Monographic Journals of the Near East General Editor: Giorgio Buccellati Afroasiatic Gnguistics Editor: Robert Hetzron, Santa Barbara Advisory Board: Ariel Bloch, Berkeley John B, Callender, Los Angeles ‘Talmy Givon, Los Angeles ‘Thomas G. Penchoen, Los Angeles Stanislav Segert, Los Angeles Volume 1 sue 5 February 1975 The Modern South Arabian Languages T. M, Johnstone ® Undena Publications Malibu 1975AFROASIATIC LINGUISTICS ‘AL includes contributions in linguistics within the vast domain of Afroasiatic (Hamito-Semitic) languages. Articles of general, theoretical interest using Aftoasiatic material, descriptive, historical and comparative studies are included. Editor: Robert Hetzron (1346 San Rafael, Santa Barbara, Ca. 93109, US.A.) Advisory Board: A. Bloch, J. B. Callender, T. Givén, T, G. Penchoen, S. Segert. MONOGRAPHIC JOURNALS OF THE NEAR EAST MINE is a system of journals on the Near East, with each journal devoted to a specialized study area, and each issue consisting of a single article. Current journals in the system are Afroasiatic Linguistics and Assur. General Subscription For a prepayment of $12.50 the subscriber selects random issues from within the entire system as desired, up to a total of 200 pages. The subscriber is also entitled to (1) periodical lists of abstracts from all journals in the system, and (2) reservation to any journal within the system, whereby issues of a given journal are sent on approval immediately upon publication (and may be returned within two weeks). Library Subscription ‘A prepayment of $12.50 for each journal in the system secures all issues ofa single volume as soon as they are Published. This subscription schedule does not allow the selection of random issues; in return, a discount is provided in the form of a greater number of pages for the basic price of $12.50 (since a volume will normally include more than 200 pages). Library subscriptions are available to both institutions and individual scholars. Individual isues are numbered sequentially within each volume. Each issue has its own pagination. A volume {s closed when a total of between 200 and 250 pages is reached. A title page and a table of contents listing all issues within each volume are sent to all subscribers atthe close of a volume. Periodicity in the order of appearance of issues is not predetermined. A volume, however, is generally completed within one year. Institutional and Professional discount of 20% on single subscriptions (higher on larger orders). Payment must accompany orders from individuals. A handling fee of 70q will be charged to Libraries if order is not prepaid. Order from: UNDENA PUBLICATIONS, P.O. Box 97, Malibu, California 90265, US.A. 1975 by Undena Publications. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photo-copy, recording, or any information storage and retrieval system, without permission in writing from the publisher.Honographic Jowrnas of the Near East Agnoasiatic Linguistics 1/5 (February 1975) THE MODERN SOUTH ARABIAN LANGUAGES T. M, Johnstone School of Oriental & African Studies The University of London ‘This is @ comparative and descriptive survey of the wodern South Arabian languages, Nehri, Harstisi, Sheri and Socotri. After a presentation of the speakers, the consonantal system is analysed with special attention to glottalized consonants, the laterals, palatal ization, the relation of J and h. “The morphology section first describes verbal derivation, tenses, conjugation and some special phenomena. Further sections are devoted to the morphology of the noun, of adjectives, numerals, denonstratives and pronouns. 'This is followed by a survey of case-marking prepositions, adverbs, conjunctions and verb modifiers. Syntax is represented by a shorter note, containing, among other things, illustrations of unreal conditional constructions. TABLE OF CONTENTS page TABLE OF CONTENTS ee ee eee eee eee eee 1, THE SPENERS. 6 eee eee eee ae 2, THE MSA PHONOLOGICAL SYSTIM 2 ee eee es 2.1, The consonants...) 21 it Piliipiiiiin: 3 2.1.1. The status of ©. .......2..00. wee ee eee 4 2.1.2. The glottalized consonants .. 2.1... eee eee wee 6 2.1.3. The laterals $andZ@ 2... . ee ee ee ee 7 2.1.4. The palato-alveolars and palatalized consonants. ........- 7 2.1.5. The consonants m, and 6 in Sheri... ....--- evaewe 9 2.1.6. The relation of £ and h in the MSA languages 12221222 10 2.2. The vowel system and stress . 1... eee tee ee tet 10 3. MORPHOLOGY. we eee pee 2 3.1. The morphology of the verb. . 1... 2. eee ee ee eee eee 12 1.1, The themes 62 eee et ee te ete ee eee ee eee 2 . Perfect, Imperfect, Subjunctive.. 2... eee 13 3.1.3. The conjugation. +... ... 0. sce cress esses ee 6B 3.1.4. Subjunctive in-leln 2222222222 i002 Pliiiiid as 3.1.5. The passive. ee ee ee 19 3.1.6. The non-occurrence of f= 2... eee te tee ee 1g 3.2. The morphology of thenom. . 22211111) Sees ae 2d 3.3. Adjectives 1. ese ese lilies rersrssteesss 2 3.4, The numrals....- Se ee wee ee tee wee ee 222 T. M, Johnstone [AAL 1/5 3.5. The demonstratives .... . ns wee eee 3.6. The personal promos TTT IIiiiiiiii Dil as 37, Particles and adverbs 1111) 11 Be eet es 16 4, NOTES ON SYNIAX. 2... ee eee ee ee eee we eee eee 28 SELECT BIBLIOGRA 6 ee 29 1. THE SPEAKERS ‘The MODERN Soun# ARABIAN (MSA) Languages, which are now confined to a relatively small area in and around Dhofar and to the island of Socotra, are the last vestiges of a group of closely related South Semitic languages which were spoken over the whole of South Arabia. ‘The modern languages exhibit certain features however which are absent from EPIGRAPHIC SOUTH ARABIAN (ESA), and it has been doubted whether they can Le considered as directly related to the old jiterary dialects. They share many distinctive features with Ethiopic and the correlation of such features will no doubt throw light on some problems of ESA morphology. ‘The main modern languages are Nehri, Sheri and Socotri. MmeuRr, whose speakers call the Language Mshtauyot, has a southern dialect, spoken in the Mukaila area and on the coast to the northeast of it, though little as far south as Mukalla itself; and a northern dialect whose main centre is in Negd in Dhofar, nanely in the high valleys around the great complex of mountains in this province of Onan.! The Northern Mehra are semi-nomadic and spend certain seasons of the year with their cael herds in the mountain area with the Sheri speaking cow- men, Some of them speak only a little Sheri but many, mainly those with property in the mountain area, are bilingual, and some who are settled on the coast speak only Sheri and Arabic. All Sections of the’Mehra may number about 15,000, but many in the south speak only Arabic. ‘The languages of the YARASTS (Hot'sé:yat, Ar, farsisé) and the pagRyrean (Ar. Baghari) are closely related to Meliri, so closely indeed that they could be terméd dialects of Mehri, depending upon the terms’of definition. However speakers of Harsiisi who number about 400- 500 appear not to be related to the Mehra, though like them they are of high social status. ‘They would seem indeed to have adopte’ a form of Mehri in the course of a long period of symbiosis with a southern group, their speech now differing fron that of their Negdi neighbours in some ways more than from the Mehri of the Mukalla area, and showing certain novel features probably adopted from Sheri in the course of their drift northeast. ‘The Batharis are of lowly social class and are not related to the Mehra, being possible aboriginals of the area. They would seem to have adopted their language after having been made the serfs of the Mehra, They are now pastoral cave dwellers and fishermen and are few in number. ‘The Seat language is spoken by a number of communities of different social status and tribal origin, numbering together about 5,000. Strictly speaking the language has no name amongst native speakers, and Sheri is merély an adjective of Shen "(the) mountain(s)'. The Arabic name for the language is not Shihri, a particularly misleading tem since it seems to relate to the place-name Shihr which is'within the general locality, but Jibal Most of my work has been on this, more conservative, dialect of Mehri. The hitherto published work is almost all on the southern dialect(s). AAL 1, 941975] Modern South Arabian Languages 3 straightforward translation of the word Sher: Siidarabische Expedition is altogether erroneot by-fomm, or to the Mehri word Shayt ‘mountain’: The nane Sxawri applied to it by the . It cannot be related either to an Arabic the Mehra call the language Sha tay. The problem is complicated by the fact that the name She'ni (pl. Ske'to), though it super- ficially means only 'mountain(man)', in fact is a depejorative term, applied by the sections of Dhofari’ society of high social standing to those people living in the mountains, who were serfs by status. However it is likely that these people, till recently serfs, are the original speakers of Sheri, which was adopted by one or more waves of invaders who’ imposed their supremacy upon the earlier inhabitants, That at least one of the socially dominant groups was of Mehri origin is suggested by the fact that these people compose poctry in a specifically poetic language which is, at base, a Sherized form of Mehri. It would seem likely that this represents the heritage from a period when Nehri was the language of this social group, though the structure of the verse is not like that of Nehri verse at all. ‘The people of the xuRIA MuRIA Islands speak a dialect of Sheri which seens to their Sheri- speaking neighbours on the mainland like "baby-language” and is only just comprehensible to them. Like the speakers of Sheri, the socorers also have no word for their language. The nane of their island, which is nomtally traced to a Sanskrit word meaning ‘blessed’, in their om language is $k'at?ré. But this is rather a vague word in Socotri which also means ‘everything, world" (donya) and moreover it has no adjectival derivate which can be correlated with the Arabic word Suqugni. ‘The native speakers, who number about 6,000, can be divided igto three or four main socio- economic groups: former slaves, concentrated mainly in Suq (Sok) but found in lesser nunbers in other large towns; the former ruling class, who are Mehri by origin but with a good deal of Arab and Socotri admixture, invariably bilingual in Arabic and Socotri; the semi-nomadic cow-herding people of the high plain and the fishing communities who probably are the aboriginals; the bedouin of the high desert plain who are few in nunber and probably only a section of the preceding group. ‘The language of the mountain people is difficult for townspeople to understand. It is probably a good deal more extensive in vocabulary, and certainly includes some words in which there occur consonants which do not occur in Socotri texts recorded either by the SAB or by myself. They were recorded by me only as word isolates, refer for the most part to legal processes, and may be borrowed from Mehri. 2. THE MSA PHONOLOGICAL SYSTEM 2.1, The Consonants ‘The consonant system of these languages bears a striking resenblance to those of the Ethiopian languages as regards the occurrence of ejectives (as against the velarized conso- nants of Arabic) and, for some languages of the group, the almost complete non-occurrence of the pharyngal ©. The system is notable also for the occurrence of the voiceless and voiced laterals 3 and 2 which are represented in the orthography of many of the Ethiopian languages and may well have been laterals in Ethiopic. ‘The total range of consonants in the MSA Languages is set out below (ves. for 'voiceless', cu. for 'glottalized’ ~ ejective, ven. for 'voiced"): AAL 1, 954 T. M. Johnstone [AAL 1/5 TABLE 1 PLOSIVE FRICATIVE NASAL LIQUID or ROLLED Ves. GL. VCD, VCS. Gh. VCD. veD. GL. vod. Labial and Labio-dental 6 6 wom Denti-alveclar t 9? nr) noe 6 8 2 ” Palato-alveolar i b&w Fon Lateral 4 z Palatal y Velar kook g x y Pharyngal hk : Glottal ” h Not all of these consonants occur in all of the MSA languages. Except in borrowed words x and y do not occur in Socotri and there is consequent loss of distinction between x/h and y/*. Thus cf, M(ehri) yayg with S(ocotri) ‘ayg ‘man'; and M yoro:bwith$otab 'to know’ In both Socotri and the Southern Nehri dialects there is loss of distinction as between 1/8, 27/78 and d/2;_thus cf. Northern M 840: Southern M éato:, S tio; NM %a-, SM do-, S do-, di- "of"; NM Paare, SM tran, S tran ‘upon'. 2.1.1 THE STATUS OF © In M and H(arsisi) the pharyngal ¢ does not occur, or strictly speaking does not occur regularly or predictably, except in borrowed words. Its non-occurrence however has a marked effect on syllabication,’and it is curious that some of these resultant features occur in the languages in which © does occur specifically. Thus in M and S(heri) forms which nomnally have prominent final syllables have prominent penultimate syllables where the final of their three radicals is *. In S, where such forms normally have short, prominent penultimate syllables, the syllable is long where the final syllable is ©. Thus contrast: TABLE 2 M Aa'gors S no! fos S 'nafog "to kick’ "no: fa "nit fa "na: fa. "to be of use! Although * occurs explicitly as a radical in § and S, it tends to be replaced by ? in both of these languages, though in neither is the appearance of ? < ‘ either regular or predictable. This tendency throws some light on the disappearance of *, except as an ortho- graphic feature, in Ethiopian languages. The following are a few examples of the occurrence of ‘ in M, contrary to expectation, in words not borrowed from Arabic: AL 1, 9071975) Hodern South Arabian Languages 5 TABLE 3 "ayban, ( "7aybox) "to gaze into the distance’ "azar (WU Adem) ‘excuse! "asg'zomn ' group of old women" Sazte:'krern (dim. of %eLi:h?) ‘Little camel calf" sa:m'doy (but unaffixed 2a:mosd) ‘he hit me deliberately" "Sayné (<8 'e'né) "tom-cat' at tare: "to get mixed up" saz 'wirg "to divert’ Sor'yer (iH 8o"yer) "to pity! In many words borrowed from O(mani) A(rabic), even those not otherwise assimilated to M phonology, ? usually occurs rather than“, or occurs as a free variant of ©, Moreover words occur often in more than one state of assimilation. Thus: TABLE 4 ragh, 7a: 68 (OA F248) "baggage" fabs, rake (OA Saks) "photograph" (H ratktad, OA Sagd) ‘marriage contract’ (Hl Adem, OA 0:C. The evidence from Ethiopic seems to indicate that NSA verbs may have retained their final vowel longer than nouns. NSA feminine nouns however have both ¢ and o feminine endings, as, e.g. M rohobert, § '$é'net ‘town’ and M raxomo:t, § axilt "gyptian vulture’ In M and H the long vowel a: is limited in occurrence to certain phonetic contexts, principally those involving © radicals, glottalized consonants and, in certain cases, the guttural conso- nants other than ©. However M and Hl differ in a mumber of respects in relation to the last group, as, c.g.: M yorbor, H yazboA "pregnant camel"; M ko:2i, Il ha:£ ‘to beckon; M xo:mod, Hxasmah ‘fifth’, fc. In § and S the range of vowels is larger, namely @, ¢, ¢, 0, 4, 0, 9, and u, all of which can occur nasalised in S under the influence of am radical. In S, although there are definable exceptions, » and e occur only in closed syllables, while © and e occur in open stressed syllables (and’before h in final unstressed syllables). In $ and S vowels are long in stressed open syllables and in final stressed VC syllables, but it would seem that length here is a phonetic rather than a phonological feature. In both § and $ however the vowel of the penultimate syllable of forms such as § £e:tay, § La:ta® (to Kill) seems to be phonologically long, under the influence of the final */y radical. In S also where b or m is elided intervocalically, the resulting vowel seems to be phonologically Jong. ‘The evidence from Mj indicates that in the MSA languages there has been regression of stress in forms which had earlier had a series of short syllables, the stressed vowel being lengthened and raised to 0: or e:. ‘The process has been modified in § however and to some extent reversed in S, Thus compare with a possible earlier takaza, M to'ko:z, S to'koz (tkoz| and S 'akoz ‘to straighten; to stick in’ &c. In § however and to a lesser extent in $ the consonantal context has a mich greater effect on the quality of the vowels of forms, and in § also on the stress pattern of forms. ‘Thus in Mj except where gutturals are involved, in the simple sound verb an earlier CaCaC(a) regularly gives CoCo:C (0:7 aw after glottalized and lateral consonants). Compare: 8 "be'nok, "to kneel" "bi'noz "to come out; receive visitors’ "bo! fon, ‘"botar "to tear’ do" for "dofor "to push" *du'tewm "dokom "to blunt’; (S) "to push’ "fe'dor "to outstrip" gh'o8 "to shatter’ *fe'nok "to wipe" Gtor "to tum aside’ go’ tum "to lop off" *gu'zum "gozom "to swear’ AAL 1, 10512 T. M, Johnstone [AAL 1/5 M 8 s (continued) anos" "Ko'nok? *hyotok? "to steal’ kobon usn "to hide" bonorb "ko'nab "to screw" bose, baat, ‘to be too tired’ batorb tab btob "to write’ kredoem "ko" dum "radon (M) 'to precede’; (S) 'to take out the animals in the morning’; (S) 'to stand guard; precede! It is difficult to generalise when so much remains to be observed but the facts indicate, as was suggested above, that in the MSA languages the stress regressed to the final syllable wherever there was No pre-final long syllable in a form, Certain exceptions to this general tendency can be defined, as for example in the forms having a final guttural radical discussed above. In $ however a contrary tendency has resulted in many forms having multiple stress, the vowels of stressed open syllables being phonetically long, but probably phonologically short. In S except in very few forms (such as those of unsuffixed subjunctive forms) stress is regularly penultimate and al1 vowels phonologically short. The quality of these phonologically short vowels in final position nevertheless shows fairly clearly that they were earlier long vowels both in $ and in 8, 3. MORPHOLOGY 41. The Morphology of the Verb 3.1.1, THE THEMES The MSA verbal system agrees with N. Semitic in that the derived themes are characterised by prefixation and infixation and by the lengthening of internal vowels; but it agrees with S. Semitic in that the prefixes and infixes may be applied to more than one kind of base, though not in all themes. Thus compare the schema of the simple and derived themes set out below: TABLE 12 qa (2) SIMPLE VERB M CoCo:c Ce:cot 4H Coco:¢ Ce:cec 8 Ca'CoC, "Cecac s "Coco "Cocec INTENSIVE-CONATIVE = M (a)Co:cec* 4H (@)Cercac* 8 (©) 'CoCoc* s "cocec “The prefix is added only to verbs whose initial radical is voiced or glottalized. AL 1, 1041975) Wodern South Arabéan Languages 13 a) @) (Table 12 continued) causative M (hayccozc q (ayccore § (e)C'Cec s "(@)CCoC REFLEXIVE M aCtaCo:C CartCoct H aCtaCorC "CatCoc S acts "Cec "CoOCaC s 2 "CoCaCat CAUSATIVE-REFLEXTVE = M $eCCo:€ $oCe:C06 nl Socco:¢ $o"Covoe 8 Foc'Cec $o'CeGse s "SoCCoC $o'Catoc® ‘These may be set out in other ways and the following schema for M shows a closer approximation to Ethiopic, given that internal gemination is not a feature of MSA derived verbs. ‘TABLE 13 (a) (b) (©) @ TEM caus. REFLEX. CAS. ‘REFL. 1. SIMPLE VERB CQCo:C/CeGoC (ho) OCoxC oC taCo#C —SaCCo:€ 2. INTENSIVE-CONATIVE (a)Co:CoC - Cazt0oC — BaCe:CaC It will be noted that there are no derived themes with a t- or an n- prefix, though the latter occurs with quadriliteral verbs. 3.1.2, PERFECT, IMPERFECT, SUBTUINCTIVE Like Ethiopic, the subjunctive of MSA differs in syllabication from the imperfect. Note the suffix -n of the imperfect in (2a), (1c), and (2d). Thus (perfect/imperf./subj.): ‘TABLE 14 Gaia) Mt boto:b/yako:tab/yakterb "to write’ bo! tob/ya'kotab/ "yoktub *kot9b/yo'hotab/t-ik'teb aw ®This is usually realized as 'CatCoC, 'CataCoC the first of which may be a better base form for M. "In this case So'CeCoCis probably the base-form, but in the examples collected the initial radical requires the occurrence of a following a. AAL 1, 105u (la.2) (2a) (ab) (ac) (2c) aaa) (2d) OME Zz OO Zz OME Ze OO 5 omnes oom zs vag T. M. Johnstone TAAL 1/5 (Table 14 continued) terberif yetborn? "to be broken’ tern/yoti! or/ yot!bor "dekon/ya'dekor/L-dd "kor "to mention’ ao :keb/ya'rakban/yaro kab "to put (a pot) on the fire’ ane:kob/ yare skobon/yareskab "nokob/yo "aukban/yo"tokab zomat/yo'zomotan/t-iz6mo£ "to put saddle-cloths on’ hhanso:m/yshanso:m/yo'hansom ‘to breathe’ anso:m/yanso:m/ ya hansom en! sém/yen' sun/" yensom® "andon/yo'neSur/'L-onSar® "to advance’ aktato: {/yoktate: fon/yokto£o:§ 'to be troubled" oktato:§/yobtate: fon/yaktestos ak'te'£e/yor'te'Lefon/yekte 'Lof "fatogar/y ftago+n/yagtergon ‘to burst, split open’ "fatgor/ yo §tegotr/yo'tagor " fotgua/yos'te'gon/yas'tegor ‘motedan/yam'tadan/L-dm'tadar ‘to ponder’ Sewbo:d/ yeSowbo:d/ya'Sakbad ‘to be hit" Sotbo:d/yo¥otho:d/ ya" Sathad Bot" bed/yoSalo:d/ys'Selbed “Yomtot/yos 'metol/t-i'emtal ‘to talk with each other" Katerbad/yoStabdan/yakterbad "to hit back, try and hit each other! $o'Lobad/ yas 'Labdan/ yoS 'tabad Be'terd/ yok Lebdan/yasterd So 'halo{/yak 'hatas/£-é'Sahag ‘to go early’ 7In 1a.2 both the $ and S pattems indicate that the MJ imperfect was earlier yeCeCo:C. Since anaptyctic vowels occur freely in MI wherever two-conSonant clusters occur, this distinction can no longer be made. 8Viz. om > am and om > un, Vie. ot > ue AL 1, 1061975] Modern South Arabian Languages 15 3.1.3. CONJUGATION ‘The verb in MSA has three numbers, singular, dual and plural. The dual has three persons including a first person.'? The revelant suffixes and affixes for the PERFECT are: TABLE 15 Mi 8 s Sing. 3m. : : - 3f not oh 2m “k “tk 28 ey ai le. -k -k 3m -0: -'0 -o 3e. ~to: ~"t0 ~oto Ze oki hi “ki le. hi Mi “bi PL. 3m V-om/v'? - vie Bf. - - - 2m kom -tum -hon 26. -tan -bon ~ton le. -on “on -on For the IMPERFECT AND SUBJUNCTIVE: TABLE 16 M § s Sing. 3m. Bf. 2m ze. le. Dual 3m. 3 Ze. le. ‘This has been noted otherwise only for Ugaritic. See also the bibliography. ‘-e:¢ is characteristic of the intensive-conative theme and of passives. \2y = an internal vowel change. ‘Depending on the type of verb, Many classes have both V and -4. AAL 1, 10716 T. M. Johnstone [AAL 1/5 (Table 16 continued) M § s PL. 3m yer 3f. for...-on 2m 2 le. Although internal ‘vowel change in M and § distinguishes persons, vowel changes also occur in M, and to sone extent in H, whenever a syllable is closed. The following exanples show that this principle operates differently in M and H, the former being more conservative in this respect. ‘The equivalent § and S forms do not show this. Thus: For the PERFECT: ‘TABLE 17 M 4 8 s Sing. 3m. hotorb botorb ba 'tob fee "tab! 3£. katobot botabort botitot ka 'toboh 2m. ho'tobh hoto:bob e'tobh ba'tobh 2£. ke! tabs hator bad ka" tab8 ka 'tobh 1c. e'tabk bato:bok beg "tobk bg 'tobke Sm. batabor batabo: foti'o ba'tobo 3£. — hatobto: botobto: botab "to ka 'tobato 2c. ka'tabké bato:b(o)ké bea "tobi bea 'tobbé le. ke'tabké kato: blalki ko "tobi ko 'tobki ym kot teub feo" tabom ita'tob "hatab 3. hetorb hotorb ha'tob *hetob 2m, ke 'tablam ——kotozbla) kom bea" tobkum bea "'tobhon 2£. fe'tabkon katz la) kon bg" tobkon ba 'tobkon lc ketorben ezozbon ba'ton fee 'tobon The subj. has the same affixes as the imperf., but the 2f.s. always has final - in M and, in all of the languages except Hl, an - prefix occurs before certain of the forms, In § and’S pl. forms show V. Thus conpare the inperf. and subj. forms conjugated below.*? ‘Though there is considerable variety in S (as in §) in the formation of simple verbs of this class, it should be pointed out that in most disyllables the penultimate syllable is promi- nent (stressed). However there are disadvantages to transcribing the base-form as ktob in this context. *Sthe verbs are M rgko:z 'to stand upright, stick in the ground’; ll Lebo:d 'to shoot, hit’; S$ na! fos 'to kick’ and S 'k*abor 'to bury’. AAL 1, 1081975) Modern South Anatian Languages ar TABLE 18 IMPERFECT: M y § s Sing. 3 Yoro: bez yatorbod ye 'nofos yo'k* abor SE. tenotkez tolorbod te'nofes ta'e? aber 2m Lerotkez teloibed ta'nofas ta'b’ obon 2. tenethez tolerbad ta'rh gos ta'k’ abin lc. erorkez elorbed ot noges atkrabor Sm. yonkazor yalbado: Yorog's0 yo'kaboro 3£. tenkozo: tabledo: tates 's0 te'k?abono 2c. — tenkezo: totbodo: taraf'so ta'krabono 1c. otkezo: etbedo: ataf's0 o'k?abono a ae yoto: bodem yo'ofos ya"? obar 3£. to'nakzon —_—talozbodan te'nofson ta'ktaboron 2m, to'nakzom 4eto:bodem ta'nofas ta'b! obor 2£. te'nakzon —taobodan to!nofson te'k?aboton lic. narozkez rato: bed na" 2of ng "b? aber SUBIUNCTIVE: M H 8 s Sing. Sm. yorkerz yotberd "yorges rik? "ber 3. tonkerz talberd "ton gos tok? "ber 2m tonkerz totberd "forges tak? "ber 2£. tonkerzi tolberd "tinges tok? "box le. otberd "L-onfes Look? "ber Du. 3m yatbodo: yoros"s0 Leth? "bero 3f. totbedot totof's0 tok? "bero ew tatbodo: taras"s0 tak? "bero le otbado: L-otes's0 Leak? "bexo mses yot'bodam yor! $08 Lik "bor 34. tot'badon tor! feson tok? 'beron 2m, tanke:zem 4ot'bodam ton fos tok? "bor 2£. tetkozon tot'bodon ton! feson tok? 'beron Lc. natkerz notberd ‘nonfos nak? "ber ‘The 2 £.s. forms with final < in MJ can be presumed to occur since the distinction between e: and : (as between o: and uz) is greatly weakened, and the mutation of e:.to i: is therefore inadequately distinctive. AAL 1, 10918 T. M, Johnstone [AAL 1/5 3.1.4. SUBJUNCTIVE IN -(o)n The verb in M'® and § has a further subj. formation which, to the best of my knowledge, is not described in the earlier literature. It is possible therefore that comparable forms do not ‘occur south of Dhofar. These forms have an ~(o)n suffix and seem to correspond formally to the MODUS ENERGICUS jussive formations in Arabic, but in M and § they occur only in sentences involving unreal conditions (for examples see Notes on syntax below). Forms of this category show in § vowel and syllabic changes, as well as suffixation of -n. In M such changes are confined to the change of final syllable -o:- to -e:- and the shortening of a long vowel on syllable closure, Thus (PERr./svBs./-n forms): TABLE 19 M tefort/yatherl/yatge:ton ‘to spit’ 8 bitz/"yokmez/yok'mizon "to jump" M se:Lom/yosto:m/yaste:man "to be safe" 8 "xelog/yxa'Lo§/yox'Ligon "to stay on after s.0.' M dot/yedle:L/yodee:ton "to know (a road)" 3 kab/'yokkab/yakbézen "to get down" M Faimo:t/yasmesn/ya%me:hon "to say" 5 “On/ya®'mer/yas'miten M zaha: §/yazho :§/yoz"hayfon "to move along sitting’ 8 e'hak? /yot'hok? /yot'hik?on "to catch up with’ M ho:xak/yaho:nok/yo'harken (2a) "to move" 8 xwrtb/yo"xorb/ya'xurban (2a) "to spoil’ M hronso:m/yo" hansom/ yo 'hansomon "to breathe’ 8 ed'hen/ "yedhan/yod'hinon "to lean on! M oxtolo: 6/yoxtolo: f/yaxtale: fan "to let s.0. down" § ax'te'Lef/yax'te'Lof/yox'te' Ligon M ' fatarat/yofte:nat/yos' tartan "to relent’ 8 "yotbor/yayte:n/yayterton ‘to gather’ (intr.: pl. forms) M $a:go:t/yake:got/yose:goton ‘to hasten’ § Sas! Ber/yo'Sa*dor/y'Sa*toron "to ask pardon’ M Somezsoh/yoine:soh/yok 'nashon "to shake hands" § 3a" fedor/yas' fedan/yod' fidaran "to outstrip’ . M ‘nk’ arbo:t? /yan'k*anbot? /yan'k*atbot’on "to hold up the hands in horror’ 8 ‘nteb! deb/yan' tobtob/yan' tobtoban'” "to drag behind" “Hl exx. are Limited to one or two forms, and I was unable to elicit further exx. (but only perf. forms) on a visit in early 1974.” It is certain however that they exist though they may be obsolescent. ‘7L€ this form is correct, then the suffix is -n, not -on. AAL 1, 101975] Modern South Arabian Languages 19 3.1.5. THE PASSIVE Passive forms occur in the MSA languages, though this is pointed out in the SAE material only for Socotri, Thus for N'® and S (PERF. /IMPERF./suBJ. Active forms are bracketed): ‘TABLE 20 M_— (xofo28/yo'n066/yors orfers/yargors/yargo:s "to kick" (fo:tay/yoto:tay/yowtery] cwtery/yauto:y/yautory "to kill" (no's? aon/yono:s?on/yors’a:n) —ans?arn/yars’orn/yans?o:n ‘to bind (hawk? a: /yahowk? a: /yo'howk?a) — awk? a: /yawe? a / yaw? a: "to put! + (hogto:t/yohogto:£/ya'haggat) —aghe:2/yagalo:£/yogot "to boil" § (no! §05/yo'nofos/"yorses) na! fis/yat'fos/yar'fos "to kick" (ler tay/ye'Lotay/"yottay) el! tiy/yal'toy/yae'toy "to kill" {e'gugen/ye'guénan/yo'gogon) _—'e'ge"fén "to trim a camel's eyelash’ (eg 'Let/yag 'Lel/'yeggot) eg'Lit/ya'ge'kol/yag'tok ‘to boil" Such forms conmonly occur in M and § only in the simple verb, though there are some examples above derived from the causative theme. 3.1.6, NON-OCCURRENCE OF t- Certain verbal forms in § and $ are characterized by the non-occurrence of the expected t- prefix in 2nd person and 3 f. forms.'® This phenomenon is confined to certain categories of fom, It occurs in both $ and § in’passive forms and in imperf, and subj. forms of causative and intensive-conative verbs. Thus from § ta'fis 'to be kicked": TABLE 21 Iupenrecr?® susguncrrve?* Sing. 3m yor! fos Yor! §o8 2m yor! §08 (£-). 24" fos 2£ yor! fis (e-)i' Ges Dual, Bf. yorga'so (t-}2%f0's0 2e yorge'so (t-)argo's0 Pl. as yor! fosan (2-]in' foson 2m yor! fos (t-) an! fos 28. yor! foson (€-)in' fosen, +H examples are not given, but occur frequently in the texts I recorded in early 1974. *90n this phenomenon in S, see also bibliography. *9Usually with the affixed particle d-, as: d-itfos etc. 215 m.sing/pl. forms often also have a prefixed £-. ALL, aL20 T. M, Johnstone [AAL 1/5 Compare also the following 3 f.s. forms as against the bracketed base-form (3 m.s. perf.): TABLE 22 5 IMPERFECT SUBJUNCTIVE (onsen) ‘meSor "e-onkor "to put, send forward’ (caus.) ("dabdat) "dakdok - "to tread’ (4-lit.) (daa) don "aden ‘to live’ (hollow) (eror) eraton feo! terin (24.5.] "to Light" (intens.-con.) Uhébos) "hubos - "to be imprisoned (pass.) 5 (eb 'né) "(tlona f-4b'ne ‘to build’ (caus. )** (tor bee) yokini'ot Le'kerbel "to crawl" (e"sutom) (te! swtmon "k-o8tom "to teach’ (intens.-con.)?? In both S and § jntensive-conative verbs show some variation in usage. It can be noted that hollow verbs in § do not exhibit this feature. 3.2. The Morphology of the Noun The substantive noun has two genders, masculine and feminine, and three numbers, singular, dual and plural. The dual is probably obsolescent except in $ hovever. FEMININE nouns, for the most part, are characterized by the endings MH o:t/-e:t, § -'ot/-'et and S -oh/-oh, or by their allonorphs. For example: TABLE 23 M — t’el(o) fost ‘cream on warm milk' nax(al£e:t ‘date palm’ Ho tot" gawe noxloleert Stel fot nox'£et Sat? 'Logoh "tom(a) Aah It may be that the -o- ending is characteristic of tri-syllabic forms, and the -e- of disyllables, but in practice it is difficult to demonstrate that this occurs consistently because of the freedom with which anaptyxis occurs at consonant juncture, simultaneously with a contrary. tendency to elide original short vowels. Thus in the above series the M forms are probably best represented phonologically as t’olo{o:t and noxZe:t respectively, though t7o8¢o:t and noxate:t also occur. ‘The sounp masculine PLURAL in MH is -é:n, in § -'dn and in S -(h)én, and also -{i}hon. The sound feminine plural in MJ is -o:tan, -'awton (less commonly -ten), S -' often ~ta); and S -(o}¢an (rarely ~itan) and -o'niton, Thus: -loti (less 221he bracketed t indicates that the form occurs with and without the personal prefix. AAL 1, 1121975] Modern South Arabian Languages a TABLE 24 (mscuLINE) + Mi gofo:n/gognizn "eyebrow" s "facor/fac'nin "bul" s "danob/'danbhén "rail! "et abon/ "ke aSnin ‘projecting rock" "sered/se'nedhon "kid" (FEMININE) : M ‘yegeno:t/yego 'nawtan ‘girl’ Yaitone:t/ a: 'terton "buttermilk: butterskin’ 7aifde:t/ 7a: "éadtan ‘upper arn’ H —_Yegono:t/yag(glo:ton ‘girl’ Yatone:t/%a'tortan ‘butter-skin' § yayb' jot/yegé'naté ‘girl’ se8i'det/*e'fodto ‘upper am" s sew! ginoh/Song'hinoten "girl, ‘*hetoh/hat'ata'nitan ‘aunt! "eeSan/to'Santon "tongue! nna himoh/ nah 'matan ‘vulture’ All the sound masculine plural affixes would seem to correlate with Ethiopic -a:n though in words borrowed from Arabic (assimilated or not) it mist be correlated with the sound plural ending of that language. ‘The DuAL in the MSA languages is formed by the affixation of ~é. However, with one or two exceptions, duals occur in M, H! and S only in association with the mumber’2, and this é is ‘thought to’be part of the numeral. Thus a Nehri understands yaygi Oro: ‘two men" as -yayg 49n0: and will syllabicate in this way. There are in M and § a few duals however which can occur without a numeral, and it is inter- esting that these, in some cases, exhibit features that occur in neither singular nor plural. Thus (sing. /dual/pl,), M yayg/"yaugé/yoyurg, S yedg/"yoti/yar, § ya'Zet/ya'joti/ *giri'. Other duals function as plurals, and are so regarded by native speakers, as, e.g fokak) ”Thalves'. (< nas’ §), & 'qubthé (< "man"; M yegge*t/yeg 'gawti/y, M ‘nas? §é. ‘As would be expected of languages of the S. Semitic group, the MSA languages have BROKEN (internal) PLURALS. Some of the plural pattems (in general terms) show more similarity to those of #thiopian languages than to those of Arabic. | Thus, M omgozomi:t/amgozo:naw ‘spine; krohor¢/k*shyor{ ‘stony ground’; aya:/ay'ayw ‘brother’; yorbor/ys'wabber ‘heavily pregnant camel"; gow{/ go "ww §z "chest" Gc. Another kind of internal plural of frequent occurrence would seem to have been developed within ‘the MSA languages. In these the i/e of the final syllable of the singular becomes o/u in the ural form. For example Mj dorhe:s? /donho:s*, S der'hes* /der'hos? ‘kid’; MI mokLeb: /mahto?b, mak'Reb/mak'tob 'unweaned camel calf": S,'*oyfeg/'oybo{ ‘kid's 'ma*rhar/'maahor, ‘denon' ; "fathi/"falho 'foal'; "k’hathan/'k’hathon ‘egg’; 'nanhen/'nanyhon ‘younger brother’, &c. AAL 1, 1132 T. M. Johnstone [AAL 1/5 3.3, Adjectives Adjectives in the MSA languages normally exhibit two genders, masculine and feminine, and two numbers, singular and plural. In $ dual forms have also been recorded but not for the other Janguages of the group. The plural of adjectives is often of common gender. There is con- siderable variation in adjective formations, but the examples listed (m.s., f.s./m.pl., £.pl. or /c.pl.) are fairly representative. ‘TABLE 25 M bo'naym, ko'naymat/icnorm, kis 'namton "generous" awborn, oWbonest/Lerbon ‘white’ u o"nom, be'nomat/hgherm, bone: "generous" ekboin, atbs 'nayt/"taben "white" § 'nistan, n'sre'nutini'stun, n's%e'niti small" lain, Litnet/Liimati ‘white’ s *beiyhan, 'kignoh/"k? éyhon, keon'heton —‘small" "tabhan, La'binah/to'bano, Labhé'naten®* ‘white! (du, m, £o'band, du. £. Lobé'naté) There are certain apparently adjectival fomations which are not inflected for gender or number, as e.g., M %etmor ‘huge’ and k’a'nayb ‘near’, S 'k’e'nib ‘near’, H{ k’o"tob ‘near’. There is a certain hazard in collecting information on $ adjectives since nominal, and verbal formations are used as adjectives, and occur side by side in the same paradigm, The occurrence of verbs in place of adjectives i$ comonly claimed as a fundamental characteristic of the Senitic languages, but js a great deal,comoner in S than in the other MSA languages. An example is: 'hiyom, ré'yemoh/'iyem, 'riyom ‘long’, which of course is 'to be long’, though positionally and functionally it does not differ from an adjective. What seems to be confined to $ is the indication of gender by vowel variation as in the examples (ms., £.s./mdu., £.du./mpl., f.pl.) in the table below. ‘TABLE 26 "zoknan, 'zoknin/zok'nani, zok'nini/'zoknon, zo'konhan "fat" "honkam, 'korbim/kar'kami, Kor'hini/kor'kamhon, fra"nakam "yellow" "Siybab, 'Siybib/Siy'babi, Siy'bibi/Siy'babhon, $a" boyob ‘old’ ‘nabthot, 'nablhit/nab'toli, nob'Liti/nob'tothon, na'bothak "ruined! 34, The Numerals ‘The numerals in the MSA languages exhibit certain interesting features. The numbers 1-10 are set out in the table below (m./£.): 281n S however many adjectives of colour have common pl. forns. AAL 1, 1141975] Modern South Arabian Languages 23 TABLE 27 M 4H 1 Past! /tayt! Pard/tiot 2 Ono: /00 "nay Goro: /B0 "not 3 Shoter0/Sa: "dayt So'layS/Sa: faye, Sa: "Oayt 4 "aanba/rabo:t o:nba/nabo:t 5 "xaymah/xammo sh *xauymoh/xemmo sh 6 hert/yote:t "hattah/yate:t 7 hho:ba/ye 'bayt ho sba/he "bay: 8 Goma :ni/Somanyirt G2mosné/Bamanest 9 ser /aat Set/sa: "ayt, S9"rayt 0 70 :Soh/a:Sone:t 20 :San/MaSoneit 5 s 1 tad /trit srad/t¥oy 2 eroh/ rat sraltrih 3 Sha'Li8/Sa'sot ‘Sike/'Sastoh 4 "onbas/Yanba"*ot "ronbo®/Yanbatah 5 xB xa "homéh/"hanoh, "hanoy 6 Bet/Btat hast, yast/"hyo?tah 7 Sor /Sba'sot "yhoo? hyob ah 8 ‘enné/Be: ‘nit ta'moni/"tomanih 9 40°/sa"ayt sa°/"se“ah 10 "°The verb forms are those (principally 3 m.s. perfect) that earlier had a final vowel which carried the stress on affixation. Similarly most prepositions earlier ended in a vowel. MAL 1, 11726 T. M. Johnstone [AAL 1/5 TABLE 32 M u 8 s sing. sn. -oh/-dah/-'ah 3 /-8 oe 3. ~ab/-das/- ~8/-e ac 2m ~ok/~éak/-usk -h/-eh -k 2. a3 /-iad/-28 -b/-8 3 le ~d/~ye/~and (ya) whlk ok du. 3c. ~hi/-dhi/-ahi ~hi/-ihi/-ohi- Bil-ebi hi ze ~ki/-dbi/-obi ~ké/-iki/-ohi Bil-ebic ahi le ~ki/-dki/-oki ~ki/~iki/-ohi Ba /-eBi aki PS. hom/=tham/-isham —-ham/tham/-ushon——-un/-ohum =(y)hon 3 san ~san/-dsen/-disan ~sen/-esan ~san 2m. izkom ~kom/-kom/-uzkam ~kwn/~okum ~kon 2 ¢. ~kan/~dben/-ithon ~kon/-dkon/-d:kon ~kon/-ekon -hon le. ~on/-dan/-den ~en/-éan/-ayn ~en/-en ~an In Socotri suffixed personal pronouns are rarely affixed to nouns and verbs. ‘Thus 'my father! is dé-ho baba 'of-me the father’, and fa:ta® tok "he has killed you’, etc. In M, Hl and § the personal suffixes are affixed to definite noms. The definite article a- is prefixed only to nouns whose initial element is voiced or glottalized consonant. A few examples will make these various points clear, Compare M bayt, "house, a’botk ‘your house"; obyu:t houses", ab'yotke 'your (s;) houses’ and ab'yotdkom 'your (pL.) houses" rogers’ the Kicked", nogéurk The kicked you", nafedshom ‘he Kicked chem’: ,¢- Yon, tusk, on you" and Léskom ‘on you', Compare also § ket ‘all,’ ket ‘all of it"; <'yet 'camels', i'ye'ted ‘his camel's <'ye"tohun Vtheir eanels'; fa:d’ (bd) ‘he shot’, Lis ‘dehun She shot then's 3.7. The Particles and Adverbs The MSA languages have in conmon the simple PREPOStTIoNs, namely b{e]- tins with’, h{e)- "to, for’, k(a)- ‘with', L(o]- ‘on, against,’ mon ‘from’, and't(o)-, the accusative matker of personal pronouns.’ Others are etymologically close, as e.g. MIS {m-)bé:n, § mun "between, ; MI raai7n, S toerA, S trae (with which compare Southern M £7a:A) ‘upon"; MM sal:]%, S $04, $ sar 'behind’; MI'onxa(:)2é, § mrin, S nhat? ‘below’; M fenozhon, gonw-, | geen, for» “ge'nc, gin-, $ "Jana ‘before’; and MJ tad, batd ‘after’. ‘This last however may be vorrowed from Arabic. Certain others do not occur in all of the languages. Thus 3S (7e)d 'to," I wle-, but unguffixed wef and M towo:Li are not comparable. , With Mi totk 'in' contrast § ak’, and with MIS ta ‘up to' contrast $ Saf, 74g. S ken ‘from’ and S fon ‘from, with"; § yal, S déok "towards' and $8 ‘an ‘from, than' do not occur in Mi § has other prepositions which do not occur (as such) in the other MSA languages, as tel 2°Qccasionally -h. 28This statement needs amplification, for which see my article listed in the bibliography. AAL 1, 1181975} Modern South Arabian Arabic 27 (M ha(:)£) Tin the company of", %er 'towards', hak*et ‘inside’, vemt (S Yamt 'direction') "towards" etc. Wi has the prepositions om-bu:n- and ma:to:d ‘after’ which do not occur in M. ‘The more important ADVERBS OF PLACE are: MHS bo(h), 5 bun ‘here’; Mi ko£ozk, § £hdn and 'Lokun, S buk ‘there’. In MH the most commonly occurring forms have a final deictic element -moh, thus M bawoh, ho'Zokomoh and H bu:mah, koLo:bamok. In & the comparable elenent is -n, these elements occur in a number of other particles. Shas also a ha particle meaning ‘here’ but this occurs most commonly in compounds, such as: %a4-ha ‘hither’ (also dé-boh); foha ‘there’, di-d-ha 'this' (see also above) and ha'to%o ‘here The -to(%0) element in the last example occurs as an element in other compound particles. ‘The common ADVERBS OF MANNER.are M we-2%0:{-molt}, Il we-t"o:-no-meh, § t'anu,S t7a(na) and *t?a 'thus'; M mezkon, H 'mokon, S 'mekon, S 'ditak’ and 'kilyho)n 'mich'; MYS 'faxoroh, S 'fahoroh "together' and Mi ke:nah, S kanak, S''fene ‘back’. ‘The 7a clement 'thus' occurs as an element in many other MS particles. The common ADVERBS OF TIME are: M S’ato:moh (viz. 4*a-no:-molt), H no:s*ora(h) (viz. no:-sozne), § na*s*anu (viz. na‘-s'a-nu) and,in some dialects na‘a, $ na‘a 'now'; M so:tor, H S*ezbor, § "dauman (from Arabic), $ 'dohor ‘always’; MA ka:wire, § 'en'fet, S gana and be? ‘before, earlier’; and M moyo:Aon, S myoto and S fot? ‘then, later’ (for which in H the prep. phrases ambu:nos’ and ma:to:des are used). The particles denoting times of the day are for the most part of nominal origin, as e.g, Mi yomo:(h), S Shen, S hex to-day"; MH yombi:, SS %om'Sén "yesterday"; Mi geshmah, SS k*a'netoh ‘tomorrow; M yobli:tok, H olli:toh, § af'*aynd "in the evening’; Mi ko-sorbeh,’$ k-'has’ag, S fo-s'abk and te-'%084,"in'the morning’; Mi yato:(h), S mon'hénom, § "Lifer ‘last night, and Mi be-hellay?, § yoso'nd ‘at night’. ‘The INTERROGATIVES of the above categories also show a fair correlation, as c.g. M Ad, H ho:nch, 5S hutun and hun, § %0% ‘where’; Mi mayt, § mit, S mé%to ‘when': M hedon, H ‘hason, § mon'ne, S 'to% ‘why’; MIS kam, S em'Se "how mich’ and M hézbo:(h), H hobo: (h), 8 yo and yot (prob. dtbott), § (%) "fo and’ %" ob "how". ‘There is relatively little correlation in the consuNCTIONS however, except where these are borrowed from Arabic. Thus M h:am, H °am, S her (and on the coast het), S ka'namme ‘if’ (real conditions); M hi:s, H %4, S has?’ and hak’ot, S fol, mo'nol, he, Sa°at’ ‘when’; but all have the common Sem. we-, though this is realised in 5 as be-. ‘The VERB MODIFIERS are of considerable interst. In M, § and S a marker, £-, is prefixed to subjunctive forms of the verb which begin with a vowel (after elision of the glottal stop, and in S also after elision of initial y). In Mehri the particle 8- and in § d- is prefixed to the imperfect when it has a present (con- tinuous) meaning, as e.g. M 3-yormer "he says, is saying’ as against yo:mot ‘he (always) says’. ‘This particle is identical in M with the relative pronoun/genitival particle. In § however both of these are now €, the particle 3 occurring only in the conditional particle a|-of] 'if (not)', Moreover the particle d- is usually prefixed to the future particle ha, as e.g. d-ha-'yoktub ‘he will write’. In Mi the future is expressed either by a verb of wishing (M 3n.s. yoko:m, Hl yexorm) or by the active participle, as for example enkotna, onkayta '(I, you c.) will cone’. In S the future is expressed by means of the auxiliary ‘aygeb 'to wish’. In S the future is expressed as shown in the preceeding paragraph, the verb being in the subjunctive. In Mj the verb may be modified by the use of the particle bex, ber whose function is similar AAL 1, 11928 T. M. Johnstone [AAL 1/5 to that of qad in Arabic. It is to be noted that § ber, S bor are, however, auxiliary verbs. Similarly the place of Mi particle %a:d 'still' yet' is’taken in S'by the ver ‘od, and in S by the verb ‘ad. Both these auxiliaries can occur together in S, as bitoh Sadoh *a{-ha ‘she has already been here." S has the interesting modifier fone ha- which, with a subjunctive verb, 'on the point of". 4. NOTES ON SYNTAX ‘The syntax of the MSA languages is well described in a fairly recent publication by E. Wagner (see bibliography) on the basis of the texts of the Siidarabische Expedition. My work up to the present has consisted of collecting in Socotra, Oman and the United Arab Emirates morpho- logical and lexical material and texts. Until the texts have been thoroughly analysed any categorical statements on syntax would be inappropriate. It would be fair to say however that while the Socotri and Nehri texts of the SAE are, broadly speaking, reliable, the sane is not true of the Sheri texts. This is even more true of the gramar by Bittner published on the basis of the SAE texts, which is quite misleading. I shall confine myself therefore to the discussion of one or two points only. Wagner (p. 127) points out that, in Harsiisi suasuncrive forms beginning with a vowel (after elision of the initial glottal stop), the subj. marker £ does not occur in the examples he is discussing. It can be confirmed that this marker occurs only rarely in H. Similarly it can be confirmed that Harsiisi normally necares with £a?(fa:), as against M of ... a(:)?, Thus Mat halk) gid ofa, HH hath) god ofa? "he is not good’. Certain kinds of coworTrowar, sentences involving unreal conditions are not attested in the M and § material collected by the SAE. The special verb forms involved are mentioned briefly above (see Table 19). The following simple examples illustrate this kind of sentence: $ d-of kun 'fogoz o?, (7]ol yos'hifon bo-yk?a'tal*)n xadi'nets to, 'If he were not lazy, he would not have stayed’ and left his work" and d-ol kun mon"Serod to, (7oL yozhtn yo b3-Sum (o)d-'Seg "If he weren't mad, he would not have come to people who were’ asleep"; M Lu(:) “oho:gas wot o:mah, f-ak™an dornal:}kal:)k tock (0)'€a? (or ... L-ankern tezk 3a), "If I had thought so, I would not have cone to you" and fu-(a]£ omzo:z 0'€a?, yak*azn Soy’donezhom 'If T did not smoke, T would have money’; H ful:) "Sant youn:zé 'mokan,'ok’a:n So:tomak ‘golem 'Tf I had a lot of money, T would buy a’pen". In Hl the perfect is often substituted for the verbal forms with suffixed -n in such sentences, and'I have no examples of -n forms except for the verb weik’a 'to be’. AAL 1, 1201975] Modern South Arabian Languages 29 SELECT BIBLIOGRAPHY ‘The publications in this field up to the end of the war were collected by Leslau in his Modern South Arabian Languages - A bibliography, The New York Public Library, NY 1946. I set out in this bibliography only the basic publications and my own publications’ all of which were after this date: BITINER, M.: Studéen zur Laut- und Formentehne der Mehri-Sprache in Siidaxabien, Vienna 1909- 1S, (KAN, Sitzungsberichte, phil.-hist. K1., Bd. 162, Abh. 5; 168, 2; 172, 5; 174, 43 176, 1; 178, 2; 178, 3). - 1 Studéen zur Shawri- Sprache in den Bergen von Dofar am persischen Heerbusen, Vienna 1216-37, (Kam ‘Sitzungsberichte, phil.-hist. KI., Bd. 179, Abh. 2; 179, 4; 179, 5; 183, 5). JAHN, Ac: Grommatik der Mehri-Sprache in Siidanabien, Vienna 1905 (KANN, Sitzungsberichte, phil.-hist. Kl., Bd. 150). JOHNSTONE, T.M.: "The non-occurrence of a ¢- prefix in certain Socotri verbal forms," B.S.0.A.8. XXX, 5, 1968. : “A definite article in the Modern South Arabian languages," B.S.0.A.S. XOTII, 2, 1970. "Dual forms in Mehri and Harsiisi," B.S.0.A.S. XXIII, 3, 1970. - : "The language of poetry in Dhofar," B.S.0.A.S. XV, 1, 1972. - : "Diminutive patterns in the Modern South Arabian languages,” J.S.S. 18, 1, 1973. - : "Folklore and folk literature in Onan and Socotra," Atabéan Studies 1, 1974. - + Harsiisé Lexicon, 0.U.P. (In the press, 1975). LESLAU, W.: Lexique Sogotté, Paris, 1938. MATTHEWS, C.J.: "Modern South Arabian determination-A clue thereto from Shari,” J.A.0.S. 89, 1, 1969. THOMAS, B.: "Four strange tongues from South Arabia-~The Hadara group," London, c. 1938 (Offprint from the Proceedings of the British Academy, xxiii, 1937). WAGNER, E.: Syntax dex Mehti~Sprache, Berlin, 1953. ‘The relevant volumes of the KAW publications of the Stidarabische Expedition are: Vol. IIT (A. JAIN), Die Nelei-Sprache in Siidarabien, Vienna 1902. Vol. IV (D.H. MULLER), Dée Meli- und Soqofré-Sprache, I, Vienna 1902 (M, texts). Vol. VI (D.H. MULLER), Pie Wehné- und Soqofri-Sprache, II, Vienna 1905 (S. texts). Vol. VII (D.H. MULLER), Die Mehri- und Soqotni-Spnache, III, Vienna 1907 (Sheri texts). Vol. IX (W, HEIN & D.H. MILLER), Mehni~ und Yadramé-Texte, Vienna 1909. AAL 1, 121SOURCES AND MONOGRAPHS ON THE ANCIENT NEAR EAST Editor: Giorgio Buccettatt ‘These two series make available orignal documents in English translation (Sources) and important studies by modern scholars (Monographs) 28a contribution to the study of history, religion, literatute, at and archaeology of the Ancient Near East Inexpensive and flexible in format, they are meant to serve the specialist by bringing within easy reach basic publications often in updated versions, to provide imaginative educational outlets for undergraduate and graduate courses, and to reach the interested segments of the educated lay audience. Bach facile i issued separately as part of a volume. One may either purchase single fascicles or subseribe to entire volumes. ‘The subscription price is $8 per volume. A volume will average 200 pages. Periodicity in the order of appearance of fascicles {snot predetermined, but a volume will normally be completed within one year. Payment must accompany orders from individuals. A handling fee of 70¢ willbe charged to Libraries if order is not prepaid. Order from: UNDENA PUBLICATIONS, P.0.Box 97, Malibu, California 90265, US. A. SOURCES FROM THE ANCIENT NEAR EAST Volume 1 Fascicle 1: R.1. Caplice, The Akkadian Namburbi Texts: An Introduction 24 pp, 804 Nomburbi ste tte of group of typical Babylonian incantations, wsed to “undo” oF ret portended evil. Consisting of bot rituals and prayers, they provide a respons to We obsred events taken by the Mesopotamian to be sgn of fture happening: Os ly are the practical correlate ofthe "omen tex,” whic Ist sh portents and ther Saniiance. The faselle by Cape Includes long introduction, which explains the ture of the texts and thelr ritual Suz om Leber, and the aniston, with note, of representative texts, ranging from a Ritual forthe Ei of a Snake Yo a Ritual i Secure Brisk Trade MONOGRAPHS ON THE ANCIENT NEAR EAST Volume 1 Fascile 1: A, Falkenstein, The Sumerian Temple City (1954) Introduction and Translation by M. del. Elis, 21 pp. 75¢ ‘An important article which fest appeared in 1954 ina French version published inthe Journal Cahiers dire Mondate. The present vetion tas been done directly from the Geran orginal, provided specially by th author forth purposes of thi seis Focusing on two Major dimensions of one of the earesthintoreal chilation, Le: the turban and the rious Simensons, the fscicle desctics tv deta, and wth reference {o primary Souess, the orgntaton of the temple ands infections withthe Polat sd socal cdr of ancient Sumer, Fascicle 2: BL Landsberger, Three Essays on the Sumerians (1943-45) Introduction and Translation by M. del. Ells, 18 pp. 70¢ ‘Tis series of artes appeared uring the Second Wold War ina Turkish journal, in 2 ual German and Turkih version~and as 2 fsa hey have not enjoyed the dilfsion hick tei imprtance would have waranted. They represent one ofthe rare attempts (Of that pant of Avjilopcal studies, Benno Landsberger, at prewnting 2 synthe of his wows om Sumerian culture; in thi sense, they consttie a cssical and éomprehensie formulation, which hs ost none of Is oral vigor and frehnes. The Wandin, edo the Gorman test, asl revs by the aor Before his death Fascicle 3: 1. M. Diakonoff, Structure of Society and State in Early Dynastie Sumer (1959) ummary and translation of selected passages by the author troduction by M. Desrochers, 16 pp SO¢ The ate peste hee apeaeé athe Engh eu fo LALDihonff's Sumer: Sue sd Sots nen Nesom, Msc 1933. Ths tosh pabieaton mares ‘nig pn in Asta fora the tegen he eon ee ak Asi to become wet know nth We toga the wy fer tae esa {sche bnteenescn snd Sit Asis tao rue ae ead ct ige enn of vt Aula fri song ing Stn rhino the Unvrty of Leni ination poids To ‘the major contributions by Diakonoff available in western languages aeASSUR [ASSUR is meant to serve the needs of the specialized feld which is closely identified with the study of Assyrian as a Gialect of Akkadian and with the history of Assyria as a special aspect of Mesopotamian civilization, from early times down to the end of the Assyrian empire. Given the intensity of linguistic and historical exchanges with neighboring regions, itis clear that the study of Assy dialect and history cannot be cartied on in isolation, without due con- sideration to influences deriving from contacts with other people. Hence, ASSUR will also accept articles which are not exclusively Assyrian in scope, as long as they are related to Assyria and useful forthe study of its language and history Editors: KH. Deller, P, Gareli, C. Saporetti (Address correspondence to Dr. C. Saporeti, Via Vasanello 20, Cassia, 00189 Roma, Italy) Volume 1 S. Parpola, The Alleged Middle/Neo-Assyrian Irregular Verb *nags ‘and the Assyrian Sound Change / > (sf rms of the alleged inegula and defective veto *nat, constructed as curing in Middle and Neo-Assyin texts are in Tact to be understood as forms ofthe ver nai-and hence “ray i 10 be stricken fom the dictionaries, The argumentation i based on fv considerations (1) Fors asigned to na and nas respectively are in perfect complementary distribution ‘mising forms of nal are covered by *nos and vce versa 2) emantclly, both verbs are ‘wed in exactly the same function. (2) The paradigm is morphologically perfect inthe seme that all forms of "ager conform to the paradigm of a ax known from the Ol! Asian ‘tod tin which no forms of *nas are attested). (4) Theres frm evidence fr the validity (ofthe change /f/> [s/n the homologies system of Neo-Asyian. (5) Wetings With Sye > and cpu sand for phonemic fa) 2nd Issue 2: C. Saporetti, Some Considerations on the Stolae of Assur ‘The publication of new tents has led to the Kentiiation of rome ofthe eponyas mentioned In the sols of Asur. An analvals ofthe data results ina negative conclusion With regard 0 the posbility of arranging the stls in groups characterized by internal chonologcl coherence. ‘The ovginal equence hax boon lot, and even within te same group there ae els dated to iparate periods, even i thy ae all Middle-Anjlan. Possibly, a subdivision may be sugested betoen the sts placed to the North, which maybe rather las, and thos placed to the ‘South, which may be dated tothe period of peatest power but thie differentiation might be scent Issue 3: P.M, Fales, Notes on Some Ninevel Horse Lists “The ate contains new documentary evidence om hors in Neo-Asyvan times, including the copies, tranltertions and Wansations of three hitherto unpublished snd four other texts {tom Nineveh. ‘The eiacussion ofthe data provides an analysis of the internal structure and the nate of the texte In One group, the horses are claslied accotding to thet coor, 2 ‘nd (possibly) age; these texts probably reproset the preliminary stig of Incoming animals ‘dawn up fr internal ase by the adminstatve unit ofthe pace in charge of hare. ‘Another prop conse of temorana on socle quantities of hora, wich ae eports ‘ent ally to the king by the same adminisvaiv unit, to provide an “ephemeral” (Le. ‘ot destined to atta ies) account of eventAFROASIATIC LINGUISTICS Volume One 1. P. Newman and R. G. Schuh, The Hausa Aspect System, 38 pp. 2 J..L. Malone, The Development of the Anomalous Syriac Verb ofkih ‘To Find’: A Case of Convergent Factors in Linguistic Change, 10 pp. 3. R. Hetaron, Bxirinsic Ordering in Classical Arabic, 25 pp. TT. Givon, Verb Complements and Relative Causes: A Diachronic Case Study in Biblical Hebrew, 22 pp. 5. T.M, Johnstone, The Modern South Arabian Languages, 29 pp. B, W. Andrzgjewski, Indicator Particles in Somali, 69 pp. 7. H. Minkoff, Graphemics and Diachrony: Some Evidence from Hebrew Cursive, 16 pp. Volume Two 1. D. R. Cohen, Subject and Object in Biblical Aramaic: A Functional Approach Based ‘on Form-Content Analysis, 23 pp. 2, C.D, Johnson, Phonological Channels in Chaha, 13 pp. R. Hetzron, The -Converb in Western Gurage (The Role of Analogy in Historical Morphology), 12 pp. 3. A. Barnea, Reference to Time, Space and Other Types of Quantfieation in the City Dialect of Gaza, 10 pp. R. Nir, The Survival of Obsolete Hebrew Words in Idiomatic Expressions, 7 pp. 4, C.T, Hodge, The Nominal Sentence in Semitic. G. Janssens, The Semitic Verbal Tense System. 5, Segert, Verbal Categories of some Northwest Semitic Languages: A Didactical Approach. 6. A.D. Corré, Waw and Digamma, A. D. Corré, A Suprasegmental Feature of Length in Semitic, 7. JL Malone, Systematic vx. Autonomous Phonemics and the Hebrew Grapheme Dagesh. T. Givén, On the Role of Perceptual Clues in Hebrew Relativization. 9. G. Buccellati, On the Akkadian “Attributive” Genitive. Descriptive flyers with complete lists of abstracts, available on request.undena publications bibliotheca mesopotamica Primary sources and interpretive analyses for the study of Mesopotamian civilization, Volume 1. Old Sumerian and Old Akkadian Texts in Philadelphia chiefly from Nippur. Part One: Literary and Lexical Texts and the Earliest Administrative Documents from Nippur. By AAGE WESTENHOLZ. xii-210 pp., 3 plates. $18.50 (cloth), $12 (paper). ‘A journal for the study of Assyrian as a dialect of assur Akkadian and of Assyria as a special aspect of Meso- potamian civilization. Editors: KH. DELLER, P. GARELLI, C. SAPORETTI. Subscription $12.50. Volume 1 includes articles by S. Parpola, C. Saporetti, M. Fales, K. Grayson, A journal devoted to theoretical articles using Afroasiatic afroasiatic linguistics material, and to descriptive, historical and comparative studies. Editor: R. HETZRON. Subscription $12.50. ‘The first two volumes include articles by P. Newman and R. Schuh, J.L. Malone, B.W. Andrzejewski, T. Givén, T.M. Johnstone, S. Segert, C.T. Hodge, D.R. Cohen, R. Nir, H. Minkoff. A. series of grammars providing concise descriptions of tie di individual languages within the Afroasiatic family, and sfromiacia: dito directed to scholars and students in the given language areas as well asin linguistics. Editors: W. LESLAU and T.G. PENCHOEN. Volume 1 (Berber). Tamazight of the Ayt Ndhir. By T.G. PENCHOEN. 124 pp. $8.50. Shorter fascicles making available original documents in sources and monographs English translation and important studies by modern on the ancient near east scholars in the field of history, religion, literature, art and archaeology of the Ancient Near East. Editor: G. BUCCELLATI. Subscription $12.50. The first two volumes include fascicles by A. Falkenstein, B. Landsberger, LM. Diakonoff, F.R. Kraus, RIL Caplice, M. Cohen, other volumes Approaches to the Study of the Ancient Near East. A Volume of Studies offered to LJ. Gelb. A collection of 27 articles on current trends and on the potential of new approaches in linguistic, literary, archaeological and historical fields. Edited by G. BUCCELLATI. 338 pp., 2 pl. (-Orientalia NS, Vol. 42, 1-2). $12.50. A Bibliography of Homeric Scholarship. Preliminary Edition 1930-1970. By DAVID W. PACKARD and TANIA MEYERS. vi-l84 pp. $2.50. Almost 4000 titles, listed alphabetically with a topical index divided into 16 major categc Professional and institutional discount of 20% on single copies (higher on larger orders). All prices are postpaid. Descriptive flyers and information on desk copies available on request. Undena Publications, P.O. Box 97, Malibu, California 90265, U.S.A.