 1940, then as "King Fuʾād I University" (*ǧāmiʿat al-malik Fuʿād*) between 1940 and 1952, and finally as

<sup>374</sup> König, "Unkempt Heritage," 455–460.

<sup>375</sup> Ḫayr al-Dīn al-Tūnisī, *Aqwam al-masālik fī maʿrifat aḥwāl al-mamālik*, ed. Muḥammad al-Ḥaddād (Beirut: Dār al-kitāb al-lubnānī, 2012), 134–135.

<sup>376</sup> Abū l-ʿAbbās Aḥmad b. al-Hāšimī, "Baʿda ġurbat al-luġa l-ʿarabiyya, aṣbaḥnā naḫšā ʿalā l-luġa l-dāriǧa," *al-Baṣāʾir* 8 (February 21, 1936), 1–2; Charlotte Courreye, "Une défense de la langue arabe dialectale dans un journal de l'Association des Oulémas algériens en 1936: démontrer l'unicité de la langue arabe par la richesse de ses parlers," *Arabica* 63, no. 5 (2016), 494–531.

"Cairo University" (*ǧāmiʿat al-Qāhira*) after the July revolution of 1952.377 As opposed to the traditional al-Azhar University, the early Egyptian University recruited a large part of its teaching staff from European countries. The latter exerted strong influence until the end of the 1920s, when Egyptian graduates returning from various European countries with doctorates were increasingly employed.<sup>378</sup> Latin was introduced into Egyptian academic curricula in this atmosphere of establishing a new, European-influenced education system and the resulting discussions on the correct education policy. This process is described in great detail in Ṭāhā Ḥusayn's treatise *The Future of Culture in Egypt* (*Mustaqbal al-ṯaqāfa fī Miṣr*). Published in 1938, this book contains a review of Egypt's cultural history as well as an entire programme of educational reform.

On the one hand, Ṭāhā Ḥusayn wishes to put Egypt on par with the colonial powers in terms of cultural history and historical importance. Emphasizing that the past is the key to the future,<sup>379</sup> he describes Egypt as a millennia-old society that maintained intensive relations with ancient Greek civilization, thus receiving as well as providing important cultural stimuli.380 With regard to the late antique and medieval periods, he emphasizes that Christian and Islamic thought displayed an equal affinity to philosophy.381 Consequently, Ṭāhā Ḥusayn argues, European and Islamic thought were not only on par, but the latter's concept of education had even been adopted by the Europeans thanks to the Arabic-Latin translation movement of the medieval and early modern periods.<sup>382</sup>

On the other hand, Ṭāhā Ḥusayn urges Egypt to implement reforms deemed necessary to make up for strongly felt discrepancies between the education systems of Europe and Egypt. Explaining why Egypt suffered from underdevelopment in this sector,<sup>383</sup> he underscores that—given Islam's long tradition of learning from other civilizations384—adopting European specificities would neither endanger the Islamic nor the national identity of Egypt.<sup>385</sup> He points to the necessity of fully understanding the factors that made modern European civilization possible.<sup>386</sup> Since culture and science were the basis for civilization and independence,387 significant


<sup>377</sup> On its foundation and history see Raʾūf ʿAbbās, *Taʾrīḫ ǧāmiʿat al-Qāhira* (Cairo: al-Hayʾa l-miṣriyya l-ʿāmma, 1995); Donald Malcolm Reid, *Cairo University and the Making of Modern Egypt* (Cambridge: Cambridge University Press, 2002).

<sup>378</sup> Donald Malcolm Reid, "Cairo University and the Orientalists," *International Journal of Middle East Studies* 19/1 (1987), 51–75, esp. 61–62.

<sup>379</sup> Ṭāhā Ḥusayn, *Mustaqbal al-ṯaqāfa fī Miṣr* (Cairo: Muʾassasat al-Hindāwī li-l-taʿlīm wa-l-ṯaqāfa, 2014), 17–20. The treatise is also available in English translation, i.e. Taha Hussein, *The Future of Culture in Egypt*, trans. S. Glazer (New York: Octagon Books, 1975).

<sup>380</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 21–26.

<sup>381</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 27–32.

<sup>382</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 33–38.

<sup>383</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 39–42.

<sup>384</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 47–53.

<sup>385</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 53–70.

advances could only be made by giving attention to primary education, which Ṭāhā Ḥusayn defines as "one of the fundamental pillars of democracy" (*rukn asāsī min arkān al-dīmūqrāṭiyya*).388 However, education continues on a secondary level, the details of which he discusses in several chapters, including the rights of teachers, the role of exams, the appropriate school books, etc.389 Turning to higher education, Ṭāhā Ḥusayn elaborates on the role of the university in educating specialized school teachers,390 and discusses the deficiencies of the existing system of higher education,391 including the religious schooling received at al-Azhar University and among the Copts of Egypt.392 Ṭāhā Ḥusayn underscores that the responsibility for promoting cultural productivity is not confined to educational institutions and elaborates on the role of intellectuals, the theatre, and the media (radio, cinema, journalism) in this regard.393 Promoting education, he concludes, should not only be a national agenda, thus addressing Egypt's responsibility towards other Arab countries.<sup>394</sup>

In the parts of this treatise dealing with secondary and higher education, Ṭāhā Ḥusayn also addresses the need to study foreign languages. He dedicates one large and a smaller subchapter to the study of Greek and Latin and their potential role for the Egyptian education system and the culture of Egypt.

The subchapter on Greek and Latin begins with a short historical sketch of the failed introduction of both languages to secondary and academic curricula. Ṭāhā Ḥusayn explains that the Egyptian minister for education from 1924 to 1926, ʿAlī Māhir Bāšā, initially pursued the idea of introducing Greek and Latin to secondary schools so as to prepare future students for university. However, ʿAlī Māhir Bāšā committed a cardinal mistake by employing Belgian and French instead of English teachers. His decision was taken in consideration of anti-colonial sentiments against the British, but had the negative effect that, because of their linguistic deficiencies in French, students were not able to follow Greek and Latin classes taught by Francophone teachers. Consequently, this project was abolished, the employed teachers either sent to the faculties of law and letters to continue their teaching of Greek and Latin there, or re-employed as French teachers in various secondary schools.

In the faculties of law and letters, Greek and Latin were only studied for a limited period. In the faculty of law, the influential French professor Léon Duguit opposed the study of Latin and ensured that students of law were soon dispensed from learning Latin. In the faculty of letters, Greek and Latin were taught for another few years. Then, however, the teaching

<sup>388</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 71, also see 71–90.

<sup>389</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 115–122, 133–138, 143–146.

<sup>390</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 243–244.

<sup>391</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 237–244.

<sup>392</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 263–284.

<sup>393</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 285–306.

<sup>394</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 207–212.

of both languages was shifted to the faculty of languages, but successfully returned to the faculty of letters in 1934. Thus, after a promising start, the teaching of Greek and Latin was abolished in secondary education and in the faculty of law. Because neither the ministry of education nor Egyptian intellectuals grasped their importance, Greek and Latin were only taught at the faculty of letters at the time of writing, i.e. in 1938.395

Against this backdrop, Ṭāhā Ḥusayn sets out to explain this failure to introduce the study of Greek and Latin to institutions outside the faculty of letters. In a first step, he ties the problem of establishing Greek and Latin in secondary education to the political fate of ʿAlī Māhir Bāšā, unfortunately without explaining the exact reasons for his downfall. With regard to the faculty of law, still dominated by Europeans, he points to the political disposition of Léon Duguit. A radical French democrat, Duguit belonged to a faction of French society that wished to promote the study of living languages at the expense of dead languages in the interest of opening up education to all social classes.396 In Ṭāhā Ḥusayn's opinion, Duguit failed to understand that the situation in Egypt differed completely from that in France. Personal animosities with Henri Grégoire, dean of the faculty of letters and fervent advocate of Greek and Latin studies, further impelled him to assert his authority. In addition, Duguit's stand on linguistic education found favour among the latter's Egyptian colleagues in the faculty of law. According to Ṭāhā Ḥusayn, they were in fear of soon being confronted with a new and better educated generation of students that would be able to read Roman law in the original language. He explains that many Egyptians who had received a European academic education obstructed the study of the classical languages in Egypt because they had been negatively influenced by inner-European debates about the status of the so-called dead languages.397 Having received only a very superficial education in Europe, most of his colleagues had not really understood the foundations of European education and had been taken in by superficialities. Consequently, they had left Europe in the conviction that technical and scientific education, not "dead" knowledge, was necessary for the advancement of Egypt.<sup>398</sup>

For Ṭāhā Ḥusayn, however, the study of Greek and Latin stood at the core of every modern education system. No system of higher education deserving that name dispensed with the study of Greek and Latin. Egypt would be put to shame abroad if it taught Roman law at its faculties without an academic staff capable of dealing with Latin texts. Scientific

<sup>395</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 171–173.

<sup>396</sup> On Léon Duguit see Marc Malherbe, *La faculté de droit de Bordeaux (1870–1970)* (Bordeaux: Presses universitaires de Bordeaux, 1996), 229–230.

<sup>397</sup> On the inner-European debate about dead and living languages, Ṭāhā Ḥusayn cites Gustave Le Bon, *Psychologie de l'éducation. L'éducation est l'art de faire passer le conscient dans l'inconscient* (Paris: Ernest Flammarion Éditeur, 1910), 13th ed. [originally published 1902], book III, chap. II, § 2 ("Les résultats de l'enseignement du latin et des langues vivantes").

<sup>398</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 172–175.

specialization, he claims, can only be achieved by studying Greek and Latin, both of which furnished the terminology of the natural sciences, technical education, and the study of living languages. He emphasizes accordingly:

I am utterly convinced, nonetheless, that Egypt will not succeed in establishing a true system of higher education and will not be able to bring about important concomitant cultural phenomena if it does not tend to these two languages.<sup>399</sup>

#### 2.6.2 NATIONALIST AND PAN-ARAB DIMENSIONS OF ENGAGING WITH LATIN

Ṭāhā Ḥusayn's argument also has a nationalist and anti-colonial note. He regarded the study of Greek and Latin not only as the basis of every true scientific education, but also as the basis of Egypt's independence. If Egypt failed to master these languages, it would not even be able to study, interpret, and understand its own history and archaeological heritage:

Those who oppose the teaching of Greek and Latin among us should observe, think, and take stock of themselves; for their opposition entails sentencing the Egyptians to ignore their history, which they cannot get to know except by drawing on foreigners.400

Seeking the advancement of Egypt, Ṭāhā Ḥusayn thus proposes to create a department for Greek and Latin studies offering academic degrees at all levels, and again emphasizes the necessity of restructuring secondary education with the aim of allowing at least a certain percentage of potential university students to prepare for the study of these languages.401

The question we have to confront and which we have to answer with honesty, openness, clarity, and simplicity, is the following: do we want to create a climate for respectable academic research in Egypt that resembles its models in every one of the highly or moderately developed European countries or not? If we decide for the second option, then the affair is already doomed to failure, then Egypt needs neither Greek nor Latin, nor a university nor faculties. [. . .] But if we decide for the first option, [. . .] then there is no alternative

<sup>399</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 173: "wa-ana maʿa ḏālik muʾmin ašadd al-īmān wa-aʿmaqahu wa-aqwāhu bi-anna Miṣr lan taẓfar bi-l-taʿlīm al-ǧāmiʿī l-ṣaḥīḥ wa-lan yaflaḥ fī tadbīr baʿḍ marāfiqihi l-ṯaqafiyya l-hāmma illā iḏā ʿaniyat bi-hātayn al-luġatayn [. . .]."

<sup>400</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 181: "fa-llaḏīna yumāniʿūn fī taʿlīm al-yūnāniyya wa-l-lātīniyya ʿindanā yaǧib an yaraw wa-yufakkirū wa-yurāǧiʿū anfusahum; li-anna maʿnā hāḏihi l-muqāwama innamā huwa l-qaḍāʾ ʿalā l-Miṣriyyīn, bi-an yaǧhalū tārīḫahum wa-allā yaʿrifūhu illā min ṭarīq al-aǧānib."

<sup>401</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 180–184.

to nurturing these two languages, not only in the university, but also in the public schools.402

Ṭāhā Ḥusayn was not the only Arab author who endowed the study of Greek and Latin with a political dimension. This has to do with the fact that the period between the 1920s and the 1960s also witnessed intensive discussions about the present state and the future of the Arabic language. In the course and in the wake of colonial rule, this issue had become highly politicized. A contemporary observer to these debates, Anwar Chejne (1923–1983) has shown that discussions revolved around three main topics.<sup>403</sup>

The first topic concerned the problem of absorbing the large number of foreign loanwords that flooded into the Arabic language due to the dominance of Western influences in many branches of human activity. This was solved partially by creating several commissions for the Arabic language in various cities of the Arab world, which then provided the forum for further discussions on the reform of the Arabic language.404 The second topic concerned problems linked to the Arabic script, which was not easy to print, deficient in its lack of a permanent vowel system, and thus difficult to read. Some reformers proposed to follow the lead of the Turkish republic that had opted for the Latinization of the Turkish alphabet in 1928–1929. Others, however, were strictly opposed to changing the alphabet, pointing to the fact that this would cut the new generations off from a 1400-year-old literary heritage including the sacred text of Islam. Both of these topics were intrinsically linked to the third topic, the extremely large gap between classical and written standard Arabic on the one hand, and colloquial Arabic on the other hand. In the quest to create a modern system of education in societies suffering from rampant analphabetism, reforming the Arabic language was an issue of utmost political relevance: reform policies would affect the relationship to current or former colonial powers and the system of national education, as well as relations to other Arabic-speaking societies.

In the debates about the reform of the Arabic language, Latin played a modest, but distinctive role. In a conference of the Academy of the Arabic language in Cairo held at the beginning of the 1950s, the intellectual Muḥammad Riḍā l-Šabībī (1889–1965) polemicized against people who drew a parallel between Latin and Arabic and accused them of using this

<sup>402</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 177: "wa-l-suʾāl allaḏī yaǧib an nulqiyahu wa-an nuǧīb ʿanhu fī ṣārāḥa wa-iḫlāṣ wa-fī wuḍūḥ wa-ǧalāʾ huwa hāḏā l-suʾāl: anurīd an nunšiʾ fī Miṣr bīʾa li-l-ʿilm al-ḫāliṣ tašbahu amṯālahā min al-biʾāt al-ʿilmiyya fī ayy balad min al-bilād al-ūrūbiyya l-rāqiyya aw al-mutawassiṭa am lā nurīd? fa-in kānat al-ṯāniya fa-qad ḫasarat al-qaḍiyya, wa-laysat Miṣr fī ḥāǧa ilā yūnāniyya wa-lā ilā lātīniyya, wa-laysat Miṣr fī ḥāǧa ilā l-ǧāmiʿa wa-ilā kulliyyātihā [. . .]. wa-in kānat al-ūlā fa-qad rabaḥat al-qaḍiyya, wa-lā budd min al-ʿināya bi-hātayn al-luġatayn lā fī-l-ǧāmiʿa waḥdihā bal fī l-madāris al-ʿāmma ayḍan."

<sup>403</sup> Chejne, *Arabic Language*, 145–168.

<sup>404</sup> Mohammed Sawaye, "Language Academies," in *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh (Leiden: Brill, 2007), vol. 2, 634–642.

comparison to classify Arabic as a language destined to disappear some day.405 This idea was taken up by Maḥmūd Taymūr (1894–1973), a renowned Egyptian author, poet, and playwright. Taymūr came from a rather wealthy Egyptian background and was well acquainted with European affairs: his brother Muḥammad had sojourned in France, whereas Maḥmūd Taymūr himself spent two years in Europe, mainly Switzerland, between 1925 and 1927, before he became involved in Egypt's literary scene. In 1947, he received an award from the Arabic Academy of Language, was then appointed a member of this academy in 1949, and was officially welcomed by no other than Ṭāhā Ḥusayn in 1950.<sup>406</sup> Maḥmūd Taymūr published a treatise on the problems of the Arabic language in 1957. A moderate reformer, he opted for a regulated introduction of neologisms, a simplification of grammar, and the introduction of permanent vowels in the Arabic script, but refused radical proposals of giving up either the entire Arabic alphabet or even written standard Arabic in favour of graphically Latinized written dialects.407 Refuting and developing al-Šabībī's idea, Maḥmūd Taymūr explains in his treatise why it is admissable to draw certain parallels between Latin and Arabic:

We can excuse people who claim that there is a parallelism between Arabic and Latin, for Latin has once been an indigenous language that was written and spoken. Then it branched out into various dialects after the Roman conquests. These developed into independent, developed, and living languages. Thus, Latin was relegated to the sphere of writing. When its derivates such as French, Italian, and Spanish won the upper hand, the horizon of its usage was constricted, it languished, dried up, and lost its liveliness. It ended up isolated between the dusty pages of old books.<sup>408</sup>

Arabic, Taymūr asserts, will be spared the fate of Latin, however, because it is superior to Latin in one important feature. Being the language of a revealed sacred book endowed with linguistic inviolability, generations of Arabic-speakers had and would continue to contribute to its conservation. Not possessing this quality, Latin failed to receive the same care:

<sup>405</sup> Muḥammad Riḍā l-Šabībī, "Bayna l-fuṣḥā wa-lahǧātihā," *Maǧallat al-risāla* 952 (February 4, 1952), 127–130; Courreye, "Une Défense," 517.

<sup>406</sup> Jan Brugman, *An Introduction to the History of Modern Arabic Literature in Egypt* (Leiden: Brill, 1984), 254–263.

<sup>407</sup> Chejne, *Arabic Language*, 165.

<sup>408</sup> Maḥmūd Taymūr, *Muškilāt al-luġa l-ʿarabiyya* (Cairo: Maktabat al-ādāb, 1957), 5–6: "fa-li-l-nās ʿuḏruhum fīmā yaqūlūn min al-muwāzana bayna l-ʿarabiyya wa-l-lātīniyya, li-anna l-lātīniyya kānat luġa aṣliyya li-l-mukātaba wa-l-kalām, ṯumma tafarraqat baʿda l-futūḥāt al-rūmāniyya lahǧāt ʿāmmiyya ṣārat fīmā baʿd luġāt mustaqilla mutaṭawwara ḥayya, wa-baqiyat al-lātīniyya luġat kitāba, iḏ taġallabat ʿalayhā muštaqātuhā ka-l-faransiyya wa-l-īṭāliyya wa-l-isbāniyya, fa-ḍāqa muḥīṭ istiʿmālihā, wa-ẓallat tataḍāʾil wa-taǧammad wa-tafqad ḥayyawiyyatihā, wa-intahā bihā l-amr ilā l-ʿuzla bayna l-ṣaḥaʿif al-maṭwiyya min al-kutub al-qadīma."

If we reflect upon this issue, it will become clear to us that Arabic differs from Latin in one essential trait, which leaves it in a place where it is safe from what happened to that language. And this is that Arabic is the language of a revealed [i.e. "heavenly" (*samāwī*)] religion of great significance [. . .]. This is truly the most important reason that protected Arabic from vanishing in the past and in the present, and it is the reason which has endowed it with the factors of remaining existent in the future. Latin, in turn, was not granted the characteristic of being a language of a revealed sacred book endowed with linguistic inviolability which would have contributed to its conservation and care. Consequently, it fell prey to the law of nature.409

This exchange of arguments on the possibility and legitimacy of drawing parallels between Arabic and Latin in the 1950s suggests that the introduction of Latin to Egyptian university curricula from the 1920s onwards had contributed to diffusing some knowledge about the history of Latin in Egyptian intellectual circles. Thanks to the intricate relationship between linguistic issues concerning the Arabic language, and political debates touching upon the topics of colonialism, nationalism, and national education, this knowledge became part of politicized linguistic debates. These debates took place in a country that, by the 1950s, had become the most important centre for the propagation of pan-Arabism. And it was Ṭāhā Ḥusayn himself who highlighted the anti-colonial and pan-Arab lessons to be learned from the history of Latin. In a speech directed at the audience of a conference organized by the Arabic academy of Damascus in 1957, Ṭāhā Ḥusayn harshly criticized reformers who argued for the textualization of regional Arabic dialects.

They attempt to expose Arabic to what has happened to Latin in the past, and want to divide the Arab world into different linguistic shades so that if the Syrian wrote, his writing would have to be translated to the Iraqi, the Egyptian, the North African, and so forth.<sup>410</sup>

<sup>409</sup> Taymūr, *Muškilāt*, 6–7: "wa-law tadabbarnā l-amr la-ẓahara lanā anna l-ʿarabiyya tatamayyaz ʿan al-lātīniyya bi-ʿunṣur ǧawharī yadaʿuhā fī maʾman min an yaǧriya ʿalayhā mā ǧarā ʿalā tilka. wa-ḏalika anna l-ʿarabiyya luġat dīn samāwī ḏī ḫaṭar [. . .]. wa-ḏālik fī l-ḥaqq aʿẓam al-asbāb allatī ṣānat al-ʿarabiyya ʿan al-zawāl fī l-māḍī wa-l-ḥāḍir, wa-sayakūn al-sabab allaḏī yamudduhā bi-ʿawāmil al-baqāʾ fī l-mustaqbal. fa-ammā l-lātīniyya fa-lam yutaḥ lahā an yakūn luġat kitāb samāwī muqaddas lahu ḥaramatuhu fī l-luġa, wa-lahu aṯaruhu fī ṣawnihā wa-ḥiyāṭatihā, wa-min ṯumma ḫaḍaʿat li-l-nāmūs al-ṭabīʿī."

<sup>410</sup> Ṭāhā Ḥusayn, "Kalimat raʾīs al-laǧna l-ṯaqāfiyya fī l-ǧāmiʿa l-ʿarabiyya maʿālī l-duktūr Ṭāhā Ḥusayn fī ḥaflat iftitāḥ al-muʾtamar," *Maǧallat al-maǧmaʿ li-l-luġa l-ʿarabiyya bi-Dimašq* 32 (1957), 25: "wa-yuḥāwilūn an yaʿraḍū l-luġa l-ʿarabiyya limā taʿarraḍat lahu l-luġa l-lātīniyya min qabl, yurīdūn an yaǧʿalū fī l-ʿālam al-ʿarabī bilādan muḫtalifatan tatakallim luġāt muḫtalifa, bi-ḥayṯu iḏā kataba l-Sūrī turǧimat kitābatuhu li-l-ʿIrāqī wa-l-Miṣrī wa-l-Ifrīqī wa-ʿalā hāḏā l-naḥwa." Translation adapted from Chejne, *Arabic Language*, 123, also see 166.

In the huge discussions about the necessity of reforming the Arabic language that took place in the decades following the establishment of Latin studies, superficial knowledge about the history of the Latin language and its relationship to the Romance vernaculars acquired the force of an argument. In these discussions, the historical fate of Latin was repeatedly juxtaposed to the potential fate of Arabic, should the latter succumb to the aims of those wishing to substitute a common pan-Arab linguistic high register (*al-fuṣḥā*) with regional or national dialects (*al-ʿāmmiyya*) not only in the oral, but also in the written sphere. In this context, the historical fate of the Latin language stood for the fragmentation and division of a supraregional linguistic unity. It could thus be used as an argument and a warning to preserve the unity of the Arab world against internal and external, i.e. colonial, forces of division. In addition, the fate of Latin served to highlight Arabic's status as the language of divine revelation, regarded in this context as a safeguard against linguistic fragmentation. Arab unity and the Arabic literary heritage, including the Islamic heritage, cannot be preserved if standard Arabic is given up in favour of the regional dialects. This is the lesson Maḥmūd Taymūr, Ṭāhā Ḥusayn, and others seem to have learned from the history of Latin, and which they propagated among all those interested in the fate of the Arabic language. Thus, although Latin played a peripheral role in the aforementioned discussions, this role sheds light on important issues of cultural identity that have been and, in some ways, still are at stake in Arabic-speaking societies.

#### 2.7 "European" vs. "Islamic" heritage? Between transculturation and cultural segregation

In spite of his diatribes against an establishment deaf to his arguments, Ṭāhā Ḥusayn and other agents promoting the study of Greek and Latin in Egypt were ultimately successful. Egypt today hosts probably the most lively departments for Greek and Latin studies in the Arab world. In and beyond Cairo University, Egypt produces a respectable number of specialists of the Latin language, albeit with a focus on classical rather than on medieval and early modern Latin texts.411

The biography and works of an Egyptian scholar of Greek and Latin, Aḥmad ʿItmān, may serve to illustrate the status and role of Latin in at least one contemporary Arabic-speaking society. When Aḥmad ʿItmān died in a car accident in Cairo on August 15, 2013, two obituaries acknowledged him as one of the most renowned classicists in the Arab world. According to Adīb Ṣaʿb, the author of his obituary in the Arabic newspaper *al-Nahār*, he had produced "pioneering works in classical Greek and Latin studies

<sup>411</sup> See König, "Unkempt Heritage," 465–470.

as well as in comparative literature."412 In the *Classical Receptions Journal*, Lorna Hardwick praised him as "an international authority on the history of classical scholarship in Egypt and on the role of the transmission of Greek texts through Latin and Arabic translations," who engaged in "cross-cultural exchange through discussion of the histories of classical scholarship and translation," also contributing to this through "his courteously ironic and even-handed dismissal of simplistic polarities between 'Orientalism' and 'Occidentalism.'"<sup>413</sup>

Born in 1945 in a village in the Egyptian governorate of Banī Suwayf, Aḥmad ʿItmān received his bachelor's degree in classical studies at Cairo University in 1965, then his PhD in Greek and Latin literature at the University of Athens in 1973. He was active as a researcher, but also as a theatre critic and playwright, as a translator, and as member and founder of various academic societies in Egypt. He wrote several studies on the Greek and Latin literature of the classical age, and translated Homer, Sophocles, Euripides, and Virgil into Arabic. He served several official functions, e.g. heading the department of classical studies at Cairo University and the Egyptian association for comparative literature for several years, as well as founding the Egyptian association for Greek and Roman studies. His obituary in *al-Nahār* concludes:

He made a good choice opting for classical studies as the topic of his academic specialization and academic work, given that Egypt is a rich source of Greek civilization. On its soil, Neoplatonic philosophy as propagated by Plotinus and his peers came into being. At the same time, it gave life to a large number of Oriental-Christian church fathers and monks. And although the greater Syrian region was also involved in Greek culture, not only from a theological, but also from a literary and philosophical point of view, this part of the Arab world has unfortunately failed to produce academic organizations dedicated to the care of Greek and Latin literature in Syria following the model of Aḥmad ʿItmān who, without contestation, has to be regarded as the pioneer and master of classical studies in the Arab sphere.414

<sup>412</sup> Adīb Ṣaʿb, "al-Duktūr Aḥmad ʿItmān sayyid al-dirāsāt al-klāsīkiyya fī l-ʿālam al-ʿarabī," *al-Nahār Online*, August 28, 2013 http://newspaper.annahar.com/ .الدكتور-أمحد-عتمان-سيد-الدراسات-الكالسيكية-يف-العامل-العريب/61792-article

<sup>413</sup> Lorna Hardwick, "In Memoriam: Professor Ahmed Etman (1945–2013)," *Classical Receptions Journal* 6, no. 1 (2014), 175.

<sup>414</sup> Ṣaʿb, "al-Duktūr": "wa-huwa faʿala ḥasanan bi-iḫtiyārihi l-dirāsāt al-klāsīkiyya mawḍūʿan li-iḫtiṣāṣihi l-ǧāmiʿī wa-ʿamalihi l-akādīmī, naẓaran ilā kawn Miṣr manbaʿan ġaniyyan li-l-ḥaḍāra l-yūnāniyya: fa-ʿalā arḍihā našaʾat al-falsafa l-aflāṭūniyya l-muḥaddaṯa maʿa Aflūṭīn wa-sawāhu, kamā našaʾa ʿadad kabīr min ābāʾ al-kanīsa l-masīḥiyya l-šarqiyya wa-l-ruhbān. wa-maʿa anna bilād al-Šām maʿniyya hiya ayḍan bi-l-ṯaqāfa l-iġrīqiyya lā min al-nāḥiya l-lāhūtiyya fa-ḥasab bal min al-nāḥiyatayn al-adabiyya wa-l-falsafiyya ayḍan, illā annahu lam yanšaʾ fī hāḏā l-ǧuzʾ min al-ʿālam al-ʿarabī, wa-yā li-l-asaf, dawāʾir ǧāmiʿiyya wa-akādīmiyyūn li-l-ʿināya bi-l-ādāb al-iġrīqiyya wa-l-lātīniyya fī bilād al-Šām,

The focus of both obituaries is on Aḥmad ʿItmān's Greek specialization as well as on the reception of "the Classics" in the Arab world in general, a topic which has already received some scholarly attention in recent years.415 However, Aḥmad ʿItmān also explored the field of Latin studies: it would have been interesting to see what the author of *Latin Literature and its Civilizational Role (Al-Adab al-latīnī wa-dawruhu l-ḥaḍārī*) 416 would have said on the entangled history of Latin and Arabic.

In the course of this chapter, we have seen that the earliest encounters of Latin and forms of Old Arabic took place in the Middle East of Antiquity, but are difficult to reconstruct in detail. The Arabic-Islamic expansion into the western Mediterranean of the seventh and eighth centuries has to be held responsible for the creation of a linguistic contact zone between societies employing Latin and/or Arabic. Linguistic interaction and interpenetration in various milieus created different forms of entanglement and even hybridization. The number and variety of Latin-Arabic milieus was enlarged and modified when European-Christian societies began expanding into Mediterranean territories hitherto under Muslim rule from the late eleventh century onwards. However, due to the increasing substitution of Latin by the vernaculars in these societies, the latter's expansion mainly contributed to a linguistic Romanization, rather than a linguistic Latinization of the Mediterranean. Latin-Arabic milieus of interpreters and translators continued to form part of trans-Mediterranean political and economic relations until about the fifteenth century, and sprouted during the so-called Arabic-Latin translation movement of the twelfth to the sixteenth century that made many Graeco-Arabic works of science and philosophy available to European-Christian intellectuals. Increasingly, however, the Latin-Arabic dyad was replaced by various combinations of Romance languages and Arabic, and successively receded into the scholarly sphere. In Christian Europe, the emerging universities provided an institutional setting that permitted parallel engagement with Latin and Arabic in various branches of learning until around the nineteenth century, when Latin was replaced by vernacular languages even in conservative branches of academia. In the Arab world, in turn, knowledge of Latin remained rudimentary until secular institutions of higher education modelled on their European equivalents were introduced to the Arab world of the late colonial period. In this context, the study of Latin received a place in the specialized niches of

ʿalā ġarār Aḥmad ʿItmān allaḏī huwa, bi-lā munāziʿ, raʾid al-dirāsāt al-klāsīkiyya wa-sayyiduhā ʿalā l-ṣaʿīd al-ʿarabī."

<sup>415</sup> See Peter E. Pormann, "The Arab 'Cultural Awakening (Nahḍa)', 1870–1950, and the Classical Tradition," *International Journal of the Classical Tradition* 13, no. 1 (2006), 3–20; Ahmed Etman, "The Arab Reception of the Classics," in *A Companion to Classical Receptions*, ed. Lorna Hardwick and Christopher Stray (Malden: Wiley-Blackwell, 2011), 141–152; Peter E. Pormann, "Classical Scholarship and Arab Modernity," in *Modernity's Classics*, ed. Sarah C. Humphreys and Rudolf G. Wagner (Heidelberg: Springer, 2011), 123–142.

<sup>416</sup> Aḥmad ʿItmān, *al-Adab al-laṭīnī wa-dawruhu l-ḥaḍārī* (Kuwait: al-maǧlis al-waṭanī li-l-ṯaqāfa wa-l-funūn wa-l-ādāb, 1989).

academic endeavours related to the study of the ancient and medieval history of the Mediterranean. During a short period, approximately between the 1930s and the 1960s, the history of Latin played a minor role as part of anti-colonial, nationalist, and pan-Arabic discourses. In the great discussions about the future role of the Arabic language for Arab societies, the history of Latin was branded as the example not to be followed. Today, Latin-Arabic milieus are confined to academic spheres both in Europe and the Arab world. The entangled history of Latin and Arabic seems to have lost any political, economic, or social significance. But is this entirely true?

Recent debates show that the history of Latin-Arab entanglement actually does play a role in discussions that try to define the role of Islam as part of the European cultural heritage. Throughout the twentieth and twenty-first centuries, scholars have debated the impact of Islam on European culture, in each case discussing the effects of the Arabic-Latin translation movement. They have continued an often rather polemic debate already led in the sixteenth century between Arabists and humanists, in which the former emphasized, and the latter denied the relevance of Arabic-Latin translations for European intellectual history.417 This debate was reinvigorated considerably when, reacting against a new educational policy formulated by the European Union that demanded a presence of Islam in European history books, the French medievalist Sylvain Gouguenheim negated the impact of Arabic-Latin translations on the development of European intellectual history in his book *Aristote au Mont Saint-Michel*, published in 2008. Gouguenheim's contribution provoked a debate led in academic publications and the American, European, and Arab press, as well as in various blogs. Among the counter-reactions, some of them equally ideological, the world-touring exhibition *1001 Inventions* defines its task as raising "awareness of the missing 1000 years of the Muslim contributions to our shared scientific and technological heritage."418 By highlighting Arabic or Islamic contributions to the history of science, this exhibition reiterates positions formulated in Europe by Arabists since Guillaume Postel, and in the Arab world by educational reformers and policy-makers since the nineteenth century. Reformers of the nineteenth century—such as Rifāʿat al-Ṭahṭāwī, Ḫayr al-Dīn al-Tūnisī,419 rather secularist Arab intellectuals of

<sup>417</sup> Klein-Franke, *Die klassische Antike*; Hasse, "Die humanistische Polemik," 65–79; Hasse, *Success and Suppression*, 137–316.

<sup>418</sup> Foundation for Science, Technology, and Innovation, "Global Impact," accessed November 7, 2017, http://www.fstc.org.uk/global-impact. For criticism of the exhibition project, see Sonja Brentjes, "Review of 1001 Inventions: The Enduring Legacy of Muslim Civilization, edited by Salim T.S. al-Hassani," *Aestimatio* 10 (2013), 119–153.

<sup>419</sup> Al-Ṭahṭāwī, *Taḫlīṣ al-ibrīz*, 15; al-Ṭahṭāwī, *An Imam in Paris*, trans. Newman, 115; Ḫayr al-Dīn al-Tūnisī, *Aqwam al-masālik*, ed. al-Ḥaddād, 35–36, 80–81; trans. Khayr al-Din al-Tunisi, *The Surest Path: The Political Treatise of a Nineteenth-Century Muslim Statesman*, trans. Leon Carl Brown (Cambridge, MA: Harvard University Press, 1967), 99–100, 138.

the twentieth and twenty-first centuries such as Ṭāhā Ḥusayn and Adonis,420 Baathist school-books,421 and fundamentalist thinkers such as Sayyid Quṭb or Maḥmūd ʿAkkām422—have all highlighted and often over-emphasized Europe's "debt" to the Arab or the Islamic world, depending on the respective author's ideological background.

It is characteristic of these discourses, both in Europe and in the Arab world, that they juxtapose, sometimes even oppose two civilizational entities. This is a typical feature also of scholarly literature on the Latin language. In Europe, Latin philologists have regarded Latin as "a ferment of European culture"423 and—with the exception of a few specialists interested in loanwords and translations424—have either dealt with Latin and Arabic from a comparative perspective,<sup>425</sup> or tended to ignore the Arabic language: there exists a philological monograph on the "Graeco-Latin Middle Ages,"426 but none on the Latin-Arabic Middle Ages. A tendency to disregard Latin-Arabic entanglement as an essential part of Euromediterranean history also exists in the Arab world. In *Latin Literature and its Civilizational Role*, for example, Aḥmad ʿItmān, the Egyptian Classicist mentioned at the beginning of this section, calls for an engagement with Latin by describing it as useful to access the history of a different civilization.

The necessity of following up [on the post-classical history of Latin] until these later periods derives from the fact that Arabic-Islamic civilization came into much closer contact with the Latin language of the later centuries than with classical Latin [. . .,] and we hope that we have succeeded in [. . .] creating, among the interested reader, an atmosphere of yearning and suspense with regard to

<sup>420</sup> Ḥusayn, *Mustaqbal al-ṯaqāfa*, 33–38; Adonis, *Violence et Islam. Entretiens avec Houria Abdelouahed* (Paris: Seuil, 2015), 219–220.

<sup>421</sup> Wizārat al-tarbiya, *Tārīḫ al-ʿArab fī l-ʿaṣr al-umawī. Al-Awwal al-iʿdādī* (Damascus [?]: al-Muʾassasa l-ʿāmma li-l-maṭbūʿāt wa-l-kutub al-madrasiyya, 1986), 167– 168; Wizārat al-tarbiya, *Tārīḫ al-ʿuṣūr al-ḥadīṯa. Al-Ṯānī l-ṯānawī l-adabī, al-ǧuzʾ al-awwal* (Damascus [?]: al-Muʾassasa l-ʿāmma li-l-maṭbūʿāt wa-l-kutub al-madrasiyya, 2001–2002), 61–62.

<sup>422</sup> Sayyid Quṭb, *al-ʿAdāla l-iǧtimāʿiyya fī l-islām* (Cairo: Dār al-šurūq, 1995), 187, 202; trans. in Sayyid Quṭb, *Social Justice in Islam* (originally published 1949), trans. William Shepard (Leiden: Brill, 1996), 284, 304; Maḥmūd ʿAkkām, "Muḥāwara bayna Šarqī wa-Ġarbī," in Maḥmūd ʿAkkām, *Fikr wa-manbar: Qaḍāyā l-insān wa-mafhūmāt al-risāla fī ḫuṭbat al-ǧumʿa*, ed. Muḥammad Adīb Yasirǧī (Aleppo: Fuṣṣilat, 2003), 432–433.

<sup>423</sup> Laetitia Boehm, "Latinitas—Ferment europäischer Kultur: Überlegungen zur Dominanz des Lateinischen im germanisch-deutschen Sprachraum Alteuropas," in *Germania latina / Latinitas teutonica*. *Politik, Wissenschaft, humanistische Kultur vom späten Mittelalter bis in unsere Zeit*, ed. Eckhard Keßler and Heinrich C. Kuhn (Munich: Wilhelm Fink, 2003), 21–70.

<sup>424</sup> For the field of loanwords, see the publications of Latham cited in this chapter, for the field of Arabic-Latin translations see the individual publications of Burnett and Hasse.

<sup>425</sup> Leonhardt, *Latein*. See index for comparisons with Arabic.

<sup>426</sup> Walter Berschin, *Griechisch-lateinisches Mittelalter: Von Hieronymus zu Nikolaus von Kues* (Bern: A. Francke, 1980).

the link between our Oriental, Arabic-Islamic civilization, and the civilization of Europe from its ancient origins to its contemporary tendencies.427

Although acknowledging links, ʿItmān juxtaposes two civilizational entities called "Oriental" or "Arabic-Islamic" on the one side, and "European" on the other side, and thus reinforces rather than breaks up cultural dichotomies. Although he is well aware that Rome ruled almost the entire area that is considered Arabic-speaking today, Aḥmad ʿItmān, "the pioneer and master of classical studies in the Arab sphere,"<sup>428</sup> did not appropriate Latin for the Arabic-Islamic sphere, but clearly marked Latin as the manifestation of a civilizational Other.

One wonders why this prolific intellectual, who—as a scholar, a translator, a playwright, and a commentator on contemporary issues of cultural policy—contributed so much to the diffusion of knowledge about classical cultures in the Arab world, could make such a clear distinction between civilizations. In a series of quasi-Socratic dialogues entitled *Our Way to Freedom* (*Ṭarīqunā ilā l-ḥurriyya*), which ʿItmān conducted with the Egyptian philosopher Zakī Naǧīb Maḥmūd on a large range of political and cultural topics, the much-respected philosopher clearly positions the Arabic-Islamic sphere at the crossroads between a Western civilization marked by rational Greek thought and a Far Eastern civilization marked by a specific kind of spirituality.429 Did Zakī Naǧīb Maḥmūd, a scholar clearly sympathetic to ancient Greek philosophy and a strong proponent of rationalist education, thus reiterate culturalist clichés known, among others, from Orientalist discourse? Or did he try to urge his readers to accept that "Arabic-Islamic civilization" cannot be separated from Greek thought? Did Aḥmad ʿItmān only reproduce or did he endorse the positions of Zakī Naǧīb Maḥmūd? Did he believe that the Greek heritage was closer to Arabic-Islamic civilization than the Latin heritage? Did he understand Latin-Christian and Arabic-Islamic civilization as two, ultimately equal continuators of ancient Greek civilization? Aḥmad ʿItmān's Graeco- and Latinophile scholarly and literary engagement remains ambiguous. It supports the same dichotomous civilizational categories defended by Sylvain Gouguenheim. Its objective, however, is not to construct a conception of European history that is "free" of any Arabic or Islamic influence, but to encourage Arab intellectual engagement with European culture and to re-define the Arab world as a cultural

<sup>427</sup> ʿItmān, *al-Adab al-laṭīnī*, 247: "wa-taʿūd ḍarūrat hāḏihi l-mutābiʿa ḥattā l-fitrāt al-mutaʾaḫḫira ilā ḥaqīqa anna l-ḥaḍāra l-ʿarabiyya l-islāmiyya qad iḥtakat maʿa lātīniyyat al-ʿuṣūr al-mutaʿaḫḫira akṯar min iḥtikākihā bi-l-lātīniyya l-klāsīkiyya. [. . .] wa-naʾmul an nakūn qad naǧaḥnā fī [. . .] ḫalq ǧaww min al-taṭalluʿ wa-ltaraqqub ladā l-qurāʾ al-muhtamīn bi-rabaṭ ḥaḍāratinā l-šarqīyya l-ʿarabiyya l-islāmiyya bi-ḥaḍārat Ūrūbbā min uṣūlihā l-qadīma ilā ittiǧāhātihā l-muʿāṣira."

<sup>428</sup> Ṣaʿb, "al-Duktūr."

<sup>429</sup> Aḥmad ʿItmān, Zakī Naǧīb Maḥmūd, *Ṭarīqunā ilā l-ḥurriyya. Muḥāwara* (Cairo: ʿAyn, 1994), 17–23.

transit zone between East and West that encompasses the best of both worlds.

The questions raised in the preceding paragraphs show that the history of Latin-Arabic entanglement forms part of the wider debate on Orientalism and Occidentalism. The short sketch of this history given in this chapter has shown that forms of Latin and Arabic have interacted and interpenetrated in a period exceeding two millennia. Again and again, processes of transculturation created new Latin-Arabic milieus, leaving us with masses of documentary evidence that defy any effort at clearly separating "Orient" and "Occident." At the same time, the history of Latin-Arabic entanglement has been characterized repeatedly by processes of cultural segregation: Latin-Arabic milieus withered away or were even consciously destroyed, although the ultimate demise of this linguistic dyad was chiefly caused by the rise of the Romance vernaculars and their replacement of Latin in most direct relations with the Arabic-speaking sphere. Finally, in spite of all historically documented milieus of Latin-Arabic entanglement, some of them promoted actively, Latin and Arabic have repeatedly been understood as markers of two distinct cultural spheres and heritages.

In conclusion, the two-pronged macro-historical approach to the history of Latin-Arabic entanglement chosen in this chapter defies simplistic interpretations that reduce this history either to a macro-history of transculturation or to a macro-history of cultural segregation. Oscillating between these two poles, the historical, geographical, and cultural scope of Latin-Arabic entanglement regularly questions the categories of geography, culture, ethnicity, and religion, which are often used to draw clear boundaries between "Islam" and "the West." It can inspire us to transcend culturalist categorizations, not by negating their existence and historical impact, but by depriving them of their status as indisputable paradigms of historical interpretation.

# Part II Latin and Arabic: Case Studies

Daniel Potthast (Ludwig-Maximilians-Universität, Munich)

**3. Diglossia as a Problem in Translating Administrative and Juridical Documents: The Case of Arabic, Latin, and Romance on the Medieval Iberian Peninsula**

It is obvious that, in multicultural environments, strategies for understanding each other are necessary in order to manage daily life. In multilingual societies, or communities in close contact with other communities using a different language, interpreting and translating become a major means of facilitating normal activities. Although research often focuses on translations of literary and scientific works, these were—in some regards—exceptions. Research on translations of non-literary texts from the economic or legal sphere can offer us insights into how languages intermingled. On the one hand, knowledge of foreign languages could be used to shape identities by differentiating between "us" and "them." On the other hand, different languages could intermingle to create hybrid spaces where new cultural milieus developed.

The medieval Iberian Peninsula furnishes us with a very interesting example, since it was there that Latin Europe met its "Other," that is, the Arabic sphere. From a contemporary view, the language situation is often understood as reflecting the segregation of two cultures: classical Arabic may have given Latin and its vernaculars some loanwords, mostly for Oriental products; Arabic literature may have been translated to transfer the knowledge of Greek antiquity to Europe. Nevertheless, the notion prevails that the terms "Christian," "European," and "Latin" somehow belong together and that they can be clearly distinguished from everything that is "Muslim" or "Arabic." This notion is wrong in many respects: first, the medieval Iberian Peninsula featured Arabicized Christians, who translated the Bible into Arabic and used Arabic in their legal documents, as well as Romance-speaking Muslims, who successively lost their knowledge of classical Arabic. Second, the different societies stood in close contact with each other, and consequently knew and influenced each other. These interactions not only took place in centres for the translation of "scientific" texts, but were also an integral feature of daily life. This means, of course, that—although translations of literary works are interesting for the history of interlingual relations—administrative and juridical documents used in daily life should be considered a major field of translation. From our perspective, it is difficult to appraise this form of entanglement between Arabic and Latin. Much of it happened only orally, and in many cases the written material is lost, because it was considered irrelevant for future generations. However, there are quite a number of medieval Iberian sources that can increase our understanding of how relations between Arabic and Latin developed, including in the legal and administrative spheres.

# 3.1 Languages of the medieval Iberian Peninsula

When we speak of Arabic and Latin, we need to include at least four language variants: classical Arabic and Latin as literary varieties on the one hand, and the Andalusī Arabic dialect and Romance as spoken varieties on the other hand, some of the latter developing to become written languages. Due to the specialization of its author, this chapter will mainly focus on Arabic. We will see later, however, that Latin was largely replaced by Castilian, and later also by Catalan, in the juridical and administrative sphere in the different Christian kingdoms of the Iberian Peninsula, whereas Romance-Arabic diglossia, in the case of Arabic speakers, lasted until their expulsion in the early seventeenth century. Dialects were written out only in a few literary genres; prominent examples of written Arabic dialect are some of the *ḫarǧa*s of Andalusī poetry, especially that of Ibn Quzmān (d. 555/1160). Apart from this, written dialect—in particular as a means of expression that was regarded as appropriate for the occasion is relatively scarcely attested. Although the Arabic-speaking Moriscos of sixteenth-century Valencia replaced classical Arabic with their own written dialect, as María del Carmen Barceló has shown,1 this was due to their social situation, which prevented them from following the traditional curriculum of Arabic-Islamic learning. Consequently, they retained the ideal that classical Arabic constituted the only written language, but failed to implement this in practice.

What is the evidence for linguistic entanglement on an everyday basis? Deeds dealing with matters of real estate provide a compelling example. Interestingly, documents of this kind are mostly preserved from the period after a region formerly under Muslim control had fallen into Christian hands. According to Islamic law, deeds only had legal force for as long as the witnesses to them were alive and, consequently, many became invalid

<sup>1</sup> María del Carmen Barceló Torres, *Minorías islámicas en el país valenciano: Historia y dialecto* (Valencia: Universidad de Valencia, 1984).

and were consequently destroyed after the witnesses' death.2 As a result of the shift from Muslim to Christian rule and the concomitant change of the legal system, however, deeds acquired a new form of relevance since, in the Christian legal system, proof of ownership had to be kept, regardless of whether the witnesses were still alive. In the case of Granada, we have a large corpus of Arabic legal deeds of different kinds (contracts, inheritance, documents, court records), all of them documenting the ownership of real estate. They were produced, with only a few exceptions, in the last decades before the Christian conquest or even in the first years after it, when Arabic was still in use as a legal language.<sup>3</sup> Some of these contracts were translated into Castilian in the sixteenth century.4

Under Christian rule, the original documents as well as their translations had to function within the framework of the new legal system. Consequently, the translation did not have to observe the Islamic form of legal validation, in which witnesses signed the deeds and later served in cases of disagreement to confirm the nature of the legal act, the deed thus serving as an *aide-mémoire*. In contrast to this, the Christians acknowledged the deed as the actual legal transaction and not only as its protocol.<sup>5</sup> Here we see a point that is pivotal for translating legal and administrative documents: since they document or even figure as a performative act, the translator must know how their different frameworks function. Translating legal and administrative documents implies not only reproducing the meaning of a text, but also showing why it is valid. In our case, the translator had to replace Arabic-Islamic<sup>6</sup> with Latin-Christian signs of validation. Mere knowledge of the languages did not suffice: the translator had to add insights into and explanations of the different legal systems and chancery practices. Ultimately, the translator needed to have received a formal


education that was difficult to access for most Arabic speakers in territories of the Iberian Peninsula ruled by Christians.

In contrast, the problem of understanding Arabic was minor: even if the role of Arabic and the extent of Arabicization in the different regions of the Iberian Peninsula at different times are disputed,7 we can assume that some knowledge of Arabic could be found in al-Andalus and the neighbouring Christian kingdoms. Among the Christian population, it probably decreased in parallel with the decline of Muslim political power. However, given the existence of the Muslim Kingdom of Granada until 1492 and an Arabic-speaking Muslim community around Valencia, where Muslims formed the population's majority until the end of the fifteenth century,8 it retained some importance. It is from the sixteenth century onwards that Arabic was understood only by Moriscos and some experts, and thus was considered a purely "Muslim" language, leading the Inquisition to confiscate Christian-Arabic texts and Arabic translations, branding them "Islamic."<sup>9</sup>

## 3.2 The role of Arabic in Christian environments

Before the sixteenth century, the situation was entirely different. Although Arabic is not normally considered to have generated an identity for Christians outside the Muslim sphere of influence, there is evidence that Arabic was known and used among such Christians. Peter I, king of Aragon (r. 1094–1104), signed a number of his charters with a monogram, adding the Arabic version of his name (*rašama*<sup>10</sup> *Bīṭruh b. Šānǧuh* / "signed, Peter, son of Sancho") on some of them.11 Take the case of Toledo, where we have more than 1,100 Arabic documents, all written after the Christian conquest in 1085.<sup>12</sup> Here we see a coexistence of different legal systems that we

<sup>7</sup> For an overview of the different hypotheses, see Otto Zwartjes, "al-Andalus," *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh, vol. 1 (Leiden: Brill, 2006), 96–101.

<sup>8</sup> Brian A. Catlos, *Muslims of Medieval Latin Christendom, c.1050–1614* (Cambridge: Cambridge University Press, 2014), 168–226; Robert Ignatius Burns, *Muslims, Christians, and Jews in the Crusader Kingdom of Valencia: Societies in Symbiosis* (Cambridge Iberian and Latin American Studies) (Cambridge: Cambridge University Press, 1984).

<sup>9</sup> For an example, see a manuscript from 1542 with an Arabic translation of the Gospels (by Isḥāq b. Bilašku, ninth century, Cordoba), and the Pauline Epistles in Madrid (Biblioteca Nacional MS. 4971, fo. 131r).

<sup>10</sup> The form is Andalusī colloquial Arabic instead of the classical Arabic *rasama*.

<sup>11</sup> Alberto Montaner, "La Historia Roderici y el archivo cidiano: Cuestiones filológicas, diplómaticas, jurídicas y historiográficas," *e-Legal History Review* 12 (2011), 51.

<sup>12</sup> (Partial) editions in Angel González Palencia, *Los mozárabes de Toledo en los siglos XII y XIII* (Madrid: Instituto de Valencia de D. Juan, 1926–1930). For a recent study of not only the Arabic documents, but also their contemporaneous Romance and Latin counterparts, see Diego Adrián Olstein, *La era mozárabe: Los mozárabes de Toledo (siglos XII y XIII) en la historiografía, las fuentes y la historia* (Salamanca: Universidad de Salamanca, 2006).

do not understand completely.<sup>13</sup> Latin, Arabic, and Castilian were used concurrently to write documents (which are preserved, as in case of the above-mentioned Arabic documents from Granada, because they prove the right of ownership of real estate). The use of Latin reached its peak shortly after the conquest, Arabic in the thirteenth century, and Castilian after that. Especially for the twelfth century, we have to suppose that a flourishing Arabicized Christian culture existed, which was strengthened at regular intervals by immigrating Christians fleeing from what remained of Muslim al-Andalus.<sup>14</sup>

The formulary of the Christian-Arabic deeds reproduced the model of contemporaneous Islamic deeds—with the exception of Islamic formulae that were replaced by more "neutral" monotheistic ones. The documents normally begin with the *basmala* and *ḥamdala* as formulae that were also used by Arabicized Jews and Christians in the East. The contracts were concluded according to the "Sunna of the Christians" (*sunnat al-naṣārā*). Wilhelm Hoenerbach argued that Islamic law and the Castilian *fuero juzgo* were compatible, since they both derived from Roman law.<sup>15</sup> Unfortunately, this hypothesis is not yet fully substantiated, and we certainly lack a complete comparison between Muslim deeds and the Christian-Arabic deeds from Toledo. In any case, Toledo was a city where Arabic was accepted as a legal language for more than two hundred years after the Christian conquest16—thus, at a time when, according to modern estimations, less than one percent of its population was Muslim.17 The documents display a slight loss of some features of classical Arabic, and also influences of Castilian, but never completely reach the language level of the dialect.18 Surprisingly, the documents show that Arabic had ceased to be used as a spoken language long before the custom of writing deeds in Arabic had stopped. Many later documents explain that their content had to be translated for the people involved in the legal transaction. This shows that, despite the

<sup>13</sup> Jean-Pierre Molénat, "Quartiers et communautés à Tolède (XIII<sup>e</sup> –XV<sup>e</sup> siècles)," *En la España medieval* 12 (1989), 163–190; Jean-Pierre Molénat, "Mudéjars et mozarabes à Tolède du XII<sup>e</sup> au XV<sup>e</sup>siècle," *Revue des mondes musulmans et de la Méditerranée* 63–64 (1992), 143–153; Christian Saßenscheidt, "Mozárabes und Castellanos im Toledo des 12. Jahrhunderts: Die Entwicklung des Toledaner Doppelalcaldentums," in *Die Mozaraber: Definitionen und Perspektiven der Forschung*  (Geschichte und Kultur der Iberischen Welt 7), ed. Matthias Maser, Klaus Herbers (Berlin: LIT-Verlag, 2011), 125–150.

<sup>14</sup> Jean-Pierre Molénat, "Los mozárabes, entre al-Andalus y el norte peninsular," in *Minorías y migraciones en la historia*, ed. Angel Vaca Lorenzo (Salamanca: Universidad de Salamanca, 2004), 11–24.

<sup>15</sup> Wilhelm Hoenerbach, "Some Notes on the Legal Language of Christian and Islamic Deeds," *Journal of the American Oriental Society* 81 (1961), 34–38..

<sup>16</sup> Toledo was conquered in 1085, but the last Arabic deed was written in 1315. See González, *Los mozárabes de Toledo*, III, 230 (doc. 939).

<sup>17</sup> Olstein, *La era mozárabe*, 121–122.

<sup>18</sup> Ignacio Ferrando, "The Arabic Language among the Mozarabs of Toledo during the 12th and 13th Centuries," in *Arabic as Minority Language* (Contributions to the Sociology of Language 83), ed. Jonathan Owens (Berlin: de Gruyter, 2000), 45–64.

extensive loss of an Arabicized Christian culture, a writing tradition in classical Arabic, based on principles also used in Islamic countries, had survived to some extent.

In the same period, the first Castilian gold coins were minted in Toledo between 1174 and 1221. They have Arabic inscriptions and follow Almoravid models. We are not obliged to consider this as evidence for the use of Arabic as an official language in Castile, since the similarity of the new coins to their Almoravid predecessors was a means to secure their acceptance. In the inscriptions, all Muslim references have been converted into Christian ones: the ruler is no longer addressed as "Commander of the Faithful" (*amīr al-muʾminīn*) as on Almoravid coins, but as "Commander of the Catholics" (*amīr al-qatūliqīn*). Instead of Muḥammad, the pope is mentioned as "Imām of the Christian Church, Pope of great Rome" (*imām al-bayʿa l-masīḥiyya bābah Rūmā l-ʿuẓmā*). The Qurʾānic quotations on Almoravid coins, "Whoso desires another religion than Islam, it shall not be accepted of him; in the next world he shall be among the losers" (Q 3:85: *wa-man yabtaġi ġayra l-islāmi dīn fa-lan yuqbala minhu wa-huwa fī l-āḫirati mina l-ḫāsirīn*), has been replaced by a Gospel quotation "In the name of the Father, the Son, and the Holy Spirit, one God! 'Who believes and is baptized, will be saved!'" (Mark 16:16: *bi-l-abi wa-l-ibni wa-l-rūḥi l-quddūsi l-ilāhi l-wāḥidi man amana wa-ʿtumida yakun sālim*).<sup>19</sup> Interestingly, this is not an ad hoc translation, but originates from an older Bible normally ascribed to a certain Isḥāq b. Bilašku (fl. 908 or 946 in Cordoba).20 Recent research has shown that this translation was a revision of an older translation.21 We do not know why that translation was created, but the inscription proves that it was known in Cordoba and also used by official persons.

#### 3.3 Arabic-speaking minorities as translators

Further examples of Arabicized communities under Christian rule come from the Muslim sphere. In the course of the Christian conquest, Muslims were subjected to Christian domination in all regions of the Iberian Peninsula. Their numbers differed, and in most regions, they were de-Arabicized relatively quickly. Instead, they used Aljamiado, that is, Romance in Arabic script. The main exception is the Kingdom of Valencia, which Jaime I of Aragon conquered in the middle of the thirteenth century. Valencia kept its Arabic-speaking, Muslim population—the Mudéjares, or later, when they

<sup>19</sup> Casto Maria del Rivero, *La moneda arábigo-española. Compendio de numismática musulmana* (Madrid: Maestre, 1933), 45–46.

<sup>20</sup> Both Munich manuscripts as well as the London manuscript have as text: "man amana wa-ʿumida yakūnu sālim" (Munich, Bayerische Staatsbibliothek Cod. Aumer 234, fo. 100v; Munich, Bayerische Staatsbibliothek Cod. Aumer 238, fo. 43v; London, British Library MS. add. 9061, fo. 76v).

<sup>21</sup> Juan Pedro Monferrer Sala, "Tres interferencias hebreas en la traducción árabe andalusí del evangelio de Marcos contenida en el ms. Qarawiyyīn 730," *Collectanea christiana orientalia* 13 (2016), 279–287.

officially became Christians, the Moriscos—for a few centuries. However, not all groups of Muslims were equally persistent in staying. While Muslims in the rural areas tended to stay, the urban centre, Valencia, quickly developed a Christian majority; the Muslim elite emigrated almost completely, while the remaining Muslim populations consisted mainly of artisans and peasants. They had their mosques, and we know of their *qāḍī*s, but they were in some regards isolated from the rest of the Islamic world.

However, a certain degree of exchange with other Muslims was retained: these Muslims were even a topic in Aragonese-Mamlūk relations, since the Mamlūks spoke up for their right to either practise their religion freely or to be allowed to emigrate to Muslim countries without hindrance. Whether their *qāḍī*s and religious scholars could compete in religious knowledge with scholars in other regions is doubtful, considering that they were appointed by the Aragonese king, whom they served as officials in the local and regional administration.22 Thus, political loyalty was of more importance than an education in line with the standard curriculum of Arabic-Islamic knowledge. The Arabic documents written by these communities show a decreasing ability to write classical Arabic, which reached its lowest ebb in the sixteenth century, when they were forced to convert to Christianity. Carmen Barceló has argued that the spoken dialect was put into writing in this period, one of the few instances when a regional Arabic dialect became a written language that totally replaced its classical variant.23 If we follow Barceló, this can only be assumed for the sixteenth and early seventeenth centuries, when writing Arabic and referring to an Arabic-Islamic frame of knowledge could arouse suspicions, so that knowledge of classical Arabic was almost completely abandoned. It is significant that a considerable portion of the documents analysed by Barceló were preserved as parts of inquisitorial archives. The documents from the preceding centuries show that classical Arabic was an ideal not often achieved. Even in the *dār al-islām*, most texts were written in some kind of Middle Arabic. Outside the *dār al-islām*, Middle Arabic may have been cruder, due to the lack of institutions teaching a formal variant of Arabic. However, as long as the opportunity existed to retain knowledge of the ideal, writers followed it as best as they could.

Let us look at some examples of how the language situation of the Mudéjares influenced the translation of documents. In the diplomatic relations between European-Christian and Arabic rulers, the translations of documents were of particular interest. During the time of the crusades, the rulers often met personally, and communicated with the help of an interpreter. Even the results of their negotiations were generally written down in a way that seemed to reproduce an oral translation. At least,

<sup>22</sup> Brian A. Catlos, *The Victors and the Vanquished: Christians and Muslims of Catalonia and Aragon 1050–1300* (Cambridge Studies in Medieval Life and Thought 4) (Cambridge: Cambridge University Press, 2007), 154–162.

<sup>23</sup> Barceló, *Minorías islámicas en el país valenciano*, 143–151.

al-Qalqašandī (d. 821/1418), author of the most important Mamlūk chancery manual *Ṣubḥ al-aʿšā fī ṣināʿat al-inšāʾ* ("The Dawn of the Night-Blind: On Chancery Practice"), attributed the stylistically poor quality of treaties concluded with the crusaders to such ad hoc translations.24 In the fourteenth and fifteenth centuries, communication took place in a more indirect way, since rulers met only rarely, tending to exchange letters instead. Although letters were always carried by ambassadors, the written document was understood as the central means of communication. Arabic letters express this explicitly by beginning with formulae such as "Your letter has arrived accompanied by your messenger" (*waṣala kitābukum ṣuḥbat rasūlikum*); the protocol of diplomatic receptions shows that the ambassador was generally understood as a mere carrier, even if he delivered more detailed information and negotiated resulting treaties. The letters were translated in the chanceries, which were also able to produce documents in foreign languages. According to al-Qalqašandī, the Mamlūk chancery had foreign-language offices for Persian, Greek, and "Frankish."25

On the Latin side, the first chancery with an Arabic office we know of is the Norman chancery on Sicily, which used Latin, Greek, and Arabic, the first attested Arabic document being dated to 1095. The Arabic office first followed Fāṭimid chancery practices,26 then switched to the practices of the Muʾminid chancery at the end of the twelfth century.27 The Aragonese chancery had an Arabic office at the latest from the thirteenth century onwards, in which mainly Jews worked in the early years. Two Arabic documents from it are preserved—both surrender treaties concluded during the conquest of the *Regnum Valenciae* that were extensively studied by Robert Ignatius Burns. He showed that the Latin and Arabic versions often did not correspond in terms of content, since the chancery lacked the ability (or the will) to translate Latin-Christian concepts into Arabic. Interpersonal relations in the European-Christian feudal system, and in Islamic international law, worked in different ways. Consequently, the Latin version of a treaty of surrender could be understood as prescribing the slow integration of a Muslim territory into the victorious Christian realm, as well as the establishment of a relationship of liege and lord between the respective Muslim and Christian. The Arabic version of the treaty of

<sup>24</sup> Daniel König, "Übersetzungskontrolle: Regulierung von Übersetzungsvorgängen im lateinisch/romanisch-arabischen Kontext (9.–15. Jahrhundert)," in *Abrahams Erbe: Konkurrenz, Konflikt und Koexistenz der Religionen im europäischen Mittelalter*, ed. Ludger Lieb et al. (Berlin: de Gruyter, 2015), 476–477.

<sup>25</sup> Abū l-Abbās Aḥmad al-Qalqašandī, *Kitāb Ṣubḥ al-aʿšā fī ṣināʿat al-inšāʾ*, ed. Muḥammad ʿAbd al-Rasūl Ibrāhīm (Cairo: Dār al-kutub al-sulṭānīya, 1913–1919), vol. 1, 165–167.

<sup>26</sup> Jeremy Johns, *Arabic Administration in Norman Sicily: The Royal Dīwān* (Cambridge Studies in Islamic Civilization) (Cambridge: Cambridge University Press, 2002), 257–280.

<sup>27</sup> Nadia Jamil and Jeremy Johns, "A New Latin-Arabic Document from Norman Sicily (November 595 h/1198 ce)," in *The Heritage of Learning: Arabic and Islamic Studies Dedicated to Professor Wadād al-Qāḍī* (Islamic History and Civilization 122), ed. Maurice Pomerantz and Aram Shahin (Leiden: Brill, 2015), 121–144.

surrender, in turn, would characterize the same document as a terminable contract, which stipulated certain duties such as peacekeeping and paying tribute. Thus, the revolts of Ibn Huḏayl alias al-Azraq that troubled Jaime I (r. 1213–1276) for such a long period are partly explained as resulting from misunderstandings between the contractors; as Burns and Paul Chevedden proved, the Latin and Arabic documents on the same procedure often failed to correspond.<sup>28</sup>

# 3.4 A letter too difficult to understand

A translator's remark on a Mamlūk letter—written in 1330 to the king of Aragon—shows that even the chancery sometimes had problems understanding Arabic texts. It says:

This is translated from a letter from the Sultan of Damascus, which was sent to the very noble king, Don Alfonso, by the Grace of God King of Aragon, of Valencia, of Sardinia, of Corsica, and Count of Barcelona. The person who translated this letter says that no one who saw it was able to say what this Arabic meant, but that it is executed with great skill, in verses of enormous subtlety of the type that is effected with Arabic grammar. In many places, he was not able to translate words, because such words do not exist in Romance, or he had to translate the meaning. This is the translation, which follows. The most difficult part is when he speaks in the third person.<sup>29</sup>

If we now look at the original letter, which is fortunately preserved in Barcelona,30 we see a normal Mamlūk letter. Its language, which follows the ideals of Mamlūk epistolography, is thus written at an elevated linguistic level. However, its vocabulary does not differ from other Arabic letters in the same archive. It is the script that constitutes the main challenge for

<sup>28</sup> Robert I. Burns, Paul E. Chevedden, and Míkel de Epalza, *Negotiating Cultures: Bilingual Surrender Treaties in Muslim-Crusader Spain under James the Conqueror* (The Medieval Mediterranean—Peoples, Economies and Cultures 22) (Leiden: Brill, 1999), 15–59, 143–192.

<sup>29</sup> John Boswell, *The Royal Treasure: Muslim Communities under the Crown of Aragon in the Fourteenth Century* (New Haven, CT: Yale University Press, 1977), 382–383: "Este es translado de una carta del Ssoldan de Damasco que fue enbiada al muy noble Rey, Don Alfonsso, por la gracia de Dios Rey d'Aragon, de Valencia, de Çardena, de Corçega, Comte Barchilona. Dize el que traslado esta carta, non sse cuyde niguno que viese esta carta que es de entendeder este arauigu della segundo la lengua espeçial, ante es fecha a gran maestria por viesos vesifagados de gran soteleza, del que la fizo en la gramateca del arauigo. En muchos logares non se pudo trasladar los viervos, ca non auie tales viervos en romançe, o ve de trasladar la entiçion. Este es el trasladu que sse ssigue. Lo mas es como quan ffabla a terçera persona." Translation by Daniel Potthast.

<sup>30</sup> Maximiliano Alarcón y Santón and Ramón García de Linares, *Los documentos árabes diplomáticos del Archivo de la corona de Aragón* (Madrid: Escuela de Estudios Árabes, 1940), 370–371 (doc. 152).

a modern reader. The letter is written in a normal Mamlūk document hand, that is, in the script *ṯuluṯ*. 31 As usual in documents from Mamlūk and other Eastern chanceries, many diacritical dots, used to differentiate between consonants, are missing. The Maġribī and Andalusī chanceries, whose documents were more common in the Aragonese chancery, were not only written in a Western script *(Maġribī)*, but were also completely vocalized and, consequently, very easy to read. As the quotation shows, the translator only had vocabulary problems, whereas the different scripts caused him no trouble.

A further example of misunderstanding is provided in a letter written in Tunis in 1360.32 Its sender writes that parts of the tribute payment can be postponed to the following year, in Arabic: *fī qābil ḏālika l-ʿām*. The translator confused the word *ʿām* ( عام( , that is, "year," with the word *ʿilm* ( علم( , that is, "knowledge," thus rendering the Castilian translation as *cosa sabida*, that is, "common knowledge," and thereby producing a senseless text. Here indeed, the script is the reason for the wrong translation, since in Maġribī script the letter *alif* has a small bottom stroke to the left that seems to connect it to the following letter, as is the case with the letter *lām*.

# 3.5 Understanding different systems of validation

Such problems of understanding the text are not well-attested. More often, the sources present us with problems of understanding different concepts used in administrative and legal documents. As we have seen, the validation of documents was important and led to the preservation of Arabic deeds from Granada. When such documents were translated from Arabic to Romance, only a limited knowledge of Islamic forms of validation was needed. The deeds were accepted as valid; their translator had only to give them an acceptable Latin form. In the exchange between Arabic and Latin speakers, the translations had to be comprehensible to members of both linguistic groups to validate the document. In Arabic administration, the *ʿalāma* served as validation. It was a calligraphically-shaped motto written above (in the Mashreq, i.e. the Middle East, and in Ifrīqiya, i.e. the eastern part of northwestern Africa) or below (in the rest of the Maghreb) the document's text. The motto's text could be a (religious) formula or a personal name—it was sometimes used by a single person and sometimes by a whole dynasty.33 In the Mashreq, most officials close to

<sup>31</sup> For the different types of letters and the scripts that had to be used in them, see al-Qalqašandī, *Kitāb Ṣubḥ al-aʿšā*, ed. Ibrāhīm, vol. 6, 189–196.

<sup>32</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 320–323 (doc. 141). For the contemporaneous translation, see Andrés Giménez Soler, "Documentos de Túnez, originales ó traducidos, del Archivo de la Corona de Aragón," *Anuari del Institut d'Estudis Catalans* (1909–1910), 243.

<sup>33</sup> During the period in question, that is, the fourteenth and fifteenth centuries, the Naṣrids in Granada used the formula "this is correct" (*ṣaḥḥa hāḏā*), the Marīnids in Fez used "it was written on the mentioned date" (*wa-kutiba fī l-taʾrīḫ* 

the sultan—for example, judges—had their own *ʿalāma*, while its use in the Maghreb was restricted to the actual ruler.34 European Christian chanceries, in turn, used a different method of validation based on seals. Arabic chanceries also knew seals—clay and lead *bullae* from different periods of Arabic history are preserved and chancery manuals mention them. However, they only served to close documents and had no authority for validating the document's authenticity. Ibn Ḫaldūn (d. 808/1406), who served as *kātib al-ʿalāma* in the Ḥafṣid chancery, wrote that the ruler's seal had only a symbolic function, as it was not used regularly.35 These differences were a topic in the written communication. Thus, the Marīnīd Sultan Abū l-Ḥasan ʿAlī wrote to Pedro IV of Aragon in 1350:

وذكرت أنك وجدت كتابنا بغري طابع مشع وقع عندك فيه ارتياب فتوقفت يف أمر ]...[ حىت تتحقق أن الكتاب الذي وصلك هو كتابنا واعلم أن العالمة اليت نوقع يف آخر كتبنا خبط يدان عالمة الصحة على ما كتبناه فوقها وأما طابع الشمع فال عربة به وإمنا عمل حياطة على الكتاب أن ال يفك فيقرأ

You mention that when you found our letter without a wax seal, you entertained doubts [about the ambassador] and let the case of […] rest, until it was attested that the letter that reached you was our letter. Know that the *ʿalāma*, which we write at the end of a letter with our own subscription, is the sign of authentication for everything that is written above it. The wax seal does not mean anything; it serves only as a device to prevent the letter from being opened and read.<sup>36</sup>

One cannot be sure that Pedro was ignorant of the function of the *ʿalāma*, as the letter states in a later passage that the Marīnīd messengers had written in another letter that Pedro had recognized the *ʿalāma* and its meaning. The doubts about the letter's authenticity probably only served to buy more time, before the king had to act in accordance with the sultan's request. That Pedro was able to feign such ignorance shows that Muslim rulers considered the linguistic abilities of the Aragonese chancery to be

*al-muʾarraḫ*), the Muʾminids in Marrakesh used "Praise to God alone!" (*al-ḥamdu li-llāh waḥdahu*), and the Ḥafṣids in Tunis used "Praise to God and thanks to God!" (*al-ḥamdu li-llāh wa-l-šukr li-llāh*). The Mamlūk sultans used their names, but letters to Christian rulers were written without *ʿalāma*.

<sup>34</sup> On most occasions, subordinates used their personal signature in their own handwriting as a means of validation; in a few cases, however, the signature was elaborated, resembling the later Ottoman *tuġras*. See Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 324–330 (docs. 142–143).

<sup>35</sup> Ibn Ḫaldūn, *Taʾrīḫ al-ʿalāma Ibn Ḫaldūn*, 7 vols. (Beirut: Dār al-kitāb al-lubnānī, 1956–1959), vol. 1, 476.

<sup>36</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 197 (doc. 99).

limited. Seen together with the examples featuring the problems of translating texts correctly, the documentation presented so far suggests, in fact, that, despite its Jewish and Muslim staff, the Aragonese lacked fundamental abilities in dealing with Arabic documents.

# 3.6 Languages of negotiations (I): Aragon–Granada

Before drawing premature conclusions, let us look at another instructive example of language knowledge in diplomatic exchange: in 1300, Aragon and the Naṣrids of Granada negotiated a military alliance against Castile. The documentation of this alliance and its negotiations is very dense.<sup>37</sup> In addition to the original Arabic treaty sent from Granada to Aragon,38 we also possess the Castilian text of the treaty sent from Barcelona to Granada as a copy in the registers.<sup>39</sup> Moreover, we have the Castilian text of a parallel treaty between Muḥammad II and Alfonso de la Cerda—pretender in Castile, who fought on the side of Aragon40—and a bilingual negotiation protocol.41 The Naṣrid royal prince came to Zaragoza for the negotiations in summer 1300. Jaime II and Alfonso de la Cerda ratified the actual treaty in autumn, and Muḥammad a few months later still, in January 1301. The protocol seems to be the result of the negotiations in Zaragoza. It is written in Arabic and Castilian—which is surprising, considering that, in the early fourteenth century, one would have expected Catalan to be used in Aragon, since Castilian replaced Catalan in Aragon a hundred years later. Although there is no concrete evidence for this, the protocol seems to have been written by a single scribe: before the first Castilian paragraph, we see a crossed-out line of Arabic script containing the words "Chapter for the King of" (*faṣl ʿan malik*). Here, the scribe seems to have started to write in Arabic before he switched to Castilian, writing: "And moreover, We, the King of Aragon mentioned above […]" (*Et otrossi nos sobredito Rey de Aragon* […])." The protocol contains almost the complete text of the later treaty, including purely formulaic sections. The introduction and clauses binding the sultan of Granada are written in Arabic, the clauses for the Aragonese king are written in Castilian, and the closing part is missing. By comparing this document to the final version of the treaty, written in the Naṣrid chancery in Granada, we see that the final document was written more carefully—it is actually very easy to decipher—and is written completely in Arabic. The Aragonese version, of which we have an archival copy, was written completely in Castilian.

<sup>37</sup> Andrés Giménez Soler, *La Corona de Aragón y Granada: Historia de las relaciones entre ambos Reinos* (Barcelona: Imprente de Casa Provincial de Caridad, 1908), 67–81.

<sup>38</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 7–10 (doc. 3).

<sup>39</sup> Giménez, *La Corona de Aragón y Granada*, 76–78.

<sup>40</sup> Giménez, *La Corona de Aragón y Granada*, 80–81.

<sup>41</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 4–6 (doc. 1).

Thus, when the alliance was concluded, two similar, but actually rather different texts in different languages, prepared by two different chanceries, were exchanged. Both texts follow the model of Arabic peace treaties: after the invocation, they start in Arabic with the phrase "So that everybody who reads this document, knows that […]," (*li-yaʿlama kull man yaqifu ʿalā hāḏā l-kitāb annā* […])<sup>42</sup> or in Castilian with "Sepan todos quantos esta carta vieren como […]." This formulary is clearly neither a contract nor a bilateral document. However, it spells out the rules for the subjects of one party and states that the other party should follow the same rules. In consequence, it represents more of a decree than a real contract, but it was a frequently used form for medieval Arabic-Latin peace treaties.43 To a certain extent, the Aragonese version transfers an Arabic model into its own language, whereas the Arabic version is adapted only in its material form: it is written on parchment. Around 1300, paper had almost completely substituted parchment in the Arabic sphere, even in the case of very important documents—all Mamlūk treaties were written on paper. Its *mise en page* is more difficult to evaluate, since no medieval treaties between Arabic rulers are preserved. However, the layout is centred on a piece of parchment that is wider than its length, and the equally wide margins on the right- and left-hand sides seem rather uncommon for Arabic documents. Most astonishing is the validation by a pending wax seal, of which only the holes in the parchment are preserved where the seal was fixed. Except for this and a few other treaties from Granada and other Maġribī chanceries, wax seals were never used by Arabic medieval chanceries.44 As a result, we can understand the treaty as a hybrid of Arabic and Latin forms of contracts. The Naṣrid chanceries validated it not only by using the *ʿalāma*, but also added a pending seal. How the Aragonese

<sup>42</sup> Al-Qalqašandī, *Kitāb Ṣubḥ al-aʿšā*, ed. Ibrāhīm, vol. 6, 342, interprets the first word as jussive *li-yaʿlam kull*, that is, "everybody knows." Since the Andalusī and Maġribī documents are vocalized, we see that the jussive was used as the final clause. Decrees from the Mashreq follow a different formulary, which introduces an equivalent part in another way, so that al-Qalqašandī, as a Mamlūk clerk, here, has only limited authority.

<sup>43</sup> Around half of all preserved Arabic peace treaties are written as decrees. Decrees had several advantages over normal contracts (*kitāb*). They did not require the presence of the contractors with whom an oath was taken, but could be decreed in the absence of the other party. Islamic law regulated them to a lesser extent, since they were administrative and not juridical acts. Since they were formulated in a way suggesting that the Muslim ruler granted the Christian ruler privileges, they demonstrated his superiority. Research on why treaties were so often written as decrees is limited: Rüdiger Lohlker, *Islamisches Völkerrecht: Studien am Beispiel Granadas* (Bremen: Kleio, 2006), for example, regards them as purely legal documents.

<sup>44</sup> The only completely preserved pending wax seal is found on the French-Hafṣīd peace treaty concluded after the Eighth Crusade (1270) to Tunis. It is edited and described in Antoine-Isaac Silvestre de Sacy, "Mémoire sur le traité fait entre le roi de Tunis et Philippe-le-Hardi, en 1270, pour l'évacuation du territoire de Tunis par l'armée des croisés," *Histoire et mémoires de l'institut royal de France, Académie des inscriptions et belles-lettres* 9 (1832), 448–477.

chancery validated it, we do not know; however, the text was written following an Arabic model—but in Castilian.

If we compare the final versions with the notes of the negotiation, we see that the Castilian version conforms almost word-for-word to the Castilian parts of the notes. The Aragonese chancery translated only the Arabic parts, but could use the rest of the earlier draft. In the case of the Arabic version, the language of the negotiation and the language of the final text do not correspond: the Arabic part of the protocol is written in a form of Andalusī dialect—it uses a conjugation in which the first person singular takes the form *nafʿal* and the plural *nafʿalū*. Moreover, it employs different conjunctions (*in kān* for conditional clauses) and non-classical prepositions (such as *matāʿ*). Otherwise, the text tries to add the case endings of classical Arabic—not written as *taškīl*, but as separate letters *alif-nūn*. The final version, in turn, is written in normal chancery style without deviation from classical grammar. Even if we do not know which language was used in the actual negotiations, we may assume that both sides had the ability to converse in the spoken varieties of Arabic and Latin, but that their knowledge of the written language was limited. The change from Latin to Romance in Iberian-Christian chanceries thus probably made it easier for the Arabic side to understand what had been written down. In contrast, Arabic linguistic conservativism may have caused the aforementioned translation problems for the Aragonese side.

## 3.7 Languages of negotiations (II): Aragon–Cairo

We find a similar example of using only an informal level of Arabic in negotiations and in drafting the text of a treaty in the case of a trade agreement between the Mamlūks and Aragon concluded in 1430.45 In addition to the actual treaty concluded in Rhodes, we also possess a draft of the same treaty written in Cairo in 1429.46 An Aragonese ambassador came to Cairo to negotiate the agreement, but when the Mamlūks demanded an additional clause, the ambassador was not authorized to accept unforeseen changes. Thus, he probably returned to Aragon with the draft. The text looks like a normal Mamlūk treaty; that is, it is written on a long paper scroll, now cut into 111 pieces, with wide spacing between the lines. Its

<sup>45</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 372–390 (doc. 153).

<sup>46</sup> Mercé Viladrich, "Jaque al sultán en el 'damero maldito.' Edición y traducción de un tratado diplomático entre los mercaderes catalanes y el sultanato mameluco (1429)," in *L'expansió catalana a la mediterrània a la baixa edat mitjana, Actes del Séminaire/Seminari organitzat per la Casa de Velazquez (Madrid) i la Institució Mia i Fontanals (CSIC, Barcelona)*, ed. Maria Teresa Ferrer i Mallol and Damien Coulon (Barcelona: Consell superior d'investigacions científiques, Institució Milà i Fontanals, Departament d'estudis medievals, 1999), 161–205.

formulary corresponds with the formulary of Mamlūk decrees (*marsūm*) 47 without the *ʿalāma,* since it is only a draft. One year later, the Aragonese ambassador did not come to Cairo; instead, the Grand Master of the Knights Hospitaller on Rhodes, Antoni de Fluvía, mediated the negotiations. The resulting treaty is very untypical for Mamlūk documents, since it is not written on a scroll, but rather on eight folia. Its scribe lacked all calligraphic abilities, so that the script looks ordinary—almost ugly—and is more difficult to decipher than is normally the case for scripts from the chancery (*dīwān al-inšāʾ*). Contrary to most other Mamlūk administrative documents, it is not validated by an *ʿalāma*, but instead by signatures whereas the text field for the testimonies is surrounded by the formula "There is no god but God" *(lā ilāh illā llāh)*—which could be interpreted as a substitute for the *ʿalāma*. An agreement between rulers could in fact also take on the form of a *kitāb*, a contract that needed validation by the signatures of witnesses. However, the combination of a decree's formulary and this kind of validation—normally used in juridical documents—is otherwise not attested. The formulary tries to reproduce a decree, but only resembles it. After the invocation, the text starts with the sentence "That it be known to everyone who sees this paper" (*an yakūna maʿlūm li-man yarā hāḏihi l-waraqa*), thus loosely reproducing the beginning of treaties from the Maghreb (*li-yaʿlama kull man yaqifu ʿalā hāḏā l-kitāb*). Consequently, the Mamlūk form of a *marsūm* was replaced by a formulary only attested in the Muslim West, probably because the Aragonese delegation had a better knowledge of the necessary formulae. The discrepancy with the expected form shows that its scribe had problems formulating classical Arabic. The text is not written in a dialect, but in a variant of Middle Arabic instead; it tries to follow the rules of classical grammar, but adds many minor errors (for example, shortened imperfect forms, different rules of congruence, etc.). These problems are explained if we look at the names of the Mamlūk delegation: Muṣṭafā Bek b. […] Murād Ḫān, Muṣṭafā Bek b. […] b. Murād, Salǧūq b. […] b. Ḫān al-Turkī, and ʿAbd ar-Raḥmān b. […].48 Obviously, the members of the delegation were Mamlūks, that is, soldiers of Turkish origin, and not scribes. Al-Qalqašandī points out that a scribe in the Mamlūk chancery at least had to know Turkish as a foreign language, since the Mamlūks had only limited knowledge of Arabic. The 1430 trade agreement shows that the delegation consisted only of policy makers without any trained scribe, resulting in a final text that did not conform to the ideals of chancery practice (*inšāʾ*). As mentioned earlier, al-Qalqašandī explained that the poor stylistic quality of treaties concluded with the crusaders resulted from the fact that they were drafted using an oral form of

<sup>47</sup> For Mamlūk decrees and their formulary, see Hans Ernst, *Die mamlukischen Sultansurkunden des Sinai-Klosters* (Wiesbaden: Harrassowitz, 1960), XXIII–XXXIX; and Donald S. Richards, *Mamluk Administrative Documents from St Catherine's Monastery* (Leuven: Peeters, 2011), 19–31.

<sup>48</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 377.

Arabic, to the effect that stylistic ideals had to be abandoned.49 The aforementioned examples reinforce this theory, showing that the knowledge of classical Arabic was certainly available in the chanceries, but not necessarily at the places where treaties were negotiated. As in the Muslim West, the persistence of classical Arabic as a written language resulted in problems of composing texts—even for the Muslims involved. Although we do not have any information about the actual process of negotiating, we have to assume that diglossia added further problems for the European Christians, because they needed not only translators and interpreters for the spoken language, but in some situations, for the written language in its varieties between classical and Middle Arabic as well.

# 3.8 The role of the diplomatic personnel

Since responsibility for the deviances from classical Arabic in the agreement lay with the members of the Mamlūk delegation, furthering our understanding of how diplomatic documents were translated requires a closer look at the people involved. The above-mentioned negotiations between Granada and Aragon in 1300 were an exception, in that they represent one of the few occasions in which high-ranking members of the elite from both sides met: Zaragoza was the residence of the kings of Aragon, so Jaime II as well as the Castilian Alfonso de la Cerda would have participated. The actual heir apparent, Muḥammad III, led the delegation from Granada.<sup>50</sup> The extant notes of the negotiations provide the sole indication that some of the persons involved were bilingual. Further information on interpreters and translators is completely lacking. The usual situation in diplomatic exchange was that a delegation from one side visited the other ruler. In the case of Aragon, these delegations mostly consisted of noblemen and merchants, as during the negotiations of 1429 and 1430 with the Mamlūks (Rafael Ferrer and Lluís Sirvent in both years, Pere de Cassaggia only in 1429).51 The delegations were accompanied by Jews and Mudéjares as translators, who occasionally also served as ambassadors without being accompanied by Christian diplomats.52 In diplomatic exchange, a small distinction—unexplained until now—was made between Mudéjares and Jews. The former only served as envoys to rulers in al-Andalus and the Maghreb,

<sup>49</sup> Al-Qalqašandī, *Kitāb Ṣubḥ al-aʿšā*, ed. Ibrāhīm, vol. 14, 70–71.

<sup>50</sup> Giménez, *La Corona de Aragón y Granada*, 67.

<sup>51</sup> Viladrich, "Jaque al sultán," 174; and Alarcón and García de Linares, *Los Documentos árabes diplomáticos*, 390.

<sup>52</sup> For an overview on persons who acted as diplomats in the exchange between Aragon and Muslim rulers, see Nikolas Jaspert, "Zur Loyalität interkultureller Makler im Mittelmeerraum: Christliche Söldnerführer (alcayts) im Dienste muslimischer Sultane," in *Loyalty in the Middle Ages: Ideal and Practice of a Cross-Social Value*, ed. Jörg Sonntag and Coralie Zermatten (Brepols Collected Essays in European Culture 5) (Turnhout: Brepols, 2016), 235–274.

whereas Jews were also sent to the Mamlūks.53 The ambassadors sent from Arabic rulers are more diverse: of course, there are also some officials, called *qāḍī*, *qāʾid*, or *fāris*, that is, members of the juridical and military elite, but we often also find European-Christian merchants and mercenaries. The Mamlūks, who sent only high-ranking officials on their relatively few missions, must be regarded as an exception.<sup>54</sup> Hypotheses that Arabic knowledge on Christian Europe, its languages, and political developments was limited in comparison to European-Christian knowledge of the Arabic sphere are totally unsubstantiated.<sup>55</sup> Many people, who knew both sides as merchants or mercenaries, populated the harbours on both sides of the Mediterranean. Even the Arabic chanceries, whose secretaries were educated very traditionally, had, as mentioned before, offices for foreign languages. The Naṣrid chancery of Granada, in particular, produced a number of Romance documents written there and sent to Christian rulers. These documents are almost completely in Romance, except for an Arabic validation: they are sealed, but also feature the usual Naṣrid *ʿalāma*. 56 In the sixteenth century, it became more and more acceptable in the Maghreb to also use Castilian as a language of diplomatic exchange.<sup>57</sup>

On the other side of the Mediterranean, the Aragonese diplomatic apparatus also featured some bilingual experts. We can assume that a number of noblemen serving as ambassadors knew some Arabic—even if we have examples where they clearly negotiated with the help of interpreters.<sup>58</sup> The Jews and the Mudéjares were consulted because of their knowledge of languages. We can suppose that the Sephardic Jews—especially shortly after the Christian conquest of the formerly Muslim territories on the Iberian Peninsula—displayed good knowledge in drafting documents of private


law, whereas their knowledge of administrative chancery practices was probably more limited, as the studies by Burns show for the thirteenth century.<sup>59</sup> We know less about the Mudéjares in the chancery. Comparatively much is known about Aḥmad b. Ḥasan b. Šuʿāʿ. He is attested as a translator in an Aragonese delegation to Granada. In the same year, i.e. 745/1344, he copied the letter of safe conduct for ʿAlī b. Kumāša, an ambassador from Granada to Aragon,60 and translated several Arabic letters.61 He was also the *qāḍī* of Játiva. As mentioned before, the *qāḍī* in Aragon was the person appointed to administrate the Muslim community, so the title reveals relatively little about his knowledge of classical Arabic and Islamic sciences. Another Ibn Šuʿāʿ, Ibrāhīm, probably a relative of Aḥmad,62 who was also *qāḍī* of Játiva, served on a diplomatic mission to Granada in 1361 and negotiated with the Marīnīds—as attested in a document arranging the release of the son of a Marīnīd *wazīr*, who was held in Aragonese custody in 1360.63 Later, he became seneschal at the court of the Aragonese queen Eleonora.

Unfortunately, none of the translations of these documents is preserved. Only the Arabic copy by Aḥmad b. Šuʿāʿ of ʿAlī b. Kumāša's safeguard is found in the Archivo de la Corona de Aragón.64 It is discernible from the original at first glance, since the complete copy is written in very straight lines, contrary to the ideals of Arabic chancery practice, where the lines are curved. The script itself is skilful and clearly readable, but differs from the normal Andalusī and Maġribī chancery style. A closer look at the script reveals a few differences to the originals that mostly contravene the rules of classical Arabic.<sup>65</sup> The document mentioned above, witnessed by Ibrāhīm b. Šuʿāʿ for the release of the Marīnīd prisoner and written in classical Arabic, follows the form of a normal testimony (*išhād*). The different witnesses' signatures obviously do not correspond with the script used for the

<sup>59</sup> Burns, Chevedden, and de Epalza, *Negotiating Cultures*, 214–216.

<sup>60</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 110–111 (doc. 56).

<sup>61</sup> Alarcón and García de Linares, *Los documentos árabes diplomáticos*, 110–113, 117–119 and 122–124 (docs. 56, 57, 60, 63). Not one of his translations has survived; we know of them only from notes on the Arabic documents.

<sup>62</sup> Barceló, *Minorías islámicas*, 372, argues that he was Aḥmad's son, which seems implausible, since he names himself Abū Isḥāq Ibrāhīm b. Abū ʿAbdallāh Muḥammad b. Šuʿāʿ.

<sup>63</sup> Pedro Longás, "Capitulaciones celebradas para el rescate de Abu Omar Muza Benibrahim, vizir de reino de Fez, cautivo en el reino de Aragón," in *Homenaje ofrecido a Menéndez Pidal: Miscelánea de estudios lingüísticos, literarios e históricos*, no editor, vol. 3 (Madrid: Librería y Casa Editorial Hernando, 1925), 551–561 (= P.PaisValenciano I 265).

<sup>64</sup> Archivo de la Corona de Aragón, Cartas árabes 55. Alarcón and García de Linares, *Los Documentos árabes diplomáticos*, 109 (doc. 55) give only a very short description of the document.

<sup>65</sup> Daniel Potthast, "Translations of Arabic Diplomatic Letters in the Aragonese Chancery," in *Dasselbe mit anderen Worten? Sprache, Übersetzung und Sprachwissenschaft; Akten des 2. Symposiums des Zentrums historische Sprachwissenschaften (ZhS), München, 11. und 12. April 2014*, ed. Peter Schrijver and Peter-Arnold Mumm (Münchner Forschungen zur historischen Sprachwissenschaft 16) (Bremen: Hempen, 2015), 183.

text of the *išhād*. The last signature by Ibrāhīm b. Šuʿāʿ seems very clumsy, as if made by a man not used to writing. Since it is difficult to appraise the quality of the writing based only on the signature of a witness—signatures on legal documents were normally written in almost unreadable scripts we would be going too far if we assumed that his Arabic writing skills were minor. However, we can be sure that the document was drafted and written by someone else—probably a member of the Marīnīd delegation.

Even if both Ibn Šuʿāʿ-s travelled to Granada, lived there at least for some weeks or months, and probably also had some private, written contact with a number of Naṣrid officials, who regularly translated Arabic letters, their knowledge of the Arabic dialect of Valencia would only have enabled them to roughly understand Classical Arabic. As we have seen, their position as *qāḍī* seems to have been in some way hereditary and probably also needed royal approval, so that formal knowledge of Arabic-Islamic scholarship was not necessary and, most likely, was not accessible in their original environment in Valencia.

# 3.9 Conclusion

To briefly summarize this broad overview of Arabic-Latin translations, we see that at least parts of the medieval Iberian Peninsula featured bilingual societies at certain times, but that the bilingualism fully incorporated only the spoken language varieties, while knowledge of the literary languages was only found among experts. Even in cases where such experts were desperately needed, as in diplomatic exchanges, they were not always available, and people had to rely on their own knowledge of the dialect to understand the literary language. This result is surprising, given that contact between Christian and Arabic societies was often intensive. For the fourteenth century, for which we possess much documentation of the diplomatic relations between the Aragonese kings and the Muslim rulers whose chanceries used Arabic, we can see that delegations travelled to and from Aragon at regular intervals of a few weeks. From a linguistic point of view, these exchanges worked: first, because the Aragonese could understand a considerable amount of the letters' content based on their knowledge of the Valencian dialect; and second, because the envoys often had a bilingual background and were thus able to solve any problems that arose orally.

Thus, we see an imbalance between the Arabic and Iberian-Christian sides, since oral and written knowledge of the vernaculars were more easily acquired than knowledge of the languages of scholarship, which could only be learned in particular places. Establishing Romance dialects as written languages simplified the Muslims' access to Christian Europe, since, unlike Latin, there were many places where they could study them.

If we take a look at the first diplomatic situation for which we have original documents that show us which languages were used, we see that the Muslims had similar problems finding experts who could understand the Iberian-Christian side, which still used Latin. In 1069, the rulers of Zaragoza and Navarra, al-Muqtadir I and Sancho IV, concluded a peace treaty.<sup>66</sup> It was completely written in Latin. Al-Muqtadir accepted it with the addition of one Arabic sentence:

أان ملتزم هذا إذا التزم األمري ساجنه سلمه هللا كلما ثبت فيه إن شاء هللا وهو املستعان

I am bound to this, if the Amīr Sānǧuh—may God preserve him—is also bound to everything that is established in it, God willing! He is the One Whose help is sought!

We can only speculate whether al-Muqtadir understood the Latin text he accepted, even if it is a very vulgar variant. However, by developing Romance into a written language, the non-Arabic textual culture of Christian Iberia became more accessible for everyone lacking a formal education in Latin, while Arabic, with its diglossia, continued to remain a more inaccessible language in its written form.

<sup>66</sup> José Maria Lacarra, "Dos tratados de paz y alianza entre Sancho el de Peñalen y Moctadir de Zaragoza (1069 y 1073)," in *Colonización, parias, repoblación y otros estudios*, ed. José Maria Lacarra (Zaragoza: Anubar, 1981), 92–93.

Benoît Grévin (CNRS, Paris)

# **4. Between Arabic and Latin in Late Medieval and Renaissance Italy**

#### 4.1 Introduction: A peculiar position in the middle of the "corrupting sea"

In the study of interaction between Arabic and Latin during the Middle Ages, the importance of certain periods and areas stands out. Three major regions of interaction are the Iberian Peninsula, Italy, and the parts of Palestine and Syria ruled by Frankish crusaders during the twelfth and thirteenth centuries. The common link between these regions is that they were all part of a broader Mediterranean space situated at the frontier of the Arabic- and Latin/Romance-speaking spheres, and thus bound to be at the epicentre of, or at least the starting point for some major interaction between the two languages. More specifically, all of them were strongholds of Christian-Latin powers at different times, ruling for some duration over Muslim populations that had originally been predominantly Arabic speaking.

The peculiarities of local or regional history certainly make any attempt at comparison between the three zones hazardous. The Iberian Peninsula presents a special case, insofar as the interaction between the two linguistic systems, Arabic and Latin/Romance, stretched from the early Middle Ages well into the modern era. When the Muslim conquerors entered Spain in 711, the local Christian elites wrote in Latin, and also governed using Latin as a prestige language. One could even argue that the entire population (with the exception of the Basque-speaking areas) was still speaking what could be characterized as a variety of evolved forms of late Latin rather than proto-Romance languages.1 By the ninth century, the Muslim presence had become pervasive in the south and centre of the peninsula. The laments of Alvarus of Cordoba (d. ca*.* 861) on the attraction of Arabic literature and the decadence of Latin studies among the Christian elites of

<sup>1</sup> On this point, see Michel Banniard, *Viva voce: Communication écrite et communication orale du IVe au IXe siècle en Occident Latin* (Paris: Institut des études augustiniennes, 1992); and Reinhard Kiesler, *Einführung in die Problematik des Vulgärlateins* (Tübingen: Niemeyer, 2006).

al-Andalus are well known. They are a testimony to the complex processes of interaction that had meanwhile occurred between a Muslim power and an appealing Arabic-Islamic culture on the one hand, and the part of the population that had so far remained faithful to the Christian religion and, to some extent, to the former prestige language, Latin, on the other hand.2 In the aftermath of the most important phase of the so-called *Reconquista* (1085–1248), the conquering Christian kingdoms of León-Castile, Aragon, and Portugal faced the opposite situation. Now vastly expanded "Christian" states, whose populations spoke various forms of Latin-derived Romance languages, but whose cultural traditions and church apparatus still depended heavily on Latin writing and Latin knowledge,<sup>3</sup> administered important Muslim minorities for whom Arabic in its Qurʾānic and classical forms was a prestige language. These Muslims maintained a tradition of communicating in local Arabic dialects, even if they increasingly tended to use Romance dialects for their internal and extra-communitarian communication. In fact, Islamicized Iberian populations had never totally ceased to use Romance languages in their daily lives. The extent to which Romance was used may be contested, but no one doubts that this widespread use of Romance had important consequences for the emergence of bilingual textual practices, from the elaboration of the *muwaššaḥ*-poems to the birth of Aljamiado.4 Thus, in some form or another, the Iberian Peninsula was to remain a potential place of interaction between Arabic and Latin (and derived languages) for almost a thousand years, from 711 to the expulsion of the Moriscos at the beginning of the seventeenth century.

<sup>2</sup> On that point, see Banniard, *Viva voce*, 459–489. See also Chapter 2.3.3 in this volume.

<sup>3</sup> Latin would be substituted more or less gradually as an official language of administration only from the thirteenth century onwards, beginning with Castile and Portugal. Due to complex sociolinguistic, symbolic, and educational factors, the Latin language resisted better in England and in France (where the shift occurred partially during the fourteenth century at the royal chancery), and even later in Italy, Germany, and Eastern Europe. See Serge Lusignan, *La langue des rois au Moyen Âge: Le français en France et en Angleterre* (Paris: Presses universitaires de France, 2004), on the rhythm of this progression, for a brief comparison between different Western European countries, and a more specific discourse on France. For the situation in England, see Serge Lusignan, "Communication in the Later Plantagenet Empire: Latin and Anglo-Norman as Regal Languages," in *The Plantagenet Empire, 1259–1453*, ed. Peter Crooks et al. (Donington: Shaun Tyas, 2016), 273–289.

<sup>4</sup> The question of the rhythm of Arabization before the *Reconquista*, and of the de-Arabization of the Muslim populations in the Christian kingdoms after the *Reconquista*, has been the object of numerous and often fierce debates. See Mercedes García-Arenal and Katarzyna Starczewska, "'The Law of Abraham the Catholic': Juan Gabriel as Qurʾān-Translator for Martín de Figuerola and Gilles de Viterbe," *Al-Qanṭara* 35, no. 2 (2014), 409–459, particularly 412–415, who prove that knowledge of Arabic was more resilient than ordinarily thought, even as late as 1500, and even in northern zones like central Aragon. For Aljamiado culture and Arabic culture in Castile in the mid-fifteenth century, see Gerard Wiegers, *Islamic Literature in Spanish and Aljamiado: Yça of Segovia (fl. 1450), his Antecedents and Successors* (Leiden: Brill, 1994). See also Chapter 2.4.2 in this volume.

As late as the fifteenth and the sixteenth centuries, during the "second wave" of Latin translations of the Qurʾān, Spanish Arabic-trained literati of Muslim background still acted as informants.<sup>5</sup> Their activities testify to the persistence of this Iberian-Arabic culture and its contribution to the elaboration of Latin knowledge of Islam at a European level.

In comparison, the history of Arabic-Latin linguistic relations in Syria-Palestine and in Italy seems rather meagre. In the first of these two cases, the only period of possible major interaction would have been the twelfth to thirteenth centuries. One should consider, however, that the histories of Latin pilgrimages, eastern Mediterranean commerce, and the Kingdom of Cyprus provide backgrounds for linguistic interaction beyond this period.6 With regard to Italy, the major institutional framework for linguistic interaction between the two languages was certainly the Norman Kingdom of Sicily. This polity officially became a kingdom only in 1130. However, its specific relation with Islam and Arabic began between approximately 1060 and 1090, during the progressive conquest of the island by a Norman aristocracy, that had installed itself in the southern parts of the Italian mainland a generation earlier.7 This realm thus included a continental part covering the entire Italian Mezzogiorno, where Arabic had been spoken only very sporadically.<sup>8</sup> Although Sicily was subjected to a process of Latin colonization that reduced the Muslim, Arabic-speaking majority to a minority during the course of the twelfth century, there remained an

<sup>5</sup> To the famous couple Yça of Segovia and Juan de Segovia (Wiegers, *Islamic Literature*) we can now add the pair Juan Gabriel of Teruel and Egidio da Viterbo. The Aragonese Muslim convert Juan Gabriel of Teruel helped Egidio create a new, sophisticated Latin version of the Qurʾān. For more on this, see Chapter 5 of this volume. On the concept of a second wave of Latin translations of the Qurʾān, see Benoît Grévin, "Les traductions médiévales du Coran: une question de cumulativité? (XII<sup>e</sup> –début XVI<sup>e</sup>s.)," *Revue des sciences religieuses* 90 (2016), 471–490.

<sup>6</sup> On the circulation of Latin travellers and pilgrims in the Holy Land after the Mamlūk "Reconquista" (and the abundant Latin literature that resulted from these exchanges), see Camille Rouxpetel, *L'Occident au miroir de l'Orient chrétien: Cilicie, Syrie, Palestine et Égypte (XIIe* –*XIVe siècle)* (Rome: École française de Rome, 2015). See also Chapter 2.4.4 in this volume.

<sup>7</sup> On the linguistic dimension of the Norman conquest of Sicily, which involved the subjection of a large Muslim population (a culturally Graeco-Arabic minority was still Christian at the time of the conquest) to Latin-Christian power, see Alex Metcalfe, *Muslims and Christians in Norman Sicily: Arabic Speakers and the End of Islam* (Abingdon: Routledge Curzon, 2003). For a general overview, see Annliese Nef, *Conquérir et gouverner la Sicile islamique aux XIe et XIIe siècles* (Rome: École française de Rome, 2011). On the Muslim presence in Italy before (a Sicilian emirate, but probably dominating an initially almost totally Greek-speaking population, rather than a Latin one), during, and after Norman rule, see Alex Metcalfe, *The Muslims of Medieval Italy* (Edinburgh: Edinburgh University Press, 2009). See also Chapter 2.4.1 in this volume.

<sup>8</sup> Metcalfe, *The Muslims of Medieval Italy*, 16–24. A number of "embryonic" emirates were founded in peninsular southern Italy during the tenth century. In contrast to the Sicilian emirate, they were never able to survive for more than a few years or decades. On the origins of the emirate of Sicily, see *Les dynamiques de l'islamisation en Méditerranée centrale et en Sicile: nouvelles propositions et découvertes récentes*, ed. Annliese Nef and Fabiola Ardizzone (Rome: École française de Rome, Edipuglia, 2014).

important Muslim, at least partly Arabic-speaking population on the island until the beginning of the thirteenth century.9 The late linguistic history of this tradition will be examined in more detail in the following pages. For now, suffice it to say that this insular population of Arabic speakers was reduced to a few marginalized enclaves during the agitated history of the following dynasty (i.e. the Hohenstaufen, 1194–1266). Only the continental city of Lucera in northern Apulia, a city-garrison specially rebuilt by Frederick II to house loyal Muslim troops transplanted from Sicily, constituted some kind of exception. One must add that the official history of Arabic-speaking Muslim "Sicilians" was radically imperilled by the fall of the Hohenstaufen and their substitution by the Angevin kings of French origin in 1266. It met its end with the destruction of Lucera by the second sovereign of the new dynasty, Charles II, in 1300.10 Based on this sketchy résumé, one could assume that the chronological span of major interaction between Arabic and Latin in Southern Italy was not much longer than in Syria-Palestine, with the very peculiar exceptions of the islands of Malta and Pantelleria.11 We shall see, however, that this view is somewhat misleading, even if it can be considered correct in terms of the concrete interaction between Christians and Muslims.

Indeed, one could claim that the years 1060–1300 assume a special significance for the history of medieval interaction between Arabic and Latin in the entire Mediterranean area. This period witnessed the apogee of several Latin-Christian political entities with partly or predominantly Arabic-speaking populations. After this period, such polities either disappeared, like the Frankish principalities of the eastern Mediterranean, or were transformed into states that were more "classical" and lacked Muslim minorities, like the Kingdom of Sicily. When this had happened, the Iberian kingdoms were left isolated in their status of Christian Latin states with legally tolerated Muslim minorities.12 Moreover, this period between 1060

<sup>9</sup> On the Muslims in Sicily during the reign of Frederick II (1194–1250), see Annliese Nef, "La déportation des musulmans siciliens par Frédéric II: précédents, modalités, signification et portée de la mesure," in *Le monde de l'itinérance en Méditerranée de l'antiquité à l'époque moderne*, ed. Claudia Moatti et al. (Bordeaux: Ausonius, 2009), 455–478.

<sup>10</sup> On the history of Lucera, see Julie A. Taylor, *Muslims in Medieval Italy: The Colony at Lucera* (Oxford: Lexington, 2003); Metcalfe, *The Muslims of Medieval Italy*, 275– 298; Nef, "La déportation des musulmans siciliens par Frédéric II"; and for the linguistic background of the inhabitants, see the literature cited above in fn. 7.

<sup>11</sup> On the peculiar political, cultural, and linguistic context of Malta at the end of Middle Ages, see the excellent contribution by Henri Bresc, "Malte et l'Afrique (1282–1492)," *Revue des mondes musulmans et de la Méditerranée* 71 (1994), 63–74, which also contains some details on the—quite different—situation in Pantelleria.

<sup>12</sup> One could also note that the historiographical tendency to belittle the coexistence of Muslim minorities and Christian majorities distinctly affected the historical perception of other areas of Latin Christendom that possessed a Muslim minority until 1300. See in particular the persistent presence of Muslim (certainly Turkic-speaking) minorities in the Kingdom of Hungary; often at the service of the king, this was a presence that lasted until the thirteenth century, despite the hostility of the papacy—a somewhat interesting parallel to the role of Muslim elites in Sicily. On this not so well-known history, partly masked by the

and 1300 also witnessed peculiar forms of interaction between these three subspaces of the great "corrupting sea:"<sup>13</sup> an element of the most brilliant manifestations of cultural interaction between late medieval Arabic and Latin cultures resulted from the possibilities proffered by pan-Mediterranean impulses echoing from Syria to Spain. A small, but highly influential section of actors at the Sicilian court, for example, originated in the metropolitan Syrian city of Antioch.14 The famous admiral George of Antioch is a testimony to the importance of a Greek-speaking emigration from Antiochia to Norman Sicily,<sup>15</sup> while Theodore of Antioch, in turn, was one of the major translators of Frederick II, and translated from Arabic into Latin, rather than into Greek.16 One could also argue that the extraordinary cultural programme of translation from Arabic into Castilian (but also into Latin) coordinated by Alfonso X of Castile was at least partly rooted in Alfonso's desire to emulate the Arabic-Latin translation programme of Frederick II of Hohenstaufen, Sicilian king as well as emperor of Germany.17 A cousin of Frederick and a pretender to the imperial throne left vacant after Frederick's death and the short-lived reign of his son Conrad IV, Alfonso developed his cultural programme not only as a tentative move to create a specific Castilian state culture, but also with a view to appropriating the universalistic ideology of the Hohenstaufen dynasty. Religious as well as political and geographical factors supported frequent collaboration between Italy and the Iberian Peninsula on Latin translations from Arabic in the late medieval and early modern era. One of the most important "Italian" Latin translations of the Qurʾān during the early sixteenth century, the

construction of a homogeneous Christian nation by traditional historiography, and partly obscured by the lack of sources on the origin of these Muslim populations (with possible connections with the Islamization of the Bulgarians of the Volga and the Eurasian commercial networks), see Nora Berend, *At the Gate of Christendom. Jews, Muslims and "Pagans" in Medieval Hungary, c.1000*–*c.1300* (Cambridge: Cambridge University Press, 2001).

<sup>13</sup> I have borrowed the expression from the title of the seminal essay by Peregrine Horden and Nicholas Purcell, *The Corrupting Sea: A Study of Mediterranean History* (Oxford: Blackwell Publishers, 2000).

<sup>14</sup> On Antioch as a link between Arabic and Latin cultures, see Charles Burnett, "Antioch as a Link between Arabic and Latin Culture in the Twelfth and Thirteenth Centuries," in *Occident et Proche-Orient: Contacts scientifiques au temps des Croisades*, ed. Baudoin van den Abeele et al. (Turnhout: Brepols, 2000), 1–78.

<sup>15</sup> Mario Re and Cristina Rognoni, eds, *Giorgio di Antiochia: L'arte della politica in Sicilia nel XII secolo tra Bisanzio e l'Islam* (Palermo: Istituto siciliano di studi bizantini e neoellenici, 2009).

<sup>16</sup> On Theodore of Antioch, see the bibliography up to 2000 in Wolfgang Stürner, *Friedrich II: Der Kaiser 1220–1250* (Darmstadt: Primus Verlag, 2000), 422–429; for new elements, see Giuseppe Mandalà, "Il Prologo delle Risposte alle questioni siciliane di Ibn Sabʿīn come fonte storica: Politica mediterranea e cultura arabo-islamica nell'età di Federico II," *Schede medievali* 45 (2007), in particular, 67–84.

<sup>17</sup> On the translations at the court of Alfonso X of Castile and León, see Leonard Patrick Harvey, "The Alphonsine School of Translators: Translations from Arabic into Castilian Produced under the Patronage of Alphonso the Wise of Castile (1221–1252–1284)," *Journal of the Royal Asiatic Society* 1 (1977), 109–117; and Vicenç Beltrán, *La corte de Babel: Lenguas, poética y politica en la España del siglo XIII* (Madrid: Gredos, 2005). See also Chapter 2.4.5 in this volume.

translation commissioned by Egidio da Viterbo, was made with the help of an Aragonese convert from Islam.18 The career of some early modern interpreters and translators from Arabic, such as Diego de Urrea, took place between the Maghreb, Spain, and Italy.<sup>19</sup>

These last considerations suggest that the history of Arabic in Italy did not cease with the end of the Norman dynasty in 1194, the death of Frederick II in 1250, or even with the dismantling of Lucera in 1300. Indeed, during the early Renaissance—understood here as the period covering the Trecento and the Quattrocento, that is, 1300–1500—Italy provides a good starting point for an alternative history of Mediterranean Arabic-Latin relations at the end of the medieval and the beginning of the early modern period. It is a history that allows us to ask more precisely how the two languages could have interacted in a Latin-Christian Mediterranean context, in the absence of important Muslim Arabic-speaking minorities. This is a period when interest in Arabic had not totally disappeared, but was receding, according to traditional scholarship. On the one hand, it is true that the conditions for accessing this language had changed considerably, now that the period of creation and the first apogee of the partly Muslim-populated Latin-Christian kingdoms during the twelfth and thirteenth centuries was over.20 On the other hand, the importance of Italy as the centre of a variety of merchant networks, its cultural ascendency, and the vast number of sources preserved on the Apennine Peninsula make it possible to outline an alternative history, which reveals a far more complex story than the traditional narrative would have us believe. This history helps to explain how a considerable number of Jewish and Christian Europeans based on the northern shore of the Mediterranean could have had access to one form of Arabic or another, and could even have tried to transmit a part of this knowledge, without ever having succeeded in perfectly mastering classical Arabic, due to sociolinguistic and pedagogical limitations. In certain ways this Italian Arabic-Latin history of the late Middle Ages and the early Renaissance period represents a missing link in the chain of a *longue durée*-history of Arabic-Latin relations that can be positioned between the "great narrative" of the medieval Arabic-Latin translations, and the birth of modern Orientalism. It shows that, in order to understand the

<sup>18</sup> See García-Arenal and Starczewska, "The Law of Abraham the Catholic"; see also Katarzyna Starczewska, *A Latin Translation of the Qur'an (1518/1621) Commissioned by Egidio da Viterbo: Critical Edition and Introductory Study* (Wiesbaden: Harrassowitz, 2018).

<sup>19</sup> On Diego de Urrea and his voyages between Spain, the Maghreb, and southern Italy, see Fernando Rodríguez Mediano and Mercedes García-Arenal, "De Diego de Urrea à Marcos Dobelio, interprètes et traducteurs des 'plombs,'" in *Maghreb–Italie: Des passeurs médiévaux à l'orientalisme moderne (XIII<sup>e</sup> –milieu XX<sup>e</sup> siècle)*, ed. Benoît Grévin (Rome: École française de Rome, 2010), 141–207.

<sup>20</sup> For Italy after the fall of the Hohenstaufen, see Benoît Grévin, "De Damas à Urbino: Les savoirs linguistiques arabes dans l'Italie renaissante (1370–1520)," *Annales HSS* 703 (2015), 607–635.

sociolinguistic and sociohistorical implications of these processes of interaction, we should study not only the most obvious forms of these relations.

## 4.2 A brief history of Arabic and Arabic knowledge in late medieval and Renaissance Italy

The history of Arabic-Latin relations in Trecento and Quattrocento Italy was considered a minor field in the country's cultural history until quite recently, especially compared to the wide-ranging and continuous study of the complex linguistic and cultural interactions between Greek, Arabic, and Latin in Norman and post-Norman Sicily.21 A series of recent studies, most notably initiated by Angelo Michele Piemontese, has helped to improve this historiographical panorama, even if numerous questions remain unresolved.22 To understand how certain forms of knowledge about the Arabic language could sporadically flourish in various contexts, and be transmitted via Latin, between 1300 and 1500 in the Apennine Peninsula, we must first examine two potential backdrops to the learning and teaching of Arabic. In this context, the persistence of residual Arabic-speaking and/or writing/reading communities on the peninsula acquires particular importance, as does the import of a knowledge of Arabic by Italian travellers stationed, at different times of their lives, in various locations of the Arabic world and Italy.

<sup>21</sup> Metcalfe, *Muslims and Christians in Norman Sicily*; Karla Malette, *The Kingdom of Sicily, 1100–1250: A Literary History* (Philadelphia, PA: Pennsylvania Press, 2005); Metcalfe, *The Muslims of Medieval Italy*; Nef, *Conquérir et gouverner la Sicile islamique*.

<sup>22</sup> On Beltramo Mignanelli, see, among others, Angelo Michele Piemontese, "La lingua araba comparata da Beltramo Mignanelli (Siena 1443)," *Acta Orientalia Academiae Scientiarum Hungariae* 48 (1995), 155–170. On Arabic and Qurʾānic culture in Giovanni Pico della Mirandola's circles, see Angelo Michele Piemontese, "Il Corano latino di Ficino e i corani arabi di Pico e Monchates," *Rinascimento* 36 (1996), 227–273. On the Arabic in the *Hypnerotomachia Polifili*, see Angelo Michele Piemontese, "Le iscrizioni arabe nella 'Poliphili Hypnerotomachia'," in *Islam and the Italian Renaissance*, ed. Charles Burnett, Anna Contadini (London: Warburg Institute, 1999), 199–220. On Arabic and other Oriental languages between Savonarola and Giovanni Pico della Mirandola, see Angelo Michele Piemontese, "Lo studio delle cinque lingue presso Savonarola e Pico," in *Europe and Islam between the 14th and 16th Centuries* (Naples: Istituto Universitario Orientale, 2002), 179–202. On the links between the Arabic teacher Moncada and the court of Urbino, see Angelo Michele Piemontese, "Guglielmo Raimondo Moncada alla Corte di Urbino," in *Guglielmo Raimondo Moncada alias Flavio Mitridate: Un ebreo converso siciliano*. Atti del Convegno Internazionale Caltabellotta (Agrigento), October 23–24, 2004, ed. Mauro Perani (Palermo: Officina di studi medievali, 2008), 151–171. On the study of the Qurʾān in the fifteenth and sixteenth centuries, see Angelo Michele Piemontese, "Il Corano in Italia umanistica," in *Bibbia e Corano: Edizioni e ricezioni*, ed. Carmela Baffioni et al. (Milan: Bulzoni Editore, 2016), 31–66. These papers, which contain rich bibliographical references, are written in a beautiful, intricate baroque Italian, sometimes with a deliberate attempt to create a labyrinth-like circular progression. Hence, some misunderstandings have occurred in their reception in non-Italophone milieus.

#### 4.2.1 FROM INSIDE: A RESIDUAL BUT PERSISTENT LANGUAGE

Arabic was already in decline in southern Italy (Sicily was the only region where it was widely spoken) when the Norman dynasty was replaced by the Hohenstaufen. The latter retained parts of the Norman political ideological heritage, since the Hohenstaufen claim to the throne was based on the marriage of the German emperor Henry VI to the Norman heiress Constance of Sicily, who was the mother of the future king and emperor, Frederick II.23 The idea that Frederick II, king of Sicily from 1198 to 1250, resided in Palermo with his court, is cherished in historiography, but mostly false. After his return from Germany, the king resided mainly on the Italian mainland, between Apulia and Campania.24 Frederick II fought several minor wars in Sicily, particularly during the 1220s. His aim was to subjugate the last pockets of Muslim resistance, after this minority had retreated to the interior of the island, e.g. to the region south of Monreale, to avoid being annihilated during the civil wars that raged during the emperor's childhood. There were traces of unrest among the residual Muslim population on the island until very late in his reign.<sup>25</sup> After 1224, however, the year that saw the foundation of Lucera,26 these movements can be considered insignificant. Around 1200, relatively little-known groups of originally Arabic-speaking (or Arabic-Greek speaking) Christian families existed, some of whom still played an important role under the rule of the Hohenstaufen dynasty, and probably maintained a working knowledge of some form of Arabic for part of the thirteenth century.27 However, no obvious testimony can help us gauge the extent to which these skills survived over time. Some of the most learned men active at the court of Frederick II originated from this mostly Palermitan milieu, like the mathematician John of Palermo, who was probably selected for a diplomatic mission to Ḥafṣid Tunis in 1240 because of his linguistic abilities.<sup>28</sup> Still, the choice of the *Mašriqī*, Theodore

<sup>23</sup> Wolfgang Stürner, *Friedrich II: Die Königsherrschaft in Sizilien und Deutschland 1194–1220* (Darmstadt: Primus Verlag, 1992), particularly 34–85.

<sup>24</sup> On this historiographical problem, see Benoît Grévin, "Linguistic Cultures and Textual Production in Palermo, from the End of the 11th to the End of the 15th Century," in *A Companion to Medieval Palermo: The History of a Mediterranean City from 600 to 1500*, ed. Annliese Nef (Leiden: Brill, 2013), 413–415.

<sup>25</sup> Stürner, *Friedrich II: Der Kaiser*, 66–74; Metcalfe, *The Muslims of Medieval Italy*, 275–280.

<sup>26</sup> Nef, "La déportation des musulmans siciliens par Frédéric II."

<sup>27</sup> On this particular milieu and its complex and frequently Graeco-Arabic linguistic roots, see Henri Bresc, "Arabi per lingua, greci per rito, i Mozarabi di Sicilia con e dopo Giorgio," in *Giorgio di Antiochia: L'arte della politica in Sicilia nel XII secolo tra Bisanzio e l'Islam*, ed. Mario Re and Cristina Rognoni (Palermo: Istituto siciliano di studi bizantini e neoellenici, 2009), 263–282; for the question of a "Mozarabic" Sicilian milieu: see Nef, *Conquérir et gouverner la Sicile islamique*, 90–123; and Giuseppe Mandalà and Marcello Moscone, "Tra latini, greci e 'arabici': ricerche su scrittura e cultura a Palermo fra XII e XIII secolo," *Segno e testo: International Journal on Manuscripts and Text Transmission* 7 (2009), 143–238.

<sup>28</sup> On John of Palermo and his links to Frederick II, see the bibliography in Stürner, *Friedrich II: Der Kaiser*, 387–397; and Mandalà, "Il prologo delle Risposte," 76, 79–86. The court's interest in sending John on a Tunisian mission is known

of Antioch, to write the official letter to the Tunisian sultan on this mission, suggests that these remaining Arabic-speaking Sicilians were not necessarily sufficiently versed in the stylistic intricacies of classical or post-classical Arabic to render the complexity of imperial Latin phraseology into fashionable courtly Arabic.29 This raises questions concerning the kind of Arabic mastered by these Latinized Sicilian courtiers with an "Arabophone" background, as well as the linguistic level required to translate highly formalized Latin into Arabic and vice versa—a problem to be examined further.

Apart from these residual elites from Palermo or other Sicilian towns, a nucleus of Muslims loyal to the throne led organized lives in Lucera from 1224 to the end of the dynasty in 1266. They managed to resist the first two Angevin kings' attempts to suppress this religious and institutional enclave until 1300.30 A papal letter provides an invaluable testimony concerning the linguistic status of this population ten years after its transplantation from Sicily to Apulia. Pope Gregory IX was quite unhappy at the prospect of a Muslim colony installed at the gates of the *Patrimonium Petri* by a Christian power that was often at odds with the papacy. In 1233 he sent a letter to the emperor and king to demand permission to send friars to Lucera with the aim of converting this "colony" to Christianity. In the missive, he underlined the fact that the Muslim population understood Italian (*italicum idioma*) very well, and could thus be subjected to preaching.<sup>31</sup> Although originating in the last extant Muslim pockets of Sicily, the Muslim population of Lucera was already bilingual. Its Arabic was perhaps already semantically contaminated to a large extent by Romance elements, in a way that would have had parallels in the earliest stages of the formation of the Maltese language. Gregory's request was denied, and the persistent adherence to Islam on the part of the troops based in Lucera was one of the factors in the renewal of the conflict between the Hohenstaufen rulers and the papacy—a conflict which ultimately led to the dynasty's downfall.

through the only preserved *registrum* of Frederick's chancery (for the end of 1239 and the beginning of 1240), see *Il registro della cancelleria di Federico II del 1239–1240*, ed. Cristina Carbonetti Venditelli (Rome: Istituto Storico Italiano per il Medioevo, 2002). On this precise embassy and the sociolinguistic questions raised by the selection of Frederick II's staff, see Chapter 4.3.1.

<sup>29</sup> *Il registro della cancelleria*, no. 575 (February 10, 1240), ed. Carbonetti, 541–542: *Magistro Teodoro de litteris scribendis regi Tunisi in Arabica*.

<sup>30</sup> On the fall of Lucera, see the recent publication by Benjamin Scheller, "Assimilation und Untergang: Das muslimische Lucera in Apulien und sein gewaltsames Ende im Jahr 1300 als Problem der Globalgeschichte," in *Europa in der Welt des Mittelalters: Ein Colloquium für und mit Michael Borgolte*, ed. Tillmann Lohse and Benjamin Scheller (Berlin: De Gruyter, 2014), 141–161.

<sup>31</sup> See *Historia diplomatica Friderici secundi*, ed. Jean Louis Alphonse Huillard-Bréholles, vol. 4 (Paris: Plon, 1855), 452 (August 27, 1233): "Gregorius etc. Federico etc. [. . .] imperialem mansuetudinem rogandam duximus et hortandam quatenus sarracenis qui Capitanate Nuceriam incolunt et italicum idioma non mediocriter ut fertur intelligunt, per tuas litteras firmiter dare debeas in mandatis ut fratres ordinis Predicatorum, pacis angelos, quos ad eos cum exhortationis verbo dirigimus, in pace suscipiant, patienter audiant et prudenter iis que pro salute sua proponuntur intendant."

The continuity of religious tradition would have assured a minimum knowledge of Qurʾānic (and thus classical) Arabic among these last, officially protected Muslims of the Sicilian kingdom. Certain high officials at the court of Frederick II, such as John the Moor, had strong links with this milieu, and were probably able to navigate to some extent between Arabic in various forms (classical and Sicilian dialectal) on the one hand, and Romance and Latin, the almost monopolistic language of the imperial Sicilian administration, on the other. The papal letter nevertheless suggests a progressive linguistic acculturation of these transplanted Sicilian Muslims, which raises doubts concerning their knowledge of classical Arabic, as well as the exact nature of the Arabic probably still spoken in this community.

Paradoxically, if Arabic survived as a written and (very probably still largely) spoken language on the island of Sicily after 1300, this was due not to the presence of a Muslim population, but to the existence of a massive network of Arabic-speaking Jewish communities.32 These communities had remained attached to the use of Arabic in the sociolinguistic form of local Judaeo-Arabic dialects after the Norman conquest. Although its exact significance is open to interpretation, the famous quadrilingual inscription carved on behalf of the priest Grysantus in memory of his mother, proves that, in Norman Palermo, Judaeo-Arabic had acquired a symbolic status of some sort, together with Latin, "standard" written Arabic, and Greek.<sup>33</sup> Furthermore, the relevance of these Jewish communities was increased by events that would lead to an even more spectacular pervasiveness of Sicilian Judaeo-Arabic practices during the thirteenth century. Following persecution by the last Almohads in the Maġrib al-Aqṣā, i.e. far western North Africa, many Jews fled to more hospitable regions during the 1230s. Frederick II, hoping to repopulate the parts of Sicily that had been economically and demographically depressed by the internecine wars between Christians and Muslims during his minority, invited them to settle on the island, which they did, in around 1239–1240.34 Jews made up an estimated five per cent of the total population of Sicily in the fourteenth and fifteenth centuries.35 We possess numerous testimonies for the use of Judaeo-Arabic, including onomastic and other clues, as in the Aragonese dynasty's use of Sicilian Jews as official translators and ambassadors to Tunis. They suggest that a notable part of the Jewish community maintained a written and oral practice of the language until its expulsion from Sicily in

<sup>32</sup> On the Jews in medieval Sicily, see Henri Bresc, *Arabes de langue, juifs de religion: L'évolution du judaïsme sicilien dans l'environnement latin, XIIe –XV<sup>e</sup> siècles* (Paris: Bouchene, 2001).

<sup>33</sup> On Grysantus and the inscription in honour of his mother, see Nef, *Conquérir et gouverner la Sicile islamique*, 101–107.

<sup>34</sup> On the migration of the Jews from *Ġarb/Garbum* (= *al-Maġrib al-aqṣā*) see Giuseppe Mandalà, "La migration des juifs du Garbum en Sicile (1239)," in *Maghreb-Italie: Des passeurs médiévaux à l'orientalisme moderne (XIIIe –milieu XX<sup>e</sup> siècle)*, ed. Benoît Grévin (Rome: École française de Rome, 2010), 19–48.

<sup>35</sup> Bresc, *Arabes de langues, juifs de religion*, 88.

1492–1493.36 From what we know about the origins of these communities, we can assume that the Judaeo-Arabic dialects spoken across Sicily were far from uniform, even if this issue still awaits further investigation. At least for some time after the immigration of the Moroccan Jews, a considerable difference must have existed between the linguistic practices of the older communities, whose language was perhaps more akin to the Judaeo-Arabic dialects of Ifrīqiya, and the newer ones, with their Moroccan dialects. As a testimony to the circulation of Arabic-speaking practices in the Latin Mediterranean world, we also know that, due to the deteriorating situation of Jewish communities in late fourteenth-century Castile and Aragon, some prominent literati with a Spanish Judaeo-Arabic background settled on the island around 1390, and even brought with them some important scientific Arabic texts.37 Until their expulsion, the Jewish-Sicilian communities maintained specific scientific and cultural traditions, which made them potential candidates for teaching Arabic to Christian clerks or humanists.<sup>38</sup> Their working scientific language, in addition to the two sacred tongues of Judaism (Hebrew and Judaeo-Aramaic), was an Arabic transliterated into Hebrew characters. To some extent, this scientific language can also be described as culturally Judaeo-Arabic. Structurally, however, it was not directly connected to the Judaeo-Arabic dialects spoken by Jews throughout the ages.39 In addition to these linguistic skills, the necessity or allure of cooperation with Christian authorities and Christian society incited them to acquire an often thorough knowledge, not only of Romance Sicilian,

<sup>36</sup> For the role of these Jewish ambassadors, sent to Muslim rulers as late as the fifteenth century, see Bresc, *Arabes de langues, juifs de religion*, 40.

<sup>37</sup> On this point, see Giuseppe Mandalà, "Da Toledo a Palermo: Yiṣḥaq ben Šelomoh ibn al-Aḥdab in Sicilia (ca. 1395–1396–1431)," in *Flavio Mitridate mediatore fra culture nel contesto dell'ebraismo siciliano del XV secolo*. Atti del convegno internazionale di studi, Caltabellotta, June 30–July 1, 2008, ed. Mauro Perani, Giacomo Corazzol (Palermo: Officina di studi medievali, 2012).

<sup>38</sup> See Chapter 4.2.3 for the career of Guglielmo Raimondo Moncada.

<sup>39</sup> On Judaeo-Arabic, see Joshua Blau, *The Emergence and Linguistic Background of Judaeo-Arabic* (Jerusalem: Ben-zvi Institute, 1999, first ed. 1965); and Geoffrey Khan, "Judaeo-Arabic," in *Encyclopedia of Arabic Language and Linguistics*, ed. Kees Versteegh, vol. 2 (Leiden: Brill, 2006), 526–536. The concept includes the spoken Judaeo-Arabic dialects from past to present, with all their peculiarities, as well as the original written Arabic production of the Arabic-writing Jewish communities, traditionally classified as Middle Arabic texts with some specific Jewish sociolinguistic idiosyncrasies. The relationship between the two poles of the concept is thus akin to the somewhat ambiguous pairing that characterizes Arabic historical dialectology and the study of "Middle Arabic writings." The Judaeo-Arabic oral dialects had some sort of impact on the written forms of Judaeo-Arabic, but they were not its matrix. The higher the textual level, the lesser the possibility that the impact of the spoken dialect could have been statistically important; paradoxically, one could even argue that stylistic and linguistic interferences with the spoken practices of the communities form part of the peculiarities characteristic of copies of scientific texts of Muslim origin in the Arabic-writing Jewish communities, thereby raising the question of the boundaries of the concept of written "Judaeo-Arabic." On such Arabic manuscripts written in Hebrew characters, recently attributed to the Sicilian communities, see Giuseppe Mandalà, "Un codice in caratteri ebraici dalla Trapani degli Abbate (vat. Ebr. 358)," *Sefarad* 71 (2011), 7–24.

but also of Latin. This is a competence attested to within the community during the times of Frederick II and Charles I of Anjou—the latter commissioned an important medical translation from Arabic into Latin from a prominent Sicilian Jewish scholar.40 We will see below the extent to which, and with what limitations, this multilingual knowledge of scientific written Arabic and oral dialectal Judaeo-Arabic in its Sicilian variant, in addition to Romance dialects and administrative-scientific Latin, was transplanted to the continental courts of Renaissance Italy during the second half of the fifteenth century.<sup>41</sup>

We cannot expect that Jewish communities in continental northern, central, or even southern Italy would have used a form of Arabic as an everyday language in the same way that their Sicilian counterparts did. However, there are indications that members of Jewish communities in Campania or Tuscany, for example, had at least some working knowledge of Arabic during the fourteenth and the fifteenth centuries. In a booklet on Arabic which is discussed below, the Sienese diplomat and merchant Beltramo Mignanelli comments incidentally on the diffusion of comparable knowledge among members of the community of Siena.42 We also have clues to some interest in Arabic in the Jewish communities of Naples.43 The knowledge available in these circles would have been linked to the general pervasiveness of Judaeo-Arabic learning in a Jewish-Mediterranean setting and its limited diffusion as a scientific, philosophical, and medical language in communities based outside the areas where Arabic had formerly been spoken.

#### 4.2.2 FROM OUTSIDE: MERCHANTS AND THEIR FELLOW TRAVELLERS

A second line of investigation concerns the networks established by Italian merchants with the Maghreb, particularly Tunis, and above all, the Mashreq, that is, Egypt and Syria-Palestine, particularly Damascus and Aleppo. These merchants generally hailed from areas controlled by the

<sup>40</sup> On this translation, executed by Farāğ b. Sālim (Latinized Faracio, Faresche, or Faragius) for Charles I of Anjou between 1278 and 1282, see Gian Luca Borghese and Benoît Grévin, "Aspects linguistiques de la diplomatie sicilienne au XIII<sup>e</sup> siècle (1220–1290)," in *Les langues de la négociation: Approches historiennes*, ed. Dejanirah Couto and Stéphane Péquignot (Rennes: Presses Universitaires de Rennes, 2017), 88–89.

<sup>41</sup> See Chapter 4.2.3.

<sup>42</sup> See the edition of this introduction to Arabic, a sort of preface to a bilingual anthology of the Psalms written in 1443 in Siena and still kept there, in Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 157: "Et etiam hebrei stolidi non tamen omnes extollunt linguam ipsorum propter primam cuius argumentum est falsissimum et sic esse probavi multis ex eorum doctoribus qui proh dolorem Senis prope me habitant et maxime arabicum scientibus qui negare non possunt nec negant."

<sup>43</sup> On this point, see Ofra Tirosh-Becker [in Hebrew], "Ha-glosot ha-'araviyot she-be- 'Makre Dardeke' be-nusah ha-'italki: mah tivan?" ["The Arabic Glosses Contained in the 'Makre Dardeqe' in its Italian Edition: What is their Nature?"], *Italia: Studi e ricerche sulla storia, la cultura e la letteratura degli ebrei di Italia* 9 (1990), 37–77.

commercial empires of Genoa and Venice, but not always. Again, the case of Beltramo Mignanelli allows a glimpse at the sort of career that acquainted adventurous Italians with Arabic, and in certain cases motivated them to try to transmit the knowledge they had acquired in the Arabic sphere to their contemporaries back home.44 Mignanelli, a merchant and diplomat from Siena, was born in 1370, and travelled extensively, perhaps first to the Maghreb,45 but above all to Egypt and in Syria, and even as far as Iraq, Iran, and Arabia. He headed diplomatic missions in the Orient on behalf of the Visconti, briefly masters of Siena at the end of the fourteenth century. Back in Italy, he made use of his knowledge to serve as a translator for the papal curia, for example, when important negotiations took place between certain Oriental churches and the restored Roman papacy during the council of Florence in 1439. Mignanelli reached an advanced age (d. 1455) and had time to write many texts related to his experiences in Islamic societies. He left concrete evidence and various traces of his knowledge of Arabic. Some of his numerous Latin historiographical and polemical works (*Ascensus Barcoch*; *De ruina Damasci*<sup>46</sup>) contain transcriptions of Arabic with a strong Egyptian flavour. Moreover, towards the end of his life, he wrote a Latin *Libellus*, which features a brief presentation of the linguistic characteristics of Arabic and an anthology of the Psalms. The latter includes three Latin versions as well as the Arabic text in Arabic characters lacking vocalization.47

Beltramo Mignanelli is exceptional in at least two ways: he did not depend on the Venetian or Genoese network for his Oriental travels, and he chose to leave some tangible traces of his knowledge of Arabic in his numerous writings, composed during his old age in Italy. However, he was certainly not the only Italian merchant of the Quattrocento who displayed an important knowledge of Arabic, which then had an effect on the contemporary production of Latin texts. With regard to historiography and literature, we know of the testimonies of Niccolò de' Conti (1395–1469), a merchant from the Venetian dependency of Chioggia. He learned Arabic

<sup>44</sup> On Beltramo Mignanelli, see Piemontese, "La lingua araba comparata da Beltramo Mignanelli," as well as Nelly Mahmoud Helmy, "Memorie Levantine e ambienti curiali: L'Oriente nella vita e nella produzione di un senese del Quattrocento: Beltramo di Leonardo Mignanelli," *Quaderni di storia religiosa* 13 (2006), 237–268; and Nelly Mahmoud Helmy, *Tra Siena, l'Oriente e la curia: Beltramo di Leonardo Mignanelli e le sue opere* (Rome: Istituto Storico Italiano per il Medioevo, 2013).

<sup>45</sup> This is claimed in traditional modern narratives, but finds almost no echo in the medieval sources. See Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 158.

<sup>46</sup> An extensive description of his life and work can be found in Mahmoud Helmy, *Tra siena, l'Oriente e la curia*, with an edition of the *De ruina Damasci* (ibid., 307– 340), and the *Ascensus Barcoch* (ibid., 341–387).

<sup>47</sup> Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 155–157, 165 (transcription of the explanations on the nature of Arabic, and reproduction of the first page of the anthology of the Psalms); and Mahmoud Helmy, *Tra Siena, l'Oriente e la curia*, 266–270 (description of the *Liber de variantibus Psalterii*). Nelly Mahmoud Helmy has announced that an edition of this text will be forthcoming.

and Persian during his travels, which took him as far as the south of modern-day Vietnam, and which even forced him to convert to Islam on his way back to Italy. Poggio Bracciolini (d. 1459) then set down his travels in writing in a humanistic Latin of the highest quality.48 Another example is Emmanuel Piloti (fl. ca*.*1371–1420/1438), a Cretan merchant of Venetian origin— Crete had been a possession of the republic since the thirteenth century. He traded on the Egyptian market for a long time. After settling in Italy in his later life, he wrote a treatise on the Orient. The original version has been lost, and the text survives only in a Middle French version.49 It seems characteristic of the relations between Arabic and Western European languages in the late medieval and early Renaissance period that the greater part of these testimonies were originally written in Latin. It is probable that only Beltramo Mignanelli had the knowledge to write directly in this language without some sort of intermediary. In the case of Niccolò de' Conti, it is certain that he dictated his history in Italian, and that it was later Latinized. The status of Piloti's original text is less certain. It seems, however, that it was originally written in Italian and only later translated into Latin as well as Middle French.50 Among these three merchants, Beltramo Mignanelli is the only one to have left tangible evidence of his degree of knowledge and the type of Arabic with which he was familiar. To a certain extent, he must have been able to navigate between a very dialectal and a more classical form of Arabic. Moreover, he was not a bad Latinist, albeit certainly not a firstrate one, according to the standards of fully blooming humanism. Speaking in general terms, however, the linguistic universe of these merchants was probably more conditioned by switching back and forth between Italian and an everyday variant of *Mašriqī*, i.e. Middle Eastern Arabic. However, when they wanted to commit their information to memory, they chose Latin, the prestige language—oral or written—in Italy of the Quattrocento, particularly with a view to the possibility of being acknowledged by the major Latinspeaking and Latin-writing institution, i.e. the papal curia.

These merchants' lives, travels, and alleged or proven linguistic capacities were certainly far from ordinary, but they were not unique. A

<sup>48</sup> On Niccolò de' Conti, see Francesco Surdich, "Conti, Niccolò de," *Dizionario Biografico degli Italiani* vol. 28 (Rome: Treccani, 1983), 457–460. The Latinized narration of his adventures is edited in Poggio Bracciolini, *De varietate fortunae*, ed. Outi Merisalo (Helsinki: Suoamalainen Tiedeakademia, 1993), 153–174.

<sup>49</sup> On Piloti, see, most recently, Antonio Musarra, "Piloti, Emanuele," *Dizionario Biografico degli Italiani* 83 (Rome: Treccani, 2015), 678–679; Damien Coulon, "Regards contrastés sur les musulmans du sultanat mamlūk par des marchands chrétiens à la fin du Moyen Âge," *Revue des Sciences Religieuses* 90 (2016), 570–579.

<sup>50</sup> On the problem of the status of Piloti's text, see the contradictory indications of Coulon, "Regards contrastés sur les musulmans du sultanat mamlūk," 570 (original treatise in Latin); Emmanuel Piloti, *Traité d'Emmanuel Piloti sur le passage en Terre sainte (1420)*, ed. Pierre-Herman Dopp (Leuven: Nauwelaerts, 1958), introduction (original treatise in Italian); and the synthesis of Musarra, "Piloti, Emmanuele" (initial version in Italian, with subsequent versions in Latin and French).

prosopographical study of Italian merchants in the Maghreb and the Orient of the Trecento and Quattrocento would certainly provide us with further clues regarding the potential extent of Arabic skills in the Italian merchant milieu of the period.<sup>51</sup> Some testimonies suggest a broader knowledge than would have been assumed traditionally, at least in specific circles. Unfortunately, however, the evidence is often ambiguous, if not profoundly frustrating for researchers interested in this issue. In one of his *Seniles*, written in 1370, Petrarch expresses disdain for Arabic poetry to his friend Giovanni Dondi dell'Orologio.52 It is obvious that he did not read Arabic poetry in the original language. Consequently, we can deduce that some (possibly Venetian?) acquaintance had passed on some kind of information about Arabic literature to him. What kind of information this could have been remains open to speculation. However, two testimonies from the end of the period attest that, at the end of the Quattrocento, it would have seemed conceivable to use the merchant networks of Venice as a logistic base to acquire a profound knowledge of Arabic texts, later transmitted back into Latin.

These testimonies provide us with information on two medical authorities, Girolamo Ramusio (d. 1486) and Andrea Alpago (d. 1521). Both stayed in Damascus for a long period, where both were employed as physicians of the Venetian consulate, with Andrea Alpago arriving in Damascus at the time of Ramusio's death. Both pursued the same objective: to revise and improve the older translation of Ibn Sīnā's (d. 427/1037) *al-Qānūn fī l-ṭibb*, i.e. the *Canon* of Avicenna.<sup>53</sup> Ramusio died before accomplishing this, but left a partial transcription of the *Qānūn* with working notes. Alpago, in turn, had time to return to Italy. From the Venetian republic, he obtained a position at the University of Padua, and, among other works, left an *Interpretatio arabicorum nominum*, that is, an explanation of the medical terms used by Ibn Sīnā/Avicenna, which was published in Venice a few years after Alpago's death.54 In the two cases of Ramusio and Alpago, the similarity of their backgrounds, their language acquisition techniques, and their objectives is striking. Literati with precise scientific objectives used the institutionalized merchant infrastructure of the Venetian republic to spend long

<sup>51</sup> See some possible interpretations in Grévin, "De Damas à Urbino," particularly 613–618.

<sup>52</sup> See Pétrarque, *Lettres de la vieillesse XII–XV*, ed. Elvira Nota, trans. Jean-Yves Boriaud, presentation Ugo Dotti (Paris: Les Belles Lettres, 2006), ep. XII, 2 [68], 92–95. On this subject, see Benoît Grévin, "Connaissance et enseignement de l'arabe dans l'Italie du XV<sup>e</sup> siècle: Quelques jalons," in *Maghreb-Italie: Des passeurs médiévaux à l'Orientalisme moderne (XIIIe siècle–milieu XX<sup>e</sup> siècle)*, ed. Benoît Grévin (Rome: École française de Rome, 2010), 103–104.

<sup>53</sup> On Alpago and his activity as a translator from Arabic and corrector of the Latin version of the *Canon*, see Giorgio Vercellin, *Il Canone di Avicenna fra Europa e Oriente nel primo Cinquecento: L'Interpretatio Arabicorum nominum di Andrea Alpago* (Torino: Utet, 1991). On Ramusio, see Danielle Jacquart, "Arabisants du Moyen Âge et de la renaissance: Jérôme Ramusio († 1486), correcteur de Gérard de Crémone († 1187)," *Bibliothèque de l'école des chartes*, 147 (1989), 399–415.

<sup>54</sup> Published in Vercellin, *Il canone*, 52–140.

BENOÎT GRÉVIN

periods of residence in Syria. Their aim was to learn a vast amount of a very specialized form of Arabic, and to elaborate working tools in Latin on the basis of their acquired knowledge. Once again, these complex intellectual operations did not result in the writing of Italian treatises, but of Latin translations and lexicons. In around 1500, it was hardly conceivable to teach medicine at an academic level in any other tongue than Latin. Furthermore, Ramusio and Alpago's objective was to complete and correct a traditional corpus that had already been widely disseminated for centuries thanks to the great wave of medieval translations. The case of these two physicians also shows that the prevalent methods of acquiring linguistic information and translating skills had undergone substantial modifications. As mentioned in the introduction to this chapter, most Latin-Christian powers ceased to control large pools of Arabic-speaking populations and associated cooperative literate elites over the course of the thirteenth century. Consequently, linguistic skills had to be sought abroad. With the collapse of the Mamlūk state in 1517, the conditions for obtaining this linguistic knowledge altered substantially once again.

#### 4.2.3 AT THE CENTRE: COURTS, HUMANIST NETWORKS, AND THEIR ATTRACTIVENESS

In Quattrocento Italy, knowledge of Arabic either persisted in partly Arabic-speaking and Arabic-writing Jewish communities in Sicily, or was acquired thanks to the linguistic exploitation of mercantile networks in the East. The Italian Quattrocento (in particular, the years 1460–1490) stands out, however, because of the increasing attention accorded to Arabic (and, in some ways, to Arabic artefacts) in a number of the most prestigious Italian courts. This attention was intrinsically linked to a shift in Italian humanism and humanist tastes.

In the Italian history of Arabic-Latin interaction, a handful of manuscripts suggest that ideological objectives and sociolinguistic possibilities converged at certain moments. This convergence resulted in the creation of artefacts that stand symbolically for the will to construct a bilingual/ bigraphical culture. Unsurprisingly, the Norman Kingdom of Sicily lived through one of these moments. A still relatively ill-studied series of preserved trilingual Latin–Greek–Arabic manuscripts containing either the Psalms or parts of the New Testament are a testimony to the desire to create ecclesiastical writings able to reflect the trilingual universe of Norman (and, in some measure, Hohenstaufen) Sicily.55 Although we do not know

<sup>55</sup> Angelo Michele Piemontese, "Codici greco-latino-arabi in Italia fra XI e XV secolo," in *Libri, documenti, epigrafi medievali: Possibilità di studi comparativi*, ed. Francesco Magistrale et al. (Spolete: Centro italiano di studi sull'Alto Medioevo, 2002), 455; Guglielmo Cavallo, "La cultura italo-greca nella produzione libraria," in *I Bizantini in Italia*, ed. Guglielmo Cavallo et al. (Milan: Schweiwiller, 1982), 525; Nef, *Conquérir et gouverner la Sicile islamique*, 214–215.

the exact context of their creation, it is clear that they originated within the boundaries of the kingdom. Other Latin-Arabic manuscripts, which have been present on Italian soil since the Middle Ages, are of less clear origins. A manuscript kept in the Municipal Library of Poppi in Tuscany, for example, also contains a bilingual version of the Psalms, together with an Arabic-Latin glossary. Although linked to the monastery of Camaldoli, it is not clear to my knowledge whether it was produced in Italy or rather in the East.56 Another manuscript in the Biblioteca Riccardiana of Florence has a somewhat less nebulous history. This voluminous Arabic-Latin lexicon, already edited in 1871, was certainly brought to Tuscany during the Quattrocento, but is clearly of Iberian manufacture, providing further proof of the continuous interaction between the three major zones of Arabic-Latin entanglement during the late Middle Ages.<sup>57</sup>

In contrast to these examples, the ms. Urb. Lat. 1384 in the Vatican Library was obviously produced in Italy as the result of an impressive textual and artistic programme.<sup>58</sup> This manuscript was conceived from the start as a partly Arabic-Latin artefact. This was not the case, however, with every Arabic-Latin Italian manuscript, as we shall see. The ms. Urb. Lat. 1384 contains three sections. The first section presents an elegantly calligraphed bilingual Arabic-Latin version of a treatise on talismanic magic, attributed to an unknown author named Ibn al-Ḥātim.59 The second section includes only the Latin translation of astronomic tables, attributed by the translator to Arabic sources. Finally, the third section contains an Arabic-Latin version of *sūra*s 21 and 22 of the Qurʾān. The text is richly illuminated, the astro-magical treatise in particular adorned with elegant medallions featuring the constellations, which are discussed in their Arabic nomenclature. The manuscript also contains a series of glossaries


<sup>56</sup> Giovanni Cipriani, "Poppi: Biblioteca comunale," in *Inventari dei manoscritti delle Biblioteche d'Italia*, vol. 6, ed. Giuseppe Mazzatinti (Forlì: Luigi Bordandini, 1896), 137; Maria Elena Cataluccio Magheri and Antonio Ugo Fossa, *Biblioteca e cultura a Camaldoli: Dal Medioevo all'Umanesimo* (Rome: Anselmiana, 1979), 442; *I Manoscritti della Biblioteca comunale di Poppi (secoli XII–XVI): Un esperimento di catalogazione diretto da Emanuele Casamassima*, ed. Guglielmo Bartoletti and Ilaria Pescini (Florence: Editrice Bibliografica, 1993), 66–67; Piemontese, "Codici greco-latino-arabi in Italia fra XI e XV secolo," 464.

<sup>59</sup> On this point, see: Kristen Lippincott and David Pingree, "Ibn al-Ḥātim and the Talismans of the Lunar Mansions," *Journal of the Warburg and Courtauld Institutes* 50 (1987), 56–58; Kristen Lippincott, "More on Ibn al-Ḥātim," *Journal of the Warburg and Courtauld Institutes* 51 (1988), 188–190; Marc Oliveras, "El *De imaginibus caelestibus* de Ibn al-Ḥātim," *Al-Qanṭara* 31, no. 1 (2009), 171–220; Grévin, "Editing an Illuminated Arabic-Latin Masterwork of the Fifteenth Century."

that explain certain Arabic terms directly transcribed in Latin characters in the magical and Qurʾānic translation. All this is complemented by three precious dedicatory introductions. Thanks to these prefaces, we are able to retrace a part of the manuscript's history, and above all, to identify both the person who designed it and the man for whom he worked. The sumptuous bilingual artefact was made at the request of Federico III, Duke of Urbino (r. 1444/1474–1482), known as an exceptional bibliophile and patron. A team of scribes and painters worked under the supervision of Guglielmo Raimondo Moncada, alias Flavius Mithridates, who was responsible for the prefaces and, most probably, for the greater part of the translation. Although Moncada was still a relatively obscure figure in the history of Quattrocento humanism during most of the twentieth century, he is now far better known.60 Born a Sicilian Jew in the small town of Caltabellotta, he converted to Christianity probably during his adolescence or young adulthood, and began building a career on the basis of his linguistic skills and cultural heritage. In the 1460s, he tried to convince powerful Christian patrons to help him establish a school for the teaching of Hebrew, Aramaic, and Arabic, the three languages with which he was acquainted thanks to the rabbinic education he had received during his childhood. It is highly probable that his mother tongue was a variety of the Sicilian form of Judaeo-Arabic, and that he lived in a bilingual Romance-Arabic community. Under the protection of cardinal Giovanni Cybo, the future Innocent VIII (sed. 1484–1492), he introduced himself to some of the major courts of central Italy. At the end of the 1470s, he worked at the court of Urbino. The ms. Urb. Lat. 1384 would have been finished in around 1482, at a time when he would have been well on his way to gaining entry to the papal court as official master of Semitic languages. A quintilingual sermon that he delivered before Pope Sixtus IV (sed. 1471–1484) on Good Friday, 1481, can be seen as the apogee of his career.61 After this performance, he remained in favour at the papal court, where he obtained an official position, but was quite soon disgraced under unclear circumstances. Guglielmo Raimondo Moncada had taken to nicknaming himself Flavius Mithridates, thus alluding to his polyglot

<sup>60</sup> The studies on Moncada/Mithridates, which were not so abundant during the greater part of the twentieth century, have increased notably, with numerous editions of the translations from Hebrew or Aramaic into Latin, and various studies. See, most notably, the proceedings of the two Sicilian congresses of 2004 and 2008 (published in 2008 and 2012), i.e. Perani, ed., *Guglielmo Raimondo Moncada alias Flavio Mitridate*; and also Perani and Corrazol, ed., *Flavio Mitridate mediatore*, with an extensive bibliography. For one of the best recent editions of Moncada's translations from Hebrew (with some of his comments on the relations between Arabic and Hebrew), see Gersonide, *Commento al cantico dei Cantici nella traduzione ebraico-latina di Flavio Mitridate: Edizione e comment del ms. Vat. Lat. 4273 (cc. 5r–54r)*, ed. Michela Andreatta (Florence: Olschki, 2009).

<sup>61</sup> The sermon is superbly edited and commented in Flavius Mithridates, *Sermo de passione Domini*, ed. Chaim Wirszubski (Jerusalem: The Israel Academy of Sciences and Humanities, 1963).

skills.62 In the years 1486–1487, he attached himself to the young polymath and philosopher Giovanni Pico della Mirandola, whom he mentored and taught in Hebrew, Aramaic, and Arabic, and for whom he executed translations from Hebrew to Latin. After a rupture with his brilliant pupil, Moncada disappeared into the papal jails, apparently in 1489.63 Important details of his early life as well as of his late years are missing, and not all of his extant manuscripts have been edited so far. However, we know enough about his career to retrace the mechanisms of his social ascension. What opened the doors of these courts and allowed him to gain access to the most fashionable intellectual milieus of his time was clearly a new type of demand for the sort of knowledge this ambitious convert could offer. The Italian courts and humanistic circles of the 1470s and 1480s were no longer satisfied with the mastery of Latin and Greek. They also began to display an interest in Hebrew and Aramaic, conceived as access gates to a broader and philologically more precise biblical culture as well as to esoteric speculations like Kabbalah. This new fascination also entailed a rising interest in Arabic. Those Sicilian Jews who lived in perpetual interaction between the three languages were potential purveyors of a corresponding intellectual culture, on the condition that they endorsed the cultural codes of the milieu that requested their skills. Converting to Christianity and, above all, learning to express himself in elegant humanistic Latin and to read Greek, gave the young Guglielmo Raimondo Moncada, alias Flavius Mithridates, the opportunity to introduce his Sicilian Jewish heritage with its Arabic component into the courts of Northern Italy. This he did zealously and without showing excessive scruples as to the actual limits of his knowledge of Arabic64 and the real nature of some of the intellectual issues connected with Aramaic.<sup>65</sup> We should not regard

<sup>62</sup> The nickname Mithridates alludes to Mithridates VI Eupatôr, King of Pontus (135–163 bc), a formidable enemy of the late Roman republic, as well as a sovereign famous for his ability to speak the numerous languages of his realm.

<sup>63</sup> On the life of Moncada/Mithridates, see the numerous résumés in the collected volumes quoted in notes 60 and 61. Important information on his Sicilian background and his first Hebrew names is provided by Angela Scandagliato, *Judaica minora sicula: Indagini sugli ebrei di Sicilia nel Medioevo e quattro studi in collaborazione con Maria Gerardi* (Florence: Giuntina, 2006). On his choice of different nicknames, see Flavius Mithridates, *Sermo de passione domini*, ed. Wirszubski, 48–49.

<sup>64</sup> His knowledge of Arabic is thoroughly questioned by Hartmut Bobzin, "Guglielmo Raimondo Moncada e la sua traduzione della sura 21 ('dei profeti')," in *Guglielmo Raimondo Moncada alias Flavio Mitridate*, ed. Perani, 173–183; and Burman, *Reading the Qurʾān in Latin Christendom*, 133–148. Others highlight his Arabic skills, e.g. Piemontese, "Il Corano latino di Ficino"; and Piemontese, "Guglielmo Raimondo Moncada alla corte di Urbino." See my comments in Chapter 4.3.2 for a possible sociolinguistic explanation not considered by these scholars.

<sup>65</sup> Moncada certainly had a taste for cryptography, symbolism, and jokes. He used the Ethiopian alphasyllabary, quite new on the Italian cultural market at the time, to give an aura of mystery to the Judaeo-Aramaic of the Targums or of other texts, when he was teaching Giovanni Pico della Mirandola, or when he was transcribing fragments of Aramaic in his quintilingual sermon for the papal court. On this point, see Flavius Mithridates, *Sermo de passione domini*, ed. Wirszubski, 35–40. For the circulation of the alphasyllabary in Italy and

this lack of scruples as a mere social tactic. It also reflects the difficulty of reconciling the insular culture he had inherited with the expectations of his new continental milieu, and in particular, the discrepancy between the Judaeo-Arabic culture of his childhood and the Arabic-Islamic character of the texts that he would have to translate.

While promoting Arabic culture in circles that were speaking and writing in Latin, Moncada pursued a two-pronged strategy. On the one hand, he created textual objects whose spectacular features were bound to attract the attention and admiration of his audience and readership—among these were the gorgeous miniatures of the ms. Urb. Lat. 1384 with its mysterious Arabic calligraphy, or his quintilingual sermon with its fireworks of Greek, Arabic, Aramaic, and Hebrew quotes scattered throughout the Latin text. On the other hand, he accomplished more substantial tasks, for example, when he tried to teach Arabic or made lengthy translations from Hebrew or Aramaic, with his many manuscripts attesting to his philological approach. Moncada left an enormous number of glosses and philological notes, predominantly in Latin, on a peculiar Qurʾān written in Hebrew characters (now ms. Vat. Ebr. 357 in the Vatican Library).66 He had probably found or stolen this manuscript, probably created in western Sicily around 1400, from one of the libraries of his former co-religionists.67 He seems to have used it not only as a personal working basis for translating the Qurʾānic text and for some Muslim Qurʾānic exegesis, but also to teach Arabic to his pupils during the 1480s, among them, probably, Giovanni Pico della Mirandola. Indeed, towards the end of the century, Arabic was not only the preoccupation of fashionable or powerful court circles, worried by the advance of Ottoman Islam or interested in esoteric or magic recipes. It had reached the status of an alternative hermetic language to be used, for example, together with Greek and Hebrew, in the cryptic

the confusion of Chaldaean with Ethiopian, see, most recently, Samantha Kelly, "The Curious Case of Ethiopic Chaldean: Fraud, Philology and Cultural (Mis)understanding in European Conceptions of Ethiopia," *Renaissance Quarterly* 68, no. 4 (2015), 1227–1264.

<sup>66</sup> On this manuscript and its links to Moncada, see Piemontese, "Il Corano latino di Ficino," 266–271, who was the first to understand the nature of the predominantly Latin annotations on the Arabic text, and their link to Moncada. See Benoît Grévin, "Le 'Coran de Mithridate' (ms. Vat. Ebr. 357) à la croisée des savoirs arabes dans l'Italie du XV<sup>e</sup> siècle," *Al-Qanṭara* 31, no. 2 (2010), 513–548, with more extensive quotations from Moncada's personal annotations. See also Benoît Grévin, "Flavius Mithridate au travail sur le Coran," in *Flavio Mitridate mediatore fra culture nel contesto dell'ebraismo siciliano del XV secolo*, ed. Mauro Perani and Giacomo Corazzol (Palermo: Officina di studi medievali, 2012), 200– 230. In the otherwise fundamental synthesis on Latin medieval Qurʾāns, that is, Burman, *Reading the Qurʾān in Latin Christendom*, 142, this manuscript seems to be the object of a misunderstanding. Burman characterizes the manuscript as a short anthology and doubts Moncada's authorship of the bulk of the Latin annotations on the Arabic text. However, the Qurʾānic text is almost wholly preserved, and Moncada's authority is beyond any doubt, given that he signs his very characteristic notes more than once.

<sup>67</sup> On Moncada's successful attempts to acquire manuscripts at the expense of his former religious community in Sicily, see Scandagliato, *Judaica minora sicula*, 466.

pseudo-inscriptions of refined texts like Francesco Colonna's *Hypnerotomachia Poliphili*, first published in 1499.68 Still, the gap between the available teaching supply and the learning demand was enormous, and would be for centuries. The avidity with which the papal court accepted the help of Leo Africanus (alias al-Ḥasan b. Muḥammad al-Wazzān al-Fāʾsī) at the beginning of the sixteenth century attests to this, as does the fact that cardinal Egidio da Viterbo (sed. 1517–1532) was forced to recruit translators from Spain (and to receive the assistance of the same Leo Africanus) in order to publish a new, glossed translation of the Qurʾān.69 A sizeable number of men with some knowledge of Arabic probably circulated in the Apennine Peninsula around 1480 or 1500. However, only a very few had the knowledge required to execute first-rate translations of complex texts such as the Qurʾān.

#### 4.3 From diplomacy to translation: Sociolinguistic problems, recurring patterns, inflexions

The Arabic culture that flourished in certain parts of the Apennine Peninsula from the Norman conquest to the "syncretic" humanism of the late Quattrocento was strongly associated with the Latin language. For reasons of prestige as well as cultural habits, the language chosen to translate or accompany Arabic texts was very seldom a form of Italian. Certainly, linguistic interaction between the vernacular language—Italian of one sort or another—and Arabic existed. A vernacular sermon, preached by Giordano da Pisa in Florence in around 1305–1306 and transcribed by attentive clerks, contains an interesting and pertinent commentary on the equivalence between the term *podestà—*the official recruited on an annual basis from outside the Italian communes to maintain civil peace—and the significance of the Arabic word *sulṭān*. 70 The tendency to privilege Latin in line with humanistic aesthetics was not mandatory for every genre: a refined text such as the *Hypnerotomachia Poliphili*, although teeming with Latin

<sup>68</sup> On this point, see Piemontese, "Le iscrizioni arabe nella 'Poliphili Hypnerotomachia.'" For the inscriptions, see Francesco Colonna, *Hypnerotomachia Poliphili*, ed. Giovanni Pozzi and Lucia A. Ciapponi (Padova: Antenore, 1980).

<sup>69</sup> On the revision of the Qurʾānic translation commissioned by Egidio da Viterbo from Leo Africanus, see García-Arenal and Starczewska, "The Law of Abraham the Catholic," 421–437, and the introduction to Starczewska, *A Latin Translation of the Qurʾān.*

<sup>70</sup> See Giordano da Pisa, *Quaresimale Fiorentino 1305–1306*, ed. Carlo Delcorno (Florence: Sansoni, 1974), the edition of the Florentine sermons preached during Lent by Giordano da Pisa in 1305–1306, sermon 44, 277. The Italian *reportatio* of the sermon specifies that Giordano da Pisa quoted a biblical sentence, or at least some words of it, in Arabic: "ma egli [podestade] è nome troppo alto: questo è il nome del soldano, podestade; onde soldano in nostra lingua è podestade. Onde dice il vangelo, il quale fu iscritto in quella lingua de' saracini, quando dice: 'Io hoe podestade di porre l'anima mia et cetera,' sì dice 'soldayn, *sulṭān*.' (Frate Giordano il disse in quella lingua egli)."

(and Greek) quotes, was written in Italian, albeit a very Latinized form of the language.71 However, due to the persistent and even growing prestige of Latin as a political, cultural, and literary medium in Italy, the status of the texts to be translated, as well as the configuration of knowledge and learning techniques, Latin was generally privileged during the process of translating or interpreting an Arabic text. Rather than being Italianized, the text was to be Latinized, or presented together with a Latin version. The possibility of an intermediate Italian version was not excluded, particularly when the key informant was a merchant, but it was far from automatic.72 In courtly or ecclesiastical contexts—one of the major humanistic courts during the Quattrocento being the papal curia—Latin was mandatory as the language of majesty or God. In this respect, the sociolinguistic situation of the Apennine Peninsula certainly differed considerably from that of the Iberian Peninsula. There, the new translation of the Qurʾān envisioned by Juan of Segovia, prepared in the years 1454–1456, included the Arabic text, a vernacular Castilian, and a Latin version.73 The Castilian version had been prepared by the Mudéjar Yça of Segovia. Juan then translated the Castilian text into Latin.74 The mechanism was not new: many twelfth- and thirteenth-century translations had been achieved with this staged, two-phase technique. However, older medieval translations generally did not conserve the vernacular intermediate version. In Italy, the creation of similar vernacular translations would only come later. The first Italian translation of the Qurʾān, for example, was not produced before the sixteenth century. It was not only very fragmentary, but also indirect, since it was based on the old Latin translation of Mark of Toledo, produced at the turn of the twelfth to the thirteenth century, rather than on the original Arabic text.<sup>75</sup>

<sup>71</sup> See Colonna, *Hypnerotomachia Poliphili*, ed. Giovanni Pozzi and Ciapponi.

<sup>72</sup> On this point, see the considerations in Chapter 4.2.2 on the literary production linked to Niccolò de' Conti and Emmanuel Piloti.

<sup>73</sup> On the translation of Juan of Segovia, almost entirely lost, see Burman, *Reading the Qurʾān in Latin Christendom*, 178–197; and see also the new discoveries presented by Ulrich Roth and Reinhold F. Glei, "Die Spuren der lateinischen Koranübersetzung des Juan de Segovia: Alte Probleme und ein neuer Fund," *Neulateinisches Jahrbuch* 11 (2009), 109–154, which is synthetized in Ulli Roth, "Juan of Segovia's Translation of the Qurʾān," *Al-Qanṭara* 35, no. 2 (2014), 555–578.

<sup>74</sup> On the milieu and activities of the Mudéjar, who helped Juan de Segovia, see Wiegers, *Islamic Literature in Spanish and Aljamiado*.

<sup>75</sup> On this translation, see *Iddio ci dia buona viaggio e guadagno: Firenze, Biblioteca Riccardiana, ms. 1910 (Codice Vaglienti)*, ed. Luciano Formisano (Florence: Edizioni polistampa, 2006), introduction: 31–34, and text: 267–281. On the translation of Mark of Toledo, see *Alchoranus Latinus quem transtulit Marcus canonicus Toletanus*, ed. Nàdia Petruus Pons (Madrid: Consejo Superior de Investigaciones científicas, 2016).

#### 4.3.1 DIPLOMACY, LANGUAGE LEVELS, AND TRANSLATION SKILLS IN THE TWELFTH AND THIRTEENTH CENTURIES

We must keep in mind the Latin-Romance sociolinguistic context of this Arabic-Latin culture. However, in a time in which speakers and writers of Latin were also literati speaking and writing in Romance, the status of Latin, as the ultimate link in the translating chain, or, more often, as a direct translation tool from Arabic, did not present any particular problem, stylistic questions aside. From the thirteenth to the fifteenth century, there was no lack of excellent Latinists on the Apennine Peninsula, even if the aesthetics of writing and speaking Latin varied greatly between the biblically flavoured phraseology employed in the chancery of the Norman and Hohenstaufen kings of Sicily and the strict Ciceronianism that was fashionable in Renaissance Italy from around 1400 onwards. The situation was quite different and much less favourable with regard to Arabic. The lexicographer and jurist Ibn Makkī (d. 505/1107–1108) had already criticized the linguistic skills of the Arabic-speaking elites in pre-Norman Sicily in the eleventh century.76 Such criticism was in line with the widespread obsession to condemn every deviation from the theoretical norms of classical *fuṣḥā*, and was certainly not specific to Sicily: the bad quality of Maghrebi Arabic, Andalusī elites excluded, was a topos during the High Middle Ages. In Sicily, however, the linguistic problem became ever more pressing towards the end of Norman rule and during the transition to the Hohenstaufen dynasty, that is from the reign of William II, starting in 1166, to the return to Italy of Frederick II in 1220. The reason for this was that the available pool of insular Muslim literati had significantly decreased. Under the Hohenstaufen dynasty, it was already difficult, not so much to find some speaker of Siculo-Arabic in the kingdom, but to find the human resources able to translate from Arabic to Latin or from Latin to Arabic according to the high standards requested both for scientific translation and for solemn political communication.

The above-mentioned diplomatic mission, directed by Frederick II to the Ḥafṣid sultanate, provides a good example of this problem of finding appropriate personnel. A series of Latin mandates, preserved in the imperial register for October 1239–May 1240, reveals the dilemma faced by the Sicilian court.77 The emperor, who sojourned in northern Italy at the time, took great care to select the personnel. The mission was to be headed by Enrico Abbate, scion of an important Sicilian family. He was to be accompanied by Oberto Fallamonaca, another insular official of Frederick II, who probably came from an Arabic-speaking Christian family and may have had

<sup>76</sup> On this point, see Annliese Nef, "L'analyse du taṯqīf al-lisān d'Ibn Makkī et son intérêt pour la connaissance de la variante sicilienne de l'arabe: propositions méthodologiques," *Oriente moderno* 77, no. 1 (1997), 1–17.

<sup>77</sup> *Il registro della cancelleria di Federico II*, no. 539–542, 512–515, and 575, 541–542.

some knowledge of written Arabic.78 The emperor's insistence on recruiting two additional literati for the operation reveals the imperial court's attention to the linguistic aspect of the negotiations. Frederick II tried repeatedly to ensure the presence of the mathematician John of Palermo in the diplomatic mission, probably because of the latter's linguistic as well as his scientific skills. He was a renowned mathematician, who translated from Arabic.79 We have already seen that the emperor selected still another man of letters of *Mašriqī*, i.e. of Middle Eastern origin, Theodore of Antioch, to write the Arabic version of the official missive addressed to the Ḥafṣid sultan.<sup>80</sup>

The reason for this peculiar choice was not only Theodore's proficiency in Arabic. The letter needed to be written in the solemn rhythmic and rhymed prose used for diplomatic and political correspondence in the Islamic sphere in this period, and elaborated according to the stylistic-rhetoric discipline known as *ʿilm al-inšāʾ*. <sup>81</sup> Theodore was apparently considered the best Arabic writer at the court, perhaps because he had received the necessary stylistic-rhetoric training in the Orient, along with his scientific knowledge. Such training, associated with the prose and poetry of classical and post-classical Arabic literary culture, would certainly have been difficult to acquire in Palermo after 1190, even in the bilingual parts of the Christian elite or among the Jewish population. This dependence on an Oriental scholar on the part of a court reputed for its linguistic universalism was not necessarily scandalous: the stylistic qualifications required to master this particular type of writing were very high, even by Muslim standards. We should consider that, in the Christian West during the same period, a man of letters would have been able to read and write Latin fluently, without automatically mastering the subtleties of the structurally equivalent rhetorical techniques used for epistolary political communication, the so-called *ars dictaminis*. 82

Indeed, a well-known but controversial testimony from the early Norman period illustrates the problems created by miscommunication in solemn literary correspondence. So far, no Arabic letter sent by a Sicilian Norman king to a Muslim sovereign has been discovered. However, a reply from the Fatimid chancery to a Sicilian royal letter has been transmitted by al-Qalqašandī's (d. 821/1418) *Ṣubḥ al-aʿšā fī ṣināʿat al-inšāʾ*, one of the major encyclopaedias on the art of composing chancery documents from the Mamlūk period. The Fatimid letter, dated around 1137, touches on many subjects. One passage shows, however, that in a preceding Norman Arabic

<sup>78</sup> On this point, see Mandalà, "La migration des juifs," 19; Mandalà, Moscone, "Tra Latini, greci e 'arabici'," 195.

<sup>79</sup> Mandalà, "Il prologo delle Risposte," 76, 79–86.

<sup>80</sup> Ibid., 67–84.

<sup>81</sup> On *ʿilm al-inšāʾ* see Chapter 1.4 of this volume.

<sup>82</sup> On the *ars dictaminis* in the context of the Sicilian Court, see Benoît Grévin, *Rhétorique du pouvoir medieval: Les Lettres de Pierre de la Vigne et la formation du langage politique européen* (Rome: École française de Rome, 2008).

missive, a linguistic error of some sort had been committed, to the effect that it had disturbed parts of the communication process. An excuse and an explanation on the part of the Normans had been provided in a subsequent letter. The Fatimid Arabic letter selected by al-Qalqašandī now put an end to the case. It assures the Norman addressees that their explanation has been accepted, and comments on the difficulties of translating from one language to the other in cases in which the target language does not possess the terminology of the original language. This, the letter asserts, can effectively prevent a writer from making himself understood—a highly interesting comment on the difficulty of translating from Greek or Latin into Arabic, as perceived in the chancery of Fatimid Egypt!83 Since the Norman part of the preceding correspondence is lost, we will never know what kind of Latin or Greek term or expression the Sicilian writer had used as part of the Arabic letter sent to the Fatimid caliph in Egypt, and what had actually caused confusion. We know for certain, however, that the letter was written during a stage of Norman-Sicilian history in which the island still featured an important Muslim and Christian elite capable of writing Arabic. That a linguistic error such as this occurred in a Western chancery reputed for its proficiency in Arabic demonstrates that, even under apparently favourable circumstances, the highly formalized communication processes of medieval political powers represented a challenge.

A linguistic history of medieval embassies has yet to be written.<sup>84</sup> There is certainly no lack of sources on peculiar Arabic-Latin exchanges during the reigns of Frederick II and the first Angevin king, Charles I. In recent years, Giuseppe Mandalà has battled some researchers' attempts to minimize the importance of direct scientific contacts between Arabic and Latin milieus during the thirteenth century. He was able to demonstrate repeatedly that numerous embassies sent by Frederick II to Egypt or to the Islamic West can be directly linked to scientific exchanges resulting in translations, not only from Arabic to Latin, but also from Latin to Arabic. This was the case, for example, in a diplomatic mission of 1227 destined to prepare the "crusade" of 1228–1229. During this mission, the Sicilian high dignitary Berardo, archbishop of Palermo, translated a Latin inscription carved at the base of one of the great pyramids for his Egyptian hosts either directly

<sup>83</sup> See on this subject Nef, *Conquérir et gouverner*, 91–93, pointing to the study Marcel Canard, "Une lettre du calife fatimide al-Ḥāfiẓ (524–544/1130–1149) à Roger II," Atti del Convegno Internazionale di Studi Ruggeriani (Palermo: Società siciliana di Storia patria, 1955), 125–146, reed. Marcel Canard, *Miscellanea Orientalia* (London: Variorum reprints, 1973), and on Jeremy Johns, *Arabic Administration in Norman Sicily. The Royal Dīwān* (Cambridge: Cambridge University Press, 2002), 259–265.

<sup>84</sup> See the suggestions of Jean-Marie Moeglin and Stéphane Péquignot, *Diplomatie et "relations internationales" au Moyen Âge (IXe –XV<sup>e</sup> siècle)* (Paris: Presses universitaires de France, 2017), 112–126. For the Sicilian kingdom during the thirteenth century, see Borghese and Grévin, "Aspects linguistiques de la diplomatie sicilienne."

or with the help of some Arabic speakers.<sup>85</sup> Later, the redaction of a Latin version of the *Quaestiones sicilianae*, based on a series of questions sent by Frederick II to the Western scholar Ibn Sabʿīn, can also be connected to Hohenstaufen diplomacy in the Maghreb.86 Finally, the Latin translation of the medical treatise *Kitāb al-Ḥāwī* or *Liber continens*, made by the Sicilian Jew Faragius (*Farağ*) for Charles I, was a direct result of a gift made by the Ḥafṣid sultan to the Angevin king of Sicily.87 A Sicilian medieval diplomatic mission could thus involve numerous linguistic operations. These included the redaction of Arabic letters potentially fraught with stylistic or linguistic deficiencies, as well as translations that have survived in the form of scientific treatises. Competent linguistic actors chosen for these tasks did not necessarily master the entire range of Arabic communication skills. Over time, southern Italian sovereigns had come to rely on minorities like the Arabic-speaking Jews for their diplomatic transactions with Tunis,88 without being able to uphold the semantic and linguistic sophistication of the Norman and Hohenstaufen embassies. The sociolinguistic conditions of these exchanges changed considerably between the eleventh and the fifteenth centuries, as did the abilities of native Italian Arabic-speakers to understand and reproduce the linguistic intricacies of Muslim elite culture.

#### 4.3.2 SOCIOLINGUISTIC LIMITS, PIVOTAL ROLES, AND WRITING STRATEGIES IN THE HEYDAY OF HUMANISM

Switching back from Norman and Hohenstaufen Sicily to central and northern Italy in the heyday of the Renaissance, a few centuries later, we can examine yet another facet of the same problem. We have already suggested that the translation accomplishments of the fifteenth century were constrained by certain technical limits, which resulted from the cultural background of the actors involved. Beltramo Mignanelli's very short treatise on Arabic in the introduction to his bilingual anthology of the Psalms provides an example. In this case, we can argue that the relative lack of theorization in this treatise does not result so much from his own lack of linguistic or conceptual skills. Quite to the contrary, it has to be regarded as the effect of a general difficulty common to all Latin literati. Since the latter were mentally conditioned to identify Latin with grammar, developing

<sup>85</sup> On this point, see Giuseppe Mandalà, "Un ambasciatore di Federico II in visita alle piramidi: Berardo arcivescovo di Palermo (a. 1227)," *Aevum* 85 (2011), 1–22.

<sup>86</sup> Mandalà, "Il prologo delle Risposte"; and Giuseppe Mandalà, "The Sicilian Questions," *Journal of Transcultural Medieval Studies* 3 (2016), 3–32.

<sup>87</sup> See Borghese and Grévin, "Aspects linguistiques de la diplomatie sicilienne," 88–89, on descriptions, fortunately preserved, of the process of redaction and production of the book in the otherwise destroyed registers of the Angevin Sicilian chancery.

<sup>88</sup> Salvatore Fodale, "Un ebreo trapanese ambasciatore dei Martini a Tunisi: Samuele Sala," *Studia historica et philologica in honorem M. Batllori* (Rome: Publicaciones del Instituto Español de Cultura, 1984), 275–280.

an autonomous description of a foreign language represented an enormous challenge.<sup>89</sup> Despite its short length and its limits, Mignanelli's brief description of Arabic is rather impressive. It takes into account the linguistic structure and semantic richness of Arabic and addresses its geographical diffusion as well as its sociolinguistic variants.90 Mignanelli's familiarity with the linguistic concepts and clichés of the Mamlūk and Timurid Orient results in some assertions that reflect an empathy with the Oriental view of linguistic culture, which is quite unusual in medieval Latin literature. His description of Arabic as a language characterized by three declensions and an overwhelming semantic richness surprises less than an anecdote, apparently related to the Mamlūk sultan Barqūq. Here, Arabic is defined as the language of religion/judgement (*judicium*: probably the Latinization of Arabic *dīn*), Persian (and Greek!) as the language of harems and women, and Turkic as the language of war and the army.91 Despite its schematic character, this functionalist description of the three major languages used in the Muslim Orient, to which Greek is added, gives a good account of how the Mamlūk elite defined the different functions of Turkic—as the language of the military caste and of war, and Arabic—as the language of the legal religious system.92 In another of his writings, the *Liber Machometi et opinio perfida iudeorum*, Mignanelli contradicts the notion, widely diffused in the Latin West, that Hebrew was the first language of humanity, that is, the language of Adam. Instead, Mignanelli argues that wiser Christians,


<sup>89</sup> On this general problem, see Benoît Grévin, *Le parchemin des cieux. Essais sur le Moyen Âge du langage* (Paris: Éditions du Seuil, 2012), 178–186.

<sup>90</sup> The text can be found in Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 155–156.

Muslims (*Sarraceni*), and Jews agreed that the language of Adam was "Chaldaean," called *soriana* (*sūryānī*) by the Muslims, an indication that he may have discussed linguistic theories with Muslim interlocutors.<sup>93</sup>

While Mignanelli had developed a remarkable empathy for linguistic concepts and clichés that circulated in the Muslim sphere during his lifetime, his transcriptions from Arabic in the *Ascensus Barcoch* and his bilingual *libellus* reveal the limits of his extensive knowledge. As already mentioned, his Latin transcriptions of Arabic phrases—in this case of messages exchanged between the Mamlūk sultan Barqūq and Tamerlane—are strongly influenced by the Egyptian pronunciation of Arabic. Because of their lack of apparent sense, scholarship has neglected these transcriptions until Angelo Michele Piemontese explained them correctly.94 The possibility of using his bilingual anthology of the Psalms, in turn, is limited by the fact that the Arabic text is not vocalized.<sup>95</sup> A third, indirect testimony gives yet another impression of the difficulties Mignanelli potentially faced during his work as a mediator between Latin and Arabic. Flavio Biondo, a first-rate humanist who wrote an interesting treatise on the exact nature of Latin as an elite or popular language during Antiquity96 and participated in the council of Florence, harshly criticized the interpreters from Arabic to Latin active during the council for their inability to master the subtleties of theological phraseology.97 Given the complexity of the matter and the total lack of qualifications on the part of the critic, this judgement was certainly ungenerous. It suggests, however, that the skills of Mignanelli and his colleague(s) were tested during the difficult negotiations. We are confronted again with the problem that institutional agents in need of linguistic mediators could not draw on an adequately trained pool of specialized linguists. This, in turn, obliged the interested institutions to rely on the stretched competencies of inadequately trained personnel, at the risk of making the latter work beyond their capacities.

With the activity of Guglielmo Raimondo Moncada, alias Flavius Mithridates, we face yet another dimension of the same problem. Unlike

<sup>93</sup> The text is edited in Mahmoud Helmy, *Tra Siena, l'Oriente e la curia*, 391–392. Quite interestingly, Beltramo uses this topos of Syriac as the first human language in a passage with a clear anti-Judaic tendency: "Et declarato hoc pura consciencia, ut reperi, transeo ad iudeos, magis malignantes et pessimos, dicentes eorum linguam ebraicam fuisse primam sub celo et ipsam Adam, primum patrem omnium, fuisse locutum." This is noteworthy because the idea that Hebrew was the first human language was in fact fairly widespread in Latin Christendom during his lifetime.

<sup>94</sup> Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 163. See the edition of *Ascensus Barcoch* in Mahmoud Helmy, *Tra Siena, l'Oriente e la curia*, 374, for the passage concerned.

<sup>95</sup> See Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 165.

<sup>96</sup> On this argument, see Flavius Blondus, *De verbis Romanae locutionis*, ed. Fulvio Delle Donne (Rome: Istituto Storico Italiano per il Medioevo, 2008).

<sup>97</sup> See Mahmoud Helmy, *Tra Siena, l'Oriente e la curia*, 212–213, particularly 213 fn. 15, with Flavio Biondo's attack on the incompetence of the translators, first during the council of Florence (1439–1441), and then in connection with successive embassies from Oriental churches in 1442.

Beltramo Mignanelli, the Sicilian convert most probably never travelled to the Arab world. Whatever Arabic he knew when he undertook his translations, he had acquired on the streets of Sicilian towns or in the *Yeshiva*, where he received the necessary schooling to become a rabbi like his father. Although a man with obvious linguistic talents, Moncada did not live up to his self-proclaimed image of a specialist of Arabic. Several scholars have noted that his translations of *sūra*s 21 and 22 preserved in the ms. Urb. Lat. 1384 are rather mediocre and contain numerous errors.98 The fragmentary character of his Qurʾānic and exegetical translations contained in the ms. Vat. Ebr. 357 makes it difficult to judge the level of proficiency of this work. However, it is clear that these texts also feature recurring linguistic problems.99 In connection with the ms. Urb. Lat. 1384, scholars such as Harmut Bobzin and Thomas E. Burman concluded that Moncada was, if not an impostor, at least not sufficiently able to shoulder the roles of a translator and a teacher of Arabic. Some features of his personality seem to confirm this intellectual's tendency to obfuscate his real competencies.100 The central part of the otherwise bilingual manuscript created for the duke of Urbino, for example, lacks an Arabic text for a simple reason: Moncada pretended to have translated these astronomical texts directly from Arabic. It has come to light, however, that they probably derived from a Hebrew intermediate created by Ibn al-Aḥdab, a Jewish scholar who emigrated from Castile to Sicily, and whose family was friends with Moncada's father.101 We also know that Moncada borrowed one of the old Latin translations of the Qurʾān at the papal library when he worked on his quintilingual Sermon. It is very possible that he used unknown preexisting Latin, Italian, or Hebrew translations as supports for his proper work on various Arabic texts, without ever admitting this.102

All of this provides a complementary explanation for some of the linguistic "deficiencies" displayed by Moncada. This explanation is interesting from a philological point of view, and should not be read as an accusation of personal mediocrity. An analysis of recurrent translation errors in the

101 On this point, see Mandalà, "Da Toledo a Palermo," 14–15.

<sup>98</sup> Bobzin, "Guglielmo Raimondo Moncada e la sua traduzione della sura 21 ('dei profeti')," 173–183; Burman, *Reading the Qurʾān in Latin Christendom*, 133–148.

<sup>99</sup> For a detailed analysis of the recurring translation problems in the mostly Latin annotations of Moncada on the Qurʾān of ms. Vat. Ebr. 357, see Grévin, "Le Coran de Mithridate," 537–548.

<sup>100</sup> On his use of the Ethiopian alphasyllabary, and his manner of dealing with the risk that speakers of Hebrew might reveal his trick to Giovanni Pico della Mirandola, see Flavius Mithridates, *Sermo de passione Domini*, ed. Wirszubski, 38 fn. 1. Moncada is too often judged from a moral point of view and should rather be examined in the light of processes of individuation and of negotiating multiple cultural identities (Jewish, Christian, Arabic-speaking, Hebrew-learned, Latin-speaking, and even libertine and queer, since the notes abound in rather spectacular sexual allusions) in late Quattrocento Italy. On his exuberant personality, see Saverio Campanini, "Guglielmo Raimondo Moncada (alias Flavio Mitridate) traduttore di opera cabbalistiche," in *Guglielmo Raimondo Moncada alias Flavio Mitridate*, ed. Perani, particularly 72–80.

<sup>102</sup> Piemontese, "Il Corano latino di Ficino," 268.

fragmentary Latin translations of the Qurʾānic ms. Vat. Ebr. 357 suggests that some of these lapses were caused by the very structure of the Arabic spoken by Moncada in Sicily. Like other Arabic dialects or varieties, Siculo-Arabic had very probably lost an important part of the conjugation system of Classical Arabic, e.g. the classical passive forms. Moncada makes repeated errors when encountering the same linguistic patterns—patterns that were probably lacking in spoken Judaeo-Arabic. Against this backdrop, it seems quite probable that his language skills and their respective limits resulted from his early linguistic conditioning. The latter had enabled him to speak and, at least partly, read some very specific forms of Judaeo-Arabic with a certain degree of fluency. However, this knowledge only gave him partial access to classical Muslim texts such as the Qurʾān, whose correct comprehension required a cultural and pedagogical immersion into Muslim culture, the lack of which Moncada could not fully compensate for. Fifteenth- or early sixteenth-century translations of the Qurʾān were accomplished with the help of a person with a Muslim background: Juan de Segovia (d. 1458) drew on the help of Yça of Segovia; Egidio da Viterbo (d. 1532) employed converts from Islam to Christianity, with Juan Gabriel of Teruel preparing parts of the translation and Leo Africanus adding some corrections. The disparity between these versions and the works of Moncada can be explained from a sociolinguistic and sociocultural point of view that considers the peculiar history of Siculo-Arabic and, more precisely, of the Sicilian variant of Judaeo-Arabic.

## 4.4 Conclusion

Much remains to be said, not only about the sociolinguistic, but also about the sociocultural biases that influenced and, in some ways, impeded the rapid development of a new culture of Arabic learning in Italy at the end of the fifteenth century, despite the increasing attention paid to this language in humanist circles. There existed, for example, a strong tendency to associate Arabic with Hebrew and Aramaic in its various forms, from Judaeo-Aramaic to Syriac. The conceptual interest inherent in this association can be regarded in some ways as a precursor to modern Semitic comparative studies. Among Italian humanists, however, this tendency resulted from a highly traditional representation of language and languages. Arabic was perceived as a sacred and scientific language, and as such became associated with the "Languages of the Cross" (Latin, Greek, and Hebrew). It represented one element in a mysterious pentacle of erudite tongues which, in the mind of humanists like Pico della Mirandola, included Latin, Greek, Hebrew, and Aramaic, the last being an important vector of texts like the *Sepher ha-Zohar*, considered at the time to be a book of prophetic valour produced in remote antiquity.<sup>103</sup> Since the time of the later crusades, the

<sup>103</sup> On the evolution of this doctrine of the linguistic "pentacle," see Benoît Grévin, "Anamorphoses linguistiques: le pentacle des langues référentielles dans

medieval church had promoted the idea that these languages should be taught in European *studia* in order to help convert infidels. This formulated objective stood in clear contradiction to what would have been the practical linguistic requirements for better communication with the Arabic-Islamic sphere. Various forms of Middle Arabic were certainly used as a daily language from the Maghreb to the Mashreq. However, one could hardly argue that the study of "Chaldaean" or biblical Hebrew would help to promote Latin ecclesiastical propaganda in the Muslim sphere. The contrast between the linguistic goals proclaimed at the council of Vienne (1311– 1312) in the famous bull that envisioned the creation of chairs of Arabic, Hebrew, and Aramaic in the most important European *studia*, 104 and the realities of communication between European Christians and the Muslim sphere, is indeed striking. We do know that the friars in the East—from the borders of Hungary to Persia—did train in the languages prevalent in Inner Asia of the thirteenth and fourteenth centuries, such as Persian and Turkic.<sup>105</sup> The famous *Codex Cumanicus*, among other testimonies, shows that contemporaries were able to conceive of a rational approach to learning these languages, and it is possible that equivalent works concerning some form of Arabic existed.106 Of greater importance, however, was the old conception of Arabic as a language of science and controversy. The Latin obsession with Hebrew cultural heritage and the traditional Jewish view of the Arabic language as a deformation of Hebrew107 combined to form a conceptual understanding of the Arabic language that inevitably led to a conflation of the study of Arabic with the emerging study of Hebrew in Christian circles. This would heavily influence the projects and achievements of fifteenth- and sixteenth-century men of letters interested in Arabic, as the works of Beltramo Mignanelli demonstrate. Although he deplored the fact that he did not know Hebrew, he returned to the subject of the interrelation of Hebrew and Arabic more than once and even left a

l'Occident médiéval," to be published in *Hiéroglossie I. Moyen Âge latin, monde arabo-persan, Tibet, Inde*, ed. Jean-Noël Robert, in print.

<sup>104</sup> Berthold Altaner, "Die Durchführung des Vienner Konzilbeschusses über die Errichtung von Lehrstühlen für orientalische Sprachen," *Zeitschrift für Kirchengeschichte* 52 (1933), 223–236.

<sup>105</sup> On the practice of Persian in Domican milieus in the Trecento as attested by biblical glosses, see Angelo Michele Piemontese, "Le glosse sul vangelo persiano del 1338 e il codex cumanicus," *Miscellanea Bibliothecae Apostolicae Vaticanae*, vol. 8 (Vatican City: Biblioteca Apostolica Vaticana, 2001), 313–349.

<sup>106</sup> On the *Codex Cumanicus*, see *Codex Cumanicus: Édition diplomatique avec Fac-Similés*, ed. Vladimir Drimba (Bucharest: Editura Enciclopedica, 2000); and *Il codice cumanico e il suo mondo*, ed*.* Felicitas Schmieder, Peter Schreiner (Rome: Edizioni di Storia e di letteratura, 2005).

<sup>107</sup> Moncada supports this idea in his notes, in line with medieval Hebrew classics such as the *Kuzari*. An example is provided in Gersonide, *Commento al cantico dei Cantici nella traduzione ebraico-latina di Flavio Mitridate*, 189. Here Moncada incidentally comments on the influence of Hebrew on Arabic: "Et certum est quod lingua Arabica conformis est lingue Hebraice in pluribus, cum multa vocabula Hebraica adhuc servaverit, sicut filia que aliquid nostre consuetudinis habet." He left similar remarks in his annotations on the Qurʾānic text of the ms. Vat. Ebr. 357.

blank space in the *Liber de variantibus psalterii* to write down the Hebrew version.<sup>108</sup> This obsession with the link between Hebrew and Arabic was overwhelming in the writings of Moncada. Incited by his own Hebrew culture and Judaeo-Arabic writing habits, he could not avoid exaggerating the linguistic proximity between Hebrew and Arabic. Indeed, the ms. Vat. Ebr. 357 contains some very interesting attempts to explain the meaning of Qurʾānic concepts on the basis of traditional Hebrew religious concepts, while some translation choices are undoubtedly influenced by the filter of a Hebrew perspective.109

This dialectic relationship between Hebrew and Arabic was emphasized in Quattrocento Italy thanks to the marginal but increasing participation of Hebrew communities in the redefinition of Humanism.110 However, if we want to understand the directions taken by Arabic studies between around 1470 and 1530, we must also regard this dialectical relationship as part of the broader cultural framework of the "linguistic pentacle." In a period that stood on the verge of enormous cultural changes brought about by the Reformation and the increasingly acknowledged political metamorphosis of Islam produced by the rise of the Ottomans, the five languages, Latin, Greek, Hebrew, Aramaic, and Arabic, boasted a significance that went far beyond their pragmatic functions. Because of their supposed or real significance as sacred languages, they represented far more than a simple means of oral and written communication.

In the preface to his translations from the Qurʾān contained in the ms. Urb. Lat. 1384, Moncada boasted about a project he had apparently discussed earlier with the duke of Urbino. He never realized this project which, given the complexity of the task, actually seems rather fantastic. It consisted in creating a gigantic quadrilingual version of the Qurʾān, translated into Latin, Hebrew, and Aramaic. The intellectual premises of this vision are not only found in the cultural background of Moncada; in fact, they cannot be understood without taking into account the familiarity of Latin elites with the idea that Arabic constituted a sort of complement to the linguistic constellation of Latin–Greek–Hebrew–Aramaic. This idea found expression in some hermetic works such as the *Hypnerotomachia*, or, on a cultural level, in some syncretic aspects of the thought of humanists like Giovanni Pico della Mirandola.111 The preliminary work preserved in

<sup>108</sup> Piemontese, "La lingua araba comparata da Beltramo Mignanelli," 157, introduction to the *Liber de variantibus*: "Et sciant me dimisse illud spatium quod inter versum et versum et differentiam et differnentiam est cupiens in eo ponere in Hebraico sicut in arabico scriptum est."

<sup>109</sup> See, for example, Grévin, "Flavius Mithridate au travail sur le Coran," 36 fn. 27, on Moncada's gloss of the Arabic term *sakīna* with the Hebrew term *šeḫīna* (ms. Vat. Ebr. 357, commentary to *sūra* II, 248, fol. 57r), with all the theological implications that such an equation could create in a mind obsessed with Kabbalah.

<sup>110</sup> On this point, see Giulio Busi, *L'enigma dell'ebraico nel rinascimento* (Torino: Aragno, 2007). See also the classic work of Moshe Idel, *La Cabbalà in Italia (1280– 1510)* (Florence: Giuntina, 2007).

<sup>111</sup> On the Arabic and the multilingual pseudo-inscriptions in the *Hypnerotomachia*, see Piemontese, "Le iscrizioni arabe nella 'Poliphili Hypnerotomachia'." On Pico

the ms. Vat. Ebr. 357, with its Qurʾānic text written in Hebrew characters, mirrors such a conception to a certain extent. It was neither isolated nor without a following, since one of the first (if not the first) long texts printed in Arabic was included in a multilingual programme just like this, albeit less ambitious in terms of the word count. In 1516, a quarter of a century after Moncada's disappearance, Agostino Giustiniani published a quintilingual Psalter, including Hebrew, Aramaic, Arabic, Greek, and Latin versions of the text.112 A generation after the sumptuously illuminated codex for the duke of Urbino, the teaching and learning of Arabic had entered a new phase, in pace with the popularization of printing. Various defining characteristics of the preceding periods were now missing: the last, at least partly Arabic-speaking organized communities of Italy had disappeared due to the expulsion of the Jews from Sicily in 1492–1493, whereas the Ottoman take-over in major parts of the Arab world led to a reorganization of European-Christian merchant networks. In spite of these significant changes, some entrenched cultural patterns were to give this interest in Arabic learning a line of continuity. It was to remain closely associated with religious controversy as well as the study of the Bible and the Qurʾān, until the beginning of a new phase of Orientalism.

and Islam, see Angelo Michele Piemontese, "Traccia araba su codice latino," *Litterae Caelestes: Rivista annuale internazionale di paleografia, codicologia, diplomatica et storia delle testimonianze scritte* 1 (2005), 41–60.

<sup>112</sup> See Piemontese, "Il Corano in Italia," 53–54. On Agostino Giustiniani, see Aurelio Cevolotto, "Giustiniani, Agostino," in *Dizionario Biografico degli Italiani*, vol. 57 (Rome: Treccani, 2001), 301–306.

## Katarzyna K. Starczewska (CCHS-CSIC, Madrid)

# **5. Beyond Religious Polemics: An Arabic-Latin Qurʾān Used as a Textbook for Studying Arabic**

The span of time between the medieval and the early modern periods witnessed at least nine Latin translations of the Qurʾān.1 As Thomas Burman notes, Latin, and thus Christian, interpretations of the sacred book of Islam oscillated between religious polemics and philological zeal.2 This chapter traces these oscillations and underlines the transition from a treatment of the Qurʾān as a source of polemical material to an understanding of its utility to master Arabic. The case study chosen to demonstrate this change of focus is a sixteenth-century translation of the holy book commissioned by the Italian cardinal Egidio da Viterbo (d. 1532), which was reworked in the seventeenth century by the Scottish Orientalist David Colville. Colville's annotations and glosses are testament to the copyist's genuine eagerness to use the Qurʾān (or various interpretations of it) in combination with other sources to develop a sense of the functioning of the Arabic language.

The reasons behind each of the Latin translations of the Qurʾān, executed between 1141–1143 and 1698, varied from country to country and from one translator (or group of translators) to another.3 Broadly speaking, we might interpret medieval renditions of the Muslim holy book as texts devised to encourage the intellectual engagement with Islam with the aim of argumentative deconstruction as well as tools for political propaganda. Conversely, early modern translations can additionally be seen as erudite endeavours aiming at mastering Arabic. Translators engaged with long-lasting polemical themes, yet did so from a distant, scholarly perspective. This chapter offers a detailed examination of the commented

<sup>1</sup> The research leading to these results was supported with funding from the European Research Council under the European Union's Seventh Framework Programme (FP7/2007–2013) / ERC Grant Agreement number 323316, project CORPI: "Conversion, Overlapping Religiosities, Polemics, Interaction. Early Modern Iberia and Beyond."

<sup>2</sup> See Thomas E. Burman, *Reading the Qurʾān in Latin Christendom, 1140–1560* (Philadelphia: University of Pennsylvania Press, 2007).

<sup>3</sup> For an overview of Latin translations of the Qurʾān, see Burman, *Reading the Qurʾān*; Burman, "European Qur'an Translations, 1500–1700," in *Christian-Muslim Relations. A Bibliographical History, Volume 6: 1500–1900*, ed. David Thomas (Leiden: Brill, 2015), 25–38; and Benoît Grévin, "Le 'Coran de Mithridate' (ms. Vat. ebr. 357) à la croisée des savoirs arabes dans l'Italie du XV<sup>e</sup> siècle," *Al-Qanṭara* 31, no. 2 (2010), 513–548, with the bibliography cited therein.

copy of a translation commissioned in 1518. Both the translation of 1518 and the commented copy of 1621 appear to lie at the intermediary point between the two approaches to the Qurʾān described above. On the one hand, the translation was produced in Iberia within the first two decades of the sixteenth century when Muslim conversion to Christianity was a highly disputed and prioritized issue. On the other hand, it was commissioned by an Italian cardinal so interested in language acquisition that, on the same trip to Spain to commission the translation, he also purchased two grammar books of the Arabic language.4 Now lost, the original translation commissioned by Egidio was equipped with many philological aids to support an understanding of the Arabic text and thus promote the learning of the Arabic language. It was copied in four parallel columns: the first contained the Arabic source text; the second was made up of the transcription of the text into the Latin alphabet, so that someone who was not familiar with the Arabic alphabet could read it fluently; the third column contained the actual translation into Latin; finally, the fourth column was filled with quotations from Muslim exegetes, elucidating the text. At a later point, the entire text was heavily corrected by a reputed authority of that time, the erudite Leo Africanus.<sup>5</sup> While we cannot know exactly how much use the commissioner Egidio da Viterbo made of these adjustments, or how much Arabic he was able to learn from it,6 the peculiar layout of the text caught the attention of another student of Arabic a century later.7 The current chapter focuses on the uses that the Scottish scholar David Colville made of this Qurʾān translation when he copied it in 1621. Let us begin by tracing the history of this Latin Qurʾān before it reached Colville's hands.

<sup>4</sup> Alastair Hamilton, "'Nam tirones sumus.' Franciscus Raphelengius' Lexicon Arabico-Latinum, Leiden, 1613," in *Ex Officina Plantiniana. Studia in memoriam Christophori Plantini (ca. 1520–1589)*, ed. Marcus de Schepper, Francine de Nave (Antwerp: Vereeniging der Antwerpsche Bibliophielen, 1989), i.e. *De Gulden Passer* 66–67 (1988–1989), 557–589, here 561–562. On Egidio's library, see Natalie Zemon Davis, *Trickster Travels. A Sixteenth-Century Muslim between Worlds* (New York: Hill and Wang, 2006), 369, fn. 3.

<sup>5</sup> See Katarzyna Krystyna Starczewska, *Latin Translation of the Qurʾān (1518/1621) Commissioned by Egidio da Viterbo. Critical Edition and Introductory Study* (Wiesbaden: Otto Harrassowitz Verlag, 2018); and Thomas E. Burman "The Latin-Arabic Qurʾān Editions of Egidio da Viterbo and the Latin Qurʾāns of Robert of Ketton and Mark of Toledo," in *Musulmanes y cristianos en Hispania durante las conquistas de los siglos XII y XIII*, ed. Miquel Barceló and José Martínez Gázquez (Barcelona: Universitat Autònoma de Barcelona, 2005), 103–117, and Burman, *Reading the Qurʾān*.

<sup>6</sup> On this topic see Katarzyna K. Starczewska, "Anti-Muslim Preaching in 16th-Century Spain and Egidio da Viterbo's Research on Islam," *Rivista di Storia e Letteratura Religiosa* 3 (2015), 413–430.

<sup>7</sup> For a more general overview see Burman, *Reading the Qurʾān,* chapter 6 "The Manuscripts of Egidio da Viterbo's Bilingual Qur*ʾ*ān: Philology (and Polemics?) in the Sixteenth Century," 149–177.

# 5.1 The history of Egidio da Viterbo's Qurʾān

As we learn from David Colville's preface to the translation, Egidio da Viterbo obtained his translated Qurʾān when he was papal legate to the king of Portugal and Castile, a fact corroborated in other sources. In April 1518, Egidio left Rome for Spain as Pope Leo X's legate to emperor Charles V, in order to ask the latter to join forces against the Ottomans.8 It must have been during this period that Egidio received the opportunity to meet Juan Gabriel of Teruel. He employed this former Muslim and Christian convert, originally named Alí Alayzar,9 but known as Iohannes Gabriel Terrolensis in the Latin translation or Joan Gabriel, to produce a new translation of the Qurʾān.10 Juan Gabriel was likely the former *faqīh* of Teruel and had probably been forced to receive baptism in 1502 along with the other Mudéjares of that city.11 In the course of his conversion, he not only changed his status from a Muslim jurist (*faqīh*) to that of a Christian, but also began instructing Catholic preachers on the tenets of Islam so that they could preach against it with greater knowledge and conviction. The Catholic preacher Joan Martí Figuerola explains in his work *Lumbre de fe contra el Alcorán* (Valencia, 1521)<sup>12</sup> that he owed his knowledge of Arabic and of the Qurʾān to the teachings of Maestre Johan (Juan) Gabriel, a convert to Christianity.13 Figuerola was an ecclesiastical figure connected to the bishop of Barcelona, Don Martín García, from whom he took over the campaigns to preach to the Moors. In his sermons, he regularly invoked Muslim sources, especially the Qurʾān. Juan Gabriel's instructional material must have gained a certain fame among the Spanish clergy, and thus it became possible for Egidio da Viterbo to employ the former *faqīh* to translate the entire Qurʾān, not into the vernacular, as that would have probably been of little value for an Italian cardinal, but into Latin. However, the original translation must have been regarded as flawed, as Egidio subsequently decided to have it corrected in Viterbo by his godson, Leo Africanus.

<sup>8</sup> Balbino Rano, "La Orden Augustiniana en la Península Ibérica durante los años 1500–1520," in *Egidio Da Viterbo, O.S.A., E Il Suo Tempo. Atti Del V Convegno Dell'Istituto Storico Agostiniano Roma-Viterbo, 20–23 Ottobre 1982*, ed. Institutum Historicum Augustinianum (Rome: Institutum Historicum Augustinianum, 1983), 32.

<sup>9</sup> Ernesto Utrillas Valero, "Los mudéjares turolenses. Los primeros cristianos nuevos de la Corona de Aragón," in *De mudéjares a moriscos. Una conversión forzada*, ed. Centro de Estudios Mudéjares (Teruel: Centro de Estudios Mudéjares, 2003), 809–826, here 820, 823, who refers to the Muslim name as mentioned in Archivo Histórico Provincial de Teruel, *Consejo de Teruel*, Carpeta Azul, doc. 274.

<sup>10</sup> Mercedes García-Arenal and Katarzyna K. Starczewska, "'The Law of Abraham the Catholic.' Juan Gabriel as Qurʾān Translator for Martín de Figuerola and Egidio da Viterbo," *Al-Qantara* 35, no. 2 (2014), 409–459.

<sup>11</sup> García-Arenal and Starczewska, "'The Law of Abraham the Catholic.'" On Juan Gabriel, see also Katarzyna K. Starczewska, "Juan Gabriel," in *Christian-Muslim Relations, Volume 6: 1500–1900*, ed. David Thomas (Leiden: Brill, 2015), 415–419.

<sup>12</sup> Madrid, Biblioteca de la Real Academia de la Historia, MS Gayangos 1922/36.

<sup>13</sup> García-Arenal and Starczewska, "'The Law of Abraham the Catholic,'" 412–414.

Two years after he commissioned the Latin translation of the Qurʾān, Egidio met his "diversely erudite"14 godson. Though his Muslim name was al-Ḥasan al-Wazzān, in Italian circles he was better known as Leo Africanus (or Yūḥannā l-Asad). When al-Ḥasan al-Wazzān was baptized by Pope Leo X in 1520, Egidio was one of his godfathers. Five years after his baptism, Leo Africanus corrected Egidio's translation of the Qurʾān at the cardinal's residence in Viterbo.<sup>15</sup> The original manuscript with Leo's corrections has not been preserved, yet there are certain clues in David Colville's copy of 1621 that allow us to distinguish between the different layers of the text and to identify more precisely which amendments were made by Leo. Surprisingly, David Colville's copy can be read quite literally between the lines: on the basis of various comparisons and analyses, I have come to the conclusion that, with a few exceptions, the main text contains the original translation, produced in Iberia, whereas the text preserved in the space between the lines are the corrections inserted by Leo Africanus. This implies that Leo Africanus was able to detect, if not all, then at least some of the negligences and errors committed by Juan Gabriel. Moreover, Leo's corrections attest to his literal understanding of Qurʾānic Arabic, and suggest that he was not able to express himself correctly in Latin. Engaging with Gabriel's translation and Leo's corrections, Colville criticized Leo for not having been able to improve the original translation. Notwithstanding his harsh criticism of Leo's contribution, Colville did not wish to leave it out. Thus, he copied the original text together with the corrections of Leo Africanus. Colville's approach resulted in the particular layout of the manuscript (see Fig. 5.1), which was copied by the Scottish scholar in the library of El Escorial and brought with him to Milan, where it remains to this day.16

What becomes apparent when reading David Colville's prologue is the authentic concern, shared by European intellectual elites, to acquire accurate instruction in Arabic. The Qurʾānic material more generally available at the time was Theodor Bibliander's edition, published in 1543 in Basel, of Robert of Ketton's twelfth-century Latin translation.17 This version, however,

<sup>14</sup> *Liber sacrosancti Evangelii de Jesu Christo, Domino et Deo nostro*, ed. Johann Albrecht Widmannstetter (Vienna: Zymmermann, 1562), fols. a\*\*\* 4a–b, describe Leo Africanus as a man of "pleasant disposition and diverse erudition" (*ingenii amoenitatem, eruditionemque variam*).

<sup>15</sup> See Davis, *Trickster Travels*; and Natalie Zemon Davis, "Leo Africanus and his Worlds of Translation," in *Translators, Interpreters and Cultural Negotiators*: *Mediating and Communicating Power from the Middle Ages to the Modern Era*, ed. Federico M. Federici and Dario Tesscini (Basingstoke: Palgrave Macmillan, 2014), 62–80. See also Katarzyna K. Starczewska, "Leo Africanus," in *Christian-Muslim Relations. A Bibliographical History, Volume 6: 1500–1900*, ed. David Thomas (Leiden: Brill, 2015), 439–449; and Katarzyna K. Starczewska, "Leo Africanus' Contribution to a Latin Translation of the Qur'ān. A Case Study of Intellectual Activity after Conversion", *SMSR* 84, no. 2 (2018), 479–497.

<sup>16</sup> Milan, Biblioteca Ambrosiana, MS D 100 inf.

<sup>17</sup> See, among others, José Martínez Gázquez, "Las traducciones latinas del Corán, arma antislámica en la Cristiandad medieval," *Cuadernos del CEMyR* 13 (2005), 11–27; and José Martínez Gázquez, "Finalidad de la primera traducción latina del Corán," in *Musulmanes y cristianos en Hispania durante las conquistas de los* 

Figure 5.1: Sūrat al-Baqara (fragment) in Milano, Biblioteca Ambrosiana, MS D 100 inf., f. 46.

was often criticized by European scholars as being unfaithful to the Arabic original: some of its parts were abridged, the titles of the *suwar* (pl. of *sūra*) and their numbering had been assigned arbitrarily by the translator; and, on the whole, it was more a rhetorical summary of the Qurʾānic content than an actual translation.18 As commissioner of the 1518 rendition of the Qurʾān, cardinal Egidio da Viterbo chose to rely on the assistance of native speakers in his quest to understand Arabic. Colville, on the other hand, was a firm believer in self-education and reluctant to admit the native speakers' linguistic superiority. This attitude led him to commit various mistakes, described in detail below. In other words, while the copyist's criticism of Leo's language skills seems to be exaggerated, it is perhaps Colville's own knowledge of Arabic that should be called into question.

# 5.2 David Colville's studies and travels

David Colville was an erudite member of the Catholic clergy, versed in several languages.19 He was probably born in 1581 near Cleish, in eastern Scotland, the youngest son of Robert Colville and Margaret Lindsay.20 In 1597 he began his education in St. Andrews, where he studied Greek, some Hebrew, and some rudiments of Chaldean and Syriac. In 1606, Colville left for Avignon, converted to Catholicism, and commenced his theological studies, which he later completed in Rome at the Scots College in 1608. Subsequently, he went to Venice and Padua to study law, and to Bologna to deepen his knowledge of medicine. In 1617 Colville reached Spain, where he worked as a librarian in El Escorial from 1617 to 1627. He later explained in his letters that it was perhaps the most tranquil period of his life, during which he was able to devote himself to his studies.21 Among other activities, he worked on the library's collections of Arabic manuscripts22 and served as a royal interpreter by appointment of Philip III and Philip IV of Spain. Furthermore, he was a professor of Hebrew, Greek, and Arabic in

*siglos XII y XIII*, ed. Miquel Barceló and José Martínez Gázquez (Barcelona: Bellaterra, 2005), 71–77. See also Cándida Ferrero Hernández and Oscar de la Cruz Palma, "Robert of Ketton," in *Christian-Muslim Relations. A Bibliographical History, Volume 4: 1200–1350*, ed. David Thomas (Leiden: Brill, 2011), 508–519.

<sup>18</sup> See Martínez Gázquez, "Las traducciones latinas del Corán," the section "Críticas de Juan de Segovia al Corán latino de Pedro el Venerable," 26–27.

<sup>19</sup> This and the following sections are based on Starczewska, *Latin Translation of the Qurʾān (1518/1621)*, xcv–cix.

<sup>20</sup> John Durkan, "Three Manuscripts with Fife Association, and David Colville of Fife," *The Innes Review* 20 (1969), 47–49.

<sup>21</sup> Gregorio De Andrés, "Historia del texto griego Escurialense (Θ. IV. V. 30) de la vida de S. Sinclética y sus traducciones latinas," *La Ciudad de Dios* 178, no. 3 (1965), 491–511.

<sup>22</sup> Braulio Justel Calabozo, *La Real Biblioteca de El Escorial y sus manuscritos árabes. Sinopsis histórico-descriptiva* (Madrid: Instituto Hispano-Árabe de Cultura, 1978), 225. Robert Jones, "Piracy, War, and the Acquisition of Arabic Manuscripts in Renaissance Europe," *Manuscripts of the Middle East* 2 (1987), 96–110.

the college attached to the monastery.23 As far as the last of these languages is concerned, it seems that Colville did not begin studying Arabic until 1621, the year he copied Egidio's Qurʾān, and it took him a mere two years to master it.24

In 1627, Colville left El Escorial for Italy, where he hoped to find well-stocked libraries filled with stimulating material. He travelled from Valencia to Genoa and then to Rome.25 In 1628, he reached Turin as an interpreter of Charles Emmanuel I, duke of Savoy. In 1629, he came to Milan with a good number of manuscripts in Greek and Arabic, copied from El Escorial, together with his commentaries and translations. In Milan, the scholar was hosted by cardinal Federico Borromeo,26 to whom he bequeathed around twenty manuscripts copied from El Escorial, half of which were in Arabic. These manuscripts are currently preserved in the Biblioteca Ambrosiana.27

What emerges from the prologue of Colville's copy of Egidio's Qurʾān, and also from his later letters, is a sense of frustration at his lack of recognition in the field of philological studies. It might have been the cognitive dissonance between Colville's self-image as a gifted scholar and the position and remuneration he was given that led him to gloss and strive to correct the Latin Qurʾān he was copying. In the prologue to this translation, Colville complains that mortals are often fooled by misconceptions: when they see a Muslim, they believe immediately that the person is fluent in Arabic. And similarly on the Arabic side: when a Spaniard is taken captive, the captors assume straight away that their prisoner can write Spanish and read Latin. Colville draws similar analogies for the Jews and Greeks, finally concluding boastfully that while he had not been born either a Jew or a Greek, he had taught himself to know these languages better than the natives.<sup>28</sup>

Colville's prologue suggests that he had great self-confidence in his abilities to master Oriental languages. Nevertheless, five years later, in a letter sent in 1626 from El Escorial to the Jesuit Guillaume Bauters, rector of the College of Leuven from 1620 to 1625, the Scottish scholar offered his services in a text-editing capacity, and gives a more balanced résumé of his skills:

<sup>23</sup> Douglas Morton Dunlop, "David Colville, a Successor of Michael Scot," *Bulletin of Hispanic Studies* 28 (1951), 39.

<sup>24</sup> Gregorio De Andrés, "Cartas inéditas del humanista escocés David Colville a los monjes jerónimos del Escorial," *Boletín de la Real Academia de la Historia* 170, no. 1 (1973), 83–155, here 86.

<sup>25</sup> De Andrés, "Cartas inéditas," 105–110.

<sup>26</sup> Enrico Rodolfo Galbiati, "L'orientalistica nei primi decenni di attività," in *Storia dell'Ambrosiana, Il Seicento*, ed. Ada Annoni (Milan: Cassa di Risparmio delle Province Lombarde, 1992), 114.

<sup>27</sup> De Andrés, "Cartas inéditas," 89.

<sup>28</sup> Cited in Starczewska, *Latin Translation*, 5, 7: "Omnia quae in hac lingua scio, absque praeceptore didici [...] Ego uera experientia didici me qui neque domo neque natione Hebreus aut Graecus sum, utramque linguam rectius calluisse Hebraeis atque Graecis natione*.*"

"However, in order to satisfy your Most Venerable Lordship's wish, I shall answer sincerely and modestly, just as a nobleman should, regarding how little I am skilled in letters and foreign languages, without any boastfulness [. . .]. I learned Latin, Greek, and Hebrew in my fatherland as an early adolescent. I also learned Chaldean and Syrian in schools there. Subsequently, I improved my Greek while studying the entire philosophical curriculum, as I listened to the text of Aristotle himself in its original Greek. Furthermore, afterwards I improved my Hebrew during my pilgrimage, so that I can translate well and explain everything concerning grammar; as for the meaning of the Bible, to tell the truth, I would never say that I do not understand the words, but I can only grasp one thousandth part of their meaning [. . .]. I read and understand the comments of the rabbis; also the Chaldean although to a lesser degree [. . .], and in the same way Syrian, which I learned from the New Testament and from lexicons and other studies. Additionally, I cannot guarantee anything with regard to the orthography of the long and short vowels in Hebrew and especially in Chaldean [. . .]. Lastly, I learned Arabic here in this house [the Escorial], from many teachers, so that I can skillfully write it; and I transcribed many [texts] with my hand, especially two dictionaries, which consisted of many very copious volumes; however, as for writings in a purer style, and which do not degenerate into common speech or any foreign influences, I could easily explain and translate them more profusely into Latin, as Greek seems to be resisting itself."29

<sup>29</sup> My English translation. Latin text cited in De Andrés, "Historia del texto griego," 499–500: "Sed ut destricte satisfaciam desiderio R. V., ingenue referam et cum modestia prout virum probum decet, quantum possunt tenues meae vires in literis et cognitione linguarum, citra omniam jactantiam [. . .] linguam latinam, graecam et hebraeam in patria sub primos annos adolescentiae ubi et chaldeam et syriacam etiam in scholis didici; graecam subinde excolui studiis et toto curriculo philosophico, quia ipsum textum Aristotelis graece in suo fonte audiui; hebraeam etiam continuo excolui tota peregrinatione ut bene interpretari ualeam et de omnibus rationem reddere quae ad grammaticam spectant; de sensu Bibliorum, ut uerum fatear, nunquam dixerim me uoces non capere sed uix millesimum sensum percipere possem; [. . .] commentaria etiam Rabbinorum lego et intelligo; chaldaica perinde sed inferiore gradu [. . .] ac eodem modo syriaca quae ex Nouo Testamento et lexicis et aliis studiis didici; nihil praeterea polliceri possim circa orthographiam uocalium longarum et breuium in hebraeis, praecipue chaldaeis [. . .]. Denique arabicam hic in domo ista ex multis magistris didici ut non inscite scribere ualeam et multa mea manu descripsi, imprimis dictionaria duo pluribus constantia tomis et copiosissima; at quae stilo scripta puriore nec degenerunt in idiotismum aut peregrinitatem aliquam, facile illa explicare et interpretari possim lingua copiosior latina, cum graeca strenue certare uidetur." Gregorio de Andrés, "Cartas inéditas," 96–97, also translated the letter into Spanish.

Therefore, according to his own testimony, Colville taught himself Arabic, relying mainly on the manuscripts that he had at hand and that he copied.30 If we are to assume that Colville commented upon the translation he was copying the very same year he began learning Arabic, his language skills are truly impressive. Several comments and annotations contain clues about which materials were at his disposal, which he used as tools to try to correct or enrich Egidio's translation. However, the accumulation of parallel translations and comparisons with the Arabic original sometimes prevented the copyist from reaching any particular conclusion, leaving him hesitant about which rendition of the Qurʾanic text was correct.

## 5.3 Colville and Erpenius's grammar

David Colville copied the Milan manuscript in order to "be more skilful in the study of this extremely difficult language."31 From my perspective, his greatest merit derives precisely from his ignorance of Qurʾānic Arabic: as Colville was not sure which version of the translation was correct—the original authored by Juan Gabriel, or the corrections by Leo Africanus—he copied both. It appears that Colville expected to improve his Arabic by copying Egidio's Qurʾān and by commenting on its contents.

Colville was clearly acquainted with the tradition of translating the Muslim holy book, at least in a vague sense. In the prologue he explains that he compared the text he was copying "with the translation of Robert the Englishman," clearly referring to Robert of Ketton.32 Much further along, he comments on verse Q 88:21–22: "So warn [them], for you are a warner / you are not someone holding power over them" (*fa-ḏakkir innamā anta muḏakkirun / lasta ʿalayhim bi-muṣayṭir*). Alongside what appears to be Leo's translation of the Arabic term "muṣayṭir," i.e. "someone holding power," as "custos"—in contrast to Juan Gabriel's "disiunctio"—Colville notes that Robert rendered it as "tu non es coactor," and put a gloss in the margin.33 Indeed, turning to Bibliander's edition of Robert of Ketton's Qurʾān, we find "Tu namque doctor es, non coactor," i.e. "For you are a teacher, not someone who coerces." Additionally, as Colville claims, there is a gloss in this place, which states "Doctor, non coactor Machumet.

<sup>30</sup> The Biblioteca Ambrosiana manuscripts transcribed by Colville that I have been able to identify are: B 349 suss., Q 114 sup., O 42 inf., M 86 suss., P 270 sup., J. 92 sgg., S 110 sup., B 134 sup., B 137 sup., B 139 sup., B 145 sup., B. 146 sup., D 141 P inf., and Z 193 sup. I would like to thank the staff of the Veneranda Biblioteca Ambrosiana for their invaluable help in locating these materials.

<sup>31</sup> "Ideoque omnia qua potui, diligentia descripsi, ut sic exercitatior fierem in studio tam difficilis linguae, ego Dauid Coluillis Scotus in coenobio D. Laurentis 1621 in Bibliotheca Regia."

<sup>32</sup> "Et tandem contuli cum translatione Roberti Angli, et indices in margine apposui ex illo. Deus bone! Quam aliena est translatio illa ab arabico ut uix unam lineam reperias quadrare textui!"

<sup>33</sup> In the original, "disiunctio *add.* custos *et* transtulit Robertus: 'Tu non es coactor' et posuit glossam in margine." Cited in Starczewska, *Latin Translation*, 723.

Figure 5.2: Erpenius' Historia Iosephi Patriarchae ex Alcorano arabicè, D2.

Sed contrarium saepissime in Alcoran docet," i.e. "Muḥammad, a teacher, not someone who coerces. But in the Qurʾān, he very often teaches the opposite."<sup>34</sup>

Bibliander's edition of Robert of Ketton's work was not the only translation of the Qurʾān that Colville consulted. At first glance, the twelfth *sūra* of the *M* manuscript is conspicuous for its elegant lettering and the absence of verse numbers. The copyist explains that this is due to the fact that he transcribed this *sūra* before the others, having found it printed and translated by Erpenius.35 Thomas Erpenius was a central figure among European-Christian scholars of Arabic. He was a Dutch professor of Arabic in Leiden as well as the founder of an Arabic press vital for the development of European Arabist scholarship.<sup>36</sup> Erpenius was able to discern the meaning of difficult Qurʾānic passages thanks to the assistance of Aḥmad b. Qāsim al-Ḥaǧarī, a Morisco exiled to Morocco from Spain and author of *Kitāb Nāṣir al-dīn*. 37 Al-Ḥaǧarī visited Erpenius and his disciple Jacob Golius in Leiden, and all three kept up a learned correspondence after al-Ḥaǧarī's return to Morocco.<sup>38</sup>

The fragment that Colville claims to have copied from Erpenius probably comes from the latter's textbook entitled *Historia Iosephi Patriarchae ex Alcorano arabicè*, published in Leiden in 1617.<sup>39</sup> As Alastair Hamilton explains, this grammar was intended for students who were already familiar with Erpenius's earlier *Grammatica Arabica*, published in 1613. It was also one of the first books to be printed in the special press, equipped with Arabic fonts, established by Erpenius in Leiden.40 The *Historia Iosephi Patriarchae* uses as chrestomathy *sūra* 12 (*sūrat Yūsuf*), which is printed together with an interlinear word-for-word translation into Latin, and another, more approachable rendition in the margins (see Fig. 5.2). What follows in the remaining part of the manual is the translation of this *sūra* by Robert of Ketton and its grammatical commentary. The final part of

<sup>34</sup> Theodor Bibliander, *Machumetis Saracenorum principis eiusque successorum uitae ac doctrina [. . .]* (Basel: Johann Oporinus, 1550), 185, republished by the working group GRAC-UMR 5037 (September 2010), 24: http://grac.univ-lyon2.fr/dialogues-de-chretiens-avec-l-islam-682831.kjsp?RH=1464270711526.

<sup>35</sup> "Azoaram istam transcripsi prius quam caeteras, quia reperi impressam et translatam ab Herpemio." Cited in Starczewska, *Latin Translation*, 261.

<sup>36</sup> Mercedes García-Arenal and Fernando Rodríguez Mediano, *The Orient in Spain*: *Converted Muslims, the Forged Lead Books of Granada and the Rise of Orientalism* (Leiden: Brill, 2013), 245.

<sup>37</sup> Pieter S. van Koningsveld, Qasim Al-Samarrai, and Gerard A. Wiegers, "General introduction," in Aḥmad Ibn Qāsim al-Ḥaǧarī, *Kitab Nāṣir al-Dīn ʿalā l-qawm al-kāfirīn / The Supporter of Religion Against the Infidels*, ed./trans. Pieter S. van Koningsveld et al., second ed. (Madrid: CSIC, 2015), 13–74.

<sup>38</sup> García-Arenal and Rodríguez Mediano, *The Orient in Spain*, 143, 423.

<sup>39</sup> Erpenius's translation of *sūra* 12 (*sūrat Yūsuf*) is also referenced in the margins of Zechendorff's Qurʾān. Roberto Tottoli, "The Latin Translation of the Qurʾān by Johann Zechendorff (1580–1662) Discovered in Cairo Dār al-Kutub," *Oriente Moderno* 95 (2015), 5–31, here 18–19.

<sup>40</sup> Alastair Hamilton, "The Qurʾān as Chrestomathy in Early Modern Europe," in *The Teaching and Learning of Arabic in Early Modern Europe*, ed. Jan Loop et al. (Leiden: Brill, 2017), 215.

the book consists of three different Latin versions of the opening *sūra* 1 (*al-fātiḥa*) of the Qurʾān."<sup>41</sup> In Erpenius's words, the textbook contains:

"a certain sample of an authentic Arabic text, extremely accurately marked with vowels, also translated into Latin word by word, and explained; I am talking about the History of the Patriarch Joseph, as it is expounded by this Ishmaelite impostor in the Qurʾān not without added lies and fables. Because truly the Arabic language, not unlike Hebrew, cannot be in any way mastered satisfactorily without the help of a text with all the vowels correctly annotated, and there is nothing that can be regarded more correct than the Qurʾān; indeed, the Arabs themselves derive almost all their understanding of grammar from that work alone; nothing was more helpful for me than showing you a chapter of the Qurʾān, easily understandable and including much material relevant for the thorough understanding of the language."42

In his preface, Erpenius explores a change in the perception of the Qurʾān, which occurred in the early modern period. He suggests that, although the text is still full of "lies and fables," it also provides excellent training material for practising standard Arabic, "with all the vowels correctly annotated." Colville seems to subscribe completely to this idea, claiming that:

"there is some benefit to be derived from the translation of both [i.e. Juan Gabriel and Leo Africanus], even when it errs. For we can gain as many benefits and experience from the errors of others as from things well done. For this reason, I have carefully written everything I could in order to be more skillful in the study of this extremely difficult language."43

<sup>41</sup> For a more detailed description of Erpenius's grammar see Hamilton, "The Qurʾān as Chrestomathy," 215–218.

<sup>42</sup> Thomas Erpenius, *Historia Iosephi patriarchae, ex Alcorano, Arabicè. Cum triplici versione Latina, & scholijs Thomae Erpenii, cujus & alphabetum Arabicum praemittitur* (Leiden: Ex Typographia Erpeniana, 1617), A2–A3: "specimen quoddam textus Arabici authentici accuratissime uocalibus insigniti, atque de uerbo ad uerbum in Latinum uersi, & explicati; Historiam inquam Iosephi Patriarchae, ut eam impostor ille Ismaeliticus in ٍ ْآن ر ُ قـ non sine admixtis mendaciis & fabulis enarrat. Cum enim lingua Arabica, non secus atque Hebraea, sine ope textus accurate uocalibus omnibus notati addisci haudquaquam feliciter possit, nec quidquam sit quod accuratione cum Alcorano certare queat; quin ex hoc fere solo tota rei Grammaticae ratio elici ab ipsis Arabibus soleat: nihil mihi potius fuit, quam ut caput aliquod eius facile intellectu & multa ad linguae solidam intelligentiam pertinentia complectens uobis exhiberem."

<sup>43</sup> "Est tamen utilitas aliqua ex utriusque translatione delibanda etiam cum errauit, cum ex aliorum erratis quandoque non minus quam ex recte gestis emolumentum experientiamque capere possimus. Ideoque omnia qua potui diligentia descripsi, ut sic exercitatior fierem in studio tam difficilis linguae." Cited in Starczewska, *Latin Translation*, 6.

The statement also justifies Colville's approach to the text: he sees himself not as a mere copyist but as a critical reader. Consequently, his annotations should be viewed as the product of this particular learning technique. However, in spite of the apparent similarity between their views regarding the utility of the Qurʾān as a help for students of the Arabic language, a comparison of *sūra* 12 (*sūrat Yūsuf*) in Erpenius's *Historia Iosephi Patriarchae ex Alcorano arabicè* and in Colville's transcription reveals both similarities and differences. For example, in Erpenius' grammar, verse Q 12:4 is translated as follows:44

َ َد ْ ُت أَح أَي َ ّن ِ ر ِ ِت إ َ َ أَب ِ ي ِيه َب ِ ُ ُف ل ُوس َ َال ي ْذ ق ِ إ ْ ِ ل ُم ه ُ تـ ْ أَيـ َ َ ر َر َم ْق ال َ َ و ْس َ َّ الشم ًا و َْكب َ َ كو َ َشر ع َ ِين ِد َاج س

Cum dixit Iosephus patri suo: "O pater mi! Vtique ego uidi undecim stellas, et solem, et lunam; uidi eos me adorantes."44

Colville, in turn, uses his particular copying system, in which the two versions of the text are maintained. In the following passage, the version *supra lineam* is presented in parentheses, whereas the underlining is maintained as in the manuscript, where it indicates the parts of the text that should be substituted by the version *supra lineam.* Thus, Colville writes:<sup>45</sup>

َ َد ْ ُت أَح أَي َ ّن ِ ر ِ ِت إ َ َ أَب ِ ي ِيه َب ِ ُ ُف ل ُوس َ َال ي ْذ ق ِ إ ْ ِ ل ُم ه ُ تـ ْ أَيـ َ َ ر َر َم ْق ال َ َ و ْس َ َّ الشم ًا و َْكب َ َ كو َ َشر ع َ ِين ِد َاج س

Quando dixit Ioseph patri suo: "O meus pater! Ego uidi undecim stellas, et solem, et lunam, uidi eos qui me reuerebantur (mihi prosternentes)."45

The translations of the following verses, Q 12:5–8, seem to have more in common, bold font having been used here to highlight parallels. In Erpenius we read:

<sup>44</sup> Erpenius, *Historia Iosephi patriarchae,* D2.

<sup>45</sup> Starczewska, *Latin Translation,* 261.

ٰ َى ل َ َ َك ع ْي ؤ ُ ُ ْص ر ْص ق َ ََّ َ ل تـ ُن َ بـ َ َال ي )12:5( ق ان َ َ ْط ِ َّن َّ الشي ًا إ ْد ُوا لَ َك َ كي ِيد َك ي َ ِ َك فـ ت َ ْو ِخ إ َك ِ ٰ ل َك َذ َ ِ ٌني )12:6( و ُّ مب و ٌُّ َد ِ ع َان ِنس ْل ِ ل ِ ِ يث َاد ِْو ِ يل ْ الَح َ ِن ت ُ َك م ِم لّ َ ع ُ يـ َ ُّ َك و ب َ يك ر َ ِ ب َ ت ْ َي َا وب َ كم َ ُ ْق ع َ ٰ ِ آل يـ َى ل َ َع ْ َك و َي ل َ ُ ع َه ت َ ْم ع ِ ُّ ن ِم ت ُ ي َ و اق َ َ ْح ِس إ َ َ و ِيم َاه ر ْ ِبـ إ ُ ْل ب َ ِن قـ ْ َك م ي َ و َ ٰ أَبـ َى ل َ َا ع أََتَّه ان ِ ف َ ْد َ ك َ ٌ )12:7( لَّق ِيم َك ٌ ح ِيم ل َ َّ َك ع ب َ ِ َّن ر إ ْذ ِ َني )12:8( إ ِ ل ِ لسائ َّ ِ َ ٌت لّ ِ آي ه ِ ت َ ْو ِخ إ َ ُ َف و ُوس ي نَّا ِ َا م ِين َِٰل أَب ب إ ُّ َ ُ أَح ُوه أَخ َ ُ ُف و ُوس َالُوا لَي ق ِ ٍني ِ َ ي ضَلٍل ُّ مب َ لَف َن ِ َّن أَب ٌ إ ة َ ُ ْصب ُ ع ْن َن َ و

tuus, et docebit te de explicatione narrationum, **et complebit gratiam suam super te et super** familia Iaacobi, sicuti **compleuit eam super** parentibus tuis antea Abrahamo et Ishaco. Etenim dominus tuus sciens, sapiens." (12:7) Certe fuerunt in Iosepho et fratribus eius signa interrogantibus. (12:8) Cum dixerunt: "Iosephus et frater eius cariores apud patrem nostrum quam nos et nos complures. Certe pater noster **in errore manifesto**. 46

(12:5) Dixit: "O filiole mi! Ne narra uisionem tuam super fratribus tuis, et struent tibi dolum, etenim Satanas homini hostis manifestus. (12:6) Et sic **eliget** te dominus

In Colville's version we read:

$$\begin{array}{c} \text{\huge{\huge{\huge{\huge{\hfil{\hfil{\hfil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phifil{\phill}}}}}}}}}}}}}}\end}}\end{}}}\end{}}}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}}\}$$

(12:5) Dixit: "O fili mi (filioli mi)! Non declares (narres) insomnium tuum super fratres tuos (fratribus tuis), quare (quia) facient traditionem tibi et dolum, et quia (certe) diabolus pro persona (hominibus) est inimicus manifestus. (12:6) Et sic audiet (**eliget**) te creator (dominus) tuus, et demonstrabit tibi solutionem (docet te ex significationibus) historiarum, **et complebit gratiam suam super te, et super** eos (familiam) Iacob, quemadmodum **compleuit eam super** patres tuos ante Abraham et Isaach; quia creator (dominus) tuus est sapiens, uidens (sciens)." (12:7) Certe fuit in Ioseph et in fratribus suis mysterium (miraculum) pro scrutantibus (rogantibus). (12:8) Et quando dixerunt: "Certe Ioseph et frater eius amatur plus a patre nostro, plusquam nos, et nos sumus congregatio, et quod (certe) pater noster est **in errore manifesto**. 47

It is tempting to assume that Colville copied the *sūra* he found in Erpenius's textbook, but then merged it with Juan Gabriel's and Leo Africanus's versions. If the copyist had truly consulted the 1617 *Historia Iosephi Patriarchae*, he would have learned from it much more than just the word-for-word translation of one *sūra*, for it contains the Arabic alphabet, information about grammar, Robert of Ketton's version of the chapter in question, and comments on Qurʾānic vocabulary and phraseology. Superficial study of Erpenius's textbook probably motivated Colville to look more closely at the Qurʾān out of philological interest, but also seems to have given him the false impression that he had already mastered the Arabic language. This impression resulted in Colville's profuse glossing of the text. In the following pages, I consider Colville's annotations containing erroneous corrections, paying particular attention to those that stand in stark contrast to the amendments made by Leo Africanus.46 <sup>47</sup>

## 5.4 Colville, Leo, and the number of verses in the Qurʾān

Over the pages of the Latin Qurʾān, Colville presents himself as a man of bold judgements. As his prologue makes clear, he is particularly inclined to disparage the skills of Leo Africanus. Dealing with verse Q 11:85, Leo rejects the word "decipiatis," correcting it with the neologism "uilatis," which, in his opinion, was closer to the Arabic meaning of the root "b-ḫs" of the Qurʾānic verb "tabḫasū," i.e. "you [plural] shall not deprive." The term "vile" seems a reasonable equivalent of "baḫs," and the form "uilatis" somehow resembles the Arabic verbal morphology. Nevertheless, Colville is appalled by Leo's liberties, as he exclaims:

"'uilatis,' id est 'uile faciatis,' inquit, ecco ridiculum glossatorem!"<sup>48</sup>

"'do not vile,' that is, 'vilify,' he says, behold the ridiculous glossator!"

Such was Colville's outrage that he even seems to have confused Latin with Italian, using the word "ecco" instead of the classical "ecce."

Colville once more vented his anger without obvious motive, this time in correct Latin, in the note that accompanies the title of *sūra* 15 (*al-ḥiǧr*), where an alternative title is provided.49 Next to the headline "de lapidibus" Leo proposes "pauimento." The Scotsman exclaims: "behold the barbarity of the corrector!" (*ecce barbariem correctoris!*). The heated remark hardly seems justified, since the word "al-ḥiǧr," translated into English as "The Rocky Tract," "The Stoneland," or "The Rock City," can with all accuracy be

<sup>46</sup> Erpenius, *Historia Iosephi patriarchae,* D3.

<sup>47</sup> Starczewska, *Latin Translation,* 261.

<sup>48</sup> Starczewska, *Latin Translation*, 257.

<sup>49</sup> Starczewska, *Latin Translation*, 285.

translated into Latin as "pauimentum," i.e. a floor composed of stones.50 There are numerous similar examples, in which Colville criticizes Leo's corrections. However, it is striking that Colville sees fit to include these alternative translations in his copy, although he considers them incorrect.

Colville appears to feel a strong need to polemicize against the glosses of the first translator, Juan Gabriel. One of many examples is his note next to verse Q 20:29 (*wa-ǧʿal lī wazīran min ahlī*). Next to the translation of the Arabic word "wazīr," rendered here as "admonitor," the copyist wrote:

"consiliarium, addit glossa quod Hispanice dicitur 'Al-guazil' sed puto eum decipi quod 'alguazil' est صيلَ الو, hoc est 'lictor' seu 'compraehensor.'"<sup>51</sup>

"Advisor, the gloss adds that in Spanish it is 'Al-guazil' ["alguacil" i.e. a minor official] but I think he is mistaken, as 'alguazil' is 'al-waṣīl,' which is 'lictor' or 'compraehensor.'"<sup>52</sup>

The question of why Colville would seek to derive the etymology of the Spanish word "alguacil" from "al-waṣīl" (i.e. "the intimate friend") remains open. He might have relied on an external source of information or simply made it up himself. In any case, from the perspective of modern etymological studies, Juan Gabriel's gloss is impeccable.<sup>53</sup>

Nevertheless, Colville leaves numerous glosses in which we can see that he does not necessarily denigrate or judge, but simply studies and collects information that he found relevant for his understanding of the Qurʾān and its characteristic form of Arabic. On numerous occasions, we see Colville comparing the Arabic text with the translation as he copies it, struggling to understand the equivalence between the original and the Latin version. We might interpret his comment on the last word of verse Q 22:28 (*li-yašhadū manāfiʿa lahum wa-yaḏkurū sma llāhi fī ayyāmin maʿlūmātin ʿalā mā razaqahum min bahīmati l-anʿāmi fa-kulū minhā wa-aṭʿimu l-bāʾisa l-faqīra*) in this way, i.e. the term "al-faqīr," i.e. a poor person. In reference to the Latin word "pauperibus" (rendered in the main text in its plural form but corrected into singular "pauperi" between the lines), he comments:

"in alio *erat* ريِ ِ َق الف et cum cesra [*?*] et in alio ُ ِري َق الف cum demma."<sup>54</sup>

<sup>50</sup> An annotation of a similar kind can be found next to the Latin heading of *sūra* 46, titled according to the manuscript "de Arditate." Colville's note in reference to the title reads: "Ecce barbariem, uoluit dicere 'Colliculorum.'" In this case, Colville's alternative is more adequate. Cited in Starczewska, *Latin Translation*, 551.

<sup>51</sup> Starczewska, *Latin Translation*, 334.

<sup>52</sup> Starczewska, *Latin Translation*, 334.

<sup>53</sup> *Diccionario de la Real Academia Española* (DRAE), s.v. *alguacil, la*: "Del ár. hisp. alwazír, y este del ár. clás. Wazīr," accessed November 24, 2017, http://dle.rae. es/?id=1ny83D5.

<sup>54</sup> Starczewska, *Latin Translation*, 355.

"In the other [copy] it was *al-faqīri* with *kasra* and in [yet] another *al-faqīru* with *ḍamma*."<sup>55</sup>

Colville's annotation is significant in two ways: primarily it provides further confirmation that he used more than one copy of the Qurʾān in Arabic. This would not have been difficult given the number of Muslim holy books kept at El Escorial.<sup>56</sup> Secondly, it proves how little Colville actually knew of Arabic grammar, since he could not distinguish which was the correct *ḥaraka*—in other words, which short-vowel ending is grammatically correct within this sentence structure.

In various annotations, Colville expresses his concern about the correct numeration of the Qurʾānic verses. The following comment combines two of Colville's obsessions—the correct count of verses and criticism of Leo Africanus. The gloss is located between the title of *sūra* 15 and its opening *basmala* and makes direct reference to the verse count that almost always accompanies the *sūra* titles. The copy states that *sūra* 15 contains ninety-nine verses, which corresponds to the verse count in modern standard editions. Colville, however, comments:

"Nevertheless I found ninety-seven, in the Arabic title it is said that there are ninety-seven, in the Latin translation ninety-nine; and the very inept corrector inserted the letter *d*, which means that there is a doubt. He should not have doubted but corrected in the right way for once, and written 'ninety-seven.'"57

At the beginning of *sūra* 18, which was supposed to contain 150 verses, the copyist acknowledges that he did not find more than ninety-eight, although in the other codex there were said to be 121.<sup>58</sup> Colville shows similar attentiveness at the beginning of *sūra* 21, where he expected to encounter 112 verses, but found only 109.59 It is worth noting that other seventeenth-century European scholars of the Qurʾān found the question of verse numbering and division particularly challenging.60 Roberto Tottoli associates the problem with the fact that Christian translators were unable to identify an undisputed standard in this matter. Interestingly, it seems

<sup>55</sup> Starczewska, *Latin Translation*, 355.

<sup>56</sup> See Justel Calabozo, *La Real Biblioteca*, 224–226.

<sup>57</sup> In the original, "reperi tamen 97, in titulo arabico dicitur quod sunt 97, in translatio latina 99; et corrector ineptissimus posuit literam 'd' qua uult significare esse dubium. Non debebat dubitare sed corrigere saltem semel recte et scribere 'nonaginta septem.'" Cited in Starczewska, *Latin Translation*, 285.

<sup>58</sup> Starczewska, *Latin Translation*, 313.

<sup>59</sup> Starczewska, *Latin Translation*, 343.

<sup>60</sup> See Reinhold F. Glei and Roberto Tottoli, *Ludovico Marracci at Work: The Evolution of his Latin Translation of the Qurʾān in the Light of his Newly Discovered Manuscripts with an Edition and a Comparative Linguistic Analysis of Sura 18* (Wiesbaden: Harrassowitz, 2016), 20–31.

that Colville also seized this opportunity to practice Arabic numerals, as he proudly copies them in their original script in various places.61

# 5.5 Colville's dictionaries

Judging by the remarks preserved in the margins of this Latin translation, it seems likely that, while copying Egidio's Qurʾān, Colville had numerous resources at his disposal. He probably started with Erpenius's abridged grammar, and, as he went along, compared aspects of the translation with various Arabic originals. Occasionally, he also cited the authority of a dictionary.

In his letter to the Jesuit Guillaume Bauters, mentioned above, Colville claimed to have copied two voluminous Arabic dictionaries at El Escorial. These dictionaries have been identified by Gregorio de Andrés as *Al-Qāmūs al-muḥīṭ*, written by Muḥammad b. Yaʿqūb al-Fayrūzābādī in the early fifteenth century, and *Tāǧ al-luġa wa-ṣiḥ*ā*ḥ al-*ʿ*arabiyya*, written by Ismāʿīl b. Ḥammād al-Ǧawharī in the eleventh century.62 The *Tāǧ al-luġa wa-ṣiḥāḥ al-*ʿ*arabiyya* was a fairly popular glossary among European Arabists, and among Arabic speakers. It was particularly famous as the first dictionary to order words according to the last letter of their root, a practice that proved to be helpful in composing rhymed poetry.<sup>63</sup>

In the pages of Egidio's Qurʾān, Colville mentions having consulted al-Ǧawharī's dictionary on three occasions in his comments on verses Q 15:74, Q 17:5, and Q 18:9. He transcribes the Arabic name "al-Ǧawharī" as "Goheri." Commenting on Q 18:9 (*am ḥasibta anna aṣḥāba l-kahfi wa-rraqīmi kānū min āyātinā ʿaǧaban*), Colville refers to the translation of the Arabic word "ar-raqīm," which Juan Gabriel translated as "flumen" (river) and Leo corrected to "riuus" (brook). However, Colville was not convinced by either of these translations. He annotates that his *(sic!)* "dictionary of al-Ǧawharī says that it was a tablet in which the deeds and the names of those who are in hell were written."64 Evidently the copyist was not aware of the *tafsīr* tradition, i.e. Qurʾānic exegesis, according to which "ar-raqīm" was the name of a river or a valley.65 It is telling that, unlike the medieval

<sup>61</sup> See, e.g., chapters VI, XX, XLIV, LII, LVI, and LVIII.

<sup>62</sup> De Andrés, "Historia del texto griego," 498–500.

<sup>63</sup> See García-Arenal and Rodríguez Mediano, *The Orient in Spain*, 254–255, 343.

<sup>64</sup> Colville comments: "istud [*?*] رقيم dictionarium meum Goheri dicit esse tabulam in qua scripta sunt gesta et nomina eorum qui sunt in inferno." Colville seems to refer to al-Ǧawharī's dictionary, *Tāǧ al-luġa wa-ṣiḥāḥ al-'arabiyya*, ed. Muḥammad Muḥammad Tāmir (Cairo: Dār al-ḥadīṯ, 2009), 461 [s.v. "raqm"], who writes: "yuqāl: huwa lawḥun fīhi asmāʾuhum wa-qiṣaṣuhum." The second part of the Latin phrase "qui sunt in inferno" seems to be a later addition. Cited in Starczewska, *Latin Translation*, 313.

<sup>65</sup> See, among others, al-Ṭabarī's exegesis of 18:9, in which he uses the word "wādin" to mean both "river" and "valley." Al-Ṭabarī, *Ğāmiʿ al-bayān fī taʾwīl al-Qurʾān*, ed. Aḥmad Muḥammad Šākir, 24 vols. (Beirut: al-Risāla, 2000), vol. 17, 602.

European scholars who worked on the Latin translations of the Qurʾān,<sup>66</sup> Colville left aside any religious interpretation authored by Muslim exegetes. It appears that he was not familiar with any work of *tafsīr* and was interested only in the literal meaning of Qurʾānic Arabic. Curiously, his consultation of al-Ǧawharī's definitions is inconsistent. As already mentioned above, he only used it on three occasions, placed relatively close together in the text. Given the fact that Colville demonstrated rather limited knowledge of Arabic at the time of glossing Egidio's translation, one wonders whether he had already copied al-Ǧawharī's text. What we know is that Colville finished the copy of this Qurʾān in 1621, and that by 1623 he had been studying Arabic for two years and had copied two dictionaries of the language. Moreover, during his stay in El Escorial, Colville had access to at least two more Arabic dictionaries. In fact, it is highly likely that he consulted the work of Pedro de Alcalá, and almost certain that he wrote a short preface to the glossary authored by Leo Africanus and Jacob Mantino.

Pedro de Alcalá—author of *Arte para ligeramente saber la lengua arábiga* and of the dictionary of spoken Arabic *Vocabulista arábigo en letra castellana*, printed in Granada in 1505—composed both works at the request of the archbishop of Granada, Hernando de Talavera (d. 1507).67 The works were intended for use by Christian clergy preaching in areas of southern Spain that remained Arabic-speaking. However, local Moriscos may also have used such resources as reference material.68 Pedro de Alcalá's dictionary was peculiar in that the Arabic words were transcribed into the Latin alphabet, as there was no printing press with Semitic types in Spain of the early sixteenth century. This characteristic might account for the confusion in Colville's notes, for, if I am not mistaken, it was Pedro de Alcalá's dictionary to which the copyist refers as "the lexicon of Granada." Referring to the Arabic term "maqīl" (يلِ ق َ م(, i.e. "resting place," Colville writes in the gloss to verse Q 25:24 (*aṣḥābu l-ǧannati yawmaʾiḏin ḫayrun mustaqarran wa-aḥsanu maqīlan*):

"'meridiem' *s.l. et* id est, inquit glossa: 'Locus in quo statur ad umbram in meridie.' Inuenique id ipse in lexico Granatenis scriptum مكيال cum caph. et hispanice dicitur: 'la sesta.'"

"'Midday,' which is, as the gloss says, 'a place in which one stands in shade at noon.' And I myself found it in the lexicon of the Granadian written 'makīlan' with a kāf, and in Spanish it is called 'la siesta.'"<sup>69</sup>

<sup>66</sup> See, e.g., Thomas E. Burman, "Tafsir and Translation: Traditional Arabic Qurʾān Exegesis and the Latin Qurʾāns of Robert of Ketton and Mark of Toledo," *Speculum* 73 (1998), 703–732.

<sup>67</sup> Petrus Hispanus, *De lingua arabica libri duo*, ed. Paul de Lagarde (Göttingen: Diederich, 1883).

<sup>68</sup> See García-Arenal, Rodríguez Mediano, *The Orient in Spain*, 39–40.

<sup>69</sup> Starczewska, *Latin Translation*, 379.

Indeed, in Pedro de Alcalá's dictionary, printed without Arabic fonts, we find an entry that states: "'sesteadero lugar para tener siesta' [i.e. "napping place to have a siesta"]: maqyāla, -āt."70 The key problem with this attribution is that Pedro de Alcalá's system of transcription does not distinguish between *kāf* (ك (and *qāf* (ق(, which he transcribes indiscriminately as *q* or *c*. 71 Had Colville been aware of this inconsistency, he would not have mentioned that Pedro de Alcalá wrote "maqīl" (يلِ ق َ م (with the letter "kāf" (ك(, simply because it was impossible to know how the author of the dictionary wanted to write it. However, the clearly Hispanic context of the gloss and the reference to the "dictionary of Granada" make plausible the assumption that Colville was using Pedro de Alcalá's book.

Another dictionary that passed through Colville's hands, but of which he made no mention while glossing Egidio's Qurʾān, is the wordbook that Leo Africanus authored together with Jacob Mantino.72 The manuscript containing this dictionary is preceded by a short description, probably in Colville's hand.73 Interestingly, in this description, the dictionary's author is said to be unknown ("Incerto authore") and, this time, there are no critical remarks regarding his knowledge of Arabic.

# 5.6 Conclusion

To conclude, let us state clearly that Colville's marginal comments on this translation have little academic value; the copyist was often wrong and excessively judgemental. However, his annotations are fairly informative as to the materials available in the Royal Library of El Escorial before the 1671 fire.74 They tell us that the scholars working there on Arabic had at their disposal not only the famous collection of Mawlāy Zīdān al-Nāṣir but also Erpenius's grammar book, Robert of Ketton's translation of the Qurʾān, and more than a few Arabic dictionaries. Furthermore, Colville himself is also representative of the change in the European approach to the Qurʾān in the early modern period. Even though he was not an unbiased reader, his interests in glossing the Muslim holy book were strictly philological and almost entirely detached from medieval polemical currents. Colville was so engrossed in the Arabic grammar—the verbal forms, the declension of nouns, the numerals—that he almost entirely disregarded the theological

<sup>70</sup> Elena Pezzi Martínez, *El vocabulario de Pedro de Alcalá* (Almería: Editorial Cajal, 1989), 471.

<sup>71</sup> See Abdelouahab El Imrani, "Lexicografia Hispano-Árabe. Aproximación al análisis de cinco diccionarios elaborados por religiosos españoles" (unpublished PhD thesis, Universidad Complutense de Madrid, 1998), 30–33.

<sup>72</sup> Madrid, Real Biblioteca del Escorial, MS 598, Manuscritos árabes. This dictionary has been described by Davis, *Trickster Travels*, 84–85.

<sup>73</sup> Madrid, Real Biblioteca del Escorial, MS 598, f. 3.

<sup>74</sup> See, among other works, Daniel Hershenzon, "Traveling Libraries: The Arabic Manuscripts of Muley Zidan and the Escorial Library," *Journal of Early Modern History* 18, no. 6 (2014), 535–558.

dimensions of the text. One is left to speculate whether Colville would have hoped to engage with the content of the Qurʾān in order to defend his own religious views once he attained sufficient mastery of Arabic. In any case, thanks to Colville's lack of language skills, we have one less refutation of the Qurʾān in Latin Christendom. However, this is not the only debt we have to Colville's ignorance. We should also be thankful that his befuddlement while copying Egidio's Qurʾān caused him to leave it very much as he had found it. Thanks to his lack of discernment we are able to consult two versions of the work: the original one by Juan Gabriel and the one authored by Leo Africanus.

# Jan Scholz

# **6. Cicero and Quintilian in the Arab World? Latin Rhetoric in Modern Arabic Rhetorical and Homiletical Manuals**

One usually does not associate Arabic rhetoric with Roman authors who wrote in Latin. Instead, one is rather concerned either with Arabic rhetoric, i.e. the autochthonous Arabic tradition of rhetoric *(ʿilm al-balāġa)*, or the reception of Greek rhetoric *(ʿilm* or *fann al-ḫaṭāba)*, in particular Aristotle's rhetoric, in the heyday of Arabic Aristotelianism between the ninth and the thirteenth centuries.1 Modern Arabic rhetorical manuals constitute an important primary source for the history of rhetorical theory in the Arabic world, which has received hardly any scholarly attention so far.2 These manuals start appearing from the end of the nineteenth century onwards, first in Lebanon, then in Egypt, where they still play an important role today. Designed either for a more general public or concretely addressing

<sup>1</sup> A note on terminology: Arab authors mostly use the expression *fann al-ḫaṭāba* (or *ʿilm al-ḫaṭāba*) when referring to rhetorical theory, i.e. to the theory of public speech. Sometimes also, the mere term *al-ḫaṭāba* is used, although it literally designates public speech and not its theory. Aristotle's *Rhetoric* is mostly referred to as *Kitāb al-ḫaṭāba*.

<sup>2</sup> The only article explicitly analysing such a modern rhetorical manual is Abdulrazzak Patel, "Nahḍah Oratory: Western Rhetoric in al-Shartūnī's Manual on the Art of the Orator," *Middle Eastern Literatures* 12, no. 3 (2009), 233–269. Patel provides a number of crucial observations, among others with regard to the influence of Cicero on al-Šartūnī's manual. Philip Halldén, "What is Arab Islamic Rhetoric? Rethinking the History of Muslim Oratory Art and Homiletics," *International Journal of Middle East Studies* 37, no. 1 (2005), 19–38, mentions some of these manuals, calling for the need to study rhetorical theory in the Arab world, but he does not provide a concrete analysis of the modern rhetorical manuals, and speaks of rhetorical theory in a general sense, without identifying the rhetorical awakening at the end of the nineteenth century. Charles Hirschkind, *The Ethical Soundscape: Cassette Sermons and Islamic Counterpublics* (New York: Columbia University Press, 2006), refers to the manuals and provides some quotations. However, he does not enter into a detailed analysis of the phenomenon. Furthermore, see Jan Scholz, "Modern Arabic Rhetorical Manuals: A Transcultural Phenomenon," in *Engaging Transculturality: Concepts, Key Terms, Case Studies*, ed. Laila Abu-Er-Rub et al. (Abingdon: Routledge, 2019), 170–184. The article Jan Scholz, "Dramatic Islamic Preaching: A Close Reading of ʿAmr Khālid," in *Religion and Aesthetic Experience: Drama – Sermons – Literature*, ed. Sabine Dorpmüller et al. (Heidelberg: Heidelberg University Publishing, 2018), 149–170, in turn, draws on modern Islamic televangelism and its modern homiletical basis.

the needs of Islamic preachers, thus designed as homiletical manuals, they are used, for instance, to teach preaching at al-Azhar University. Available in the bookshops of Lebanon, Egypt, and other Arab countries, they sell quite well, as a bookseller in Cairo told me.

Because they unite the different rhetorical traditions mentioned above, these manuals constitute a transcultural phenomenon.<sup>3</sup> For the history of Latin-Arabic entanglement, they constitute an interesting object of research for different reasons: they not only draw upon Arabic and Greek, but also upon elements of Roman, i.e. Latin rhetoric, even though to a lesser degree. I will explore this entanglement through an examination of writings on rhetorical performance, this being one field in which Latin-Roman authors were particularly prolific. Graeco-Roman rhetorical tradition forms part of the (so-called) 'Occidental' philosophical tradition. Since it is often considered an important element in the construction of what is defined as 'European' or even 'Western' thought, one can regard it as a sort of cultural marker.4 In his article "Rhetoric and *ʿilm al-balāġa*," William Smyth goes as far as to state that, "[t]raditionally, rhetoric has formed one of the bases of Western culture."<sup>5</sup> The role rhetorical traditions occupy in acts of cultural differentiation is also evident in the recurring distinction between Arabic rhetoric on the one side, and European, Western, or Occidental rhetoric on the other side, e.g. in manuals.6 Against this backdrop, the reception of Latin rhetoricians in the Arabic context also assumes relevance in terms of cultural identity construction, a topic to be explored at the end of this chapter.

The chapter is structured as follows: Firstly, I will differentiate between the different understandings of rhetoric, Greek (and later Graeco-Roman, or today Western) and Arabic rhetoric.<sup>7</sup> In a second step, I will explain why it seems fruitful to search for Latin or Roman influences in those sections of Arabic rhetorical manuals dealing with performative questions.8 Subsequently, I will discuss whether it is justified to speak of a Roman influence on Arabic rhetorical manuals by tracing processes of transmission that reveal strong links between Egyptian and Lebanese manuals. In a last step, I will show that some passages by al-Ǧāḥiẓ also play an important role in the discussion of the performative aspects of rhetoric. As references to al-Ǧāḥiẓ are particularly prominent in Muslim rhetorical manuals, such references can be understood as a kind of cultural marker.

<sup>3</sup> Scholz, "Modern Arabic Rhetorical Manuals."

<sup>4</sup> William Smyth, "Rhetoric and *ʿIlm al-balāgha*: Christianity and Islam," in *The Muslim World* 82, no. 3–4 (1992), 242–255.

<sup>5</sup> Smyth, "Rhetoric and *ʿIlm al-balāgha*," 242.

<sup>6</sup> See for instance Thomas Bauer, "[Rhetorik, außereuropäische] V. Arabische Kultur," in *Historisches Wörterbuch der Rhetorik*, ed. Gert Ueding (Tübingen: Niemeyer, 2007), vol. 8, 111–137; Smyth, "Rhetoric and *ʿIlm al-balāgha*," 254.

<sup>7</sup> Particularly in this part of the chapter, but also in other parts of the article, I draw on observations outlined in Scholz, "Modern Arabic Rhetorical Manuals."

<sup>8</sup> Roman influence also plays a role in parts of rhetorical manuals that deal with other questions. However, given the importance of performative aspects within rhetorical theory, this chapter will focus on the latter.

# 6.1 Different understandings of rhetoric

Although it is common to use the term rhetoric with regard to both the Arabic *(ʿilm al-balāġa)* and the Greek *(fann al-ḫaṭāba*9*)* traditions of rhetoric, these two variants are only partly comparable. Consequently, using the term rhetoric in a general sense can easily lead to misunderstandings. Originally, rhetoric refers to the theory of public speaking, literally of the public speaker (Greek: *rhḗtōr*).<sup>10</sup> The tradition of Graeco-Roman rhetoric treats all aspects of public speech relating to the speaker, the speech, and the audience. Therefore, it is particularly important to stress that the Graeco-Roman tradition reflects not only upon how the text should be structured, written, and stylistically elaborated, but also considers performative questions, i.e. the question of how a speech should be delivered. It is therefore common to distinguish between text-oriented and performance-oriented parts of Graeco-Roman rhetorical theory. The former, for instance, deal with figures of style (or tropes), linguistic embellishments, rhetorical argument, and reasoning, all of which are regarded as serving the aim of persuading the audience. The performance-oriented parts discuss how to use and modulate the voice, as well as how to employ gestures, facial expressions, and body language in order to express different emotions and affect the audience.

Arabic rhetoric *(ʿilm al-balāġa)*, in turn, is primarily a tradition of literary rhetoric. It reflects upon the text-oriented parts; figures of style, semantic questions, appropriate expressions, linguistic embellishments, etc., but—except for some rather marginal considerations at the beginning of the tradition—does not attribute much importance to the performative aspects of speech.11 In view of this difference between the Greek and the

<sup>9</sup> A note on vocalization: It has been stated that "*ḫaṭābī* [and thus *ḫaṭāba*] refers to the logical rhetoric of *falsafa* [renvoie à la rhétorique-logique de la *falsafa*], *ḫitābī* [and thus *ḫiṭāba*] to the pragmatic rhetoric of *balāġa* [à la rhétoriquepragmatique de la *balāġa*]. (Pierre Larcher, "Eléments de rhétorique aristotélicienne dans la tradition arabe hors la *falsafa*," in *La Rhétorique d'Aristote: Traditions et commentaires de l'antiquité au XVIIe siècle*, ed. Gilbert Dahan and Irène Rosier-Catach (Paris: Vrin, 1998), 241–256, here 254.) However, the difference in vocalization is not always as clear as Larcher suggests. The modern rhetorical manuals often use the vocalization *ḫaṭāba*, and they indeed explicitly link to the Greek tradition. However, this understanding is not simply a rhetorical-logical one, but also pragmatic. I have therefore opted for the vocalization *ḫaṭāba* throughout this chapter.

<sup>10</sup> Gregor Kallivoda et al., "Rhetorik," in *Historisches Wörterbuch der Rhetorik*, ed. Gert Ueding (Tübingen: Niemeyer, 2005), vol. 7, 1423–1740, here 1424.

<sup>11</sup> The need for this differentiation is not always emphasized. It is quite common to simply speak of Arabic rhetoric without insisting that it is a literary tradition. See, for example, Bauer, "[Rhetorik, außereuropäische] V. Arabische Kultur"; Muhsin J. al-Musawi, "Arabic Rhetoric," in *Encyclopaedia of Rhetoric*, ed. Thomas O. Sloane (Oxford: Oxford University Press, 2001), 29–33. Other authors do make this distinction, e.g. Kristina Stock, *Arabische Stilistik* (Wiesbaden: Reichert, 2005), 4; Renate Würsch, "Rhetorik und Stilistik im arabischen Raum," in *Rhetorik und Stilistik (Rhetoric and Stylistics): Ein internationales Handbuch historischer und systematischer Forschung—An International Handbook of Historical and Systematic Research*,

Arab traditions, one might suppose that the former would supplement the latter. In pre-modern Arabic writings on rhetoric, however, the two traditions generally stayed separate. This is mainly because—following the late Alexandrian tradition—Greek rhetoric was received as a branch of logic in the Arabic context.12 Consequently, it was mostly separated from the context of the *bulaġāʾ*, i.e. those practicing *balāġa*. Furthermore, it was not conceived as a theory of public speech that would allow speakers to enhance their rhetorical performance.

This changed in the course of the late modern period.13 From the second half of the nineteenth century onwards, rhetoric as a theory of public speech became gradually more important in the Arab world. Some manuals serve to illustrate this. A rhetorical manual designed for Christian preachers, written in the eighteenth century by Ǧarmānūs Farḥāt, the Maronite bishop of Aleppo, at some point before his death in 1732, has been republished in different editions.14 One of the most important authors of what is later referred to as the "awakening of rhetoric" (*ḫaṭāba*),15 is the Jesuit

ed. Ulla Fix et al. (Berlin: de Gruyter, 2009), 2041. In Anglophone studies, the distinction is sometimes made by referring to literary rhetoric as rhetoric, and by using the term oratory to refer to a theory of public speech. For instance see, Seeger A. Bonebakker, "Aspects of the History of Literary Rhetoric and Poetics in Arabic Literature," *Viator* 1 (1970), 75–95. This distinction is unsatisfying, however, for two reasons: first, the term rhetoric in its original understanding refers to a theory of public speech; it is only with the literarization of rhetoric, and the increasing concentration of rhetorical theory on literary texts, that the term is *also* used in the sense of literary rhetoric. On this, see Julia Schmid, "Rhetorik und Stilistik in der Literaturwissenschaft," in *Rhetorik und Stilistik*, ed. Ulla Fix et al. (Berlin: de Gruyter, 2009), 1887; Elias Torra, "Rhetorik," in *Einführung in die Literaturwissenschaft*, ed. Miltos Pechlibanos et al. (Stuttgart: Metzler, 1995), 101. Second, etymologically speaking, the terms rhetoric and oratory can be regarded as synonyms, deriving respectively from the Greek term *rhḗtōr* and its Latin counterpart, *orator*.


Louis Cheikhô (1859–1927). His *Book on the Science of Literature (Kitāb ʿIlm al-adab)*, published in three editions, treats rhetoric *(ʿilm al-ḫaṭāba*) in the second volume.16 While the first of these modern Arabic rhetorical manuals, treating rhetoric as a theory of public speech, were written by Christian authors, the interest in rhetorical theory soon gained relevance beyond this sphere. Cheikhô, for instance, states that—in his times—rhetoric began assuming importance in Muslim intellectual circles, and cites important intellectuals of the late nineteenth and early twentieth centuries, such as Ǧamāl al-Dīn al-Afġānī (1838/39–1897), Muḥammad ʿAbduh (1849–1905), and Muṣṭafā Kāmil (1874–1908), as examples of Muslims interested in the art of public speech.17

An article by Muḥammad Kurd ʿAlī (1876–1953), published in 1918, "*Al-Ḫaṭāba ʿinda l-Ifranǧ*"18 (Rhetoric among the Europeans) illustrates the new importance of rhetoric. Drawing on two French rhetorical manuals by Maurice Ajam (1861–1944)<sup>19</sup> and Silvain Roudès (dates unknown),<sup>20</sup> Kurd ʿAlī introduces the reader to many French orators, briefly sketching their techniques for preparing, rehearsing, and delivering their speeches. Kurd ʿAlī does not provide many details on ancient rhetoricians, but relates a story that goes back to Cicero:21 The Roman Sulpicius Galba, when he practiced his performance at home, worked himself into such a state of excitement that, when he eventually left his house, his eyes shot arrows. We also find the amusing detail that, when Galba went to the forum to deliver his speech, his secretaries, who followed him, still suffered from

17 Šayḫū [Cheikhô], *ʿIlm al-ḫaṭāba,* 238.

<sup>16</sup> It has been published in three editions: Luwīs Šayḫū [Louis Cheikhô], *Kitāb ʿIlm al-adab: Maqālāt li-baʿḍ mašāhīr kuttāb al-ʿarab fī l-ḫaṭāba wa-l-šiʿr*, vol. 2: *ʿIlm al-ḫaṭāba* (Beirut: Maṭbaʿat al-abāʾ al-mursalīn al-yasūʿiyyīn, 1889); Luwīs Šayḫū, *Kitāb ʿIlm al-adab*, second ed., vol. 2: *ʿIlm al-ḫaṭāba* (Beirut: Maṭbaʿat al-abāʾ al-yasūʿiyyīn, 1913); Luwīs Šayḫū, *Kitāb ʿIlm al-adab*, third ed., vol. 2: *ʿIlm al-ḫaṭāba* (Beirut: Maṭbaʿat al-abāʾ al-yasūʿiyyīn, 1926). Patel dated this manual by referring to the second edition from 1913, apparently overlooking the first edition from 1889. On this basis, he concluded that Saʿīd al-Šartūnī's rhetorical manual (*Al-ġuṣn al-raṭīb fī fann al-ḫaṭīb*) is "one of the first, if not the first known, work devoted entirely to the art of oratory in the *nahḍah* period." See Patel, "Nahḍah Oratory," 261. In so doing, he contradicts Muḥammad ʿAbd al-Ġanī Ḥasan, *Al-Ḫuṭab wa-l-mawāʿiẓ* (Cairo: Dār al-maʿārif, 1980), who states that Cheikhô wrote the first book on rhetorical theory during the *nahḍa.* This misunderstanding is probably based on the fact that Ḥasan does not provide the title or the date of the work by Cheikhô he refers to, whereas Patel overlooks the first edition, concluding that Cheikhô's book is from 1913. See Patel, "Nahḍah Oratory," 239. Contrary to what Patel states, Cheikhô is indeed the first author in the *nahḍa* of a book devoted to rhetorical theory.

<sup>18</sup> Muḥammad Kurd ʿAlī , "Al-Ḫaṭāba ʿinda l-Ifranǧ," *Al-Muqtabas* 95 (1914), accessed September 30, 2017, https://ar.wikisource.org/wiki//95\_العدد**/**المقتبس**\_**مجلة .الخطابة\_عند\_اإلفرنج

<sup>19</sup> Maurice Ajam, *La parole en public: physio-psychologie de la parole, rapport du langage intérieur avec la parole, étude des procédés oratoires depuis l'Antiquité, esquisse d'une méthode scientifique d'art oratoire, enquêtes psychologiques sur la parole en public* (Paris: Chamuel, 1895).

<sup>20</sup> Silvain Roudès, *L'Orateur moderne: L'éducation de la parole, ou l'art d'apprendre à parler en public* (Paris: Pancier, 1909).

<sup>21</sup> Cicero, *Brutus. Orator*, trans. George L. Hendrickson and Harry M. Hubbell (Cambridge, MA: Harvard University Press, 1939), 80–81.

the punches and slaps he had given to them while rehearsing his speech.22 In his narration of the story, Kurd ʿAlī chose a rather ironic tone. However, the fact that he focused on the importance of delivery is significant: from the nineteenth century onwards, the increasing importance of rhetoric as a theory of public speech (*fann al-ḫaṭāba*) in the Arab world would not be limited to the text-oriented parts of speech, but would include a number of reflections on performative aspects.

This new interest is reflected in the publication of a number of Arabic rhetorical manuals, particularly since the beginning of the twentieth century. While some of these manuals deal with public speech in a general sense,23 a large number, addressed at Islamic preachers, were used for training purposes at al-Azhar University from around 1918 onwards.<sup>24</sup> Among the first manuals is *The Art of Rhetoric and the Preparation of the Orator* (*Fann al-ḫaṭāba wa-iʿdād al-ḫaṭi*̄*b*) by ʿAlī Maḥfūẓ, <sup>25</sup> written between 1926 and 1942.26 Abū Zahra, who taught rhetoric at the "Department of Principles of Religion" (*uṣūl al-dīn*) and at the Faculty of Law at Cairo University,27 published his manual *Rhetoric: Its Principles and its History during its Most Flourishing Ages Among the Arabs* (*Al-Ḫaṭāba: Uṣūluhā wa-tārīḫuhā fī azhar ʿuṣūrihā ʿinda al-ʿarab*) in 1934. To this day, it is still one of the most successful rhetorical manuals and has been published in several editions.

Both types of manuals—those addressing public speakers in a more general sense and those designed for preachers—belong to the tradition of *al-ḫaṭāba*. The term is used to distinguish the Graeco-Latin tradition of rhetoric, which includes reflections on performative aspects, from the more text-oriented Arabic tradition known as *ʿilm al-balāġa.* <sup>28</sup> As these new manuals deal with the art of public speech as developed in Greek antiquity and, in particular, in Aristotle's *Rhetoric*, it is hardly surprising that these manuals draw extensively upon the work of Arabic Aristotelians such as al-Fārābī (d. 339/950), Ibn Sīnā (d. 427/1037), and Ibn Rušd (d. 595/1198). Although these authorities play an important role, they do not necessarily constitute the primary source of information on the performative aspects of public speech. This is where Latin rhetoric comes in.

<sup>22</sup> Muḥammad Kurd ʿAlī, "Al-Ḫaṭāba ʿinda l-Ifranǧ."

<sup>23</sup> Among the most important are Niqūlā Fayyāḍ, *Al-Ḫaṭāba* (Cairo: Idārat al-Hilāl, 1930); Abū Zahra, *Al-Ḫaṭāba*; Aḥmad Muḥammad al-Ḥūfī, *Fann al-ḫaṭāba*, fifth ed. (Cairo: Nahḍat Miṣr, 2007 [1949]).

<sup>24</sup> Hirschkind, *The Ethical Soundscape*, 44, 48.

<sup>25</sup> In the title, the term *ḫaṭīb* and *ḫaṭāba* can be understood as referring particularly to the liturgical Friday preacher and the activity of preaching. However, many manuals address general rhetorical aspects and instruct the preacher in other regards besides the liturgical Friday sermon.

<sup>26</sup> ʿAlī Maḥfūẓ, *Fann al-ḫaṭāba wa-iʿdād al-ḫaṭīb* (Cairo: Dār al-iʿtiṣām, 1984). Although published posthumously in 1984, it must have been written between 1926 and 1942 as can be deduced from the Introduction, 7–12.

<sup>27</sup> Ibrāhīm Ḫalīl Ibrāhīm, "Al-Šayḫ Muḥammad Abū Zahra," accessed October 18, 2016, http://www.misralbalad.com/page.php?id=58084.

<sup>28</sup> Nonetheless, they do also refer to the tradition of Arabic rhetoric *(ʿilm al-balāġa)*: see Scholz, "Rhetorical Manuals."

# 6.2 Roman rhetoricians and the issue of performance

The conception of rhetorical theory as a theory of public speech goes back to Aristotle (384–322 bce), who laid the theoretical foundations of this discipline. He became a timeless authority in this field and the "measure for the time to come."29 Rhetorical literature produced in Europe today, for example, often relies essentially on the categories established by Aristotle in the fourth century bce. 30 In his *Rhetoric*, Aristotle emphasized that rhetorical theory should attribute much importance to the delivery of a speech, and "no treatise has yet been composed on delivery."<sup>31</sup> Despite his insistence on rhetorical delivery, Aristotle's treatment of this issue is rather short. He does address several important aspects of delivery, namely the role of the voice and the importance of gestures. Moreover, he compares the public speaker with an actor in the theatre, introducing a comparison that has subsequently played an important role in the whole so-called Western tradition of rhetorical theory. Ultimately, however, Aristotle does not provide many details on the orator's performance and mainly focuses on the voice.32 In view of Aristotle's limited reflection on performative questions, one cannot expect his Arabic-Islamic commentators, al-Fārābī, Ibn Sīnā, and Ibn Rušd, to have devoted considerably more attention to this issue, especially since they were primarily interested in logical questions, not the art of public speaking. Roman authors, in particular Cicero (106–143 bce) and Quintilian (ca. 35–ca. 100 ce), developed much more detailed reflections on the subject.33 Since they built on and expanded Aristotle's theoretical framework, it is common to speak of a Graeco-Roman tradition of rhetorical theory.

<sup>29</sup> Gregor Kallivoda et al., "Rhetorik," 1484. Trans. Jan Scholz.

<sup>30</sup> For an example of the role of ancient rhetoric, see Tim-Christian Bartsch et al., *Trainingsbuch Rhetorik*, third ed. (Paderborn: Schöningh, 2013).

<sup>31</sup> Aristotle, *Art of Rhetoric*, ed. and trans. John Henry Freese (Cambridge, MA: Harvard University Press, 1926), 346–347.

<sup>32</sup> See Volker Saftien, "Rhetorische Mimik und Gestik: Konturen epochenspezifischen Verhaltens," *Archiv für Kulturgeschichte* 77 (1995), 201. He exaggerates, however, when stating that, for Aristotle, *hypókrisis* meant only the voice. The comparison with the actor is already present in Aristotle; in fact, the term *hypókrisis* refers to it. See Bernd Steinbrink, "Actio," in *Historisches Wörterbuch der Rhetorik*, vol. 1, ed. Gert Ueding (Tübingen: Niemeyer, 1992), 43.

<sup>33</sup> Cicero's rhetorical works are *On invention* (*De inventione*), *The Best Kind of Orator* (*De optimo genere oratorum*), *Topics* (*Topica*), *On the Orator* (*De oratore*), *On Fate*  (*De fato*), *Stoic Paradoxes* (*Paradoxa stoicorum*), *Divisions of Oratory* (*De partitione oratoria*), *Orator* (*Orator*), and *Brutus* (*Brutus*). The *Rhetoric: For Herennius* (*Ad C. Herennium de ratione dicendi*), the oldest preserved book on rhetoric in Latin, was formerly attributed to Cicero, but is now attributed to an anonymous author. Quintilian's main work on the subject is Quintilian: *The Institutio Oratoria*, ed. and trans. Harold E. Butler, 4 vols (Cambridge MA: Harvard University Press, 1920–1922). See also the more recent edition and translation: Quintilian, *The Orator's Education*, ed. and trans. Donald A. Russell, 4 vols (Cambridge, MA: Harvard University Press, 2001–2002). For Quintilian's statements on rhetorical performance, see: Quintilian, *Institutio Oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter III, 243–317.

Based on the preceding overview, and considering that the aforementioned ancient authors remain relevant in modern times, an important question arises: to what extent did Roman authors leave a mark on modern Arabic rhetorical manuals with regard to the treatment of performance? This question is not only crucial in order to understand the transcultural character of modern rhetorical theory in the Arab world; it also links to issues which have gained momentum thanks to the so-called "performative turn." In the twentieth century, the performative aspects of speech have gained increasing attention. This was largely, but not exclusively, due to technological developments such as radio broadcasting, cinema, and television. In the Arab world, radio broadcasting began in the 1920s and state-owned national broadcasting in 1934. Foreign films arrived in the 1920s, increasingly complemented by local production since the early 1930s.34 Television was introduced in the 1960s.<sup>35</sup>

When searching for a Roman influence in modern Arabic rhetorical manuals, one must bear in mind that the works of Cicero or Quintilian do not yet seem to have been translated into Arabic.<sup>36</sup> However, since Arab authors could have had access to English or French translations of these authors, a direct influence obviously cannot be ruled out. Ancient Roman authorities are named in many rhetorical manuals that often rely on *ʿIlm al-ḫaṭāba*, which the Jesuit Louis Cheikhô published at the end of the nineteenth century.37 His status as a pioneer in the field of modern Arabic rhetorical manuals is closely associated with his studies in France and the fact that Jesuits historically attributed much importance to rhetorical theory. Born in 1859 in Mardīn (close to the Syrian border in modern-day Turkey), Louis Cheikhô went to Lebanon in 1868, and began his novitiate in the Jesuit seminary Lons-le-Saunier/France in 1874. There, he studied rhetoric in the third year.<sup>38</sup>

In his work on rhetorical theory, Louis Cheikhô explains that rhetorical theory *(ʿilm al-ḫaṭāba)* began with the Greeks and continued with the Romans. Although he insists on Aristotle's great importance, he also points to the role of rhetorical theory in the writings of the sophists Prodicus of

<sup>34</sup> Walter Armbrust, "The Formation of National Culture in Egypt in the Interwar Period," *History Compass* 7, no. 1 (2009), 155–180, here 161; *The Oxford History of World Cinema*, ed. Geoffrey Nowell-Smith (Oxford: Oxford University Press, 1996), 662.

<sup>35</sup> Rasha A. Abdulla, "An Overview of Media Developments in Egypt: Does the Internet Make a Difference?," *Global Media Journal, Mediterranean Edition* 1 (2006), 91.

<sup>36</sup> For a list of classical, medieval, and early modern Latin texts available in Arabic translation, see, Daniel G. König, "The Unkempt Heritage: On the Role of Latin in the Arabic-Islamic Sphere," *Arabica* 63, no. 5 (2016), 419–493, here 453, 471–473.

<sup>37</sup> Among these are Fayyāḍ, *Al-Ḫaṭāba*, 5, 7, 22, 119, 127; Maḥfūẓ, *Fann al-ḫaṭāba*, 21; al-Ḥūfī, *Fann al-ḫaṭāba*, 199, 204; Muḥammad Maḥmūd Muḥammad ʿImāra, *Al-Ḫaṭāba bayna al-naẓariyya wa-l-taṭbīq* (Cairo: Maktabat al-īmān, 1997), 238; ʿAbd al-Ǧalīl ʿAbduh Šalabī, *Al-Ḫaṭāba wa-iʿdād al-ḫaṭīb*, third ed. (Cairo: Dār al-šurūq, 1987), 151, 159–161. Cheikhô explicitly names the ancient Roman authorities in Šayḫū [Cheikhô], *ʿIlm al-ḫaṭāba*, 194, 229–230.

<sup>38</sup> Camille Hechaïmé, *Louis Cheikho et son livre "Le christianisme et la littérature chrétienne en Arabie avant l'Islam:" étude critique* (Beirut: Dar el-Machreq, 1967), 37.

Ceos (ca*.* 465–ca*.* 395 bce), Protagoras (ca*.* 490–ca*.* 420 bce), and Gorgias (ca*.* 485–ca*.* 380 bce). Then he lists the most important Roman successors: Cicero, Quintilian, and Longinus. Clearly relying on Cheikhô's pioneering work, later authors of rhetorical manuals depict the history of rhetorical theory *(ʿilm* or *fann al-ḫaṭāba)* in the same vein.39 It is thus obvious that rhetorical manuals in Arabic aim at conveying a certain historical understanding of the development of the so-called Graeco-Roman rhetorical tradition.

Against this backdrop, one might expect the rhetorical manuals to draw directly upon the various authorities mentioned. One does indeed find explicit references, primarily to Aristotle and to his Arab commentators, and occasionally to the Latin rhetoricians as well—not only in Cheikhô's, but also in other manuals. However, citations or borrowings are rarely marked explicitly. In a number of cases, Arab authors rely on Latin works, either quoting them literally or paraphrasing their ideas, but failing to name these works or their authors. I will illustrate this by discussing several passages in which the Arabic manuals obviously draw upon the works of Latin rhetoricians.

Following the Graeco-Roman tradition, rhetoric is regarded as an art (Greek *téchnē*, Latin *ars*). This means that public speech follows rules that can be learnt. Rhetorical performance is not a question of talent alone, but of training as well. The example of the Greek orator Demosthenes (384– 322 BC) often serves to illustrate this point. Not being particularly gifted, and suffering from a narrow and weak voice, it was his persistent training that allowed him to become one of the most notable orators of his time. To train his voice, he pronounced long speeches while holding pebbles in his mouth. To train his lungs, he would speak while climbing a mountain or hill. This account can be found in the works of both Cicero and Quintilian.<sup>40</sup> It was retold by ʿAlī Maḥfūẓ and Muḥammad Abū Zahra who, however, neither quoted the account verbatim nor provided references to the original work(s).41 The same applies to a story about Demosthenes's training in front of a mirror to improve his bodily delivery, told by Quintilian42 and retold by ʿAlī Maḥfūẓ. 43

Cicero reports another anecdote about Demosthenes, which frequently serves in modern manuals of rhetoric to emphasize the importance of delivery. When asked about the most important facet of public speech,

<sup>39</sup> See e.g. Abū Zahra, *al-Ḫaṭāba*, 10–11.

<sup>40</sup> Cicero, *On the Orator: Books 1*–*2,* ed. and trans. Harris Rackham and Edward W. Sutton (Cambridge, MA: Harvard University Press, 1942), 190–193; Quintilian, *Institutio Oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter III, 54, 270–273.

<sup>41</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 20; Abū Zahra, *al-Ḫaṭāba*, 21, 50. Abū Zahra (p. 21) refers to the book *Tārīḫ al-ḥaḍāra* in his first mention of Demosthenes. It is a translation of Charles Seignobos's *Histoire de la civilization*.

<sup>42</sup> Quintilian, *Institution Oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter III, 68, 280–281; see also Quintilian, *The Orator's Education*, ed. and trans. Russell, vol. 4, 120–121.

<sup>43</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 20.

Demosthenes replied, "Delivery." When asked which the second and third most important facets were, again he replied, "Delivery."44 One of the first Arabic rhetorical manuals, written by the Lebanese author Saʿīd al-Šartūnī, reproduces this account, again without citing the source.<sup>45</sup>

The discussion of developing one's own skills to overcome a weakness or defect, which has been outlined so far, shows the importance these manuals attribute to delivery. This emphasis results from the conceptualization of rhetorical theory as elaborated within the Graeco-Roman tradition. Roman rhetoricians, in particular, theorized upon the possibilities of affecting the listener emotionally by means of the oral and bodily performance, and consequently took great pains to explain how a successful orator uses both voice and gestures to transmit his emotions to the audience. Arabic rhetorical manuals often reproduce these explanations. They not only adopt the general concept of affecting the listener emotionally by means of voice and gestures, but also include a number of details that confirm their indirect or direct dependency on a work of Roman rhetoric. Cicero, for example, emphasizes in *On the Orator*: "[e]verything depends on the countenance, while the countenance itself is entirely dominated by the eyes;"46 [. . .] "the whole delivery is an expression of the soul, and the facial expressions, an image of the soul, where the eyes indicate the state of the soul."47 Another passage is found in his *Orator*, a later work on rhetoric: "as the face is the image of the soul, so are the eyes its interpreters, in respect of which the subjects under discussion will provide the proper limits for the expression of joy or grief."<sup>48</sup> Similarly, Quintilian states that gestures appeal to the eye and the voice to the ear, "the two senses by which all emotion reaches the soul."49 And elsewhere he asserts, "But of the various elements that go to form the expression, the eyes are the most important, since they, more than anything else, reveal the temper of the mind."<sup>50</sup>

Arabic rhetorical manuals contain very similar assertions. ʿAlī Maḥfūẓ writes: "delivery is particularly important because through it, he [the orator]

<sup>44</sup> Cicero, *On the Orator: Book 3. On Fate. Stoic Paradoxes. Divisions of Oratory*, ed. and trans. Harris Rackham (Cambridge, MA: Harvard University Press, 1942), 168–169.

<sup>45</sup> Al-Šartūnī, *Al-Ġuṣn al-raṭīb fī fann al-ḫaṭīb*, 44.

<sup>46</sup> Cicero, *On the Orator: Book 3*, ed. and trans. Rackham, 176–177.

<sup>47</sup> My translation into English follows Cicero, *De oratore: Lateinisch–deutsch*, ed. and trans. Theodor Nüßlein (Düsseldorf: Artemis & Winkler, 2007), 418–421. Rackham's translation avoids the term "soul" here, see Cicero, *On the Orator: Book 3*, ed. and trans. Rackham, 176–177: "For delivery is wholly the concern of the feelings, and these are mirrored by the face and expressed by the eyes." [Animi est enim omnis actio, et imago animi vultus, indices oculi.]

<sup>48</sup> Cicero, *Brutus. Orator*, trans. Hendrickson and Hubbell, 350–351.

<sup>49</sup> Quintilian, *Institutio Oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter III, 14, 250–251. Russell's newer translation avoids the term "soul" and instead translates *animus* as "mind." See Quintilian, *The Orator's Education*, ed. and trans. Russell, vol. 4, 90–91.

<sup>50</sup> Quintilian, *Institutio Oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter III, 75, 284–285.

transports his feelings to the soul of the listener and moves his affects";51 the orator's delivery "illustrates what is in the soul;"<sup>52</sup> "The face as well as the gaze should be like a mirror of the soul, illustrating its emotions."<sup>53</sup> This corresponds to Louis Cheikhô's understanding.<sup>54</sup> Although the respective sources are not cited, both the analogies regarding the conceptualization of performance and the chosen wording clearly indicate that the Latin authors served as a model.

A necessary condition for a good delivery is that the speech is memorized. Consequently, Latin rhetoricians treated *memoria* as a section of rhetorical theory in its own right. Modern Arabic rhetorical manuals also deal with the memorizing of a speech (Arabic *ḏākira*, sometimes *ḥāfiẓa*) in sections addressing aspects of performance. Again, the influence of Graeco-Roman conceptions is clearly visible: the Arabic manuals regularly emphasize that a speech learnt by heart will have greater effect than a speech read from a sheet of paper, thus insisting that the orator should memorize it.<sup>55</sup> But while the Latin authors elaborated on this topic in great detail—Cicero, for example, even developed a proper mnemonic technique—the treatment of memory is comparatively short and general in the Arabic manuals. However, attentive reading shows that the latter rely on the Roman authors: in the Roman rhetorical tradition, memory is repeatedly referred to as a treasure-house (*thesaurus*). This metaphor is first used in the *Rhetorica ad Herennium*, originally (wrongly) attributed to Cicero, and later in Cicero and Quintilian as well.56 Among the Arabic manuals, Saʿīd al-Šartūnī uses the Arabic equivalents *ḫizāna* (storage, treasure-house) and *kanz* (treasure) when speaking of memory.57

Although the quoted Arab authors do not explicitly refer to the Roman authorities in the above-mentioned passages, it is impossible to ignore the many parallels—ranging from the role of the face and the gaze to express emotions via their conceptualization as mirrors of the soul to the metaphor of the treasure-house. This also applies to the conceptual idea, so prominent in the Graeco-Roman tradition, that a listener is emotionally affected by the orator's bodily performance. To illustrate this idea, Graeco-Roman

<sup>51</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 64: "Šaʾnuhu [šaʾnu l-adāʾi l-ḫaṭābiyyi] fī l-ḫaṭābati ʿaẓī mun li-annahu bi-ḥusni l-adāʾi yanqulu ilā nafsi l-sāmiʿi mašāʿirahu wa-yuḥarriku ahwāʾah." Trans. Jan Scholz.

<sup>52</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 65: "bayān mā fī l-nafs." Trans. Jan Scholz.

<sup>53</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 67: "Wa-yaḥsunu bi-l-waǧhi wa-l-naẓari an yakūna ka-mirāʾati li-l-nafsi fī bayāni ʿawāṭifihā." Trans. Jan Scholz.

<sup>54</sup> Šayḫū [Cheikhô], *ʿIlm al-ḫaṭāba,* 142: "Wa-yaḥsunu [. . .] bi-l-waǧhi wa-l-naẓari an yakūnā ka-marāʾati l-nafsi fī bayān ʿawāṭifihā."

<sup>55</sup> For instance Maḥfūẓ, *Fann al-ḫaṭāba*, 64.

<sup>56</sup> [Pseudo-Cicero], *Rhetorica ad Herennium*, ed. and trans. Harry Caplan (Cambridge, MA: Harvard University Press, 1954), 204–205; Cicero, *On the Orator: Books 1*–*2*, ed. and trans. Rackham and Sutton, 14–15; Quintilian, *Institutio oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter II, 1, 212–213. The Latin term *thesaurus* is obviously of Greek origin (θησαυρός). However, the metaphor referring to memory as a treasure house does not stem from Aristotle's *Rhetoric*.

<sup>57</sup> Al-Šartūnī, *Al-Ġuṣn al-raṭīb*, 45.

JAN SCHOLZ

rhetorical theory regularly compares the orator to the actor; emphasizing, however, that the orator should not only act, but also ensure that the emotions he expresses are truly his own.58 In order to successfully transmit his emotions, the speaker is advised to first affect himself with the emotions he wants to express. This aspect had not yet been developed by Aristotle, who only commented, "the hearer suffers along with the pathetic speaker."59 Cicero and Quintilian, in turn, addressed the topic of self-affectation explicitly and in more detail.60 The most famous version of the concept within the Graeco-Roman tradition was formulated by Horace (65–8 bce).<sup>61</sup> Most Arabic rhetorical manuals formulate the concept in general terms, for instance: the speaker should "affect himself until the sign of his straight excitement becomes evident in his voice, his gestures, and his facial expressions";62 or that "only the self-affected can affect others."63 Although kept in general terms, such passages are clearly influenced by the Graeco-Roman concept. We find clear proof of such an influence in the rhetorical manual by Aḥmad Muḥammad al-Ḥūfī, first published in 1949. The manual quotes Horace literally: "*Iḏā aradta minnī an abkiya fa-ʿalayka an tabkiya awwalan.*" ("If you would have me weep, you must first weep yourself.")<sup>64</sup>

#### 6.3 Channels of transmission and the interdependence of Egyptian and Lebanese manuals

The preceding elaborations show that we can find a number of quotations—some marked, others not—which lead back to Roman authors. Since many manuals neither mark their quotations nor cite their sources,

<sup>58</sup> For the relevant passages, see Steinbrink, "Actio."

<sup>59</sup> Aristotle, *On Rhetoric: A Theory of Civic Discourse*, trans. George A. Kennedy (Oxford: Oxford University Press, 1991), 235 (II 7, 5).

<sup>60</sup> Cicero, *On the Orator: Books 1*–*2*, ed. and trans. Rackham and Sutton, 332–335; Quintilian, *Institutio oratoria*, ed. and trans. Butler, vol. 3, Book VI, chapter II, 25–36, 431–437; see also: Quintilian, *The Orator's Education*, ed. and trans. Russell, vol. 3: Books 6–8, 58–61.

<sup>61</sup> Rüdiger Campe, "Affizieren und Selbstaffizieren: Rhetorisch-anthropologische Näherung ausgehend von Quintilian, *Institutio oratoria*, VI 1–2," in *Rhetorische Anthropologie: Studien zum homo rhetoricus*, ed. Josef Kopperschmidt (Munich: Fink, 2000), 138.

<sup>62</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 66: "Wa-anna yataʾaṯṯara ḥattā yaẓhara aṯara l-infiʿāli l-muʿtadili fī ṣawtihi wa-išāratihi wa-malāmiḥi waǧhih." Trans. Jan Scholz.

<sup>63</sup> Abū Zahra, *al-Ḫaṭāba*, 58: "Inna lā yuʾāṯṯiru illā l-mutaʾaṯṯir." See also: al-Šartūnī, *Al-Ġuṣn al-raṭīb*, 48.

<sup>64</sup> Al-Ḥūfī, *Fann al-ḫaṭāba*, 128. The Latin original: "Si vis me flere, dolendum est/primum ipsi tibi." See Quintus Horatius Flaccus (Horace), *Ars poetica: Die Dichtkunst. Lateinisch/Deutsch*, ed. and trans. Eckart Schäfer (Stuttgart: Reclam, 1998), 10. In English, this passage is sometimes translated as "If you would have me weep, you must first express the passion of grief yourself." See Quintus Horatius Flaccus (Horace), *The Works of Horace*: *Translated Literally into English Prose*, trans. Christopher Smart, (Philadelphia: Whetham, 1836), 2: lines 102–103. The earliest Arabic translation of Horace known to me is: Hūrātiyūs [Horatius Flaccus, (Horace)], *Fann al-šāʿir* (Cairo: Al-Hayʾa al-miṣriyya l-ʿāmma li-l-taʾlīf wa-l-našr, 1970).

establishing concrete dependencies first requires detecting literal quotations and paraphrases of Roman authors in the Arabic manuals, then comparing them to the original statements, not limiting oneself to the sections on performative aspects. This done, one would have to take into account that authors of Arabic manuals quoting or paraphrasing Roman authors may have had recourse to earlier Arabic as well as to European works on rhetoric. Thus, reconstructing channels of transmission would entail establishing interconnections first between various Arabic manuals by taking note of their different publication dates, then between Arabic and European manuals, taking into account which manuals in which European languages would have been available to the respective authors. Such a study would be highly valuable for a better understanding of the modern history of rhetorical theory in the Arabic context. However, given the number of works to consider, producing such a study would require considerable effort.

Since it is impossible to pursue this objective in the current chapter, the following elaborations will focus on the connection between the Egyptian and Lebanese manuals. A closer look shows that the former rely on the latter, suggesting that the Egyptian reception of Roman authors depends on previous Lebanese engagement with them. Since the use of Lebanese manuals is not always indicated by the authors of Egyptian manuals, it is again necessary to establish interdependencies by highlighting obvious parallels. I will base my discussion on the Egyptian manuals by ʿAlī Maḥ fūẓ, written between 1926 and 1942, and Abū Zahra, published in 1934, which figure among the most influential Egyptian rhetorical manuals. Both manuals shall be related to the Lebanese manuals by Louis Cheikhô, first published in 1889, then republished in a third edition in 1926, and Saʿīd al-Šartūnī, published in 1908.

ʿAlī Maḥfūẓ does not provide any sources for his discussion of performative aspects. It is plausible to assume, however, that he draws on the Lebanese manuals, which were published before his manual was written. In fact, ʿAlī Maḥfūẓ quotes Saʿīd al-Šartūnī literally, but without marking the quotation, when he points to the importance of the voice during delivery. In the following quote I have used square brackets to mark the amendments to al-Šartūnī's text made by Maḥfūẓ and to indicate slightly differing formulations in the footnotes. The amendments are not marked in the English translation.

Li-l-ṣawti fī l-ḫaṭābati l-taʾṯīru l-akbaru[,] li-annahu65 al-mutarǧimu ʿan maqāṣidi l-ḫaṭībi wa-l-kāšifu ʿan aġrāḍihi li-muṣāḥabatihi l-alfāẓi 66 ka-l-šāriḥi lamma urīda bi-hā mimmā lā tastaqillu bi-l-kašf ʿanhu [, liannahu l-ṭarīqu ilā qalbi l-sāmiʿi wa-l-mumaṯṯilu li-ṣūrati l-maʿānī amāmahu]. Wa-ṭabaqatu l-ṣawti wa-l-lafẓu wa-hayʾatu l-waǧhi

<sup>65</sup> Šartūnī: *fa-huwa* instead of *fa-innahu*.

<sup>66</sup> Šartūnī: *li-annahu yaṣḥabu l-alfāẓ* instead of *li-muṣāḥabatihi l-alfāẓi*.

wa-ḥarakātu l-ǧismi kulluhā tataḍāfaru ʿalā bayāni mā fī l-nafsi[,] wa-taṣwīri mā bi-l-ḫāṭir.<sup>67</sup>

The voice has the greatest effect in public speech, because it translates the speaker's purposes and uncovers his intentions, because it accompanies the words. It is like the commentator when something is intended by the words, which is not independent from its uncovering [i.e. which needs to be uncovered, in order to be understood]. [The voice has the greatest effect] because it is the way to the heart of the listener and illustrates the meanings' form in his presence. The register of the voice, the wording, the mien of the face, and the movements of the body, they are all tightly interwoven in the explanation of what one bears in the soul and the illustration of what one bears in mind.

In the following passages, Maḥfūẓ's manual also relies repeatedly on that of al-Šartūnī. 68 In addition, Maḥfūẓ also quotes Louis Cheikhô literally, for example when he discusses the voice, insisting on its moderation (*iʿtidāl*) and the need for variety (*tafannun*), and underlining that "every letter has to get its own right," i.e., must be pronounced properly. Moreover, his remark that "the wide place and the abundance of listeners need a more precise and stronger voice," is taken literally from Louis Cheikhô.69 Here again, the quotation is not marked. These quotations—to which one could add others—confirm the assumption that ʿAlī Maḥfūẓ engaged intensively with the Lebanese manuals.

It is difficult to explain why ʿAlī Maḥfūẓ did not mention the two Lebanese authors. One might suppose that it was because the two Lebanese manuals were written by Christians, but there is no evidence to corroborate such an assumption, especially since Abū Zahra, a conservative Muslim scholar,70 explicitly honours Cheikhô's role as an intellectual pioneer, responsible for what he calls an "awakening of rhetoric."71 This suggests that, in 1934, when Abū Zahra's manual was published, Cheikhô's faith did not impair his intellectual reputation among conservative Muslim scholars. Assuming that this was any different for ʿAlī Maḥfūẓ would be speculation.

In sum, both Maḥfūẓ's and Abū Zahra's manuals show that Muslim authors of rhetorical manuals, writing in Egypt between the 1920s and 1940s, had recourse to Lebanese manuals written by Christian authors around the turn of the nineteenth to the twentieth century. Although this intellectual

<sup>67</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 65; al-Šartūnī, *Al-Ġuṣn al-raṭīb*, 46.

<sup>68</sup> This is the case, for example, when he emphasizes the need for good pronunciation (*ḥasan al-lafẓ*). See Maḥfūẓ, *Fann al-ḫaṭāba*, 65. Al-Šartūnī, in turn, uses the term *nuṭq faṣīḥ* (clear articulation). See al-Šartūnī, *Al-Ġuṣn al-raṭīb*, 46.

<sup>69</sup> Maḥfūẓ, *Fann al-ḫaṭāba*, 65; Šayḫū [Cheikhô], *ʿIlm al-ḫaṭāba,* 141. Some passages are slightly paraphrased. However, this is evidently a literal quotation.

<sup>70</sup> Ibrāhīm, "Al-Šayḫ Muḥammad Abū Zahra."

<sup>71</sup> Abū Zahra, *Al-Ḫaṭāba*, 14.

transfer from Christian to Muslim contexts was not always acknowledged explicitly, apparently it was not regarded as problematic. To further explore this issue, I would like to discuss another aspect of Arabic rhetorical theory and the recourse to a particular rhetorical tradition. The following section will highlight how Abū Zahra's manual deals with the remarks of al-Ǧāḥiẓ (d. 255/869) on performance. Al-Ǧāḥiẓ was a Muslim intellectual who stands at the beginning of the Arabic tradition of rhetoric. While further studies would be needed to provide a satisfying answer, his reception might be interpreted as revealing a certain desire, on Abū Zahra's part, to assign a more prominent place to the Muslim author, who, in many cases, did not play a comparable role for earlier Christian authors discussing performative aspects.72

#### 6.4 Entangled legacies: The use of al-Ǧāḥiẓ vis-à-vis the Graeco-Roman tradition

While it is generally true that the Arabic tradition of rhetoric *(ʿilm al-balāġa)* has concentrated on the text-oriented parts of rhetoric without attributing a central role to performative questions, one cannot claim that the tradition of Arabic rhetorical theory did not feature performative reflections at all. Particularly in the early ages of Arabic rhetoric, some reflections on performance indeed existed. The most important author of these is al-Ǧāḥiẓ. While his statements on this topic are not particularly extensive, they nonetheless occupy an important place in some of the modern Arabic rhetorical manuals, as I will outline in the present section. The role of al-Ǧāḥiẓ links to the issue of Latin-Arabic entanglement, because the modern manuals—particularly those published in Egypt—refer not only to the Roman authors, but combine these references with additional references to al-Ǧāḥiẓ. From an external perspective, one could regard the Roman authors as Western authorities, and al-Ǧāḥiẓ as an Islamic authority. Although Abū Zahra does not use these categories, it is striking that, in his rhetorical manual, al-Ǧāḥiẓ gains considerable importance as a reference point with regard to performative aspects. Several later manuals follow in Abū Zahra's footsteps when quoting al-Ǧāḥiẓ.

Differentiating between the various ways of producing meaning, al-Ǧāḥiẓ conceptualizes gestures as "associates" (*šurakāʾ*, sg. *šarīk*) of words. A gesture can translate, i.e. reproduce meaning by different means, accompany, and even substitute for a word. The ways to express meaning with the help of the eyes, the eyebrows, and the extremities, are thus—in al-Ǧāḥiẓ's view—"a great help" for the orator.73

<sup>72</sup> However, one must be careful with premature conclusions, given that the Lebanese author Niqūlā Fayyāḍ, *Al-Ḫaṭāba*, 13, 23, 25, 26, 27, 36, 51, 52, 73, 111, for example, refers to al-Ǧāḥiẓ several times, although generally not in reference to performative aspects.

<sup>73</sup> Abū ʿUṯmān ʿAmr b. Baḥr al-Ǧāḥiẓ, *Al-Bayān wa-l-tabyīn*, ed. ʿAbd al-Sallām Muḥammad Hārūn, seventh ed., 4 vols (Cairo: Maktabat al-Ḫānǧī, 1998), vol. 1,

While al-Ǧāḥiẓ mentions quite a number of elements that enhance vocal performance, including gestures and facial expressions, his view on performative aspects is less elaborate than that offered by Roman authors. Quintilian, for instance, enters into much more detail, elaborating on gestures, the eyes, the gaze, facial expressions, the moving of one or both eyebrows, tears, eye-lid positions, and movements of the head, vigorous movements of the arms, single gestures, and movements of the hands, the shoulders, pointing to one's chest, clapping hands, walking during the speech, etc.74

To understand why Roman authors provide much more detail on body language, we must consider the respective cultural contexts. The theatre played an important role in ancient Greece and Rome; accordingly, Aristotle and the Latin authors reflected upon the actor's performance and the effect it had on the audience from an aesthetic point of view. In consequence, these authors emphasized the degree to which body language and voice modulations could express different emotional states.75 It is not surprising that the orator was thought of in comparable terms: he could achieve an emotional effect on the audience by making use of these performative elements. This is not the case in the Arab tradition, where theatre did not play a major role until the nineteenth century. There were some theatrical traditions, such as street theatre and shadow plays, but the cultural elite did not attach a degree of social importance to it that could be compared to the attention showered upon the theatre in the Graeco-Roman world or in modern Europe.

When al-Ǧāḥiẓ provides some outlines concerning bodily performance, he does so—to speak anachronistically—from a semiotic perspective. He is well aware that gestures play an important part in conveying meaning, and says so. However, he does not discuss the ways in which gestures can affect the listener and how these can help in conveying emotions. The few remarks he offers on gestures are quite general and focus on the possibilities of expression.76 His aim is not to provide a detailed and systematic theorization of how an orator is able to affect the listener.<sup>77</sup>

76 His minimal attention to the subject does not lead to the conclusion that gestures and facial expressions were not used in practice, or would not affect listeners. The difference between authors writing in Arabic and other authors lies in the way the authors theorize upon these aspects.

<sup>78.</sup> The section is partly translated into English in: Abū ʿUṯmān ʿAmr b. Baḥr al-Ǧāḥiẓ, *The Life and Works of Ǧāḥiẓ: Translations of Selected Texts*, trans. Charles Pellat and D. M. Hawke (Berkeley, CA: University of California Press, 1969), 103.

<sup>74</sup> Quintilian, *Institutio oratoria*, ed. and trans. Butler, vol. 4, Book XI, chapter III, 61–III, 136, 278–317.

<sup>75</sup> Aristotle, for instance, explicitly hints at the effect of gestures when he "implies that acting out a role [by means of gestures] will help to induce the concomitant feelings." See Aristotle, "Poetics," ed. and trans. Stephen Halliwell, in *Aristotle: Poetics, Longinus: On the Sublime, Demetrius: On Style* (Cambridge, MA: Harvard University Press, 1995), 89 note c.

<sup>77</sup> Another feature that seems to have influenced the reflections on gesture in the Graeco-Roman tradition is the central role played by the concept of persuasion. The notion of persuasion, around which the Graeco-Roman rhetorical tradition

That the Graeco-Roman tradition attached such a high degree of importance to performative aspects is one of the reasons why this tradition was increasingly acknowledged and received in the Arabic context from the end of the nineteenth century onwards. The Graeco-Roman tradition provided reflections and instructions on various facets of bodily performance that the autochthonous tradition did not offer. However, yet to be discussed below is the extent to which a distinction between autochthonous and foreign elements makes sense. It is obvious, in any case, that these reflections and instructions became increasingly important in the modern Arab world of the late nineteenth and early twentieth centuries.

Asserting that the Graeco-Roman tradition put more emphasis on performative aspects certainly does not imply that the Arabic tradition lacks a detailed treatment of performance. Treating this facet of the Arabic tradition in terms of a deficiency or a shortcoming would be decidedly essentialist as well as Eurocentric. The Arabic tradition did not "fail" to meet a need, e.g. because of a lack of effort or theoretical sophistication. An elaborate theory of rhetorical performance did not develop because the need for one did not arise. It is only from a Western rhetorical perspective that the need for a detailed treatment of rhetorical performance was conceived. On what basis could one conclude that the same need existed in a different socio-political, socio-economic, and/or cultural environment? Theatre did not play the same role in Arab society, which is why a need to reflect on oratory performance along the lines of the Graeco-Roman tradition did not arise. Consequently, a different perspective on oratory practice emerged and developed.

evolved, does not play a comparable role in Arabic rhetoric (*ʿilm al-balāġa*). Aristotle defines rhetoric as "an ability in each [particular] case, to see the available means of persuasion," see Aristotle, *On Rhetoric*, trans. Kennedy, 36 (I 2,1), amendment in the original. Arabic rhetoric, in turn, discusses the correct conveyance of a message. In order to attain this goal, "the speech has to conform to the requirements of the situation with concomitant linguistic purity" (*muṭābaqat al-kalām li-muqtaḍā l-ḥāl maʿa fasāḥatihi*). See Masʿūd b. ʿUmar al-Taftazānī, *Muḫtaṣar al-saʿd: šarḥ talḫīs kitāb miftāḥ al-ʿulūm*, ed. ʿAbd al-Ḥamīd Hindāwī (Beirut: al-Maktaba al-ʿaṣriyya, 2003), 31, quoted in: Bauer, "[Rhetorik, außereuropäische] V. Arabische Kultur," 111; Antonella Ghersetti, "Quelques notes sur la définition canonique de *balāġa*," in *Philosophy and Arts in the Islamic World*, ed. Urbain Vermeulen and Daniel De Smet (Leuven: Peeters, 1998), 58. On the role of persuasion in Arabic rhetoric, see also Geert J. van Gelder, "The Apposite Request: A Small Chapter in Persian and Arabic Rhetoric," *Edebiyât* 12 (2001), 1–13. Nonetheless, definitions of Arabic rhetoric (*ʿilm al-balāġa*) exist, which include the notion of persuasion explicitly, as for instance the anonymous definition quoted by al-Subkī: *balāġa* is "attaining one's demands and persuading the listener," see Bahāʾ al-Dīn al-Subkī, *ʿArūs al-afrāḥ fī šarḫ talḫīṣ al-miftāḥ*, in *Šurūḥ al-talḫīṣ* (Cairo: al-Maṭbaʿa l-amīriyya, 1317/1899), 1:124–125, cited in van Gelder, "The Apposite Request," 6. Moreover, Merlin Swartz, "Arabic Rhetoric and the Art of the Homily in Medieval Islam," in *Religion and Culture in Medieval Islam*, ed. Richard G. Hovannisian and Georges Sabagh (Cambridge: Cambridge University Press, 1999), 36–65, here 36, states that "in the case of the homily, persuasion constituted its very *raison d'être*." While it seems plausible to assume that this was the case in practice, in the theoretical discussions the notion of persuasion hardly plays any role at all.

In the Arabic-Islamic sphere of pre-modern times, the most important place of rhetorical activity was the Friday sermon. Here, the focus lay on the preacher's calm and dignified attitude. A homiletical manual for Friday preachers written by Ibn al-ʿAṭṭār (d. 724/1324) in the fourteenth century illustrates this. The homiletical manual collects different performative instructions from earlier writings, and therefore offers a concise overview on how the issue of rhetorical performance was treated in connection with preaching. The author explicitly advises the preacher to avoid any greater use of gestures. Rather, one hand should lean on the sword or staff, the other on the support (*qāma*) of the pulpit (*minbar*). The preacher should stand in an upright position and keep his head and his body calm.78 He should radiate sobriety (*sakīna*) and dignity (*waqār*) in his performance. The rejection of extensive use of gestures is also evident in the work of Ibn al-Ǧawzī (d. 597/1201), a famous preacher in twelfth-century Baghdad. He actively discourages any kind of theatrical performance or rapid gestures, emphasizing that a preacher should always remain sober.79 It is certainly difficult to deduce from such instructions that the practice of preaching in the pre-modern Arabic-Islamic sphere always adhered to this normative framework. Quite the contrary, Ibn al-Ǧawzī's admonitions prove that vivid rhetorical performances existed in Arabic-Islamic preaching, but that an excess of "theatrical" elements was connoted negatively: a dignified performer avoided vivid gestures. Particularly with regard to the liturgical Friday sermon, such historical advice for preachers is generally valid even today. While one would need further research to prove this hypothesis, it seems as if the Friday sermon, because of its ritual rules, has "conserved" a different conceptualization of public speech.

In addition to the absence of theatre, this different conceptualization of public speech, in which gestures can compromise a speaker's reputation, is probably responsible for the fact that Arab authors accorded less attention to performative issues than Roman authors.80 It goes without saying that an extensive study of public speech in Arabic history would need to analyse a much larger corpus of sources, thus allowing for a more detailed elaboration of the relationship of norms and practice in the field of rhetorical performance.

<sup>78</sup> ʿAlāʾ al-Dīn Abū l-Ḥasan ʿAlī b. al-ʿAṭṭār, *Kitāb Adab al-ḫaṭīb: Awwal kitāb ufrida fī ādāb ḫaṭīb ṣalāt al-ǧumʿa*, ed. Muḥammad b. al-Ḥusayn al-Sulaymānī (Beirut: Dār al-ġarb al-islāmī, 1996), 131.

<sup>79</sup> ʿAbd al-Raḥmān b. ʿAlī Ibn Ǧawzī, *Kitāb al-Quṣṣāṣ wa'l-mudhakkirīn: Including a Critical Edition, Annotated Translation and Introduction*, ed. Merlin L. Swartz (Beirut: Dar el-Machreq, 1971), ٩٥ (Arabic text) / 174 (translation).

<sup>80</sup> It should be noted that overly vivid gestures could compromise the orator's esteem in European venues as well. However, apparently, the use of gestures was not only theorized upon much more prominently, but also better accepted within the European tradition of public speech. It goes without saying that such considerations remain preliminary without further analysis of available sources on the topic.

In this light, it is interesting to note that the positions of al-Ǧāḥiẓ feature less prominently in the Lebanese rhetorical manuals written by the Christians Cheikhô and al-Šartūnī, while they play a dominant role in manuals written by Muslim authors in Egypt. Abū Zahra's manual, among the most popular specimens of this genre, illustrates this: Abū Zahra adds a footnote to the heading of the chapter on gestures (*al-išārāt*), in which he quotes the deliberations of al-Ǧāḥiẓ on the relationship between gesture and word (*lafẓ*).81 Obviously, Abū Zahra's remarks on performative aspects are not limited to al-Ǧāḥiẓ, but include a number of observations. He stresses, for instance, that the performative elements "are the silent speech, the language of general understanding, and often the voice of emotions, as well as the expression of feelings *(ʿibārat al-wiǧdān)*."<sup>82</sup> He underscores the importance of both intentional and unintentional gestures in public speech and criticizes the use of superfluous gestures. The speaker should, for instance, not wipe his forehead continuously, as some lawyers are apt to do, without there being any perspiration to wipe off. Similarly, he should not lift his tarboosh, because such gestures "do not point to any meaning." Instead, the gestures should follow the speech, and should not be over-abundant. The amount of gesturing "depends on the manner of the speaker, his respectful attitude (*mahāba*) and his pleasing appearance (*ruwāʾ*)."83 Here, the repeated association of gesture and meaning indicates that the author thinks along the lines of al-Ǧāḥiẓ, whose quotation he puts in a most prominent place. At the same time, however, Abū Zahra's remarks follow the line of thought characteristic of Roman authors, even if a literal quotation is not to be found in this part of the manual.

This entanglement of traditions is a characteristic of modern Arabic rhetorical manuals. On the one hand, they receive, process, and digest different elements from the Graeco-Roman tradition and, on the other hand, they draw upon and refer to Arab authors. Abū Zahra's emphasis on al-Ǧāḥiẓ might be understood as a kind of Arabic-Islamic counterbalance to the influence exerted by and attributed to Graeco-Roman authors.

#### 6.5 Conclusion

Modern Arabic rhetorical and homiletical manuals of the late nineteenth and twentieth centuries boast many explicit references to Aristotle, but contain relatively few and only implicit clues to an impact from Roman or Latin rhetoric. The preceding analysis, with its focus on the performative aspects of rhetoric, shows clearly, however, that the authors of modern Arabic rhetorical manuals drew not only upon Aristotle's *Rhetoric* and his Arabic commentators, but also on a variety of Roman authors, including

<sup>81</sup> Abū Zahra*, Al-Ḫaṭāba*, 133.

<sup>82</sup> Abū Zahra*, Al-Ḫaṭāba*, 133.

<sup>83</sup> Abū Zahra*, Al-Ḫaṭāba*, 133–134.

Cicero, Quintilian, and Horace. The metaphors of the eye as a transmitter of the speaker's emotions, the face as their mirror, the memory as a treasure-house, and finally Horace's remarks on the best and most powerful methods of affecting an audience, all stem from this tradition. While the many, partly explicit references to central figures of the Graeco-Roman rhetorical tradition allow us to corroborate Roman influence, many rhetorical manuals written by Muslim authors, such as Abū Zahra, also draw upon al-Ǧāḥiẓ. In doing so, such manuals revive an early authority of the Arabic rhetorical tradition, which had lost importance over time.<sup>84</sup>

Against this backdrop, we must speak of an entanglement of different traditions: the Graeco-Roman on the one hand, the Arabic on the other. Greek rhetoric plays a crucial role because of the impact of Aristotle's *Rhetoric* on Arabic Aristotelianism, whereas the influence of Roman authors is also clearly visible. Historically, this "awakening of rhetoric" should be understood as a facet of the *nahḍa*, a period of cultural revival in the Arab Middle East in the nineteenth and early twentieth centuries, parts of which involved a conscious engagement with cultural traditions regarded as European. As Daniel König outlines in chapter 2.6.1 of this volume, the introduction of Greek and Latin studies in Egyptian universities constitutes one of the most prominent examples of Latin-Arabic entanglement in the modern Arab world. The interest in Greek and Roman rhetorical theory not only fits the *Zeitgeist*, but it shows the extent to which the academic orientation towards Greek and Roman cultural heritage could assume importance within practical life.

Because of the historical differences between Arabic and Graeco-Roman rhetoric, Arabic rhetoric can function as a cultural marker. It has made, and still can make, a difference whether one draws upon Arabic or Graeco-Roman authorities. At least one author of an Arabic rhetorical manual, ʿAbd al-Raḥmān Ǧīra,85 goes as far as to criticize the historical engagement of Muslim authors with Aristotle's *Rhetoric*, and blames this engagement for the decline of public speech. In his view, the Arabs offered a practical, the Greeks a theoretical view on the issue. However, given that Greek rhetoric developed in a pagan society, it had little to offer to a Muslim.<sup>86</sup> Ǧīra hopes that one day public speech will be practised as if Aristotle has had no place in intellectual history.<sup>87</sup> This example, in which the reception of a "foreign" tradition is not welcomed and the author refuses to engage with types of what he regards as Western rhetoric, shows that

<sup>84</sup> Bonebakker, "Aspects of the History of Literary Rhetoric," 76, insists that a theory of oratory once existed, which later became incorporated into *balāġa*. He is not very explicit, but seems to refer to the early reflections we find in al-Ǧāḥiẓ. It is according to this perspective that the modern reference to al-Ǧāḥiẓ constitutes the revival of an old element.

<sup>85</sup> ʿAbd al-Raḥmān Ǧīra, *Al-Ḫaṭāba wa-iʿdād al-ḫaṭīb*, third ed. ([Cairo]: Maṭbaʿat wizārat al-awqāf, al-idāra l-ʿāmma li-l-marākiz al-ṯaqāfiyya, n.d). Also see Scholz, "Modern Arabic Rhetorical Manuals."

<sup>86</sup> Ǧīra, *Al-Ḫaṭāba*, 14.

<sup>87</sup> Ǧīra, *Al-Ḫaṭāba*, 42.

referring to Graeco-Roman or later European authorities can also constitute a political message. This applies not only to authors who refuse to engage with the Graeco-Roman tradition, but also to those welcoming such engagement. Other Arabic rhetorical manuals repeatedly associate rhetorical theory with democratic structures. Šalabī, for instance, not only insists that Greek culture "constitutes the foundation of human thought in different regards."88 He understands public speech, and thus rhetorical theory, as an integral element of ancient Greek democracy, and refers to the importance of public speech in European history, most significantly during the French revolution.89

From a transcultural perspective, it is important to acknowledge distinctions between "the autochthonous" and "the foreign." In some cases, the manuals do not explicitly address the origin of different influences, in others the recourse to Aristotle is understood as a recourse to a "foreign" tradition. It is important for the researcher not to reproduce conservative Arabic-Islamic positions by perpetuating a differentiation between Graeco-Roman and Arabic rhetoric. A clear distinction between the two is only partly tenable from a historical point of view: not only is it impossible to completely separate medieval Arabic from ancient Greek thought; it is also extremely difficult to unambiguously allocate modern Arabic rhetorical manuals either to the Graeco-Roman or the Arabic tradition. Although the manuals are associated more with *ʿilm al-ḫaṭāba* (or *fann al-ḫaṭāba*), the Graeco-Roman tradition, than with *ʿilm al-balāġa*, the Arabic tradition, it would be wrong to understand the Arabic rhetorical manuals as products of a purely Graeco-Roman tradition. Quite the contrary is true: the manuals draw upon the Graeco-Roman tradition, but in doing so continue a long process of entanglement, which leads back to the Arabic reception of Greek rhetoric approximately a millennium earlier. As a consequence, it is hardly possible to describe the engagement of Arabic rhetorical manuals with Aristotle in terms of one culture dealing with the product of another. The engagement of Arabic-Islamic authors with Aristotle's *Rhetoric* obviously produced results, which were Greek and Arabic at the same time.

As I have shown in this chapter, the modern Arabic rhetorical manuals also include Latin elements, which contribute to the elaboration of performative theory. Again, however, an interest in the performative aspects cannot be reduced to the reception of a Latin tradition represented by such authors as Cicero and Quintilian. Rather, performative aspects are also treated with reference to al-Ǧāḥiẓ, not only in Abū Zahra's manual but also in many later manuals, which I have not considered in this chapter. Arabic rhetorical manuals that combine Aristotle, Cicero, Quintilian, and al-Ǧāḥiẓ cannot be allocated to one single cultural tradition, but are characterized by the entanglement of different cultural traditions, even if the

<sup>88</sup> Šalabī, *Al-Ḫaṭāba*, 142.

<sup>89</sup> Šalabī, *Al-Ḫaṭāba*, 70, 144.

authors who serve as sources for these traditions are not accorded the same weight.

From the perspective of transcultural studies, such forms of entanglement constitute a methodological challenge. A researcher engaged in "cultural fossil hunting" can either decide to search for bigger concepts, in order to assign them to a specific cultural tradition, or to search for "smaller fossils" and traces that help to establish connections and influences. While this is the normal everyday business of academic research in the field of reception history, and while this is also what I have done in this chapter, there are disadvantages to this methodological practice. When referring to a certain cultural heritage, particularly when this heritage belongs to an allegedly "other" tradition, one must be aware of the challenges implied in the act of classification. What is implied when a European researcher seeks Greek and Roman elements in modern Arabic rhetorical manuals, exerting some effort in order to identify them? Given that these ancient elements are found only in Arabic rhetorical manuals from the late nineteenth century onwards: do Graeco-Roman elements thus constitute a marker of "modernity"? While this may be the case in a certain sense, one has to be careful—not only because of a long history of ancient Greek influences on medieval Arabic texts, but also because Arabic intellectual history is also marked by Latin influences. The emphasis on performative elements that is a characteristic of Arabic rhetorical manuals of the nineteenth and twentieth centuries is inspired by Latin-Roman rather than by ancient Greek authors. Given the long, mainly European and American scholarly tradition of identifying Europe as the exclusive heir to Graeco-Roman antiquity, this search for Roman elements has peculiar implications in that it makes Arab modernity dependent on what is defined as a "European" import. Thus, the search for Roman elements in Arabic rhetorical manuals can contribute to the perpetuation of cultural dichotomizations, which the transcultural approach actually tries to overcome.

Is it legitimate to stress the Graeco-Roman heritage in Arabic manuals? Doing so certainly highlights the cultural heterogeneity of modern Arabic rhetorical culture and illuminates the entanglement of different literary traditions. This is why I have chosen to build on this differentiation in this chapter, not least because distinguishing between Western and Arabic rhetoric is a well-established tradition in the realm of Arabic and Islamic Studies. However, as I have similarly put forward in another article,90 in view of the aim to overcome culturalist dichotomizations, one must acknowledge the historical dimension of this process of entanglement: at present, the mingling of different traditions in modern Arabic rhetorical manuals is approximately one hundred years old. One can indeed still stress these Roman elements today, marking them as Roman, and setting them off from Arabic elements, as one can generally make efforts to distinguish the different cultural influences that make up cultural artefacts

<sup>90</sup> Scholz, "Modern Arabic Rhetorical Manuals."

of any kind. Such a procedure certainly leads to a better understanding of how different traditions merge. At the same time, however, one has to be aware that such an understanding also entails the risks of culturalization outlined above. In his important article on Sāʿid al-Šartūnī's rhetorical manual, Abdulrazzak Patel uses the term "Western rhetoric" to designate those elements commonly regarded as part of the Western tradition. Good reasons exist for doing so. Nonetheless, in the twenty-first century, at a point in time, at which Arabic rhetorical manuals form a genre that is over a century old, it seems more appropriate to highlight the entanglement of different traditions without having recourse to the dichotomizing distinction between Western and Arabic elements. In doing so, one can raise awareness of the fact that entanglements are an integral feature of cultural processes. The category "Western" may have a certain didactical function, but its use obscures the understanding of modern Arabic rhetorical manuals as the product of entangled traditions. These can be classified as "Arabic," "Greek," "Roman," "Latin," and even "Western" in a modern European sense, if one chooses to focus on a question of origins. However, if one deems this question secondary or even irrelevant, one could also claim that these manuals are just concerned with different facets of rhetoric.

# Bibliography


Aillet, Cyrille. *Les Mozarabes. Christianisme, islamisation et arabisation en péninsule Ibérique (IXe* –*XIIe siècle)*. Madrid: CSIC, 2010.

Ajam, Maurice. *La parole en public: physio-psychologie de la parole, rapport du langage intérieur avec la parole, étude des procédés oratoires depuis l'Antiquité, esquisse d'une méthode scientifique d'art oratoire, enquêtes psychologiques sur la parole en public*. Paris: Chamuel, 1895.

ʿAkkām, Maḥmūd. "Muḥāwara bayna šarqī wa-ġarbī." In ʿAkkām, Maḥmūd. *Fikr wa-manbar: Qaḍāyā l-insān wa-mafhūmāt al-risāla fī ḫuṭbat al-ǧumʿa*, edited by Muḥammad Adīb Yasirǧī, 429–438. Aleppo: Fuṣṣilat, 2003.

Alarcón y Santón, Maximiliano, and Ramón García de Linares. *Los documentos árabes diplomáticos del Archivo de la corona de Aragón*. Madrid: Escuela de Estudios Árabes, 1940.

Al-Bakrī. *Kitāb al-Masālik wa-l-mamālik*. Edited by Adrian P. van Leeuwen and André Ferré. Tunis: al-Dār al-ʿarabiyya li-l-kitāb, 1992.

Albarus Cordubensis. *Indiculus luminosus*. Edited by Ioannes Gil. Corpus scriptorum muzarabicorum 1. Madrid: CSIC, 1973.

Alberigo, Giuseppe, and Joseph Wohlmuth, eds. and trans. *Konzilien des Mittelalters, vol. 2: Vom ersten Laterankonzil (1123) bis zum fünften Laterankonzil (1512–1517)*. Paderborn: Schöningh, 2000.

Alexander de Villa-Dei. *Das Doctrinale des Alexander de Villa-Dei*. Edited by Dietrich Reichling. Berlin: A. Hoffmann, 1893.


Al-Ǧawharī. *Tāǧ al-luġa wa-ṣiḥāḥ al-*ʿ*arabiyya*. Edited by Muḥammad Muḥammad Tāmir. Cairo: Dār al-ḥadīṯ, 2009.

Al-Ḥimyarī. *Kitāb al-Rawḍ al-miʿṭār fī ḫabar al-aqṭār*. Edited by Iḥsān ʿAbbās. Beirut: Maktabat Lubnān, 1975.

Al-Ḥūfī, Aḥmad Muḥammad. *Fann al-ḫaṭāba*. 5th ed. Cairo: Nahḍat Miṣr, 2007 [1949].

Al-Ḫušanī. *Tārīḫ quḍāt al-Andalus / Historia de los jueces de Córdoba por Aljoxani*. Edited and translated by Julián Ribera. Madrid: Imprenta Ibérica, 1914.

Al-Idrīsī. *Nuzhat al-muštāq fī iḫtirāq al-āfāq / Opus geographicum sive Liber ad eorum delectationem qui terras peragrare studeant*. Edited by Alesio Bombaci et al. 2 vols. Cairo: Maktabat al-ṯaqāfa l-dīniyya, 2002.

Al-Khalili, Jim. *The House of Wisdom. How Arabic Science Saved Ancient Knowledge and Gave Us the Renaissance*. New York: Penguin Press, 2011.


Boldrini, Sandro. *La prosodia e la metrica dei Romani*. Rome: Carocci, 1992.


———. "Michael Scot and the Transmission of Scientific Culture from Toledo to Bologna via the Court of Frederick II Hohenstaufen." *Micrologus* 2 (1994): 101–111.

———. "Petrarch and Averroes: An Episode in the History of Poetics." In *The Medieval Mind. Hispanic Studies in Honour of Alan Deyermond*, edited by Ian MacPherson and Ralph Penny, 49–56. Rochester: Tamesis Press, 1997.

———. "Translation from Arabic to Latin in the Middle Ages." In *Übersetzung: Ein internationales Handbuch zur Übersetzungsforschung*, vol. 2, edited by Harald Kittel, 49–56. Berlin: de Gruyter, 2007.


Carusi, Enrico, ed. *Il diario romano di Jacopo Gherardi da Volterra dal 7 settembre 1479 al 12 agosto 1484*. Rerum italicarum scriptores 23/3. Città di Castello: Scipione Lapi, 1904.

Cataluccio Magheri, Maria Elena, and Antonio Ugo Fossa. *Biblioteca e cultura a Camaldoli: Dal Medioevo all'Umanesimo*. Rome: Anselmiana, 1979.


*Chronicon Salernitanum. A Critical Edition with Studies on Literary and Historical Sources and on Language*. Edited by Ulla Westerbergh. Stockholm: Almqvist & Wiksell, 1956.


Cipriani, Giovanni. "Poppi: Biblioteca comunale." In *Inventari dei manoscritti delle Biblioteche d'Italia*, vol. 6, edited by Giuseppe Mazzatinti, 128–150. Forlì: Luigi Bordandini, 1896.

Clackson, Sarah Joanne. "Papyrology and the Utilization of Coptic Sources." In *Papyrology and the Study of Early Islamic Egypt*, edited by Petra M. Sijpesteijn and Lennart Sundelin, 21–44. Leiden: Brill, 2004.

*Codex Carolinus*. Edited by Wilhelm Gundlach. Monumenta Germaniae Historica, Epistolae 3, Epistolae Merowingici et Karolini Aevi I, 469– 657. Berlin: Weidmann, 1892.


Fayyāḍ, Niqūlā. *Al-Ḫaṭāba*. Cairo: Idārat al-hilāl, 1930.


Ferguson, Charles A. "Diglossia." *Word* 15 (1959): 325–340.


Fischer, Klaus-Dietrich, and Ursula Weisser. "Das Vorwort zur lateinischen Übersetzung von Rhazes' Liber continens (1282). Text, Übersetzung und Erläuterungen." *Medizinhistorisches Journal* 21, no. 3–4 (1986): 211–241.

Flavius Blondus. *De verbis Romanae locutionis*. Edited by Fulvio Delle Donne. Rome: Istituto Storico Italiano per il Medioevo, 2008.


Floristán, José M. "Diego de Urrea (c. 1559–octubre de 1616), traductor de árabe, turco y persa en la corte de España: nuevas noticias biográficas." *Boletín de la Real Academia de la Historia* CCX/2 (2013): 228–273.


Folkerts, Menso. "Adelard's Versions of Euclid's Elements." In *Adelard of Bath: An English Scientist and Arabist of the Early Twelfth Century*, ed. Charles Burnett, 55–68. London: Warburg Institute, 1987.

Foundation for Science, Technology, and Innovation: "Global Impact." Accessed November 7, 2017. http://www.fstc.org.uk/global-impact.


———. "Linguistic Cultures and Textual Production in Palermo, from the End of the 11th to the End of the 15th Century." In *A Companion to Medieval Palermo: The History of a Mediterranean City from 600 to 1500*, edited by Annliese Nef, 413–436. Leiden: Brill, 2013.


Ibn Ḥawqal. *Kitāb Ṣūrat al-arḍ*. Edited by J. H. Kramers. Leiden: Brill, 1938.


———. *Al-Sifr al-ṯānī min kitāb al-muqtabis [al-Muqtabis II-1]*. Edited by Maḥmūd ʿAlī Makkī. Riyad: Markaz al-malik Fayṣal li-l-buḥūṯ wa-ldirāsāt al-islāmiyya, 2003.

———. *Crónica de los emires Alḥakam I y ʿAbdarraḥmān II* [*al-Muqtabis II-1*]. Translated by Mah*̣*mūd ʿAlī Makkī and Federico Corriente. Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, 2001.

———. *Cronica del Califa ʿAbdarraḥmān III an-Nāṣir entre los años 912 y 942*  [*al-Muqtabis V*]. Translated by María Jesús Viguera and Federico Corriente. Zaragoza: Anubar ediciónes, 1981.

Ibn Ḥazm. *Al-Taqrīb li-ḥadd al-manṭiq wa-l-madḫal ilayhi bi-l-alfāẓ al-ʿāmmiyya wa-l-amṯila l-fiqhiyya*. In Ibn Ḥazm, *Rasāʾil Ibn Ḥazm*, edited by Iḥsān ʿAbbās, vol. 4, 93–133. Beirut: al-Muʾassasa l-ʿarabiyya li-ldirāsa wa-l-našr, 1983.

———. *Ǧamharat ansāb al-ʿArab*. Edited by ʿAbd al-Salām Muḥammad Hārūn. 2 vols. 5th ed. (Cairo: Dār al-maʿārif, n.d.).


Idel, Moshe. *La Cabbalà in Italia (1280–1510)*. Florence: Giuntina, 2007.


Jacoby, David. "The *fonde* of Crusader Acre and its Tariff. Some New Considerations." In *Dei gesta per Francos. Crusade Studies in Honour of Jean Richard*, edited by Michel Balard, Benjamin Z. Kedar, and Jonathan Riley-Smith, 277–293. Aldershot: Ashgate, 2001.

Jacquart, Danielle. "Arabisants du Moyen Âge et de la Renaissance: Jérôme Ramusio († 1486), correcteur de Gérard de Crémone

(† 1187)." *Bibliothèque de l'école des chartes* 147 (1989): 399–415.

———. "Le latin des sciences: quelques réflexions." In *Les Historiens et le Latin médiéval*, edited by Monique Goullet, 237–244. Paris: Publ. de la Sorbonne, 2001.


Janson, Tore. *Prose Rhythm in Medieval Latin from the 9th to the 13th Century*. Stockholm: Almqvist & Wiksell, 1975.


Karabacek, Joseph. "Kritische Beiträge zur lateinisch-arabischen Numismatik." *Numismatische Zeitschrift* 2 (1870): 455–493.


*Language Sciences (ICHoLS VIII), Fontenay-St. Cloud, France, 14–19 September 1999*, edited by Sylvain Auroux, 47–61. Amsterdam: John Benjamins, 2003.

———. "Eléments de rhétorique aristotélicienne dans la tradition arabe hors la falsafa." In *La Rhétorique d'Aristote: Traditions et commentaires de l'antiquité au XVIIe siècle*, edited by Gilbert Dahan and Irène Rosier-Catach, 241–256. Paris: Vrin, 1998.

———. "Moyen arabe et arabe moyen." *Arabica: Revue d'études arabes et islamiques* 48 (2001): 578–609.


———. *La formation de la langue persane*. Leuven: Peeters, 1995.


———. "Un recit de la conquête de l'Afrique du Nord." *Arabica* 1 (1954): 41–42.


Lewis, Bernard. *The Muslim Discovery of Europe*. New York: Norton, 2001.

Lindholm, Gudrun. *Studien zum mittellateinischen Prosarhythmus: Seine Entwicklung und sein Abklingen in der Briefliteratur Italiens*. Stockholm, Almqvist & Wiksell: 1963.


Lyons, Jonathan. *The House of Wisdom. How the Arabs Transformed Western Civilization*. New York: Bloomsbury Press, 2009.

Macdonald, Michael C. A. "ARNA Nab 17 and the Transition from the Nabataean to the Arabic Script." In: *Philologisches und Historisches zwischen Anatolien und Sokotra. Analecta Semitica in Memoriam Alexander Sima*, edited by Werner Arnold et al., 207–240. Wiesbaden: Harrassowitz, 2009.

Maḥfūẓ, ʿAlī. *Fann al-ḫaṭāba wa-iʿdād al-ḫaṭīb*. Cairo: Dār al-iʿtiṣām, 1984.

Mahmoud Helmy, Nelly. "Memorie Levantine e ambienti curiali: L'Oriente nella vita e nella produzione di un senese del Quattrocento: Beltramo di Leonardo Mignanelli." *Quaderni di storia religiosa* 13 (2006): 237–268.

———. *Tra Siena, l'Oriente e la curia. Beltramo di Leonardo Mignanelli e le sue opere*. Rome: Istituto Storico Italiano per il Medioevo, 2013.

Maillard, Clara. *Les papes et le Maghreb aux XIIIème et XIVème siècles. Étude des lettres pontificales de 1199 à 1419*. Turnhout: Brepols, 2014.


———. *The Kingdom of Sicily, 1100–1250: A Literary History*. Philadelphia: University of Pennsylvania Press, 2005.

Malherbe, Marc. *La faculté de droit de Bordeaux (1870–1970)*. Bordeaux: Presses universitaires de Bordeaux, 1996.


———. "Un codice in caratteri ebraici dalla Trapani degli Abbate (vat. Ebr. 358)." *Sefarad* 71 (2011): 7–24.

Mandalà, Giuseppe, and Marcello Moscone. "Tra latini, greci e 'arabici': ricerche su scrittura e cultura a Palermo fra XII e XIII secolo." *Segno e testo: International Journal on Manuscripts and Text Transmission* 7 (2009): 143–238.


*Legitimationsstrategien und Erinnerung*, edited by Matthias Becher and Jörg Jarnut, 221–241. Münster: Scriptorium, 2004.


Metcalfe, Alex: "De Saracenico in Latinum transferri. Causes and Effects of Translation in the Fiscal Administration of Norman Sicily." *al-Masāq* 13 (2001): 43–86.


Montaner, Alberto. "La Historia Roderici y el archivo cidiano: Cuestiones filológicas, diplómaticas, jurídicas y historiográficas." *e-Legal History Review* 12 (2011): 1–62.

Müller, Christian. *Der Kadi und seine Zeugen: Studie der mamlukischen Ḥaram-Dokumente aus Jerusalem*. Abhandlungen für die Kunde des Morgenlandes 85. Wiesbaden: Harrassowitz, 2013.


———. "La déportation des musulmans siciliens par Frédéric II: précédents, modalités, signification et portée de la mesure." In *Le monde de l'itinérance en Méditerranée de l'antiquité à l'époque moderne*, edited by Claudia Moatti, Wolfgang Kaiser, and Christophe Pebarthe, 455–478. Bordeaux: Ausonius, 2009.


Nöldeke, Theodor, and August Müller, eds. *Delectus veterum carminum arabicorum*. Porta linguarum orientalium 13. Berlin: Reuther, 1890.


Francesco Magistrale, Corinna Drago, and Paolo Fioretti, 445–466. Spolete: Centro italiano di studi sull'Alto Medioevo, 2002.


———. "Traccia araba su codice latino." *Litterae Caelestes. Rivista annuale internazionale di paleografia, codicologia, diplomatica et storia delle testimonianze scritte* 1 (2005): 41–60.


*introductio*. Paris: apud Dionysium Lescuier, 1538.


Quṭb, Sayyid. *Al-ʿAdāla l-iǧtimāʿiyya fī l-islām*. Cairo: Dār al-šurūq, 1995.


*(XIIIe –milieu XX<sup>e</sup> siècle)*, edited by Benoît Grévin, 141–207. Rome: École française de Rome, 2010.


.الدكتور-أحمد-عتمان-سيد-الدراسات-الكالسيكية-في-العالم-العربي/61792-article Saftien, Volker. "Rhetorische Mimik und Gestik: Konturen epochenspezifischen Verhaltens." *Archiv für Kulturgeschichte* 77 (1995): 197–226.


Neuwirth, Nicolai Sinai, and Michael Marx, 649–698. Leiden: Brill, 2011.


Swartz, Merlin. "Arabic Rhetoric and the Art of the Homily in Medieval Islam." In *Religion and Culture in Medieval Islam*, edited by Richard G. Hovannisian and Georges Sabagh, 39–65. Cambridge: Cambridge University Press, 1999.


In *Acteurs des transferts culturels en Méditerranée médiévale*, edited by Rania Abdellatif et al., 146–155. Munich: Oldenbourg, 2012.

Vercellin, Giorgio. *Il Canone di Avicenna fra Europa e Oriente nel primo Cinquecento: L'Interpretatio Arabicorum nominum di Andrea Alpago*. Torino: Utet, 1991.

Vernet, Juan. *Lo que Europa debe al Islam de España*. Barcelona: Ariel, 1978.


Wiegers, Gerard. *Islamic Literature in Spanish and Aljamiado. Yça Gidelli of Segovia (fl. 1450), His Antecedents and Successors*. Leiden: Brill, 1994.


# About the Authors

**Daniel G. König** holds the chair for the history of religions at the University of Konstanz. Trained as a Latin medievalist historian and an Arabist, his research focuses on the various dimensions of Christian-Muslim relations in the wider Mediterranean sphere, including different forms of exchange, hybridization, and mutual perception, themes also addressed in his monograph *Arabic-Islamic Views of the Latin West* (Oxford: OUP, 2015)

**Benoît Grévin,** former member of the Ecole française de Rome, holds the post of research director at the French National Centre for Scientific Research (CNRS). His research engages with the linguistic and rhetorical cultures of the medieval period from a historical perspective with a particular focus on political and pragmatic writing. Among his fields of interest are the study of Arabic in Latin milieus at the end of the medieval period as well as a comparative approach to the linguistic cultures of the Latin and the Arabic-Islamic sphere. Both themes are treated in his monograph *Le parchemin des cieux. Essai sur le Moyen Âge du langage* (Paris: Seuil, 2012).

**Daniel Potthast** is project manager of the Arabic Papyrology Database (https://www.apd.gwi.uni-muenchen.de/apd/project.jsp) and works as a lecturer and researcher at the Institute for Near and Middle Eastern Studies (LMU Munich). In 2011, he defended his doctoral thesis on Andalusī polemics against Christianity at Erlangen University, published under the title *Muslime und Christen im Andalus – Mozarabische Christen und ihre Literatur nach religionspolemischen Texten des zehnten bis zwölften Jahrhunderts* (Wiesbaden: Harrassowitz, 2013). He is interested in medieval Arabic-European entanglements, Arabic diplomatics, practices of medieval diplomacy and Christian-Muslim relations.

**Katarzyna K. Starczewska** is a Juan de la Cierva Fellow at the CCHS-CSIC in Madrid, Spain. She earned her PhD in 2012 at the Universitat Autònoma de Barcelona. Since then she has been working on a number of topics which revolve around the theme of explaining Muslim culture and Arabic language in Latin. Her interests include networks of early modern translators whose target language was Latin, the reflection of Latin interreligious polemics on early modern treatises in "national" languages, and the earliest grammar books for the Arabic language written in Latin. She is the author of *Latin Translation of the Qurʾān Commissioned by Egidio da Viterbo* (Wiesbaden: Harrassowitz, 2018).

**Jan Scholz** participated in a research project on Islamic sermons at the Cluster of Excellence "Asia and Europe in a Global Context" at Heidelberg University. There, he defended his doctoral thesis on Islamic preaching in contemporary Egypt, for which he conducted fieldwork in Egypt between 2011 and 2015. In 2012–2013, he worked for the German Orient Institute Beirut (OIB). As an analyst at the Landeskriminalamt (State Criminal Police Office) of Thuringia, he currently focuses on Islamist extremism, Salafism, Jihadism and deradicalisation.

# **HEIDELBERG STUDIES ON TRANSCULTURALITY**

**5**

As linguistic systems comprising a large variety of written and oral registers including derivate "languages" and "dialects," Latin and Arabic have been of paramount importance for the history of the Euromediterranean since Antiquity. Due to their long-term function as languages of administration, intellectual activity, and religion, they are often regarded as cultural markers of Europe and the (Arabic-)Islamic sphere respectively. This volume explores the many dimensions of Latin-Arabic entanglement from macro- and microhistorical perspectives. It questions the binary opposition of "Islam" and "the West" and highlights the linguistic dimension of Christian-Muslim relations.