The Marvels Found in the Great Cities and in the Seas and on the Islands

6 A Representative of ʿAǧaʾib Literature in Syriac This volume presents the original text, accompanied by an English transla� on and commentary, of a hitherto unpublished Syriac composi� on, en� tled The Marvels Found in the Great Citi es and in the Seas and on the Islands. Produced by an unknown East Syrian Chris� an author during the late medieval or early modern period, this work off ers a loosely organized catalogue of marvellous events, phenomena, and objects, natural as well as human-made, found throughout the world. The Marvels is a unique composi� on in that it bears witness to the crea� ve adop� on by Syriac Chris� ans of the paradoxographical literary genre that was very popular among Arab and Persian Muslim writers. The East Syrian author blends together inherited Chris� an, borrowed Muslim, and local oral tradi� ons, providing his audience with a fascina� ng panorama of imaginary geography.


Marvels
The Arab takeover of the Near East changed the social and cultural landscape of the region, giving birth to a range of new polities dominated by the religion of Islam. The contribution of the Christian 'minority', 1 including that of Syriac-speaking Christians, to the formation of Islamic culture, is a much-discussed and well-documented topic. A relatively less-studied subject, however, is the opposite direction in the never-ceasing process of 5 See Dubler (1986); Bosworth & Afshar (1985); Rodinson (1978); Ron-Gilboa (2017). For a detailed inventory of these works in the Arabic literary tradition, see Demidchik (2004, 119-219). By using the term 'ʿaǧāʾib literature' I do not imply that there is such a formally defined literary genre, but imagine rather a cluster of the works of diverse

Marvels
There is a rich and developed tradition of the paradoxographical mode of writing in Arabic and Persian literatures, the earliest preserved specimens of which, such as Buzurg Ibn Šahriyār's ʿAǧāʾib al-hind and Abū Dulaf's second risāla, go back to the ninth and tenth centuries. This tradition finds its culmination in the most famous composition of Muslim paradoxography, ʿAǧāʾib al-maḫlūqāt wa-ġarāʾib al-mawǧūdāt, the encyclopaedia of natural history produced in Arabic by Zakariyā al-Qazwīnī (1203-1283). 6 As for its purpose, this literature had not only erudite or entertaining functions, but a religious dimension as well, as it sought to instil in its audience the feeling of amazement and, thus, advance it in the knowledge of God 'through observing his creations, and contemplating the wonders of his works'. 7 As I am going to demonstrate, the Syriac Christian compiler of the Marvels appropriated this mode of writing creatively, both following its conventions and adjusting it to the needs of his audience.
genres, often closely related to geographical literature, that have in common a pronounced tendency to present natural and human-made phenomena and objects that are meant to evoke the feeling of astonishment. For an argument against the existence of a literary genre of ʿaǧāʾib, see von Hees (2005); cf. also Bellino & Mengozzi (2016, 433).

Textual tradition
The Syriac text of the Marvels is attested in the following four manuscripts, described here in chronological order:  Sauget (1981, 49-62). Of the manuscritps produced by Yaldā, at least twenty four have survived. On Yaldā and the Naṣrō family of Alqosh, to which he belonged, see also Wilmshurst (2000, 247-50); Murre-van den Berg (2015, 93-94).

Marvels
were commissioned from different scribes of Alqosh for the church of Mary in Karsabā (also spelled as Karsāpā) during the seventeenth century. 10 Borgia sir. 39 is an anthology of writings, mostly comprising hagiographical works, with the addition of some parabiblical and other texts. The text of the Marvels is located on fols. 111r-115v: it follows an edifying story about a boy killed by his teacher (fols. 109v-111r), and is followed by the Fables of . The title of the work as well as its concluding sentence are rubricated. Throughout the text, rubricated subtitles and introductory words, such as 'a marvel' or 'marvels', are used as well. L = London, British Library, Or. 4528; paper, 17 × 10 cm, 248 fols., 1 col., 1737. 11 The manuscript is written in a partially vocalized East Syrian script. According to the information provided in the extended colophon (fols. 244r-245r), it was produced by the priest Introduction place where the manuscript was produced, we can safely identify this hierarch as Šemʿōn XIV Šlemōn (1700-1740). This indicates that ʿAbdīšōʿ was a member of the East Syrian community.
As for its content, Or. 4528 is an anthology of diverse works that includes parabiblical texts, such as Pseudo-Basil's History of Joseph, as well as hagiographical, exegetical, chronological, and some other compositions. The text of the Marvels is located on fols. 73v-78v: it follows a brief question and answer addressed to an unnamed doctor (fol. 73v) and is followed by the Story of Arsanis (fols. 78v-81r). The title of the work is rubricated.
Throughout the text, rubricated subtitles and introductory words, such as 'there is', 'another' or the first words of a unit, are used as well. and 8: they follow a folio containing a section of an unidentified zoological work (fol. 6) and are followed by a hexaemeric composition, 17 also partially preserved (fols. 9a-13b). Throughout the text, rubricated subtitles and introductory words, such as 'a marvel' or 'marvels', are used.

B1
The two textual witnesses V and L should be grouped together since they are almost identical in regard to the sequence of units as well as their content, except for minor textual variants. Most significant among the differences are the absence of unit I.40 in L, and the relocation in L of unit I.23 after I.19, and of unit I.16 after I.64. It should be pointed out that although the manuscript containing L is somewhat younger than that of V, the former is not derived directly from the latter, as one can conclude from some cases where L offers better readings than V. 18 To this group of manuscripts also belongs B2, comprising two disparate folios.
As far as their content allows us to judge, this version contains the same units in precisely the same order as V, that is: fols.  In their discussion, Bellino and Mengozzi seem to hold the opinion that the Arabic version is a full or abridged copy of the aforementioned Christian Arabic Vorlage. In light of the much 27 I.e., the oasis of Tayma in north-west Arabia.

Marvels
longer Syriac version of the work, however, the abridged and derivative character of the Arabic version becomes evident. What remains unclear is who and for what purposes produced this truncated translation of the Marvels. One possibility is that it was made impromptu by Gabriel Quryaqoza, or one of his associates, for the sake of inclusion in the anthology that he was preparing for Sachau. Another possibility is that this version was already in existence before this project. Since there is no textual evidence so far that would confirm the latter, the former scenario appears to be more likely.

Language
In the absence of any unequivocal textual evidence that would

Content
In what concerns the general structure of the Marvels, it is rather simple, being made up of a sequence of units of varying length, in which various wondrous objects or events are described. These units are grouped roughly into three main parts, that is, the marvels of the cities, marvels of the seas, and marvels of the Another paratextual feature of the work is the marker 'marvel/marvels', also written in red ink, that introduces most of the textual units.
As one reads through the text of the Marvels, it becomes apparent that its compiler derived this diverse material from a wide range of sources, written as well as oral. It is, however, only on rare occasions that he indicates the origin of individual units. 32 It should be pointed out that the scribes of all four of the manuscripts of the Marvels resorted to the paratextual strategy of rubricated titles, subtitles, and introductory words ('a marvel', 'marvels', 'another', and others) in order to make it easier for readers to navigate the text.

Marvels
Thus, only in one case, I.48, a written source, 'the book of Mār Basil', that he used is explicitly mentioned (on this, see below).
In what concerns oral sources, the compiler discloses them more often: by providing names of his informants, which include Joseph, the brother of ʿAbdīšōʿ of Elam (I.12), a certain Anās (I.40/II.31), Qūryāqōs of Mosul (I.41, I.47), and Rabban Emmanuel (II.47), or by referring to them in general terms ('merchants' in I.42, 'they' in I.46 and I.49). Yet, there might be even more oral accounts, incorporated into our work, whose origin is not made explicit by the compiler. One such case is, probably, the cluster of four textual units dealing with the region of Bitlis in . Given the narrow territorial focus of these units, as well as the fact that none of them seems to find close parallels in the major Arabic and Persian paradoxographical or geographical compositions, it seems justified to regard them as the elements of local folklore that reached the compiler of the Marvels not via literary sources, but through the process of oral transmission. 33

Christian background
In what concerns the composition's milieu and the confessional identity of its author, there is no doubt that it was produced by a Christian. Setting aside the fact that it is preserved in Syriac, the language used during the Middle Ages exclusively by Christians, the work features several distinctively Christian themes and images.
To  Daǧǧāl, who is often represented in Arabic sources as chained to a mountain on a distant island in the sea.
It remains an open question as to how exactly the Syriac Christian compiler of the Marvels became acquainted with this rich assortment of Muslim geographical and anthropological lore.
Although the possibility of his dependence on some specific ʿaǧāʾib composition cannot be ruled out completely, it does not seem particularly likely, as I was unable to discover an Arabic or Persian work that would contain all these traditions. A more plausible scenario, thus, is that he culled these traditions while perusing a number of existing Muslim geographical and paradoxographical works, such as those by Ibn Ḫordāḏbeh or al-Qazwīnī, and reworked them according to his taste, as he rendered this material into Syriac. 46 At the same time, a possibility of the oral transmission of at least some of these traditions, which one can easily imagine becoming a part of the folklore of the Islamicate world, should also be taken into consideration.

Date and milieu
The anonymous and compilative character of the Marvels makes the task of determining the date of its composition particularly challenging. The terminus ante quem of ca. 1680, provided by the date of the oldest textual witness in which it is attested (i.e., Borgia sir. 39), could be pushed to a somewhat earlier date, given the fact that this manuscript does not present an autograph. The terminus post quem is, however, much more difficult to establish. In what concerns the milieu, from which the composition originates, it may be said with a fair degree of confidence that it was produced by a member of the East Syrian Christian 46 For examples of the use of Muslim Arabic works on natural science by such Christian authors of 'Syriac Renaissance' as Barhebraeus and Jacob bar Shakko, see Takahashi (2004, 293-97;2002;2006).

Marvels
community, most likely, in Northern Mesopotamia. As for a more exact location of the place where the compiler of the Marvels might have carried out his work, it cannot be established with absolute certainty. It should, however, have been one of the urban centres, where a literate member of the East Syrian community might be exposed both to Arabic (and/or Persian) Muslim culture as well as to oral traditions from such regions as Central Anatolia (cf. I.12), Kurdistan (cf. I.24-27), and the Persian Gulf (cf. I.40). In a manner of speculation, a city like Mosul, the commercial hub of north-eastern Iraq, seems to be a likely option.
In that direction points also the fact that one of the compiler's informants, the priest Quryāqos (I.41, I.47), is said to be a native of this city. However, other cities with significant East Syrian communities, such as Alqōš or Urmia, cannot be ruled out completely.