ISBN 3-928943-96-0

Book reviews

Zülfü Selcan: Grammatik der Zaza-Sprache. Nord-Dialekt (Dersim-Dialekt). Berlin: Wissenschaft und Technik Verlag, 1998. xiii + 730 pages. ISBN 3-928943-96-0.

Zazaki (also known as Dimili) is a northwest Iranian language spoken across a large area of central Anatolia centered on the towns of Tunceli, Erzincan, and Bingol. Estimates of the number of speakers range from one to four million, making Zazaki the second largest minority language in Turkey. Despite the size of the speech community, Zazaki has until very recently been extremely poorly documented. The neglect of Zazaki is in part due to the policies of the Turkish government, which has consistently obstructed research on all its minority languages, but also to the misconception that Zazaki is a "Kurdish dialect." The present book goes a long way to putting the record straight on this and many other issues and is therefore of considerable import, for general linguists, scholars of Iranian languages, and all those interested in the minority languages of Turkey. The book under review can claim to be the first comprehensive grammar of Zazaki, though it shares this honor with another book, Paul (1998b). Fortunately Paul's book concentrates on the southern dialects of Zazaki, while Selcan's book covers the northern dialects, so the amount of overlap is limited. Furthermore, the two books embody quite different approaches: Paul's book is a classical corpus- based grammar while Selcan, a native speaker active as an author in Zazaki for 25 years, relies extensively on his own intuition. Selcan's book calls itself a "grammar of the Zaza language," but it is in fact much more than that: "compendium," or "handbook" would have been equally fitting titles. The book begins with a 100-page critical review of "all previous known research" (p. 7) on Zazaki. The reason for an overview of this depth is not simply linguistic, but in large measure political. As mentioned above, there is a widely held misconception that Zazaki is a dialect of some other language, most particularly of Kurdish.

Linguistics 39-1 (2001), 181-197 0024-3949/01/0039-0181
© Walter de Gruyter


182 Book reviews

This view has been all-too-willingly accepted by Kurdish nationalists, who have used it to justify extending their territorial claims to include the Zazaki speech zone. Selcan vehemently rejects what he refers to as the "Kurdocentric" viewpoint, and the entire 100-page section can be considered a rebuttal of that standpoint, meticulously cataloguing over a century of European (including Russian) research on the subject, before deconstructing most of the politically tainted efforts of Kurdish and Turkish writers (pp. 64-103). The sheer weight of evidence Selcan has amassed to prove that Zazaki is a language in its own right is impressive (see Paul 1998a; Gippert 2000 for further justification). But this section makes heavy reading, at times repetitive and polemic. One wonders, for example, whether the patently ridiculous attempts of one scholar to prove that Zazaki is a "Turkic language" are really worth a nine-page criticism (pp. 95-103). Although this section will leave little doubt in the reader's mind that Zazaki is not a Kurdish dialect, it is far less clear what, if any, the political implications of Zazaki's genetic affiliation should be. People's loyalties are not solely determined by the linguistic affiliation of their mother tongue. Ethnic, political, and religious factors also play a prominent role, as demonstrated by examples such as Hindi and Urdu. In the case of Zazaki, it has been noted that the religious split among the Zazas (Alevi Islam versus Sunnite Islam) is at least as important as language in shaping the Zazas' self-identification (see e.g. Firat 1997; Paul 1998b; xiii). Thus while on purely linguistic grounds Selcan is right to emphasize the independence of Zazaki, the fact remains that many Zazaki speakers do identify themselves as Kurds and have even been active in Kurdish nationalism (Van Bruinessen 1997: 209). For example, the writer of the most widely-used Zazaki dictionary (Malmisanij 1992) has no qualms about referring to Zazaki in the foreword of his dictionary as a "dialect spoken in Kurdistan." It is very significant that Selcan makes no reference to this dictionary, or its author, anywhere in his book. The simple fact is that, in spite of the linguistic evidence, some Zazas do consider them selves Kurds, a fact that deserves more mention than it gets in Selcan's book (just a footnote on page 36). These comments aside, Selcan has compiled the most comprehensive survey of the literature on Zazaki available, which will remain an invaluable source for future reference. The grammatical description begins with a detailed description of sources (see below), and of autonyms used by Zazaki speakers (a very complex issue), before moving on to a dialectological survey of Zazaki (pp. 123-136). The detailed phonology section (pp. 137-222) covers traditional segmental phonology and phonotactics, details of allophonic variation and discussion of dialectal differences, and an extensive section

Book reviews 183

on stress (pp. 192-203), as well as information on the relative frequency of individual phonemes in texts. A quirky feature of this section is Selcan's notion "minimal pair": he only accepts a contrast as phonemic if it occurs in two words which are SYNTACTICALLY AND SEMANTICALLY cornmutable. He would not, for example, consider a preposition and an imperative verb form as candidates for a minimal pair. This is an original, and in many respects perfectly logical, extension of the notion of functional phonemic contrast, though I am unaware of any theory that applies it consistently. One of the immediate results of this approach to Zazaki phonology is to reduce the number of phonemes. For example, Selcan does not count the aspirated/nonaspirated distinction in the voiceless stops as phonemic, although they are considered so by Paul for the northern dialects (Paul 1998b: 183). The bulk of the grammar is made up of the section on "Morphology and syntax" (pp. 228-696), beginning with a discussion of the classification of parts of speech based on traditional German grammar. There is no clear section on word formation or derivation (except pp. 571-575 on derived adjectives); "morphology" appears to be restricted to the expression of inflectional categories. The Izafe construction (section 17.1.2), one of the most fascinating aspects of Zazaki grammar, is for some reason treated in chapter 17, "Definiteness." Izafe is the traditional term in Iranian philology for the vocalic particle by which posthead nominal modifiers arc linked to their head nouns. In Persian, the Izafe particle is invariant; in Kurmanji Kurdish it inflects for gender and number of the head noun; but in Zazaki, it inflects for
(i) gender and number of the head;
(ii) category of the modifier (adjective vs. noun);
(iii) syntactic function of the entire NP in the clause.
Chapter 18 deals with case, including a row of suffixes/clitics somewhat confusingly called "postpositions" on page 273, but later "secondary case" (p. 291). An interesting feature of Zazaki is the importance of the feature [+animate] in the inflection of masculine nouns: in direct object function, inanimate masculine nouns take no oblique suffix, while animates do (p. 279):
(1) televe kitav ceno 'the pupil takes the book' (kitav-0, masc.)
(2) televe malim-i vineno 'the pupil sees the teacher' (malim-ob., masc.) Intriguingly, the constraint on inflecting inanimate masculines is only operative with direct objects of present tense verbs. Unfortunately Selcan refers to this particular SYNTACTIC function as obliquus, that is, the name of a particular morphological case. Later it transpires that masculine

Book reviews 184

inanimates CAN take the oblique case (e.g. as subject of ergative constructions, or as genitive attributes, cf. p. 284). This chapter also includes a more detailed discussion of ergativity in Zazaki, another typologically interesting aspect of the language. The next major division is the discussion of verbs and related topics (pp. 340-546). The Zazaki verb system is morphologically considerably more complex than that of Persian, or of Kurdish. Verbs in Zazaki have a morphological passive (restricted to transitive verbs) and in some tenses also inflect for the gender of the morphological subject (third person only). For example u manen-o 'he stays' versus a manen-a 'she stays'. Using a list of 533 basic verbs, Selcan undertakes a detailed classification of the verbs into eight conjugation classes and two transitivity classes (pp. 364-374), noting correlations between the two.

Like other Iranian languages, Zazaki also makes extensive use of so-called preverbs, particles of various provenance that modify the semantics of the basic stem. Section 21 deals with preverbs in some detail, and Selcan presents a highly original analysis whereby the preverbs are likened to vectors in a grid of spatial orientation (p. 414), expressing horizontal, vertical, and rising and falling motion. Also noteworthy is his explanation for the order of preverb relative to verb stem (they occur both before and after the verb stem). He links this to a more general principle of the Zazaki clause according to which elements that express the end result of a state of affairs are postpredicate, while those that contribute to a particular state occur before the verb (p. 433). The description of tense follows traditional German grammar, likewise the terminology. Notable is the lack of a formal category "future." Interestingly, Selcan describes tense from the point of view of entire clauses and explicitly includes temporal and modal adverbs as part of the inventory of the tense system. At this point, however, a major weakness of the book becomes apparent (see below): there are very few examples that go beyond a single sentence. Rather, Selcan presents a series of examples and then gives short rules for explaining the combinability of various tense forms with various types of adverbs, summarizing the results in tables (e.g. p. 449). This seems to me to be taking the principle of "segmentation and classification" to an extreme. A reader wishing for a clear statement regarding, say, the difference between the use of the imperfect and the preterit, perhaps illustrated with some longer text passages, will be disappointed (see for example the description of the semantics of the imperfect, pp. 458-460). Section 28 (p. 547) is headed "Adjectives." It begins with an attempt to distinguish adjectives from adverbs, focusing mainly on the lack of inflectional potential of adverbs, for example comparative and case (p. 547). However, the argument is considerably weakened when we learn

Book reviews 185

later (p. 564) that adjectives in the northern dialects do not have a special comparative form. In fact, the comparative construction conforms to hat has been suggested is an Anatolian areal type (see Haig forthcoming). It is somewhat irritating to find in this section yet another treatment of the Izafe construction under the heading "declination of the adjective" (pp. 549-552). The extensive tables here repeat information already given on pages 256-257 and 284. In fact the contents of the "Adjective" section are altogether rather surprising, including for example the extensive section on numerals and expressions for dates and times, pages 586-613. The remaining sections are "Adverbs," "Adpositions," "Negative expressions," "Terms of address," "Interjections," and an extensive section on various types of ideophone. The section on adpositions is the most disappointing part of the book. The definition of this word class ("words that express the relationship of one word to another") obviously cannot be taken literally. Closer inspection reveals that Zazaki uses a variety of structural means for expressing relationships covered in English by prepositions, but Selcan is committed to a purely linear classification based on a three-way distinction between preposition, postposition, and circumposition. He fails to make the crucial distinction between genuine basic prepositions and strongly grammaticalized Izafe constructions. Because in both cases the first element occurs before the lexical head, both wind up as "prepositions." Thus ve 'with, through, towards' is a simple preposition (ve cti 'with the stick') and is simply preposed to its noun, while seweta, glossed 'because of, is actually the first element of an Izafe construction, as in sewet-a Sileman-i 'because of Sileman', lit. 'the reason-of Sileman' (p. 649). This presentation obscures the interesting fact that Zazaki has very few simple prepositions; it would appear that at least the northern dialects are moving toward a postpositional type. Another unfortunate feature is the use of the term "circumposition," a relic of Kurdish linguistics. What Selcan refers to as circumpositions could in my opinion be better analyzed as NP/PP + the secondary case clitics -ra, -ro, -de, etc. Consider for example (p. 653) bin-e dare-ra, 'from under the tree'. Dare is 'tree' and bin is a noun meaning 'base'. Selcan analyzes this construction as a circumposition around a noun, that is, bin-e [dare] ra. But the more logical analysis is [bin-e darej-ra, that is, an Izafe construction consisting of two nouns, meaning 'the base of the tree', plus a secondary case clitic indicating 'source'. Despite the lack of analytical finesse, this section is amply illustrated so that the reader is in a position to draw her own conclusions. Much the same can be said of the section on Konjunktionen (pp. 661-676), essentially an inventory of conjunctions illustrated with extensive examples. Although the book gives an extremely comprehensive account of Zazaki grammar, the material is not always presented in reader-friendly

Book reviews186

fashion. I would estimate that 60-70% of the grammar consists of tables and lists. Likewise, the terminology is distinctly idiosyncratic. The "continuative" is hardly a "mood," and it actually looks like a cleft construction (pp. 478-480, 530-531). What Selcan refers to as a "relative pronoun" (p. 334) is in fact the pronominal head of a relative construction, as in English the one who is coming. The organization of the near 500-page section on "Morphology and syntax" is likewise confusing: it appears to be based on word classes (noun, adjective, verb, etc.), each of which gets a major chapter, but other same-level chapters are distributed among them. The result is that morphological and syntactic topics are mixed throughout the book. For example the Izafe construction is tucked away in chapter 17, "Definiteness," the syntax of relative clauses in chapter 33, "Conjunctions." In fact, the book simply lacks a roper section on syntax dealing with complex clauses, deletion rules, word-order variation, etc. Syntactic typologists will not find it easy to work with this grammar, but admittedly, Selcan did not write the grammar primarily for syntactic typologists. Finally, a critical word on the use of sources, and the objectives of the grammar. Selcan claims to have 250 hours of recorded material at his disposal (p. 118; the demographic details of the informants are listed on pp. 704-705). Yet apart from in the dialectology and phonology sections, virtually no reference is made to this corpus. Most example sentences in the grammar are not sourced, so I assume they are constructed by the author himself. Where examples are sourced, they are mostly from written sources, often texts written by Selcan himself under the pseudonym Zilfi, a fact that severely diminishes the book's value for historical and comparative purposes. More worrying is the fact that not a single extended text sample of authentic spoken Zazaki appears in the entire book. The whole grammar is based on mostly short, and presumably constructed, sentences. Typical examples are 'The child gives the book to the teacher' (p. 323), or 'The wolf eats the abandoned lamb' (p. 346). As has been pointed out many times (see e.g. Chafe 1994: 84), such sentences (e.g. with two or more definite full NPs) are extremely rare in natural discourse. The importance of basing grammars on authentic texts is being increasingly recognized, for both discourse and functionally based grammatical theories, but also as an integral part of the documentation of poorly documented languages such as Zazaki (see Himmelmann 1996 for explicit justification). Surely a book of this length could have accommodated some representative samples of actual language usage (compare the 70 pages of texts in Paul's [1998b] Zazaki grammar). The above comments on sources also raise the issue of the objectives of the grammar. For example, Selcan describes in detail how mathematical equations such as

Book reviews 187

"the square root of 25 equals five" (p. 609), or the numeral 1,002,003,000 are expressed in Zazaki. Now Zazaki has never been a language of education, so such expressions can hardly be considered part of established usage. What the author appears to be doing here is not so much describing actual usage but making recommendations of how a hypothetical "standardized Zazaki" should be. In this connection we could also note the complete absence of any reference to Turkish influence on Zazaki, although this is undoubtedly a feature of the modern language. In other words, the grammar is, at least in parts, prescriptive rather than descriptive. Writing a prescriptive grammar is of course a perfectly legitimate exercise for a native speaker, but the reader could expect a clearer statement on the aims of the grammar, and a clearer distinction between descriptive and normative sections. These critical comments are not intended to detract from my overall positive assessment of the book. Grammatik der Zaza-Sprache is a truly monumental achievement, which will undoubtedly prove an essential source for Zazaki for many years to come. Selcan has done an excellent job of compiling, analyzing, and presenting a vast amount of data, rounded off with maps and extensive indices and bibliographies. The book is attractively and professionally finished, with very few typos or other errors (and is, incidentally, very good value for money). I sincerely hope that Selcan will complement this impressive achievement with a collection of spoken Zazaki narratives in the near future. in chapter 33, "Conjunctions." In fact, the book simply lacks a proper section on syntax dealing with complex clauses, deletion rules, word-order variation, etc. Syntactic typologists will not find it easy to work with this grammar, but admittedly, Selcan did not write the grammar primarily for syntactic typologists. Finally, a critical word on the use of sources, and the objectives of the grammar. Selcan claims to have 250 hours of recorded material at his disposal (p. 118; the demographic details of the informants are listed on pp. 704-705). Yet apart from in the dialectology and phonology sections, virtually no reference is made to this corpus. Most example sentences in the grammar are not sourced, so I assume they are constructed by the author himself. Where examples are sourced, they are mostly from written sources, often texts written by Selcan himself under the pseudonym Zilfi, a fact that severely diminishes the book's value for historical and comparative purposes. More worrying is the fact that not a single extended text sample of authentic spoken Zazaki appears in the entire book. The whole grammar is based on mostly short, and presumably constructed, sentences. Typical examples are 'The child gives the book to the teacher' (p. 323), or 'The wolf eats the abandoned lamb' (p. 346). As has been pointed out many times (see e.g. Chafe 1994: 84), such sentences (e.g. with two or more definite full NPs) are extremely rare in natural discourse. The importance of basing grammars on authentic texts is being increasingly recognized, for both discourse and functionally based grammatical theories, but also as an integral part of the documentation of poorly documented languages such as Zazaki (see Himmelmann 1996 for explicit justification). Surely a book of this length could have accommodated some representative samples of actual language usage (compare the 70 pages of texts in Paul's [1998b] Zazaki grammar). The above comments on sources also raise the issue of the objectives of the grammar. For example, Selcan describes in detail how mathematical equations such as

"numeral 1,002,003,000 the square root of 25 equals five" (p. 609), or the are expressed in Zazaki. Now Zazaki has never been a language of education, so such expressions can hardly be considered part of established usage. What the author appears to be doing here is not so much describing actual usage but making recommendations of how a hypothetical "standardized Zazaki" should be. In this connection we could also note the complete absence of any reference to Turkish influence on Zazaki, although this is undoubtedly a feature of the modern language. In other words, the grammar is, at least in parts, prescriptive rather than descriptive. Writing a prescriptive grammar is of course a perfectly legitimate exercise for a native speaker, but the reader could expect a clearer statement on the aims of the grammar, and a clearer distinction between descriptive and normative sections. These critical comments are not intended to detract from my overall positive assessment of the book. Grammatik der Zaza-Sprache is a truly monumental achievement, which will undoubtedly prove an essential source for Zazaki for many years to come. Selcan has done an excellent job of compiling, analyzing, and presenting a vast amount of data, rounded off with maps and extensive indices and bibliographies. The book is attractively and professionally finished, with very few typos or other errors (and is, incidentally, very good value for money). I sincerely hope that Selcan will complement this impressive achievement with a collection of spoken Zazaki narratives in the near future.

Universitat Kiel

GEOFFREY HAIG


References

Chafe, Wallace (1994). Discourse, Consciousness and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago: University of Chicago Press.
Firat, Gulsun (1997). Soziookonomischer Wandel und ethnische Identitat in der kurdisch-alevitischen Region Dersim. Saarbrucken: Verlag fur Entwicklungspolitik.
Gippert, Jost (2000). The historical position of Zazaki revisited. Paper presented at the First International Workshop on Kurdish Linguistics, 12-14 May, Kiel.
Haig, Geoffrey (forthcoming). Linguistic diffusion in modern East Anatolia: from top to bottom. In Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics,
Alexandra Aikhenvald and R. M. W. Dixon (eds.). Oxford: Oxford University Press.
Himmelmann, Nikolaus (1996). Zum Aufbau von Sprachbeschreibungen. Linguistische Berichte 164, 315-333.
Malmisanij (1992). Zazaca-Turkce sozluk/Ferhenge Dimilki-Tirki. Istanbul: Deng.
(Orginally published 1987 in Uppsala.)
Paul, Ludwig (1998a). The position of Zazaki among West Iranian languages. In Proceedings of the 3rd European Conference of Iranian Studies, 11-15.09.1995, Cambridge,
Nicholas Sims-Williams (ed.), 163-176. Wiesbaden: Reichert.

188 Book reviews

—(1998b). Zazaki, Grammatik und Versuch einer Dialektologie. Wiesbaden: Reichert. Van Bruinessen, Martin (1997). Kurden zwischen ethnischer, religioser und regionaler Identitat. In Ethnizilat, Nationalismus, Religion und Politik in Kurdistan, Carsten Borck, Eva Savelsberg, and Siamend Hajo (eds.), 185-216. Munster: Lit.