Version classiqueVersion mobile

A Handbook and Reader of Ottoman Arabic

Esther-Miriam Wagner

II. Reader

7. Selections from Arabic Garshūnī manuscripts in the British library

Michael Erdman

Texte intégral

1In the history of writing and literacy in the Middle East, Arabic written in Syriac characters, known as Arabic Garshūnī, presents us with an interesting, yet often forgotten, example of cultural adoption and adaptation. Arabic Garshūnī, similar to other allographic traditions, did not have a standardised orthography on which authors and scribes might base their writings. Nonetheless, the general need for language to function as a means of communication and wide dissemination of information implied that certain patterns were adhered to across the Christian Arab world. Previously, the corpus of Arabic Garshūnī manuscripts was limited to Levantine and southern Turkish sources, but an increase in our access to digitised manuscripts from Iraq and other regions has helped to broaden our understanding of this particular means of recording and reproducing cultural heritage (McCollum 2014, 16–19).

2Within the patterns referred to above, the use of Syriac graphemes to represent Arabic sounds can be broken down into three separate categories: those for which there is a one-to-one correspondence between Arabic and Syriac graphemes; those cases in which Syriac lacks a unique means of representing an Arabic phoneme or grapheme; and a third subset in which the Syriac script represents Arabic phonemes through the use of diacritics. It should be noted that the distinction between phoneme (a unique sound in a phonetic system) and grapheme (a unique letter in a writing system) is important here. The decision to match a grapheme to a grapheme, a phoneme to a phoneme, or a phoneme to a grapheme (and vice versa) tells us as much about the copyist’s grasp of Classical Syriac and Classic Arabic as it does about their particular dialect of spoken Arabic (McCollum 2014, 227).

3The first category of mappings presents the least difficulties. Here, a one-to-one correspondence is established and is easily recognizable. Thus, the Arabic letter bāʾ, for example, is represented by the Syriac letter bēt. Within this category, however, we also find that the core Arabic graphemes function as representatives of the Arabic graphemes based on them, regardless of pronunciation. In this respect, the Syriac yōd is used for both the Arabic yāʾ and the Arabic ʾalif maqṣūra, despite the latter’s pronunciation as an ʾalif.

4The second group of graphemes are slightly more problematic, but they do reveal the pre-modern scribe or writer’s understanding of phonetics. Take, for example, the velarised consonants, for which there are two graphemes in Syriac and four in Arabic. In general, those who wrote in Garshūnī sought to replicate sounds by both the ṣādē with a dot over it and the ṭēt with a dot under it. This raises the question of vernacular pronunciations of these letters among the Arabic-speaking Christian communities of the Middle East. In particular, it focuses our attention on the merging of the velarised phonemes in some dialects, such as Lebanese, which are still present in others, such as Najdī and Khalījī Arabic.

5Finally, the third collection of graphemes is the most unstable: those that can be represented fully in Syriac with the help of diacritics, the most common of which is the rukākā, a dot below the letter. Here, the Arabic ghayn is rendered with the Syriac gāmal and a dot below, the traditional Syriac means of rendering the voiced pharyngeal fricative. Occasionally, a conscientious scribe would also use a qūshāyā, or a dot above the letter, to indicate that it was to be read as the corresponding non-spirantized letter in Arabic. The problem with this group of graphemes is that the usage of the dot is far from routine. The reader is thus left asking herself if this phenomenon—which rarely impedes comprehension—is a reflection of vernacular phonology or simple laziness on the part of the scribe.

6A final remark must be made on additional markers used in Arabic texts. The hamza, although a separate letter according linguists, never features in Garshūnī texts. When it would sit on an ʾalif, waw or yāʾ in Arabic, the basic grapheme is used. Ḥarakāt may or may not be included in a text and almost always follow the Arabic system, rather than either of the two Syriac systems in use. Similarly, two dots over the letter assist us in determining whether a final is intended to be a tāʾ marbūṭa or a final h. Lastly, the shadda occasionally appears in its Arabic form. On other occasions, it shows up as a tilde over the doubled consonant or a neighbouring one. Gemination was rare enough in Classical Syriac that it did not merit its own special diacritic.

7The full listing of the most common orthography is found in the table below:

8In traditional Syriac texts, similar to those in Hebrew and Arabic, the letters are also assigned numerical values. These numbers are often denoted by a line over the individual graphemes. This tradition was carried over into many of the Garshūnī texts used in this section.

9The traditional Syriac system of numeration is as follows (Healey 2005, 93):

Syriac Grapheme Numeral

Syriac Grapheme Numeral

1 ܐ

03 ܠ

2 ܒ

04 ܡ

3 ܓ

05 ܢ

4 ܕ

06 ܣ

5 ܗ

07 ܥ

6 ܘ

08 ܦ

7 ܙ

09 ܨ

8 ܚ

001 ܩ

9 ܛ

002 ܪ

01 ܝ

003 ܫ

02 ܟ

004 ܬ

British Library Or MS 4435; 12r


Arabic Transcription


10(1) Section 18, regarding why Gabriel was sent and not another angel. (2) We say that it was the same as in the case of Daniel, (and) when Gabriel came to Zachariah. (3) He similarly presaged the pregnancy (of Elizabeth) with John and he also brought good tidings to the Virgin. (4) So, if it had been someone else who brought good tidings to Zachariah, it would also have been someone else who brought good tidings to the Virgin. And someone else would have been believable to the listeners in any case. (5) But first He sent [him] to Daniel, and second to Zachariah, and third to Mary. (6) This proves the words of the angel, because previously he brought good tidings to Daniel and Zachariah. (7) Some people said that Gabriel is the head of the lower stream of His angels. (8) And He sent this one because his name was already associated with Daniel in the Old [Testament] (9) in order to shut the mouths of Jews who were saying that perhaps his name and task were not to be taken seriously; (10) and secondly so that he [Gabriel] would leave the Old [Testament] and enter the New [Testament]. (11) And for that reason, He sent him to Zachariah and Mary.

British Library Or MS 7205, 1v

11A book of Christian theology in questions and answers


Arabic Transcription


12(1) In the name of the Father, the Son and the Holy Ghost, the Sole God, amen. (2) The Book of Teaching for a Christian, by means of Questions and Answers between the teacher and the student. (3) Question: Oh, my brother, for some time now, I have been doubtful about the Teachings for Christians. (4) If you are in possession of any knowledge about it, impart upon me something from this honourable, blessed knowledge. (5) Answer: Yes, on [my] head and [my] eye, I will teach you what God has given me (6) and has bestowed upon me from the garden of his abundance. (7) I shall teach [you] that Christian knowledge is teaching composed of the Word of the Holy Gospels (8) and from the epistles of the Apostles. (9) And as a medium of this, the Word of the Lord [Messiah] taught what was necessary for this, (10) and what was essential, for the issue of eternal salvation. (11) Question: Tell [me], who are they who should learn this useful teaching? (12) Answer: All Christians are required to learn it. (13) Adults among them and children; men and women; for all desires and reasons.


13The above extracts come from two Garshūnī Arabic manuscripts housed at the British Library. I have sought to mirror the texts as closely as possible, and have therefore left in as many idiosyncrasies as can be reflected in a word-processed document.

14Information about the provenance of these manuscripts is scant at best. For the most part, British Library records provide only the title of the work, its pagination, and the date of its acquisition. Garshūnī manuscripts were sourced from across the northern Middle East, including modern day Syria, Turkey, and Iraq. As such, they represent the copying traditions of these communities.

15The first extract, Or. 4435 is a collection of stories to be told at Christian festivals. I have chosen a short extract explaining angelic visitation. The manuscript itself was likely copied in the 19th century in the vicinity of Malatya, Turkey (Margoliouth 1899, 42). More information can be gleaned from the catalogue of Forshall and Rosen (1838) for the second extract, Or. 7205. This Catechism in the form of questions and answers, we learn from the catalogue, is likely to have been penned in the 15th or 16th century. An addition at the back of the manuscript tells us that it was purchased by Father Elyas from Father Suleiman of Mosul in 1799. From this we know that the work was likely still in use until the end of the 18th century (Forshall and Rosen 1838, 101).

16Most of the unique attributes of Garshūnī mapping can be seen in both manuscripts. Consider, to start with, the repurposing of the ṭēt, equivalent of the Arabic ṭāʾ, as a ḍād, which is seen only in the extract from Or. 7205:

17Or. 7205

18Line 6

fayṭihi [fayḍihi] ‘his abundance’

19Line 10

ṭarūrī [ḍarūrī] ‘necessary’

20Next, we find in the two texts the use of the gāmal to reflect both the Arabic jīm and ghayn:

21Or. 4435

22Line 2

Gibrāʾil [Jibrāʾil] ‘Gabriel’

23Line 4

fa-law kān gayroh [fa-law kāna ghayrihi] ‘if it were not him’

24Or. 7205

25Line 5

gawāb [jawāb] ‘answer’

26Line 13

al-ṣigār [al-ṣighār] ‘the small ones’

27Finally, the following examples demonstrate the lack of transference of complete Arabic orthography into Arabic Garshūnī, with an example of a lack of hamza:

28Or. 4435

29Line 11

li-agal [li-ajli] ‘because’

30Or. 7205

31Line 8

rasāil [rasāʾil] ‘letters’

32As a final remark, the texts under examination, along with many of the other Arabic Garshūnī texts in the British Library collections, do not demonstrate usage of Syriac lexical items in any notable proportion. Nonetheless, it is interesting to point out the carry-over of some of the biblical names in their Syriac orthography, such as

33Or. 4435

34Line 2

Dāniyāl, which we can compare to the Arabic(in Arabic Garshūnī orthography) and the Syriac.

Table des illustrations

Fichier image/jpeg, 18k
Fichier image/jpeg, 41k
Fichier image/jpeg, 22k
Fichier image/jpeg, 14k
Fichier image/jpeg, 44k
Fichier image/jpeg, 38k
Fichier image/jpeg, 977 octets
Fichier image/jpeg, 1013 octets
Fichier image/jpeg, 1,1k
Fichier image/jpeg, 1,5k
Fichier image/jpeg, 1,1k
Fichier image/jpeg, 1,2k
Fichier image/jpeg, 1,1k
Fichier image/jpeg, 972 octets
Fichier image/jpeg, 1,1k
Fichier image/jpeg, 983 octets
Fichier image/jpeg, 993 octets
Fichier image/jpeg, 1,0k


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search