The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9: Part 1 9781463234560

Citation preview

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9

Analecta Gorgiana

1092 Series Editor George Anton Kiraz

Analecta Gorgiana is a collection of long essays and short monographs which are consistently cited by modern scholars but previously difficult to find because of their original appearance in obscure publications. Carefully selected by a team of scholars based on their relevance to modern scholarship, these essays can now be fully utilized by scholars and proudly owned by libraries.

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9

Part 1

By

Hidemi Takahashi Jos J.S. Weitenberg

Y W 2012

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com G&C Kiraz is an imprint of Gorgias Press LLC Copyright © 2012 by Gorgias Press LLC Originally published in 2010 All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2012

‫ܚ‬

ISBN 978-1-4632-0140-1

Y W

ISSN 1935-6854

Reprinted from the 2010 Piscataway edition.

Printed in the United States of America

______________________________________________________________________

THE SHORTER SYRIAC-ARMENIAN GLOSSARY IN MS. YALE SYRIAC 9 PART 1

HIDEMI TAKAHASHI, THE UNIVERSITY OF TOKYO AND

JOS J. S. WEITENBERG, LEIDEN UNIVERSITY

1. INTRODUCTION

A

mong the more neglected fields of research in Syro-Armenian studies is that of the relevant lexicographical material, which includes a number of extensive Syriac-Armenian lexica written in Syriac characters. At least three manuscripts are known to have survived that contain such lexica: (1) Harvard University, Houghton Library, Syriac 54 (373 folios, dated 1657–60),2 (2) Cairo, Centre Franciscain d’Études Orientales Chrétiennes, Syriac 11 (333 pages, dated 1665/6, trilingual in Syriac-Arabic [Garshuni]Armenian),3 and (3) Yale University, Beinecke Rare Book and Manuscript Library, Syriac 9 (lexicographical material on p. 4–199, 231–241). A detailed investigation of this material may be expected to yield many results relevant to the study of both Syro-Armenian relationships and Armenian dialectology, but that is an undertaking which is likely to take years to complete. It seems advantageous, therefore, to begin with the smallest item in this group, namely, the shorter glossary

1

found detached from the main lexicon on p. 231–241 of Ms. Yale Syriac 9. It is this glossary which is presented and commented upon in what follows.

2. ON MS. YALE SYRIAC 94 The presence at Yale University of a manuscript containing a Syriac-Armenian lexicon was known to Goshen-Gottstein, who mentions it in a note following his description of Harvard Syriac 54. More recently, this manuscript was made more widely known in the checklist of Syriac manuscripts at Yale published by Depuydt.5 What Goshen-Gottstein and Depuydt either did not know or fail to mention is the fact that Ms. Yale Syriac 9 is one of the six Syriac manuscripts acquired by Yale’s Beinecke Rare Book and Manuscript Library in 1967–68 from Mrs. Melâhat Menemencioglu, a Turkish scholar of French literature resident at the time in Besançon.6 Like Paris syr. 395– 398, which were sold to the Bibliothèque nationale by the same Mrs. Menemen-

________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 68

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

cioglu, and the now missing manuscript of the Chronicle of the Year 1234, these six manuscripts formerly had their home in Istanbul, in the manuscript collection of Petros (Fehim Beg) Eronur, the son of a Syrian Orthodox bishop who became the Master of Ceremonies (teúrifatçÕ) at the Ottoman Court.7 In the catalogue of that collection by Yuতanon Dolabani, the manuscript that is now Yale Syriac 9 is described as containing a Syriac-TurkishArmenian lexicon.8 Most, if not all, of the manuscripts in the collection of Fehim Beg seem to have been acquired in Urhoy (Urfa, U৘ha) by his father, Bishop Timotheos Paulos (bishop of Urhoy 1883– 87, patriarchal vicar in Istanbul 1887– 1913). One item (now Yale Syriac 11), at least, once belonged to the Church of Saints Peter and Paul in Urhoy, so that other items in the collection may also come from there. Many of the manuscripts deposited in the Church of Saints Peter and Paul in Urhoy are, in turn, reported originally to have come from the Monastery of Mor Abতay near Gargar (Gerger, on the bend of the Euphrates between Urhoy and Melitene/Malatya),9 and a number of manuscripts dating from around 1600 in the collection of Fehim Beg are certainly connected with the region around Gargar and Mor Abতay,10 a region that once had a significant number of Armenian speakers, or at least bearers of Armenian names, belonging to the Syrian Orthodox Church,11 so that there is a good chance that Yale Syriac 9 also originates from that area. It will be remembered that the other two SyriacArmenian lexica have similar origins: Harvard Syr. 54 was copied in the Monastery of Mor Abতay, while the copyist of the Cairo lexicon was a native in the region of Gargar, although the copying of the latter took place in Amid (DiyarbakÕr). There appears to be no dated colophons or notes in Yale Syriac 9, but

the manuscript was judged to date from the seventeenth century by Dolabani, that is to say, a date similar to the other two lexica. Yale Syriac 9 is an octavo-size manuscript, measuring 15.5 cm by 10.5 cm, with wooden covers, and with 288 written pages. The Western (W) page numbering in pencil starts on the front and the Syriac (S) on the back of the first folio, but the Western numbering inadvertently skips two pages, while one folded sheet (4 pages), part of the section for letter ҵƝ in the Syriac-Armenian lexicon, was misplaced between the time of the Syriac and Western numbering by someone incapable of distinguishing Syriac ҵƝ (ĺ) and lƗmad (ĵ), so that the two series of numbering now correspond as follows: S1–51 = W2–52; S52–173 = W57–178; S174–177 = W53–56 (misplaced sheet); S178–210 = W179–211; S211–212 = (skipped in W); S213–286 = W212–285; (S unnumbered) = W286. The Syriac page numbering is followed in this paper. The main Syriac-Armenian lexicon occupies p. 4–199 of the manuscript. The text of the lexicon in the Yale manuscript was evidently copied from an exemplar whose beginning had been lost and whose first leaves were in disarray.12 Some of the stanzas in the poem preceding the lexicon (pp. 1–3) are taken from Barhebraeus. As indicated by Dolabani (though not by Depuydt), the miscellaneous items in the rest of the manuscript include a number of stories, as well as pieces attributed to Ephrem and Mark the Ascetic. The last part of the manuscript (pp. 262–284) seems to have been copied by a different hand from that of the lexicon. Many of the texts in the latter half of the manuscript are accompanied by marginal glosses giving the Armenian renderings (in Syriac characters) of words occurring in the Syriac text.13

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 69

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

3. SYRIAC-ARMENIAN GLOSSARY The Syriac-Armenian glossary has no words marking its beginning and end, and takes up just over nine pages of the manuscript, starting near the bottom of p. 231 (Fig. 1) and ending near the middle of p. 241. The bulk of the glossary consists of a list of verbs in imperative forms. It has not been possible so far to detect any rationale according to which these verbs are arranged. Near the end of the glossary, starting on p. 239, we are given various forms of the verb “to do.” This is then followed by a list of some basic words, such as personal and demonstrative pronouns, prepositions, conjunctions and common adverbs. The whole of this glossary is presented in Part 2 of this paper in transcription. Each entry in the glossary is indicated by the page and line number and its position within the line.14 In each case, the transcription of the Syriac is followed by a translation of the Syriac in brackets and, separated by a colon, by the transcription of the Armenian counterpart. This is then followed, separated by a dash, by the equivalent of the Armenian word in a more standard form, either in Modern Western Armenian (= mod.; in which case the verbs are normally given in their infinitive forms) or in Classical Armenian (= class.; in which case the 1sg. pres. indicative forms are used),15 followed by the meaning of the Armenian word given there in guillemets («»). Further comments are given as necessary, together with cross-references (the latter usually separated by a further dash).

4. ON THE TRANSCRIPTION In the Latin-alphabet transcription of the Syriac-language material that follows, the ‘classical’ pronunciation is followed for

ease of identification, although the way the Armenian vowels are represented in the glossary makes it clear that the author(s) and the copyist of the glossary pronounced the Syriac words according to the later West Syrian tradition. In the transcription of the Armenian material in the glossary, circumflex accents are used to indicate the presence of the letters Ɨlap, waw and yǀd employed as matres lectionis; i.e. “â” indicates an instance where a consonant with ptƗ‫ۊ‬Ɨ (º› ª or º) § is followed by Ɨlap; “î” where a ¥ is followed by consonant with ‫ۊ‬bƗ‫܈‬Ɨ (º) yǀd, etc. It is to be noted that the use of the circumflex accent here has no bearing on the length of the vowel. The relatively rare Ņ combination of zqƗpƗ with waw (ĭš), where the waw may on occasion represent a consonant (i.e. -ʏʕ-, so probably in 241.1.1 and perhaps also in 239.1.3), has nevertheless consistently been rendered by Ņ “ô” (e.g. 237.20.2 ƢƀƐƌŴ = ʛʏʍʒʂʗ; ō ñ ƃ/kônĞîr Ņ 239.22.3 ĭĭĭ/wôw = ʏʕ). Where a vowel sign appears attached to an Ɨlap in the middle of a continuous sequence of letters (which typically occurs when the group of letters would normally be written as two words in Armenian), the “hiatus” will ł be ł indicated by a hyphen (e.g. 232.5.3. ƢƀŌ ū õ ŤźƉ: ł łł ma৬-aঠîr; 232.21.3. ŧĿŤʼn ūĥ ñ űƤōſĥ: îšadâg-erâ = ʌʂʎɸʖɸʆ ɿʗɾɳ). It is not always possible to decide whether a word division is intended or not when the letter preceding the Ɨlap is one that does not connect with the following ł letter (i.e. whether ŧĿĥʼn ŁŴŨĥ Ő Ľł should be transcribed as “ৢâbût-erâ” or “ৢâbût erâ”). Lines modelled on the Syriac lineae occultantes are placed above or below the letters in the glossary to indicate the Armenian schwa sound (e.g. in 232.15.1, 232.16.1). These lines are indicated by “ԥ” in the transcription below. In some cases, the lines are drawn above or below several consecutive letters, making it unclear with

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 70

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

which letters the scribe intends us to read the schwa. In such cases, the words are first transcribed as written in the manuscript, followed, as appropriate, by the likely intended pronunciation (e.g. 235.6.1. “kԥsԥpԥsdâ [lege kԥspԥsdâ]”). Special diacritical points (almost always in red ink) are used in the glossary for certain Armenian consonants. The following signs are used in the transcription to indicate the letters carrying these diacritical points: g: gƗmal [Ĩ] with point above [usually = ʆ] ঠ: gƗmal with point below [invariably = ʉ (with the possible exception of 236.5.3)] ÷: gƗmal with point in centre of letter [= ʒ (232.5.1, 232.6.2, 232.6.3, 233.8.3, 233.14.3, 234.11.2, 235.7.1 [dial.], 237.12.2, 237.16.3, 238.7.2); = ʊ (233.9.1, 233.20.1, 235.1.2, 235.15.3, 236.4.1, 237.3.3, 237.4.1, 238.3.1, 238.18.2); = ʐ (232.11.2, 233.1.1, 235.21.1, 236.12.2, 239.11.2, 240.9.2, 240.11.1, 240.16.1); = ʁ (233.4.2, 235.23.2, 237.2.1); = ʉ [?] (235.15.3, 235.19.1)] Ī: zayn [Į] with point above [almost invariably = ʅ, except at 240.6.4 (knĪ/ ʛɸʍʘ); note also 234.1.1 (Īârar = Arab. ‫ڲ‬arar)] k: kƗp [Ĵ] with point above [usually, = ɺ or ʛ] ঴: kƗp with point below [usually, = ʄ, but corresponds to standard ʉ at 233.9.1, 237.4.1 (÷î঴kdâ, ÷঴kdâ/ ʊ(ɼ)ʉʛʏʖɾɳ), 236.21.2 (ko঴Ğîr/ ɺʏʉʘʂɳʗ), 238.15.3 (ma঴trâ/ ʋɼʉɸɻʗɾɳ); note also use for Arab. প (Υ) at 232.16.3, 234.13.1, 234.22.3, 238.10.2; and correspondence to standard Arab. q (ϕ) at 240.11.1 (în÷঴dâr), 240.17.2 (înš঴dar)]

৖: pƝ [Ļ] with point above [= ʚ (237.7.2, -š৖â/ʎʚɾɳ); = ɹ (234.16.2, ৖anĞrĞû/ ɹɸʗʈʗɸʘʏʙɳ)]. ৔: pƝ with point inside circle [= ʚ (232.7.3, ৔orĞâ/ʚʏʗʈɾɳ, but cf. 238.20.1, porĞâ); = ɹ/ʚ (236.4.1, ÷an৔â/ʊɸʋɹɾɳ/ ʊɸʋʚɾɳ, but cf. 233.20.1, ÷ânpakpîr)] f: pƝ with point (sign ‫ )ޔ‬below (?) [= ʝ (only once at 232.21.3, waqԥf/ʕɸɺɿʝ, in a later addition; see footnote ad loc.)]. Ğ: semkat [ĸ] with point above [= ʘ (231.24.1 et passim); = ʈ (232.7.3, 233.5.3, 234.16.2, 236.2.1, 236.14.3, 240.4.4)]. š: šƯn [ŀ] with three points above or below (a device no doubt inspired by Pers./Turk. ̧) [= ʁ (237.16.2); = ʒ (232.20.1); = ʐ (239.11.1, 240.15.2, etc. in negative verbs; also 240.8.3 (înšes/ ʂʍʐʑɾʔ), 240.22.1 (înšsor/ʂʍʐʏʗ); and in unidentified words at 234.20.2, 236.6.1, 237.7.3, 237.8.1, 238.2.1, 238.2.2]. The ‘letters’ ÷ and š are apparently freely interchangeable, to judge from the use of both for the negative particle ʐ, and from those instances where a word written with š in one place appears elsewhere (often in the main lexicon in Yale Syriac 9) with ÷ (e.g. 240.8.3. înšes, but 6.8.1. în÷es, for ʂʍʐʑɾʔ; 234.19.1. špla঴Ğû, but 103.15.1. ÷৖lâ঴, 151.12.2. gû÷plâ঴Ğnâ). It seems unlikely that any phonetic or phonemic difference can be posited between the sounds represented by ৖, ৔ and p (i.e. pƝ without diacritical point).16 The letters ‫ۊ‬Ɲt, ‫ܒ‬Ɲt, ҵƝ, ‫܈‬ƗdƝ and qƗp are used on rare occasions in Armenian words, often to represent equivalent Arabic consonants in loanwords, but sometimes without any immediately apparent reason.17

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 71

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

ত [į]: in loanwords: 232.7.1 (তsebâ < Arab. ‫ۊ‬isƗb); 233.6.2 (šarতâ < Arab. šar‫ ;)ۊ‬233.20.3 (তasûdĞîr < Arab. ‫ۊ‬asnjd); 235.4.2 (paতmâ, prob. < Syriac pa‫ۊۊ‬em); but see 238.13.1 (তepdar); 232.2.1 (তas৬îr); 232.14.1 (তazârâ). – Cf. use of “h” for Arab. “ত” in: 233.16.1 (hadrâ, cf. Arab. ‫ۊ‬a‫ڲڲ‬ara, ‫ۊ‬Ɨ‫ڲ‬ir); 235.3.2 (habsâ, prob. < Arab. ‫ۊ‬abasa); 238.13.2 (hûkûm-erâ < Arab. ‫ۊ‬ukm). ৬ [İ]: in loanwords: 233.12.1 (mû৬î‫= ޏ‬ Arab. mu‫ܒ‬Ưҵ); 235.18.2 (zab৬ = Arab. ‫ڲ‬ab‫ ;)ܒ‬235.19.2 (mûsl৬ = Arab. musalli‫ܒ‬/musalla‫ ;)ܒ‬237.18.1, 238.1.2 (bâ৬lâ, prob. < Syriac ba‫ܒܒ‬el or Arab. ba‫ܒܒ‬ala); 238.20.3 (das৬ûr, with unwarranted ‫ܒ‬Ɲt for Pers./Arab. dastnjr); 232.5.3 (ma৬- = Pers. mƗt). – Used for Armenian ɻ/ʖ: 232.9.4, 233.18.1, 236.2.1 (৬arĞ- = ɻɸʗʈ-); 239.8.3 (৬nâzâ = ʖʍɸɽɾɳ, but 231.25.2, dnâzâ)18; 236.5.1 (৬ûrâ = ɻʏʙʓɿ, but 237.14.1, tûrâ); see also 232.2.1 (তas৬îr). – Cf. use of “t” for Arab. “৬” in: 235.8.1 (bîrtil = Arab. bir‫ܒ‬Ưl). ‫[ ޏ‬ĺ]: 236.23.3 (tabe‫ޏ‬: Arab. tƗbiҵ); 240.23.1, 241.8.2 (yâ‫ޏ‬neh/ya‫ޏ‬neh: Arab. yaҵnƯ); 240.12.1 (‫ޏ‬â÷ab: Arab. ҵa÷ab, but 240.14.2. a÷âb, 235.6.2. a÷bîĞîr); 238.8.1 (‫ޏ‬îzâtworâ < Arab. ҵizza [?]); 240.19.2 (‫ޏ‬âparîn, with unwarranted ҵƝ for Pers. ƗfirƯn, Turk. âferin) ৢ [Ľ]: in loanwords: 234.14.2 (ৢûrât < Arab. ‫܈‬njra); 240.14.3 (aৢlâ < Arab. a‫܈‬lan); 232.14.2 (ৢâbût, prob. erroneously for Arab. ‫ܔ‬Ɨbit [Turk. sabit]); in genuine Armenian words: 233.22.1 (ৢîrdâharĞîr), 236.4.2 (-ৢîrdîr = ʔʗʖɼʗ); 236.9.2 (haৢîr = ʇɸʔʂɳʗ, cf. 237.10.2 hasîr); cf. use of “s” for Arab. “ৢ” in: 234.13.2 (sabrîr < Arab. ‫܈‬abr)19; 235.9.3 (঴âlîsâ, cf. Arab. ‫ې‬alla‫܈‬a).

q [ľ]: 236.4.3 (qonahrĞîr = ʄʏʍɸʗʇʘʂɳʗ); 237.21.1 (qûqan); 237.23.1 (qrbâ) The letter ‫ܒ‬Ɲt with point in the circle [ı] (here, = ৰ), which is often used for Arabic ‫ڲ‬Ɨd (ν) and ܲƗҴ (υ) in Garshuni, does not occur in our glossary, but does occur in the main lexicon in Yale Syriac 9: e.g. 8.8.3: emayyeq [I mock]: ৰnâzîm = ʖʍɸɽɼʋ (cf. 231.25.2); 150.13.1: nešgƝ men šgƗyƗ [he errs, as in “error”]: gû঴alৰnâ.

5. OBSERVATIONS ON THE ARMENIAN DIALECT OF THE GLOSSARY 5.1. General Linguistic Classification In what follows a preliminary overview is given of the place of the Armenian material found in the glossary in the general Armenian linguistic context. The overview is largely limited to phonetic aspects and does not aim at exhaustivity; a more detailed study is required for a full description of the dialect in question. A complete picture of the dialect can only be gained with the examination also of the related material in the other SyriacArmenian lexica. It will be seen that many of the characteristics mentioned below were already noted by Margoliouth in his treatment of the lexicon which is now at Harvard.20 In a general sense, the language of the glossary belongs to the Western branch of Modern Armenian. It shared in the youngest common innovations of this group, such as the elision of medial vowels in polysyllabic words (e.g. 238.18.2. ÷ardrâ: mod. ʊɸʗʖɸʗɾɳ «construct!»),21 and the use of ʆʏʙ, ʆɿ in the formation of the present indicative (e.g. 239.14.2, tûn genîs: ɻʏʙʍ ʆɵɿʍɼʔ «you do»). These two features go back to the Early Middle Armenian period (ca. 10th–12th centuries).

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 72

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

From a more precise, dialectological, point of view the language of the glossary is most likely a member of Dialect Group 5.22 This classification is based on the consonant system of the glossary (see 5.2 below), where aspiration is apparently no longer phonemic, producing a binary opposition D – T‫ ޏ‬against classical ʖ – ɻ – ʀ, as well as on its vowel system (see 5.3 below) and a number of morphological features. Classical T Cilician D Tigranakert D

D T‫ޏ‬ T T‫( ޏ‬Group 4) T‫ ޏ‬T‫( ޏ‬Group 5)

The vowel system of the glossary, in particular the representation of classical ɸ as either “a” or “ä” (as in Tigranakert), speaks in favour of its identification with Group 5, while its identification as a member of Group 4 (which also includes literary Modern Western Armenian) seems to be ruled out by the fact that the glossary does not show the typical Cilician reflex “o” for classical ɸ.23 The main dialects that belong to Group 5 are the dialects of Malatya, Tigranakert (DiyarbakÕr), Urfa, and, in the west, Rodosto (Tekirdaۜ) and Smyrna (Izmir). Therefore the data of the glossary should be compared with these dialects rather than with the Modern Western literary language. In what follows, the western Group 5 dialects are left out of account; the dialect of Tigranakert [= Tigr.] will be adduced in most cases for comparative purposes. 5.2. Consonant System 5.2.1. Obstruents 5.2.1.1. Overview Class. voiceless ʆ Gloss. g Class. voiced ɺ Gloss. k Class. Aspirated ʛ Gloss. k

ʖ d/(৬) ɻ t/(৬) ʀ t

ʑ b ɹ p/(৔/৖) ʚ p/(৔/৖)

ʅ ʊ: Ī/(z) ÷ ʈ ʒ: Ğ/(s/z) ÷/š ʘ ʐ: Ğ/(s) ÷/š/(š)

The use of the Syriac letter “৬” for ʖ and ɻ has been noted in Section 4 above. The varying outcomes of the classical apicodental (ʅ, ʈ, ʘ) and palatodental (ʊ, ʒ, ʐ) fricatives require further study. We may in many cases be dealing with the results of scribal error (omission of diacritical points), as, for example, in those cases where we find class. ʐ in the initial position in negative verbs represented by “š” (239.13.1 etc.) as opposed to “š” (i.e. š with three points; so 239.11.1 etc.; alongside “÷” in 239.11.2, 240.9.2).24 On the other hand, the orthography with “š” in the equivalent of class. ɸʐʛ «eyes» is no error; it represents dialectal ɸʎʛ (233.3.3, 235.14.2).25 The use of “s” for ʘ instead of “Ğ” (s with point) seems in many cases to be erroneous (232.20.3, 233.12.2, 235.23.3). It is known, on the other hand, that the dialectal (Tigranakert) outcome of ʘ may sometimes be “s”. This may be the case with “Ğԥkâ” in 233.5.3 versus “skâ” in 236.20.3 (class. ʈɺɼʋ «to bend, cast»; cf. Tigr. [‫ۑ‬sgil]). Another instance of variation in orthography is found in the class. group ʗʈ [rj]: porĞâ (238.20.1; class. ʚʏʗʈɼʋ «to try, test»); arsagâ (236.12.3); arzgâ (237.23.2; class. ɸʗʈɸʆɼʋ «to release»). 5.2.1.2. Examples of Regular Correspondences (a) Classical voiceless: p (ʑ): badâ: ʑɸʖɾɳ «wall!» (238.21.2); t (ʖ): dâr: ʖɸɳʗ «bring!» (232.9.1); k (ʆ): gartâ: ʆɸʗɻɸɳ «read!» (231.23.2); c (ʅ): ĪeĪâ: ʅɼʅɾɳ «beat!» (233.19.2); þ (ʊ): ÷ardrâ: ʊɸʗʖɸʗɾɳ «construct!» (238.18.2). (b) Classical voiced: b (ɹ): pâĞ: ɹɸɳʘ «open!» (233.4.1); d (ɻ): tûn: ɻʏʙʍ «you» (239.14.2); ৬ûrâ: mod. ɻʏʙʓɿ «the door» (236.5.1); g (ɺ): kowâ: ɺʏʕɾɳ «praise!» (232.21.1); j [dz] (ʈ): Ğîzî: mod. ʈɼɽʂ «to you [pl.]» (240.4.4); Ƶ [dzj] (ʒ): arâšâ: ɸʓɸʒɾɳ «precede!»

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 73

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

(232.20.1); ÷râĞû: «irrigate!» (232.5.1)

mod.

ʒʗɸʘʏʙɳ

(c) Classical aspirated: p‫( ޏ‬ʚ): porâ: ʚʏʗɾɳ «dig!» (234.3.2); t‫( ޏ‬ʀ): tûঠ: ʀʏɳʉ «let!» (234.3.1); k‫( ޏ‬ʛ): kerâ: ʛɼʗɾɳ «scratch!, rub!» (233.3.2); c‫[ ޏ‬tsh] (ʘ): imperatives in -Ğu: -ʘʏʙ (passim) [but note ঴zgâ/঴îzgâ: ʄʘʆɾɳ «shut!» (233.17.2, 238.11.3)]; þ‫ޏ‬ [tsjh] (ʐ): ÷ar÷ârâ: ʐɸʗʐɸʗɾɳ «torture!» (232.11.2). 5.2.1.3. Clusters (a) Classical voiceless obstruents in clusters usually show the same development as in other positions: rk (ʗʆ): ergîr: ɼʗʆʂʗ «earth» (233.9.2); ৘k (ʓʆ): bargîr: mod. ʑɸʓʆʂɳʗ «sleep!» (238.6.1); rc (ʗʅ): gorĪâ: ʆʏʗʅɾɳ «destroy!» (232.10.1); rt (ʗʖ): ÷ardrâ: ʊɸʗʖɸʗɾɳ «construct!» (238.18.2); át (ʉʖ): gâঠdûdâ: ʆɼʉʖʏʖɾɳ «defile!» (233.8.2); áþ (ʉʊ): ঴ঠ÷îr: ʄʉʊʂɳʗ «pity!» (237.3.3); nk (ʍʆ): dngâ: ʖʍʆɾɳ «plant!» (235.3.3); but note: st (ʔʖ): nîst: ʍʂɳʔʖ «sit!» (233.10.2); sp (ʔʑ): spânâ: class. ʔʑɸʍɸʍɾɳ (237.16.1). (b) Classical voiced obstruent after “r” shows the same development as in other positions: rb (ʗɹ): sԥrpâ: ʔʗɹɾɳ «wipe!», cf. Tigr. [sԥrp‫ޏ‬il] (233.2.3); rd (ʗɻ): gartâ: ʆɸʗɻɸɳ «read!», cf. Tigr. [gärt‫ޏ‬äl] (231.23.2); rƵ (ʗʒ): tr÷â: ʀʗʒɾɳ «wet!, soak!», cf. Tigr. [t‫ۑޏ‬rþ‫ޏ‬il] (232.6.2); rj (ʗʈ): ৔orĞâ: ʚʏʗʈɾɳ «try!», cf. Tigr. [p‫ޏ‬orc‫ޏ‬-] (232.7.3); arsagâ/arzgâ: ɸʗʈɸʆɾɳ «release!», cf. Tigr. [arc‫ޏ‬ԥgil] (236.12.3, 237.23.2). (c) In the group “nasal + obstruent” the opposition between voiced and voiceless has become lost since the classical period. The glossary seems to show a voiced realisation: gendĞîr (235.6.3),

gendĞû «make alive!» (239.9.1): cf. class. ʆɼʍɻɸʍʂ «alive», but Tigr. [kİnt‫ޏ‬äni]; ঴ԥndâ: ʄʍɻɸɳ «laugh!», cf. Tigr. [xndäl] (233.5.2, 234.15.3); ঴on÷eĞîr: ʄʏʍʒɼʘʂɳʗ «be weary!», cf. Tigr. [xonƵän] (233.14.3); dân÷wîr: ʖɸʍʒʏʙʂɳʗ «be tortured!, suffer!», but Tigr. [danþ‫ޏ‬il] (238.7.2). 5.2.1.4. Exceptions The following are exceptions to the general representations given above. (a) Classical voiceless: [k = class. ʆ]: šîkâ: ʎʂʆɸ «there is not» (240.15.2; < ʐʂ + ʆɸʌ; cf. the regular gâ: ʆɸʌ «there is» 240.12.2; cf. Tigr. [igä]); zârk: ɽɸɳʗʆ «strike!», mod. ɽɸʗʆɼʃ, Tigr. [zärgil] (232.3.1). (b) Classical voiced: [g = class. ɺ]: gîdĞîr: ɺʂʖʘʂɳʗ «know!», mod. ɺʂʖʍɸʃ, cf. Tigr. [k‫ޏ‬inäm] (238.7.1); hogâ: ʇʏɺɸɳ «take care!», cf. Tigr. [hok‫ޏ‬äl] (239.4.1). (c) Classical aspirates: In clusters with aspirate ʘ the preceding stops may shift their articulation base: dagĞû: ʖɸʛʘʏʙɳ «heat!» (238.17.2); logĞîr: ʃʏɺʘʂɳʗ, mod. ʃʏɺɸʍɸʃ «to bathe, wash» (233.3.1); as expected, on the other hand, are: awdĞîr: ɸʙɼʖʘʂʗ «announce!» (236.24.2); hešdĞû: ʇɼʎʖʘʏʙɳ «make easy!» (237.19.3); hakĞû: ʇɸɺʘʏʙɳ(ʗ) «clothe!» (235.2.3, 238.3.1). 5.2.2. Classical Spirants ʔ ɽ ʎ ʁ ʕ/ʙ ʝ ʄ ʉ ʇ ʌ s/(ৢ) z š ÷/š w p/f ঴/(q) ঠ h/(ত?) ø/y/h Examples: s (ʔ): sîrâ: ʔʂʗɾɳ «love!» (235.5.3); haৢîr: mod. ʇɸʔʂɳʗ «reach!» (236.9.2); z (ɽ): zârtarwîr: mod. ɽɸʗɻɸʗʏʙʂɳʗ «be adorned!» (232.1.1); š (ʎ): îšadâg-: ʌʂʎɸʖɸʆ «remembrance» (232.21.3); ž (ʁ): ÷oঠwâ: ʁʏʉʏʕɾɳ

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 74

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

«collect!» (235.23.2); šam-: ʁɸʋ «prayer(time)» (237.16.2); v/w (ʕ/ʙ): wâzâ: ʕɸɽɾɳ «run!» (236.5.2); zârtarwîr: mod. ɽɸʗɻɸʗʏʙʂɳʗ «be adorned!» (232.1.1); ÷oঠwâ: ʁʏʉʏʕɾɳ «collect!» (235.23.2)26; f (ʝ): sapar: Arab. safar «journey» (234.20.1); waqԥf: Arab. waqf «donation» (232.21.3, in a later addition to the glossary); x (ʄ): go঴â: ʆʏʄɾɳ «trample» (231.25.3); qonahrĞîr: mod. ʄʏʍɸʗʇʘʂɳʗ «humble!» (236.4.3); á (ʉ): kâঠâ: ʛɸʉɾɳ «pick!» (233.12.3); h (ʇ): hrâ: ʇʗɾɳ «push!» (232.1.2); তas৬îr (232.2.1): cf. hasdîr «be seated!» (236.7.3); y (ʌ)27: îšâ < class. ʌʂʎɼʋ «to remember» (234.14.3); ûlâĞîr < class. ʌʏʙʃɸʍɸʋ «to become idle» (238.22.3); yûr: ʌʏʙʗ «to where?» (240.8.1); hanzanĞû < ʌɸʍʈɸʍʈɼʋ «to ł bring to one’s side» (236.14.3); ha-îr (ƢōſĥĬ) < ʇɸʌʂʋ «to look at» (234.2.2). 5.2.3. Classical Nasals and Liquids ʋ m

ʍ n

ʃ l

ʓ r

ʗ r

Examples: m (ʋ): maঠâ: ʋɸʉɾɳ «sift!» (237.22.1); n (ʍ): nedâ: ʍɼʖɾɳ «throw!» (238.6.2); l (ʃ): latîr: ʃɸʀɼʗ «clothing [pl.]» (236.1.1); ৘ (ʓ): râzî-: ʓɸɽʂ (Arab. rƗ‫ڲ‬in) «satisfied» (235.12.1); borâ: ʑʏʓɸɳ «cry out!» (233.6.3); r (ʗ): -ৢîrdîr ʔ(ʂ)ʗʖɼʗ «hearts [pl.]» (236.4.2). The glossary does not express the opposition ʓ : ʗ, although the dialects of Group 5 (and Group 4) preserve this opposition. 5.2.4. Other Consonantal Features The nature used in the determine preserved responsible

of the Syriac writing system glossary28 makes it difficult to whether double stops were in the speech of those for the glossary: oঠâ < class.

ʏʙʉʉɼʋ «to straighten» (235.17.3); spânâ: Middle Armenian ʔʑɸʍʍɼʃ «to kill» (237.16.1). Seemingly unmotivated duplication of stops is a frequent feature of Group 5 dialects (e.g. Tigr. [vazzil] «to run», ʕɸɽɼʃ; ԥssor «today», ɸʌʔʜʗ; [lap‫ޏ‬p‫ޏ‬il] «to lick», ʃɸʚɼʃ).29 This feature likewise is not indicated in the glossary. Mention may be made here of further irregularities, often shared with other dialects, in the treatment of consonants: lman- for ʍʋɸʍ-: lmanĞû: ʍʋɸʍʘʏʙɳ (235.4.2; also 233.23.3, 239.6.1, 241.9.2); men(Ğ)- [< menĪ(Ğ)-] for ʋɼʅ(ʘ)-: menĞû < ʋɼʅʘʍɼʃ (238.16.1)30; panĞr- for ɹɸʗʈʗ-: ৖anĞrĞû: ɹɸʗʈʗɸʘʏʙɳ (234.16.2)31; -ndrfor -ʍʗ-: mandrĞû < ʋɸʍʗɼʃ (237.1.3), ĪândrĞû < ʅɸʍʗɸʘʍɼʃ (238.1.1); -hr- for -ʗʇ- [metathesis]: qonahrĞîr: ʄʏʍɸʗʇʘʂɳʗ (236.4.3); wl- for lw-: wlâ: ʃʏʙɸɳ (237.12.1; also 237.18.2); -rš- for -ʗʇ-: oršnâ < ʏʗʇʍɼʃ (237.9.1); -št- for -ʖ- [?]: gîštaĞîr < ɺʂʖʍɸʃ [?] (236.1.3); addition of initial h-: hԥndrâ: ɿʍʖʗɾɳ (236.13.3); omission of -r- [-š- for -ʗʁ-]: šaštûr: ʎɸʗʁ ʖʏʙɳʗ (235.8.3); [-d- for -ʗʖ-]: adorĞîr: ɸʗʖʏʗʘʂɳʗ (236.1.2); absence of preconsonantal -n-: anûk: mod. ɸʍʏʍʛ (240.1.3, 234.9.2); mîk-genîk: mod. ʋɼʍʛ ʆɵɿʍɼʍʛ (239.18.1, but cf. 240.1.1, mînk); see also 233.1.1; 237.9.2. 5.3. Vowel System 5.3.1. Classical vowel “a” (ɸ) The representation of Classical Armenian “a” (ɸ) in the glossary is by either “a” or “â”. A tentative distribution is given below. Because the sign “â” also renders classical Ɲ (ɾ) (= open [İ]) we may assume that Classical Armenian ɸ shows a twofold representation in the glossary: either “a” or a palatal variant “ä”. The precise distribution

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 75

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

and phonological opposition require further study. This type of palatalisation is common in the Tigranakert dialect.32 It is not related to the palatalisations that result from Aþa৘yan’s Law. (a) Classical initial ɸ is always rendered with “a”, whether stressed or not (Tigr. usually has “ä” in these instances): azâdâ: ɸɽɸʖɾɳ «free!», Tigr. [azäd], but [äzädil] (232.19.1); arâšâ: ɸʓɸʒɾɳ «precede!» (232.20.1); adâ: ɸʖɾɳ «hate!» (234.6.1); ɸʉɸɳ «grind!», Tigr. [äáäl] (238.20.2); absprâ: ɸʑʔʑʗɾɳ «request!», Tigr. [äb‫ۑ‬sb‫ۑ‬ril] (233.16.3); ašk: class. ɸʐʛ, ɸʎʛ «eyes», Tigr. [äšk‫( ]ޏ‬236.18.1); ar: ɸɳʓ «take!», Tigr. [a৘] (233.23.1). (b) Classical interconsonantal ɸ is rendered with either “a” or “â”, in stressed and in unstressed position: kâঠâ: ʛɸʉɾɳ «pick!», Tigr. [k‫ޏ‬äáil] (233.12.3); kalâ: ʛɸʃɾɳ «walk!», mod. ʛɸʃɼʃ, class. ʛɸʌʃɼʋ, but Tigr. [k‫ޏ‬İáäl] (235.5.2); bâঠeĞû: ʑɸʉɼʘʏʙɳ «cool!», Tigr. [bäá] (233.13.1); karozâ: ʛɸʗʏɽɾɳ «preach!», but Tigr. [k‫ޏ‬äroz] (233.15.3); tâdâ: ɻɸʖɾɳ «judge!», Tigr. [t‫ޏ‬ädil] (238.12.1); tâঠâ: ʀɸʉɾɳ «inter!», Tigr. [t‫ޏ‬äáil] (234.12.3); dâঠâ: ʖɸʉɾɳ «cauterise!», Tigr. [däáil] (239.4.2); lâĞ : ʃɸɳʘ «weep!», Tigr. [iläl] (233.2.2); pâĞ: ɹɸɳʘ «open!», Tigt. [p‫ޏ‬anal], [p‫ޏ‬ac‫]ޏ‬ (233.4.1); bâk: ʑɸɳɺ «kiss!», Tigr. [bäk‫ޏ‬nil] (237.13.2). (c) In final (= stressed) position classical ɸ is rendered with “â” (in all these instances Tigr. has “ä”): knâ: ɺʍɸɳ «go!» (235.11.1); šîkâ: ʐ + ʆɸʌ «there is not» (240.15.2); orsâ: ʏʗʔɸɳ «hunt!» (233.15.2); aঠâ: ɸʉɸɳ «grind!» (238.20.2). 5.3.2. Classical vowels “e” (ɼ) and “Ɲ” (ɾ) In all “Western” dialects classical stressed “e” (ɼ) has developed into [jİ], while

unstressed “e” (ɼ) has merged with “Ɲ” (ɾ) into an open “İ”. In the glossary, classical stressed “e” (ɼ) is rendered with “î” (from an historical point of view this attests the development ɼ e Æ jİ Æ i), while unstressed “e” (ɼ) has merged with classical “Ɲ” (ɾ) into [İ], and is rendered with “e” or “â”. (a) Stressed ɼ: lîĞ: ʃɼɳʘ «fill!» (class. ʃʍʏʙʋ) (232.5.1); pîr: ɹɼɳʗ «bear!» (class. ɹɼʗɼʋ) (232.12.2); îl: ɼɳʃ «go up!» (ɼʃɸʍɼʋ) (232.17.1); dîs: ʖɼɳʔ «see!» (ʖɼʔɸʍɼʋ) (234.2.3); îd: ɼʖ «backwards» (236.18.2); gîr: ʆɼɳʗ «eat!» (237.11.2); perîm: class. ɹɼʗɼʋ «I bring» (232.12.3); îs genîm: mod. ɼʔ ʆɵɿʍɼʋ «I do», Tigr. [jİs gİnim] (239.13.2).33 (b) Unstressed ɼ: This vowel does not show diphthongisation, in particular not in initial position: epâ: ɼʚɾɳ «cook!», Tigr. [İp‫ޏ‬il] (233.7.1); ergîr: ɼʗʆʂʗ «earth» (233.9.2); ertmĞîr: mod. ɼʗɻʋʘʂɳʗ «swear!» (236.19.3). In other positions: heĪîr: ʇɼʅʂɳʗ «ride!» (232.5.2); kerâ: ʛɼʗɾɳ «scratch!» (233.3.2); perîm: ɹɼʗɼʋ «I bring» (232.12.3); but note: awîrâ: ɸʙɼʗɾɳ (which represents dialectal ɸʙʂʗɾɳ) «destroy!» (232.8.2). (c) In the glossary, ɼ appears as “â” before class. “á” (ʉ): gâঠdûdâ: ʆɼʉʖʏʖɼʋ «to defile», but Tigr. [giád] «dirt» (233.8.2); nâঠeĞû: mod. ʍɼʉɸʘʍɼʃ «to oppress» (237.3.1); aঠâw: mod. ɼʉɸʙ «it became» (240.11.2); aঠîr-â «it was»: mod. ɼʉɼʗ ɾ (participle + ɾ) (240.11.3). Note also the same change before “৘”: hârû hârû: ʇɼʓʏʙ ʇɼʓʏʙ «far away» (240.15.1); târ: ɻɼʓ «yet» (241.8.1). (d) Classical vowel “Ɲ” (ɾ): In classical texts ɾ only appears in stressed (i.e. final) position. The glossary has “â” in absolute

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 76

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

final position in imperatives and in the 3sg. “he is”; other instances of classical absolute final ɾ are rendered with “-eh” and “-ê”. In pre-consonantal position classical ɾ is rendered with “e”. As has been noted above, the phonetic results of palatalised ɸ (rendered in dialectological studies as ɸࡇ “ä”) and final ɾ (open “İ”) are both rendered by the glossary as “â”; it is not immediately clear how this situation relates to the other renderings of ɾ. Absolute final ɾ as “-â”: maঠâ: ʋɸʉɾɳ «sift!» (237.22.1) and passim; idem, as “-eh”/“-ê”: îrneh: mod. ʂʗʋɾ «from him», Tigr. [irnİn] (240.9.3); kîzneh: mod. ʛɼʘʋɾ, ʛɼʍɾ «from you», Tigr. [k‫ۑޏ‬znİn] (240.10.2)34; tê: class. ʀɾ «or» (240.18.4); preconsonantal classical ɾ: asbes: class. ɸʌʔʑɾʔ «thus» (241.6.1); tûk genek: mod. ɻʏʙʛ ʆɵɿʍɾʛ «you [pl.] do», Tigr. [t‫ޏ‬uk‫ ޏ‬gİnİk‫( ]ޏ‬239.16.1). 5.3.3. Classical vowel “u” (ʏʙ) Classical “u” (ʏʙ) is rendered by “û” both in stressed and in unstressed position: mûd: ʋʏʙɳʖ «enter» (237.11.3); tûn: ɻʏʙʍ «you [sg.]» (239.14.2); în÷û: ʂʍʐʏɶʙ «why?» (240.16.1); ûšâĞîr: ʏʙʎɸʘʂɳʗ (ʏʙʎɸʍɸʃ «to be late») (231.24.1); ûzâ: ʏʙɽɾɳ (mod. ʏʙɽɼʃ «to will, want», class. ʌʏʙɽɼʋ) (232.4.3). – Before “á” (ʉ), unstressed ʏʙ appears as “o”: oঠrtâ: ʏʙʉʉʏʗɻɾɳ «straighten!» (233.1.3); oঠrgâ «send!» (mod. ʏʙʉɸʗʆɼʃ, class. ʌʏʙʉɸʗʆɼʋ) (237.15.1); oঠâ: ʏʙʉʉɼʃ «to direct, correct» (235.17.3).

and ʏʕ, while the unstressed “o” remains unchanged.35 (a) Stressed ʏ: tûঠ: ʀʏɳʉ «come on!» (233.7.2); pû঴-: ʚʏʄ «loan» (235.11.2); asûr: ɸʔʏʗ «to this» (240.4.2); asûnĞ: ɸʔʏʍʘ «to these» (240.5.3); wûr mîn: ʏʗ ʋʂʍ «which one?» (240.7.1); wûw-â: ʏɶʕ ɾ «who is?» (dialectal vov ʕʜɶʕ). (b) Unstressed ʏ: kowâ: ɺʏʕɾɳ «praise!» (232.21.1); borâ: ʑʏʓɸɳ «cry out!» (233.6.3); norâ: ʍʏʗɾɳ «renew!» (233.13.2); orsâ: ʏʗʔɸɳ «hunt!» (233.15.2); karozâ: ʛɸʗʏɽɾɳ «preach!» (233.15.3).36 5.3.5. Classical vowel “i” (ʂ) Classical “i” (ʂ) is rendered with “î” in stressed and in unstressed position. Stressed ʂ: nîst: ʍʂɳʔʖ «sit!» (233.10.2); tîr: ɻʂɳʗ «put!» (235.22.3). Unstressed ʂ: sîrâ: ʔʂʗɾɳ «love!» (232.6.1); în÷û: ʂʍʐʏɶʙ «why?» (240.16.1). 5.3.6. Armenian schwa “Ω” (ɿ) The Armenian schwa may remain unexpressed, be indicated by lines drawn above or below consonants (here “ԥ”), or, in more restricted cases, be expressed by a full vowel (“a”, “e”, “î”). These devices may occur within a single lexeme: e.g. 232.18.3. anrkâ, but 238.23.2. anîrkâ (ɸʍɸʗɺɾɳ «revile!», cf. Tigr. [anԥrk‫ޏ‬il]); 235.12.3. hԥrâ, but 232.1.2. hrâ (/ʇɿʗɾɳ/ «push!»).

5.3.4. Classical vowel “o” (ʏ)

5.3.6.1. Armenian schwa as “ø” or “ԥ”

Parallel to classical “e” (ɼ), the classical vowel “o” (ʏ) appears in Western dialects as [wo] in stressed position, while it remains “o” in unstressed position. In the glossary, stressed “o” [wo] further develops to “û”, except in the words ʏʗ

(a) The schwa may remain unexpressed in clusters of any type: krâ: /ɺɿʗɾɳ/ «write!» (ɺʗɼʃ) (232.2.2); ঴mâ: /ʄɿʋɾɳ/ «drink!» (ʄʋɼʃ) (232.2.3); ঴ostwanîr: /ʄʏʔʖɿʕɸʍʂɳʗ/ «confess!» (ʄʏʔʖʏʕɸʍʂʃ) (232.4.2); dngâ: /ʖɿʍʆɾɳ/ «plant!» (ʖʍʆɼʃ) (235.3.3);

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 77

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

lmanĞû: dial. /ʃɿʋɸʍʘʏʙɳ/ «compare!» (standard mod. ʍʋɸʍʘʍɼʃ) (235.4.2). (b) The schwa may equally be found expressed by “ԥ” in clusters of any type: mԥdmdâ: /ʋɿʖʋɿʖɾɳ/ «meditate!» (ʋʖʋʖɼʃ) (232.20.2); sԥrpâ: /ʔɿʗɹɾɳ/ «wipe!» (ʔʗɹɼʃ) (233.2.3); Ğԥkâ: /ʈɿɺɾɳ/ «throw!» (ʈɺɼʃ) (233.5.3; cf. 236.20.3. skâ, Tigr. [ԥsgil]); sԥdgâ: /ʔɿʖʆɾɳ/ «clean!» (ʔʖʆɼʃ) (234.3.3); ঴ԥngâ: /ʄʍʆɾɳ/ «incense!» (ʄʍʆɼʃ) (234.8.3). 5.3.6.2. Armenian schwa as full vowel in the glossary The presence of a full vowel as a representative of schwa is rare but (in the case of “î” and “â”) also morphologically motivated. (a) Armenian schwa as “î”: The use of “î” seems to be incidental in a few lexemes: mîgradâ «cut (hair)!» (ʋʆʗɸʖɼʋ) (234.8.1); wî÷ârâ «pay!» (ʕʊɸʗɼʋ) (235.1.2; cf. 235.14.3, w÷ârâ); anîrkâ (238.23.2; see above); note also îngîr «friend» (ɿʍʆɼʗ) (237.6.1). Its use is structural in the following examples: erîsîd: mod. ɼʗɼʔɿɻ «your face» (238.9.1); ašgîd: dial. ɸʎʛɿɻ «your eyes» (235.14.2). Here the junction between noun and possessive is not “ԥ” as in most dialects, but “i”. This conforms to the situation in Tigranakert.37 (b) Armenian schwa as “â”: It is doubtful whether the glossary uses “a” for schwa other than in the definite article. In the case of 237.11.1. asâ (mod. ɿʔɾɳ, ɿʔɼʃ «to say»), we may be dealing with the dialectal variant ɸʔɼʃ (but Tigr. has ɿʔʔʂʃ). The suffixed definite article -ɿ appears in the glossary as a full vowel “-â”: ৢîrdâ- «the heart»: ʔʂʗʖɿ

(233.22.1); ৬ûrâ «the door»: ɻʏʙʓɿ (236.5.1; cf. 237.14.1); bâdâ- «the wall»: ʑɸʖɿ (236.11.1); kûhîdâ «with you»: ʛʏʙ ʇɼʖɿ (240.13.1).38 The definite article also appears as a full vowel “İ” (ɾ) in Tigranakert and “a” in Urfa.39 (c) Armenian schwa as “û”: This representation is the result of incidental assimilation and is found in: mûhûrâ: ʋʜʇɿʗɾɳ (ʋʜʇʗɼʃ «to seal») (238.10.3)40; zûrûĞâ: mod. ɽɿʗʏʙʘɾɳ (ɽɿʗʏʙʘɼʃ «to tell, speak», class. ɽɿʗʏʌʘ «story»; the glossary represents ɽʏʙʗʏʙʘɾɳ in conformity with many Anatolian dialects, e.g. Polis, Sivas, but cf. Tigr. ɽɿʗʏʙʘ) (238.4.2). (d) Armenian schwa as “e”: The use of “e” for Armenian schwa in the glossary is doubtful. It may occur in: 233.1.2. han÷eĞû: ʇɸʍɺɿʘʏʙɳ «rest!» (ʇɸʍɺɿʘʍɼʃ), cf. 233.13.1. bâঠeĞû: ʑɸʉɼʘʏʙɳ «cool!» (ʑɸʉɼʘɿʍɼʃ, with full ɼ); 237.10.1. pernâ: ɹɿʓʍɾɳ «take!, hold!» (ɹʓʍɼʃ), cf. Tigr. ʚɿʓʍʂʃ. The literary Western variant of the classical ɸʓʍɼʋ «to do» is ɿʍɼʃ. The glossary consistently writes this verb with “e-”. This may represent the Group 5 variant ɾʍɼʃ (Tigr. ɾʍʂʃ). Thus we find: 2 sg. imperat.: erâ «do!»; West: ɿʗɾɳ ; Tigr. [İrİ] (237.13.1) participle: enoঠ «doing»; West: ɿʍʏʉ; Tigr. [İnoá] (239.19.3) 1 sg. pres.: genîm «I do»; West: ʆɵɿʍɼʋ; Tigr. [gİnim] (239.13.2) 3 sg. pres.: genâ «he does»; West: ʆɵɿʍɾ; Tigr. [gİnİ] (239.10.1) 1 pl. pres.: mîk-genîk «we do»; West: ʆɵɿʍɼʍʛ; Tigr. [gİnink‫( ]ޏ‬239.18.1) 2 pl. pres.: tûk genek «you do»; West: ʆɵɿʍɾʛ; Tigr. [gİnİk‫( ]ޏ‬239.16.1) 1 sg. aor.: erî «I did»; West: ɿʗʂ; Tigr. ɐʗɾʘʂ (239.12.3)

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 78

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

2 sg. aor.: erîr «you did»; West: ɿʗʂʗ; Tigr. ɐʗɾʘʂʗ (239.12.1) 3 sg. aor.: erâĞ «he did»; West: ɿʗɸʙ; Tigr. ɾʗɸࡇʘ (239.9.3) 1 pl. aor.: erâk «we did»; West ɿʗʂʍʛ; Tigr. ɐʗʂʘʂʍʛ (239.18.2) 5.4. Morphology and Syntax A number of points relating to the morphology of the Armenian material in the glossary have already been touched upon. A brief list is provided below of some of the notable features. (a) 1 pl. aorist ending in -âk: 239.18.2 (erâk: ɿʗʂʍʛ, dial. -ɸʍʛ) (b) past participle in -man: 235.23.1 (tԥrmân), as in Tigr. and Urfa41 (c) present tense marker gû- (g- before vowels): 239.6.1 (gûlmanĞnâ: ʆɿ ʍʋɸʍʘʍɾ); and passim in the main lexicon in Yale Syriac 9 (d) future tense marker d-: 239.19.2 (denîk: ʑʂʖʂ ɿʍɼʍʛ); cf. Tigr. [mԥn], Urfa [bԥd(ԥr)] 42 (e) ego «go!»: 236.24.3 (araštûn-ego: ɸʓɸʒ ʖʏʙʍ ɼʆʏ); cf. Tigr. [İgo] (f) plural in -ʖɸʛ: 239.22.1 (gendâk «women»)43 (g) pronominal ablative ending in -n-: 240.9.3 (îrneh), 240.10.2 (kîzneh); cf. Tigr. [irnİn, k‫ۑޏ‬znİn] (h) ʇɼʖ with genitive (and article): 240.13.1–4 (kûhîdâ, îrhîdâ, îm hîdâ) 5.5. Loanwords The Armenian vocabulary found in the glossary includes a significant number of

loanwords from Arabic, Persian and Turkish (see indices 3–5 in Part 2). The very fact that much of the Armenian material in the lexicon in Harvard Syriac 54 consists of translations from the Arabic parts of Bar Bahlul’s lexicon indicates that the milieu in which the Syro-Armenian lexica were produced was one that was not without acquaintance with literary Arabic, while the presence of a number of Arabic interlinear glosses in the main part of the lexicon in Yale Syriac 9 shows that the copyist of our manuscript was himself capable of writing Arabic. It may be noted, on the other hand, that a number of Arabic loanwords in the glossary occur in forms that clearly reflect the spoken dialectal, as opposed to literary, forms of the words, e.g. ÷ûwâb- < ÷awƗb «answer» (232.7.2, 234.18.2); তsebâ < ‫ۊ‬isƗb «calculation, consideration» (232.7.1), with the imƗla typical of northern Mesopotamian and Anatolian dialects (“Ɨ” > “Ɲ” in the neighbourhood of -i-, e.g. ktƝb < kitƗb «book»);44 meretâ, perhaps < mƯrƗ‫ܔ‬ (232.16.2; again with imƗla). In the treatment of interdentals, we find both of the two types of development reported for the area (৮, ঎, ঌ/਌ > s, z, ਌, in certain dialects of Batman and Mardin provinces; but > t, d, ঌ, in DiyarbakÕr);45 ৢâbût < Arab. ‫ܔ‬Ɨbit «stationary», cf. Turk. sabit (232.14.2); kzmat/঴zmat < ‫ې‬idma (frequent variant: ‫ې‬i‫ڴ‬ma) «service», cf. Turk. hizmet (233.21.2, 238.10.2); Īârar < ‫ڲ‬arar «harm», cf. Turk. zarar (234.1.1); râzî- < rƗ‫ڲ‬in (defined form: rƗ‫ڲ‬Ư), «satisfied», cf. Turk. razÕ (235.12.1); but hadrâ, prob. < ‫ۊ‬Ɨ‫ڲ‬ir «ready», cf. Turk. hazÕr (233.16.1); also meretâ (see above).

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 79

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

NOTES 1 What is presented here is the first part of a revised and expanded version of a paper prepared by Takahashi and read (in absentia) at the 9th symposium of the Canadian Society for Syriac Studies. The original paper consisted of the introductory material and a transcription/ translation of sample excerpts from the glossary. The paper has been expanded to include the transcription/translation of the rest of the glossary and the indices, as well as comments on the phonological and dialectological aspects of the Armenian material. The introductory material, including the phonological and dialectological discussions, appears here; the remainder of the paper, consisting of the transcription/translation of the glossary and the indices, has been submitted to Hugoye: Journal of Syriac Studies for publication. In the preparation of the paper, Takahashi has been largely responsible for the Syriac material and the preliminary identification of the Armenian material, while Weitenberg has been responsible for the more advanced treatment of the Armenian material. Notes of thanks are due here to the staff of the Beinecke Rare Book and Manuscript Library for their kindnesses during Takahashi’s visits to the library in the summer of 2009, to Prof. Luk Van Rompay for his suggestion that Takahashi contact Weitenberg and for his assistance in doing so, and to Dr. Masayuki Ueno (The University of Tokyo) for his valuable suggestions and his help with the proofreading. 2 See Moshe H. Goshen-Gottstein, Syriac Manuscripts in the Harvard College Library. A Catalogue (Missoula: Scholars Press, 1979), 59–60. The existence of this lexicon, which was then in the possession of James Rendel Harris, was first indicated by David Samuel Margoliouth, whose project of publishing the entire lexicon saw little progress, but who has left us some useful observations in the one article that he published on the subject (D. S. Margoliouth, “The Syro-Armenian Dialect,” Journal of the Royal Asiatic Society 1898: 839– 861; cf. Sebastian Brock, “Armenian in Syriac Script,” in Armenian Studies. In Memoriam Haïg Berbérian, ed. Dickran Kouymjian

[Lisbon: Calouste Gulbenkian Foundation, 1986], 75–80). 3 See S. Brock, “Catalogue des manuscrits syriaques, éthiopiens et arméniens du Centre d’Etudes Orientales Chrétiennes du Mouski, Le Caire. A) Syriac Manuscripts,” Studia Orientalia Christiana Collectanea 18 (1985) 213–218, here 215f.; also id., “Armenian in Syriac Script,” 77. One correction may be made to the information given in the articles just cited; Melke b. Nicodemus ( ƑƟ ƎŨĥ ųƄƇƉ ĸŤƊƣ ĸŴƊſĪŴƠƀƌ), the copyist of the Cairo manuscript, was not “from Zalmaz in Gargar”, but was “from the environs (acmƗl) of Gargar (ƢūƢū ĵŤƊƕĥ ƎƉ)” (with thanks to Mr. Benjamin Trigona-Harany for forwarding to me the images of the manuscript just as the present paper was going to press). 4 Digital images of the whole of this manuscript have now been made available on the website of the Beinecke Rare Book and Manuscript Library (see http://beinecke. library.yale.edu/digitallibrary/). 5 Leo Depuydt, “Classical Syriac Manuscripts at Yale University: A Checklist,” Hugoye: Journal of Syriac Studies (http:// syrcom.cua/Hugoye/) 9/2 (July 2006). 6 A more detailed account of the six manuscripts sold to Yale by Menemencioglu will appear in H. Takahashi, “Also via Istanbul to Yale: Mss. Yale Syriac 7–12,” in In the Shadow of the Pyramids: Festschrift for Dimitri Gutas, ed. D. Reisman & F. Opwis (Leiden: Brill, forthcoming [2010]). 7 See, for example, Hanna Dolapönü [Dolabani], Tarihte Mardin. Itr-el-nardin fi tarih Merdin, trans. Cebrail AydÕn (Istanbul: Hilal MatbaasÕ, 1972), 166; Samuel Akdemir, østanbul mozai÷inde Süryaniler/The Syrian Orthodox Community in the Mosaic of Istanbul, vol. 1 (Istanbul: østanbul Süryani Ortodoks Kilisesi ve Yönetim Kurulu, 2009), 101–103. 8 Filuksinǀs Yuতanon DǀlabƗnƯ, M‫ۊ‬awyono da-ktobe sri‫ܒ‬e d-botay arke d-dayroto wa-dҵidoto suryoyoto da-b-Madn‫ۊ‬o (Fihris ma‫ܒې‬nj‫ܒ‬Ɨt al-suryƗnƯya), ed. Grigǀriyǀs Yuতanon Abrohom [IbrƗhƯm] (Aleppo: Mardin Publishing House, 1994), 19–20. 9 IƥnƗ৬iynjs AfrƗm al-awwal Barৢaum, alLuҴluҴ al-man‫ܔ‬njr fƯ taҴrƯ‫ ې‬al-ҵulnjm wa-l-ƗdƗb assuryƗnƯya, 4th ed. (Holland [Glane/Losser]: Bar Hebraeus Verlag, 1987), 26, no. 23; = Ignatius

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 80

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

Aphram I Barsoum, The Scattered Pearls, trans. Matti Moosa, 2nd ed. (Piscataway: Gorgias Press, 2003), 15–16, no. 23. 10 Fehim Beg 2 = Yale Syriac 7 (copied in the Monastery of Mor Abতay in 1590); Fehim Beg 3 = Istanbul, Meryemana 7 (copied in the Monastery of Mor Barৢawmo in 1574); Fehim Beg 6 = Paris syr. 395 (notes of ordinations for Vank and Abnj GhƗlib, ca. 1624); Fehim Beg 7 (copied in Ma‫ޏ‬arrat Mor Zakkay in 1590/1). 11 See Hubert Kaufhold, “Notizen zur späten Geschichte des Barsaumô-Klosters,” Hugoye 3/2 (July 2000); cf. Muzaffer øris, Bütün yönleriyle Süryaniler (Istanbul: Ekol YayÕmcÕlÕk, 2003), 120–136. 12 The lexicon begins, immediately after the ornamental drawing marking the beginning of ł the work, with the entry “ĖķŤƏł ŴƍƆ Ŋ Ł .ŴƀŎ ŨƦƣĥʼn ”, and runs to “ƁƍūŁĥ” (end of p. 5) in alphabetical ō order. It then returns to “ĿŴƊŹĥ” (p. 6), goes as ł far as “ĵĥƦʼn ƣŎ ĥʼn ” (p. 13), and jumps to “IJűƌŁĥ” (p. 14). From that point the rest of the lexicon is in regular alphabetical order. 13 On pages 200, 218, 220, 224, 241–243, 247, 248, 251–253, 258, 260, 261, 263, 267– 269, 279, 281, 282, 284. 14 The same three-number system is also used below for the indication of entries in the main lexicon on p. 4–199 of Ms. Yale Syriac 9. 15 In the body of the glossary the identification is normally with the modern Western Armenian lexemes and forms, the literary language to which the language of the glossary stands closest. Classical forms, on the other hand, are normally used in the observations on the language of the glossary in Section 5 below, where the historical development is of concern to us. 16 The two signs ৖ and ৔ occur only on two occasions each in the glossary. The sign ৖ occurs frequently in the main lexicon in Yale Syriac 9; ৔ is rarer, but does occur in at least about ten cases (22.20.2, 33.3.1, 33.10.2, 33.13.1, 43.4.2, 95.11.1, 133.3.1, 177.6.6, 197.4.2). 17 Cf. Margoliouth, “Syro-Armenian Dialect” (as n. 2 above), 858–860. 18 With such use of ‫ܒ‬Ɲt for “d” (ʖ), one might compare the frequent use of ‫ܒ‬ƗҴ for “d” in Ottoman Turkish (e.g. ύΎρ alongside ύ΍Ω for mod. Turk. da÷ «mountain», ίϮϗϮρ alongside ίϮϗϭΩ for dokuz «nine»).

19 The same verb is found written with “ৢ” in the page of the Harvard lexicon reproduced in Margoliouth, “Syro-Armenian Dialect,” opposite p. 850, left column, l. 1; cf. ibid., p. 853. 20 See Margoliouth, “The Syro-Armenian Dialect” (as n. 2 above). 21 The elision is more marked in the glossary than in literary modern Western Armenian; see, e.g., 232.4.2 (঴ostwanîr: ʄʏʔʖʏʕɸʍʂɳʗ), 233.18.3 (bašrtrâ: ʑɸʎɸʗɸɻʗɾɳ), 235.1.3 (m঴tarâ: ʋʄʂʀɸʗɾɳ), 235.23.2 (÷oঠwâ: ʁʏʉʏʕɾɳ), 236.24.2 (awdĞîr: ɸʙɼʖɸʘʂɳʗ); but note the absence of elision in 233.9.2. ergîr bakanâ «worship!» (cf. class. ʑɸɺɸʍɼʋ, mod. ʑɸɺʍɼʃ), probably due to ecclesiastical usage helping preserve the classical pronunciation. 22 On the classification of Armenian dialects, see Andrzej Pisowicz, “Matériaux pour servir à la recherche du consonantisme arménien. Continuation dialectale des occlusives et affriquées de la langue classique,” Folia Orientalia 17 (1976) 197–216; Bert Vaux, Phonology of Armenian (Oxford: Clarendon, 1998), 7–12; cf. Robert H. Hewsen, Armenia. A Historical Atlas (Chicago: Chicago University Press, 2001), 227 (map no. 218). On the dialects of Group 5, in particular the Tigranakert (DiyarbakÕr) dialect, see Bert Vaux, “The Armenian dialects of Tigranakert/ Dikranagerd and Urfa,” in Armenian Tigranakert/Diarbekir and Edessa/Urfa (UCLA Armenian History and Culture Series. Historic Armenian Cities and Provinces 6), ed. Richard Hovannisian (Costa Mesa: Mazda Pubishers, 2006), 191–207, with extensive literature. The data on the Tigranakert dialect in the discussion below are taken from Anahit H. Haneyan, Tigranakerti barba‫۾‬Ϸ [The Dialect of Tigranakert] (Erevan: Haykakan SSH GA Hratarakþ‫ޏ‬ut‫ޏ‬yun, 1978). 23 On the dialect of Cilicia, see Hraþ‫ޏ‬ya Aþa৘yan, Kҵnnutҵyun Kilikiayi barba‫۾‬i [Study of the Dialect of Cilicia] (Erevan: Erevani Hamalsari Hratarakþ‫ޏ‬ut‫ޏ‬yun, 2003). 24 The negative particle is regularly represented by ÷-/÷î in the main part of the lexicon in Yale Syriac 9. 25 Cf. Margoliouth, “The Syro-Armenian Dialect” (as n. 2 above), 848. 26 In 234.15.2, pîbrâ < class. ɹɼʙɼʓɼʋ «to nail», the sequence ʙʓ created by the elision of

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 81

The Shorter Syriac-Armenian Glossary in Ms. Yale Syriac 9 Part 1 _________________________________________________________________________________________

the middle syllable is then assimilated to ɹʓ. The use of “î” for ɼ here in unstressed position is unexpected. 27 On the dialectal development of initial Classical Armenian ʌ (y), see J.S.S. Weitenberg “Diphthongization of initial E- and the development of initial Y- in Armenian,” in Evidence and Counter-Evidence. Essays in Honour of Frederik Kortlandt, 1. Balto-Slavic and Indo-European Studies, Studies in Slavic and General Linguistics 32, ed. A. Lubotsky et al. (Amsterdam-New York: Rodopi, 2008), 609–616. 28 The one instance in the glossary where a double consonant is likely to be explicitly indicated as such occurs in a loanword that involves an element subject to exceptional orthographical treatment in the original language (240.20.4: šallâ, probably < Arab. in šƗҴa AllƗh). 29 Haneyan, Tigranakerti barba‫۾‬Ϸ (as n. 22 above), 55–57. 30 Cf. Margoliouth, “The Syro-Armenian Dialect” (as n. 2 above), 848. 31 Cf. Margoliouth, “The Syro-Armenian Dialect” (as n. 2 above), 854. 32 Haneyan, Tigranakerti barba‫۾‬Ϸ (as n. 22 above), 32–33. 33 The form “îs” is consistent with the material of the glossary, but does not agree with Tigr. [jİs]. 34 The “î” in kîzneh is unexpected. 35 The imperative -ego ɼɺʏɳ «come!» in 236.24.3 is an incidental secondary development from ɼʆʏʙ (cf. Tigr. [İgo]).

36 In gâঠdûdâ (class. ʆɼʉʖʏʖɼʋ «to soil», 233.8.2), the irregular reflex of unstressed “o” is influenced by the basic form ʆɼʉʖʏʖ «dirty». 37 See Haneyan, Tigranakerti barba‫۾‬Ϸ (as n. 22 above), 37; e.g. ʔʂʗɻʂʔ «my heart». 38 Cf. Margoliouth, “The Syro-Armenian Dialect” (as n. 2 above), 848. 39 See Haneyan, Tigranakerti barba‫۾‬Ϸ (as n. 22 above), 37; Vaux, “The Armenian dialects of Tigranakert/Dikranagerd and Urfa” (as n. 22 above), 192. 40 Cf. the parallel phenomenon in 238.13.2. hûkûm- < Arab. ‫ۊ‬ukm (cf. Turk. hüküm), although the additional vowel may in this case be due to the influence of Turkish. 41 See Vaux, “The Armenian dialects of Tigranakert/Dikranagerd and Urfa” (as n. 22 above), 192. On instances of participles in -man in Harvard Syriac 54, see Margoliouth, “The SyroArmenian Dialect” (as n. 2 above), 850–851. 42 See Vaux, “The Armenian dialects of Tigranakert/Dikranagerd and Urfa” (as n. 22 above), 193. 43 Cf. Margoliouth, “The Syro-Armenian Dialect” (as n. 2 above), 844 (text, l. 4), 846 (a঴prdâk «brethren»: ɼʉɹɸʗʛ, etc.). 44 See Otto Jastrow, “Das mesopotamische Arabische,” in Handbuch der arabischen Dialekte, ed. Wolfdietrich Fischer and Otto Jastrow (Wiesbaden: Harrassowitz, 1980), 140– 154, here 140 (also, libr. cit. 55). 45 See Jastrow, “Das mesopotamische Arabische” (as n. 44 above), 142 (also, libr. cit. 26, 50, with the literature cited there).

_________________________________________________________________________________________ Journal of the Canadian Society for Syriac Studies 10 (2010) ʊ Page 82



  7KH6KRUWHU6\ULDF$UPHQLDQ*ORVVDU\LQ0V