Author: Kim, Ronald I
Date published: April 1, 2011
The Neo-Mandaic Dialect ofKhorramshahr. By CHARLES G. HABERL. Semitica Viva, vol. 45. Wiesbaden: HARRASSOWITZ VERLAG, 2009. Pp. xxxiv + 378. euro78.
Modern or Neo-Aramaic (NA) studies have seen a remarkable upsurge over the past two decades, with grammatical descriptions of previously undocumented varieties now appearing every year. Yet even specialists in this flourishing field have remained for the most part in the dark about the least known NA language, Modern or Neo-Mandaic (NM). Although Western study of the religion and language of the Mandaeans dates back to the seventeenth century, it was not until the mid-twentieth century that reliable texts in the modern language were made available to the wider scholarly public. The late Rudolf Macuch, who devoted much of his life to Mandaean studies, first published materials in the NM dialect of Ahwäz in his Handbook of Classical and Modern Mandate (1965); this was followed by two further volumes (Macuch 1989, 1993), based on fieldwork conducted in the 1950s and again in 1990.
In the present volume Charles G. Haberl offers a descriptive grammar of the NM dialect of Khorramshahr (formerly Muhammara) in KhOzestan, southern Iran, compiled on the basis of extensive fieldwork with NM-speaking immigrants to the United States, above all Mr. Nasser Sobbi of Flushing, New York. It constitutes the first book-length treatment of NM since Macuch 1993, and the first to adopt a normalized system of phonemic transcription.
The grammar follows the usual organization for NA dialect descriptions. After an introductory chapter giving the social and linguistic background of NM and its remaining speakers, and tracing the long and convoluted history of Mandaean studies in the West, successive chapters treat phonology, nouns, pronouns, and verbs. A brief conclusion is followed by ten text samples, including a speech delivered by Mr. Sobbi at the opening of the 1999 ARAM meeting at Harvard (with handwritten transcription in Mandaic script); Sobbi's account of meeting Lady Drower in the 1930s; several traditional jokes; and Sobbi's rendering of Macuch's version of the famous story of the Bridge of Shushtar in the dialect of Ahwäz, part of which was published in 1965. The volume concludes with an extremely useful lexicon, which gives full concordances with the texts, a bibliography, and an index.
Häberl has succeeded admirably in presenting a generally clear picture of the phonology, morphology, and morphosyntax of the Khorramshahr NM dialect, with frequent examples drawn from the texts. Except for certain phonetic and grammatical details and lexical items, Khorramshahr NM agrees in all respects with the dialect of Ahwäz. The Middle Aramaic (MidAr.) consonants *h and *ï have merged with *h resp. *?, as in, e.g., MidAr. \Azy 'see', \ibd 'do, make' > NM Vfey, V?£<1 The NM noun phrase has been heavily influenced by Persian, having adopted numerous loanwords, which take the Persian-origin plural suffixes -fh)a or -an (pp. 130-31); the indefinite marker -i (phrase-final, e.g., qazgän honin-i 'a small pot'; pp. 146-47); and the Persian cardinal numerals, although the inherited Mandaic numerals survive beside them (pp. 149-50; cf. V.3 emmä man vs. V.ll sad man '1OO maunds'). Gender and number agreement in adjectives has been almost entirely lost: other than the passive participles (pp. 252-54), the only exception seems to be horina Other', fern, horetta.
The verb by contrast is more conservative, notwithstanding the existence of numerous phrasal verbs with y?i?d 'do', "Wroro 'become', V?/i£> 'give', etc., most of them calqued on Persian (pp. 226-29). NM has famously preserved the MidAr. perfective (i.e., the suffixed conjugation), which otherwise survives today only in Modern West Aramaic, as well as reflexes of the reflexive formations in -t- (at least the eîppfel conjugation; pp. 221-25) and the enclitic personal pronouns in copular usage (e.g., ssbir-na ? am good'; pp. 229-30). Another archaism is that the imperfective can have past as well as present reference (pp. 242-43); in at least one instance, the imperfect is specified by prefixing the perfect of V/wy: ?.4 t&wini qsmtallen 'we were playing'. In contrast, Turoyo and the Northeast NA (NENA) dialects have generalized an invariable form of the perfect of "VAtvy- as a past-tense marker: NENA -wa-, Turoyo (Midyat) -wo- < MidAr. *hwa; Turoyo (Midan) -wa - -way- < MidAr. *hway- (Kim 2010: 232-33). These facts support the view that Mandaic, at the southeastern edge of the onetime Aramaic-speaking area, has been isolated from other NA varieties for over a millennium (pp. 11-13; cf. Kim 2008: 525-27).
However, Häberl's analyses of the data are not always persuasive, and many of the forms and patterns observed by him are open to alternative interpretations. Particularly inadequate is the treatment of the Khorramshahr NM phonemic and prosodie systems, and their relation to each other. Determining the vowel phonemes of any language is far from simple, but from listening to the digital files of Häberl's text samples (available online from the website of the Semitisches Tonarchiv at the University of Heidelberg, http://www.semarch.uni-hd.de/), I believe that Jastrow (1992) is right to argue for an underlying length difference of /i/, /e/, /u/, /ö7, /ä/ vs. /ï/ (~ /ë/), /ü/ (~ /ö/), /ä/ (so also Malone 1997: 142). Häberl's own discussion (p. 76) points to a length distinction between /ë/ and /5/ in open accented syllables (from MidAr. diphthongs) and the short vowels in cemma 'nine hundred', satonni 'they heard him', as do the basic stress placement rules (pp. 77-78), where he distinguishes between "tense" and "lax" vowels. (I assume that "long" on p. 78, 1. 8 is for "tense.") The vague statement that /a/ and /ä/ (in Häberl's notation, /o/) "often" (p. 76) or "generally" (p. 77) merge in open accented syllables also belies a regular alternation between [a:] - [D:] - [o:] in open and [a] - [se] in closed syllables, as observed by Jastrow.
As in most other NA varieties, the NM long vowels were shortened in unstressed open syllables and alternated with short vowels in closed syllables, giving rise to alternations such as sanara 'tomcat' - sanarta 'queen cat' (p. 121), gatel - gztila, gztilen 'killed' (p. 253). Allophonic fluctuation in the short(ened) vowels, in both closed and open syllables, is responsible for most inconsistencies in the transcription, e.g., welät 'country, city' vs. hisan 'horse' (p. 119), or m-Eräq 'Iraq' (11.23), muhal-ye '(that) is absurd' (V.24, X.8), belwä 'trouble' (Vn.9) vs. Iraq, mohäl, balwä in the glossary; note also the apparently random variation between 2 and e in stressed closed syllables, particularly in 3pl. imperfectives with object suffixes (see, e.g., -zl(l)- in tables 25, 3 1 , 35 vs. -elfl)- in table 40). Stress was assigned on the final or penultimate syllable of the stem, which judging from forms such as gaîélnâkon ? will kill you (pi.)' excluded the imperfective lsg. ending -nä (pace p. 78); cf. NA forms from further west such as Turoyo ko-hoze-no-lax ? (m.) see you (f.)', JiIu i-pátex-?a ? (m.) open'. The apocope of final vowels or syllables is postlexical, i.e., follows stress assignment (p. 79); hence the contextual forms qagatel, qamtallen beside qagatelna ? kill', qzmîallénni 'we play' (see chapter 5, p. 181 and passim).
In several places, Häberl seems to confuse the synchronie interpretation of the NM facts with diachronic developments. Thus, pace the discussion on p. 51, the phonemic status of /w/ and /v/ is in no way dependent on the historical distinction between II- or ??-weak roots, whose endings have been synchronically underivable since ancient times (pace pp. 73-74), and ?- or ??-v roots with ? < MidAr. -b (cf. Turoyo V/Wi 'put on' < MidAr. V/M, subj. 3sg.m. lowas). NM has no synchronie process by which *[??] becomes [??] (pp. 76-77); the forms in question rather reflect a conditioned phonetic development of MidAr. *ït, *üt > ekî, ?&?, hence MidAr. copula *ït-, *mït- 'dead' > NM ekî-, meet- (pp. 186-87; cf. the Christian NA dialects of Iranian Azerbaijan, where MidAr. *ï, *ü > [ij], [uj] (Urmia) > [iç], [??] (Salamas); Polotsky 1961 : 11-15; Jastrow 1997: 352]).The use of amer in introducing direct quotations (p. 249) can be a relic of older participial usage, rather than a living function of the subjunctive; conversely, amsabba Märe 'praised be my Lord!' (p. 254) with its D-stem participle is surely a fossilized expression from the classical language.
Among other noteworthy points, unassimilated zmhadletton (1.4), wnarlonnan (1.5) for zmhaîeîon, (3)malhnnan (pp. 82-83) may belong to a higher register suitable for Sobbi's speech at the ARAM conference; and the spirantization of b, g, d before -tar (p. 89) is simply an example of dissimilation (cf. Malone 1997: 148). Pace Häberl (p. 129 n. 151), it is simply impossible that the plural morpheme -an-, which occurs throughout Aramaic and has parallels elsewhere in Semitic (cf. Syriac rabb-an-ë 'masters', Turoyo -on-e, NENA -an-e; Moscati et al. 1964: 88; Lipiñski 1997: 239-40), could have anything to do with Middle and Modern Persian -an; the latter in any case is not "the default plural suffix" (p. 131) but, as in Persian, occurs only on nouns with human referente. Enclitic -/- (p. 157) was originally proper to indirect objects, hence its broader use with Vftnr 'say' and V?/i& 'give' (see the paradigms in table 29, p. 1 85).
The volume is beautifully produced, as are all publications in the Semitica Viva series, but contains a disappointingly high number of misprints and other errors, not all of them self -correcting. I have noted the following: p. xxv, 11. 16-17: the imperative is preserved in all NA dialects; p. 45, 1. 10: "pharyngealized voiced alveolar fricative"; p. 48, bottom: for "or ft" read "ft, or Í"; 1. 18: "a /a/, and a /o/"; p. 59 (2x): "principal"; p. 66, 1. 19: delete "the labial-velar approximant /w/"; p. 71, 1. 12 (2.186): [mo.'hom.bro]; p. 70, 1. 19 (2.196): hiele; p. 72, 1. 5: add IM; p. 74 §184.108.40.206: ie in biela 'house' is a diphthong, not two vowels in hiatus; p. 79, 11. 13-8: the initial stress on ana, ani is regular, cf. gabrâ, babà, qallä on p. 78; p. 83, 11. 20, 28: [e]; 1. 23: "advisors"; 1. 28: delete "accented"; p. 85, 1. 7: "[se] or even [e]"; 1. 3 from bottom: "fourth"; p. 94, 1. 7 from bottom: "first person singular"; p. 96, 1. 6 (2.362): mujur-na; 1. 19: "newly open syllable"; p. 97, 1. 18 (2.371): lábgasyon; p. 98, 1. 9 (2.379): ['ho.ki.(mo:]; p. 101, 1. 8 from bottom: "approximant"; p. 103, 1. 7: "q > ? / V _"; 11. 3-4 from bottom (2.428): combine; p. 105, 1. 2 from bottom (2.448): kustä; p. 107, 1. 3 from bottom (2.480): [rA.'wo:. ho]; p. 124, 1. 2 from bottom: "proselyte"; p. 136, bottom: but cf. bzdq zl=foray=i tuì=i 'he put his leg under him' (3.447, VE.2); p. 139, 1. 5: 3PL; p. 143, 11. 9-10: "city"; p. 145, 1. 14 (3.474): abad; p. 146, 1. 4: "predicative"; p. 149, 1. 21: orbin; p. 151, 1. 2: "ES.PL" is left unexplained; p. 162 n. 165, 1. 3: "dieselben"; p. 166, 1. 18: "son-RES"; p. 168, 1. 23: "pronouns"; p. 172 §4.5.2: hem 'which' is an adjective; p. 174, 1. 6 from bottom: "principal"; p. 176, 1. 2 from bottom: "perfective"; p. 180, Table 22: 2sg. perf. C1S1C2 VC3-?; p. 182, 1. 7: "second plural"; pp. 190-91: this is contradicted by Table 39 on p. 194, where both hoz-i andhozy-i 'see him!' are given; perhaps Uienszbfy)-i 'he baptized him' but ssby-a, -u 'he baptized her, them'? (cf. p. 219, 11. 2-3 with n. 181); p. 193, Table 38, Verb + C-: IpI. C1S1C2I-Hi-; p. 207, Table 58: imperfective 2sg. q-smsiel-t, 3pl. q-amsfel-enl; p. 208, Table 59: imperative 2sg. C^C21C2-;, 2pl. C,a.C2-yon; Table 60, Verb + V-: 3sg. m. C1IC^y)-; p. 209, 1. 1: C^C2-; p. 213, 1. 1: for "e.g." read "cf."; p. 221 n. 182: "cites"; p. 222 n. 183: "cites", "additional"; p. 223, 1. 2 from bottom: "break.tG.PFV"; p. 230, 1. 15: y^fahim-en; 1. 18: "predicate"; p. 233, Table 86: the lsg. with short copula should presumably be gzîel-?a, and add lsg. f. fpfetil e&-e, gatil tow-it, gatil qz-hawi-na; p. 235, 1. 4: for "All three" read "Both"; p. 241 , 11. 8-1 1 : "progressive present (§220.127.116.11)" is missing; p. 252, 1. 1 1 : "break.tG.PFV"; p. 253, 11. 9-8 from bottom: why final -n in 1 sg. copula -nani; p. 263, 11. 5-6: "if Ihad had money"; 1. 18: "noting"; (1.14) tamminr, (??.16) 3mhasselna,$. 284 n. 224: "byform." A few entries in the lexicon are mislabeled, e.g., qahäzyä and hazinu under h-z-wly (p. 323) should be "imperf." viz. "subjv."
References to incorrect or nonexistent sections, or to falsely numbered examples, may betray lastminute changes to the text: p. 50, 1. 15: §18.104.22.168; p. 59, 1. 9: §22.214.171.124; p. 63, 1. 20: §126.96.36.199; p. 64, 1. 14: §2.5.2; p. 67, 1. 5: §188.8.131.52; 1. 7: §2.5.6; 1. 19: §184.108.40.206; p. 75, 1. 4 from bottom: §220.127.116.11; p. 90, 1. 8: §18.104.22.168; p. 92, 1. 7 from bottom: §2.5.6; p. 93, bottom: §22.214.171.124; p. 115, 1. 17: "As with gatl- above" (§126.96.36.199 apparently for nonexistent §188.8.131.52.1); p. 117, 1. 21: "As with gatal- above" (§184.108.40.206 apparently for nonexistent §220.127.116.11.2); p. 120, 1. 8 from bottom: §18.104.22.168; p. 123, 1. 2: §22.214.171.124 (apparently for nonexistent §126.96.36.199.1); p. 134, bottom line: (3.398); p. 136, 1. 10: (3.404), (3.405); p. 141, 1. 13: §188.8.131.52; p. 144, 1. 20: (3.449); p. 152, 11. 17, 19, 21: (3.484), (3.485), (3.486); p. 160, 1. 10: (4.19); p. 228, 1. 18: §3.5.2. The comments on p. 202, bottom and p. 215, 11. 8-10 refer to verbal roots (resp. "Vzbn and "V/ir&) that are introduced only later on.
Häberl's use of nonstandard grammatical terminology can also be distracting, e.g., "propretonic" (p. 73), "abstraction" (pp. 80, 93, 127), "oral depletion" (for "debuccalization"; pp. 89-90), "persévérant" (pp. 98, 101), "noun substantives" and "noun adjectives" (109 n. 148); "ground stem" (i.e., Grundstamm; p. 199); "modifies" (for "governs"; p. 234, 1. 12). Nouns are not "inflected" for gender (pace pp. 109, 1. 12 and 146, 1. 6 from bottom); on p. 247, 1. 1, "syntactically" should be "semantically."
Finally, the grammatical presentation could benefit at certain points from additional cross-references and fuller explanations. Thus the vocalization of a word-final glide in contextual forms such as elli, rehhu(toelya 'where', rehwä 'wind') must be inferred on p. 70 (cf. p. 114,11. 16-18); examples (2.62), (2.129), and (2.28 1) should refer to the dissimilation rule on p. 85, and (2.74) and (2.146) to the umlaut in §184.108.40.206 (pp. 83-4); and in §220.127.116.11 (p. 84), ¿mbasqart and smzaharla should be compared with, e.g., qamhadeînâ ? speak', amhasselna ? would get', with unlowered e. The distribution of short and long copula forms is left unclear on pp. 230-31, since the rule given on p. 231, 11. 12-14 is demonstrably false: cf. ?.3 /M ganzìbr-ì=ye 'he wasn't a ganzibra'. On p. 233, table 86, the 3sg. f. gatil-i is given without explanation of the contracted ending, which comes only on p. 253. On p. 246 under §18.104.22.168-3, it might be simpler to state that the subjunctive occurs in subordinate clauses and when governed by a verb such as ?-bäyi 'want' or qz-mbasqer 'know'. The impersonal passive (pp. 254-55) could be rendered with the English equivalent, e.g., "They sent me this letter." Also missing is a list of prepositions, and especially of those that take the formant -d- before the person/number markers - all the more surprising in view of Haberl 2007, which argues for qam - qamd- 'to, for' (e.g., qamd-i 'to, for him') as the source of the obligatory -d- with loanwords.
Despite these shortcomings, there is no question that Häberl has made an invaluable contribution to NA studies with the publication of this grammar, and helped to put Neo-Mandaic firmly on the map of living Semitic languages. The tragic events of recent decades have scattered the Mandaeans, including the few remaining speakers of NM, throughout the Middle East and the rest of the world, and the prognosis for the future of the language is unfortunately very poor. Under these circumstances, the collection and analysis of data must be of the utmost priority. The present volume is a welcome step in that direction, and offers the clearest picture to date of the structure of this fascinating language.
Häberl, Charles G. 2007. The Relative Pronoun d- and the Pronominal Suffixes in Mandaic. JSS 52: 71-77.
Jastrow, Otto. 1992. Review of Macuch 1989. BSOAS55: 544-46.
_____ . 1997. The Neo-Aramaic Languages. In The Semitic Languages, ed. Robert Hetzron. Pp. 334-77. London: Routledge.
Kim, Ronald I. 2008. Stammbaum or Continuum? The Subgrouping of Modern Aramaic Dialects Reconsidered. JAOS 128: 505-31.
_____ . 2010. Towards a Historical Phonology of Modern Aramaic: The Relative Chronology of Turoyo Sound Changes. In Camsemud 2007; Proceedings of the 13th Italian Meeting of Afro-Asiatic Linguistics held in Udine, May 21st-24th, 2007, ed. Frederick Mario Pales and Giulia Francesca Grassi. Pp. 229-38. Padua: S.A.R.G.O.N. Editrice e Libreria.
Lipinski, Edward. 1997. Semitic Languages: Outlines of a Comparative Grammar. Leuven: Peelers.
Macuch, Rudolf. 1965. Handbook of Classical and Modern Mandaic. Berlin: deGruyter.
_____ . 1989. Neumandaische Chrestomathie mit grammatischer Skizze, kommentierter Übersetzung und Glossar. Wiesbaden: Harrassowitz.
_____ . 1993. Neumandäische Texte im Dialekt von Ahwäz. Wiesbaden: Harrassowitz.
Malone, Joseph L. 1997. Modern and Classical Mandaic Phonology. In Phonologies of Asia and Africa, vol. 1, ed. Alan S. Kaye. Pp. 141-59. Winona Lake, Ind.: Eisenbrauns.
Moscati, Sabatino; Anton Spitaler; Edward Ullendorff; and Wolfram von Soden. 1964. An Introduction to the Comparative Grammar of the Semitic Languages: Phonology and Morphology, ed. Sabatino Moscati. Wiesbaden: Harrassowitz.
Polotsky, H. J. 1961. Studies in Modern Syriac. JSS 6: 1-32.
RONALD I. KIM
ADAM MICKIEWICZ UNIVERSITY