KURDISH LEXICOGRAPHY A SURVEY AND DISCUSSION

MICHAEL CHYET Berkeley

The goal of a dictionary is to reflect accurately the in question as it is used by native speakers both in speech and in writing. In the case of , the northern dialect of Kurdish, the spoken language is far more developed and varied than the written language. In practice, this means that most speakers of Kurmanji are illiterate, at least in Kurdish: if they have any formal education, it is either in Turkish, Arabic, or Persian. Only the Kurdish minority in Armenia has had the opportunity to be educated in Kurmanji. The situation for , the central dialect of Kurdish, on the other hand, is more balanced in this respect: many more native speakers of this dialect are also literate in it, and have consequently incor- porated a more technical vocabulary into their everyday speech. Kurmanji lexicographers are faced with a dilemma: when dealing with a language that has yet to develop a technical vocabulary, they have the task of providing what is used by people, and on the other hand feel a duty to provide the missing technical vocabulary. Modem lexicographers strive to present a work which is descriptive, i.e., a realistic reflection of the language as it is used by native speakers. However, when it comes to technical vo- cabulary known only by a small literate intelligentsia, the imposition of such vocabulary items on a populace unfamiliar with them, in the hope that some day such terms will gain currency and be accepted by the general population, is in danger of making the dictionary into a prescriptive, rather than descriptive, work. I have been working for the past eight years on the compilation of a comprehensive Kurmanji Kurdish-English dictionary. This mammoth project includes a critical look at the various Kurdish dictionaries already in existence (Kurdish-Russian, Kurdish-Turkish, Kurd- ish-Arabic, Kurdish-French, Kurdish-German) - taking care to weed out old errors - plus ad- ditions from my own fieldwork, both abroad and with immigrant communities in this coun- try. Moreover, I have culled many words from my own of literary, folkloristic, and journalistic sources. Besides reflecting modem usage, I am attempting wherever possible to give accurate etymologies for the words in the dictionary, an undertaking which will link Kurdish with the wider field of Iranian linguistics (Iranistics). At UC Berkeley, Professor Martin Schwartz has been of invaluable assistance in this endeavor. Unfortunately, Kurdish lexicography is not a matter of simply compiling the data from all the existing dictionaries into one big lexicon. Originally, it was my goal to do just this, augmenting the result with my own field notes. I soon learned, that all too many of the existing Kurmanji dictionaries are full of inaccuracies and mistakes: by listing entries of such questionable value, one would be running the risk of giving new life to old mistakes. Moreo- ver, the published dictionaries tend to be too limited in scope. Most deal only with the vo- cabulary of one region - an aspect of advantage mainly for dialect geography. Only the newer ones have tackled modem journalistic vocabulary. Some dictionaries borrow liberally from earlier dictionaries and works, reviving old errors in the process. Although Kurdish can be written in three different alphabets - between which there is a simple one-to-one correspondence - only one dictionary (Omar) uses both Latin and Arabic script, thereby making the work easily accessible to Kurds from Turkey as well as to those from Iraq and Iran. Three aspects of Kurdish are rather unevenly treated in the dictionaries: emphatic consonants, gutturals, and aspirated consonants. Regrettably the emphatics (akin to Arabic s, d, z, t) are largely undocumented in the existing dictionaries. The older dictionaries in Arabic script hint at this feature by employing the corresponding Arabic letters occasion- ally. My informants from Iraqi Kurdistan generally have an emphatic where the old diction- aries use these Arabic emphatics. In her texts from Amadiya and Jebel Sinjar, Joyce Blau regularly indicates this feature by underscoring the consonant in question. Although this is- sue has been the subject of detailed study in Arabic, the only work I know of which deals with it in Kurdish is the dissertation of Margaret Kahn, author of The Children of the Jinn. Some work on the same feature in the neighboring Neo-Aramaic dialects has been done by the late Irene Garbell and Robert Hoberman. This seems to be an areal feature. As for the gutturals, many Kurds refuse to accept the fact that these "Arabic sounds" exist in their language, and consequently neglect to include them in their writing system. The fact is that not only the Kurdish of this region, but also the Turkish and the Neo-Aramaic dialects spoken here exhibit guttural sounds. What's more, the sounds represented by q and x are also found in Arabic, but are not rejected on the same basis. This entire argument is un- scientific at best, and is really a political statement which has no place in a scholarly discus- sion of phonetics and orthography. These guttural sounds are an integral part of the Kurdish language of today, and should be recognized as such, as they already have been by Soviet scholars, as well as by Margaret Kahn in her doctoral dissertation. Professor Otto Jastrow, a specialist in Neo-Aramaic and Arabic, believes that this feature was borrowed into Kurdish from Aramaic before the Islamic conquests brought Arabic to Kurdistan. In any case, those who write Kurdish in Latin script, with the notable exception of the Soviet- trained scholars, tend to ignore the gutturals for the purposes of writing. The Soviets write these sounds as h'- the latter borrowed from the fonner two created Those who e'-x, (x) Bedirxan, A by analogy. write in Arabic script generally have no problem writing E I in the appropriate places. There is a great deal of regional variation regarding these sounds. The word for 'forehead' is e'ni - eni - h'eni and even heni depending on where the speker hails from. Moreover, it is a shib- boleth of both Yezidi speech and the Sorani subdialect of the Arbil region to transpose t and, such that the name Haci 'Ali becomes 'Aci Hali. If these sounds did not exist in Kurdish, how could we explain this very Kurdish dialectal feature? Finally, the aspirated/ non-aspirated consonantal pairs (p'-p/ t'-t/ k'-k/ V'-V) are regularly distinguished by the Soviet scholars and in a few works by modem linguists. The Soviet scholars, many of whom also know Armenian, have no doubt been influenced by the existence of this feature in Armenian as well. This distinction is generally ignored in modem Kurdish publications, with the notable exception of Musa Anter's Kurdish-Turkish diction- ary, in which only the pair aspirated k'/non-aspirated k is distinguished - and in Baran Rizgar's Kurdish-English-Kurdish dictionary. In the Arabic script, no way has been devised to distinguish these consonantal pairs. Nevertheless, for my informants from Iraqi Kurdistan, who are most comfortable using the Arabic script, the distinction is real, and has a phonemic importance. For example, they distinguish kitik "dried figs" (with non-aspirated k) from k'itik "cat" (with aspirated k'). It should be noted that the earliest collectors of Kurdish texts, among them Oskar Mann, Albert Socin and M. Auguste Jaba, while failing to distinguish these various conso- nantal niceties, went overboard in trying to record the most infinitesimal gradation of vowel length. The same can be said for contemporary texts in Arabic, Neo-Aramaic, Turkish, and the like. Another problem with the existing dictionaries (particularly Anter, Gewrani, Toei, Maqdisi, and izoli [lst ed.]) is the plethora of misprints in them. Because of this extremely common phenomenon, it is sometimes unclear whether what appears to be a variant form is in fact a typo, or simply due to regional variation. It is, of course, beyond the scope of most dictionaries to bother about indexing typographical errors as if they were real words. One effective solution to the problems outlined above - both phonetic inexactitude and the frequent occurrence of misprints - can be suggested: reliable informants. With the recent influx of refugees from Iraqi Kurdistan, there are more native speakers of Kurdish in 110