CC BY-NC-ND
57
Book reviews / Рецензии
Alexei Kassian Russian State University for the Humanities
The Indo-European Elements in Hurrian
by Arnaud Fournet & Allan R. Bomhard.
La Garenne Colombes / Charleston, 2010. 166 pp. W/o ISBN
0. The following review is based on two essential premises:
1) Publications dedicated to the Hurrian language are relatively scarce, meaning that any professional investigations in this field are welcome.
2) As a scientist, I suppose that any ideas have the right to life as long as their author can justify them by means of the scientific method.
1. The book under review is written by Arnaud Fournet, a French linguist specializing in comparative research on languages represented by limited textual corpora, and Allan Bomhard, an American macro-comparativist, whose papers and monographs on the Nostratic problem are well-known. As is noted in the Preface, Fournet generally takes the responsibility for Hurrian data, whereas Bomhard is responsible for IE matters.
The central statement of the authors (pp. 147-159) is that Hurrian (scil. Proto-Hurro-Urartian?) is a close relative of the Proto-IE language. F&B do not articulate it explicitly, but they probably imply that Hurrian is a member of the Nostratic macrofamily and, within this macrofamily, Hurrian is closer to IE than, e.g., Proto-Uralic. Their conception can thus be summarized in the form of the following tree:
"Proto-Asianic"
Indo-Hittite
Narrow IE
The book under review consists of several chapters. Ch. 1 is a general introduction. Ch. 2-6 (pp. 15-78) provide a grammatical sketch of Hurrian, accompanied by suggested IE etymological grammatical com-paranda and excurses into the comparative grammar of IE. Ch. 7 (pp. 79-106) contains the Hurrian vocabulary — a wordlist compiled from various sources (on which see below). Ch. 8 (pp. 107-146) is the etymological dictionary of Hurrian with IE cognates of Hurrian roots listed throughout. The last Ch. 9 (pp. 147-160) describes the grammatical features of the postulated "Proto-Asianic" language.
2. Synchronic portions. The book contains both a compilative synchronic grammar of Hurrian and a number of comparative ideas which are interspersed with each other, making the task of the reviewer somewhat difficult. In this section I provide some remarks on the synchronic portions of the book (although, running ahead, F&B's comparative studies, from my point of view, require more attention).
p. 2. The Hurrian loanword in Sumerian tabira 'metallurgy' is to be read tibira, tabira 'sculptor', scil. 'metal furniture-maker, coppersmith' (= Akk. qurqurru 'metal-worker, esp. coppersmith'), the more etymol-ogically correct variant tabira is mostly attested in personal names, see Wilhelm 1988: 50 ff.; ePSD; CDA: 291; CAD G: 137 ff.
p. 3. The suggested etymology of PN Kikkuli (fn. 4) seems totally improbable.
p. 4. The authors note that, up to now, two main sources of the Hurrian language have been the Mi-tanni letter and the Boghazkoi Bilingual. I suppose that Ugaritic vocabularies (the bilingual fragment of HAR-ra hubullu series, the quadrilingual version of Sa Vocabulary and the trilingual version of Sa Vocabulary) must be added to this list, since they provide a lot of important and partly exclusive lexicographic information. The first two of these have been taken into
Anatolian
Journal of Language Relationship • Вопросы языкового родства • 4 (2010) • Pp. 199-211
account by F&B, since these vocabularies were already known to E. Laroche and, therefore, their data were included into Laroche's Glossaire (GLH). But it should be more correct, however, to use Huehnergard's new edition of the quadrilingual Sa Voc. (Huehnergard 1987/2008) instead of the older one by Laroche. The trilingual Sa Voc. contains some Hurrian hapax legomena belonging to the basic vocabulary (e.g., 'new', 'louse', 'rain', 'smoke', 'tooth'; see the edition in Andre-Salvini & Salvini 1998; Andre-Salvini & Salvini 1999), but this text apparently remains unknown to F&B.
Generally speaking, F&B restrict themselves to two main lexicographical sources, as is noted on p. 79: Laroche's GLH (1980) and the data of Boghazkoi Bilingual from Catsanicos 1996 and Neu 1988. Other sources became mapped out; in particular, it concerns the multivolume series Corpus der Hurritischen Sprachdenkmäler, on which F&B make the following unjustified remark (p. 4): "these monographs seem to have an extremely limited diffusion in libraries".
p. 9 ff. It might be more correct to explicitly discriminate between a formal synchronic phonological description (which does indeed permit to postulate two rows of obstruents like /t/~/d/ or /t:/~/t/J) and a more sophisticated phonetico-phonological description which contains three rows (e.g., t: ~ t ~ d) and reveals some features of Proto-Hurro-Urartian phonology.
p. 10. The inventory and phonetic realisation of Hurrian labial consonants is a rather intricate question, since during the millennium that attested Hur-rian was spoken there were a number of Hurrian dialects which not only used different cuneiform orthographies, but could also undergo different phonetic changes. Maybe some of F&B's considerations about /f/ will became useful in future discussions, but I suppose that any extended discussion about Hurrian labial sounds must begin with a table like the one offered by Diakonoff (Diakonoff & Starostin 1986: 14). In any case, Wilhelm's (2004: 99) or Khacikyan's (Xa™KHH 2010b: 131) analyses seem more correct linguistically.
p. 18. "As discussed above, there is no real support for the existence of [...] /o/ in the available documentation". This is a very strange statement. First, the only remark "above" about /o/ that I have managed to find is a short paragraph at the bottom of p. 14 which does not explain the problem. Next, the second passage about /o/ is situated "below" — on p. 27. Over just a few sentences, F&B attempt to show that in the Mi-tanni letter the cuneiform sign U meant [?u], while the
1 Or even one row /t/, if one treats intervocalic geminates as biphonemic combinations /t/+/t/.
sign U was plain [u]. I am not able to fully understand their conception, because in the same paragraph the authors actually gloss Mitanni U as [?u]. The traditional view, however, is that Mitanni U covers [o], while Mitanni U = [u] & [u] (proposed already by F. Bork and E. Speiser in the 1930s). It is possible that the orthographic system of the Bo.Bil. represents the same opposition U ~ U, cf. Wilhelm 1992: 124 f., although this requires an additional investigation. In any case, I see no reasons to reject the traditional interpretation of Mitanni U as [o] and Mitanni U as [u] / [u].
Unfortunately, such criticisms can be easily multiplied.
3. Etymological portions. It is well known that Hurrian is not an isolated language, but has a close relative — the Urartian language. It is also reliably established that Urartian is not a direct descendant of Hurrian, but that these languages represent two separate branches of a common proto-language (Proto-Hurro-Urartian), see, e.g., Хачикян 2010a. One serious methodological flaw of F&B is that they exclude Urar-tian data from the comparison, do not attempt to reconstruct Proto-Hurro-Urartian forms or grammatical features, and confine themselves to Hurrian data from GLH and Bo.Bil. This stands in sharp contrast, e.g., with Diakonoff & Starostin's (1988) approach.
The second dramatic failure of the authors is related to the question: what do we mean when we say that the genetic relationship between language X and Y is "proven"? It is not always stated explicitly, but intuitively understood by professional comparativists that two languages can be considered genetically related, if there exists (1) an appreciable number of etymological matches between their basic vocabularies, and (2) an appreciable number of etymological matches between their main grammatical exponents (number, case, person), see Campbell & Poser 2008: 4; Бурлак & Старостин 2005: 7-24. Following Бурлак & Старостин 2005 (pace Campbell & Poser 2008) I believe that the former condition is strong, while the latter can serve as additional proof. If our comparison is based on cultural words (e.g. 'a k. of vessel') or roots with abstract semantics (like 'to break' or 'to swell'), it is easy to "establish" a genetic relationship between any two languages among the world with any system of phonetic correspondences that one likes.2 Bellow we
2 Cf. E. Helimski's review of Bomhard's Towards Proto-Nostrat-ic: A New Approach to Comparison of Proto-Indo-European and Proto-Afroasiatic (Helimski 1987/2000, more detailed in Xe^MMCKMM
1989/2000), where Helimski, using Bomhard's methods, has established such "regular correspondences" as IE §- ~ Semitic *b-and IE *t- ~ Semitic *b.
will see whether F&B get caught in this pitfall or not. As for grammatical exponents, such comparisons are standardly monoconsonantal; therefore, one must demonstrate some kind of isomorphisms on paradigmatic sets of grammatical exponents between two languages, since isolated monophonemic comparisons are possible between various unrelated languages.
p. 32. The adjectival suffix =and can hardly be postulated for Hurrian: pis=and= is a finite verbal stem (Wegner 2007: 186 f.), while aw=and=alli can theoretically be a Hittite-Luwian formation based on the Hur-rian root.
p. 33. The verbal suffix =Vsk is scarcely attested (in two stems only?), and its meaning and function are unknown.
-K- is, in reality, a rather common "global" diminutive morpheme attested in various families across Eurasia (e.g., in Proto-Altaic or Proto-North Caucasian).
p. 37. Additional examples for the hypothetical *Hn > Hurr. nn are needed. Hurr. =nV= is a nominal morpheme whose function has not been determined; "article" is its somewhat jargon-like designation in secondary literature (see, e.g, the discussion in Wegner 2007: 61). Its plural absolutive form is =na, not **=nna. These facts make the comparison with the IE demonstrative pronoun *no, *ne, *H\o-no (with a fictitious IE laryngeal) totally improbable. This is a significantly typical example of F&B's etymological approach; in the future, I will not comment upon every such case.
p. 38. Here for the first time, a Hurrian ghost morpheme "Accusative -n(i), -ne, -an" appears. Of course, on the following pages this "accusative" in -n will play a vital role in the comparison with Proto-IE, but I have no idea of how the author came by this enigmatic Hurrian case exponent.3
p. 41 ff. The author undertakes the comparison of two case/number systems — Hurrian and Proto-IE. Out of twelve slots only two Hurrian morphemes have reliable IE cognates: the Hurr. plural exponent =as (~ IE pl. *-s) and the Hurr. ergative exponent =s (~ IE nom. *-s). All other Hurrian case markers remain without IE etymology.
No. 1: the IE deictic stem *ol- does not have any plural semantics.
No. 4: the Hurrian accusative in -n(i) does not exist. The IE accusative ending is *-m, not **-n. The Greek and Hittite accusatives in -n are the result of the development *m# > *n# which is regular for these languages (but it is not the case of Hurrian, where final m is not prohibited). The Gothic accusative form hanan
3 Note that the "article" =nV= is not used in the absolutive singular (see, e.g., Wilhelm 2004: 107).
'rooster' is formed from a n-stem and cannot prove that the Proto-Germanic accusative ending was *-n.
No. 7: An enigmatic "Hittite enclitic particle -t/da" 'towards' is quoted, which does not exist.
No. 11: There is no absolutive marker -i with a locative meaning in Hurrian. Such locative forms from Bo.Bil. actually contain the exponent -ni which is not the "article" in the absolutive case, but the same case marker -ni as in the ablative-instrumental case (see, e.g., Wilhelm 1993: 105 ff.).
This is by no means an exhaustive list of criticisms; summing up, the Hurrian-IE nominal grammatical comparison is a complete failure.
p. 55 ff. The Hurrian-IE personal pronouns comparison is likewise unsuccessful, since the authors do not manage to quote even a single unequivocal parallel between Hurrian and Proto-IE personal pronouns.
p. 66 ff. The same concerns verbal personal exponents. No persuasive matches between Hurrian and Proto-IE.
Thus, the monophonemic grammatical comparison does not prove the Hurrian-IE relationship. Let us now look at the root etymologies (ca. 125 entries) proposed by the authors (p. 107-143). Below I will not analyze all of the etymologies. For one thing, these include connections between phantom Hurrian and phantom Proto-IE forms; e.g., Hurr. hab=an (p. 113) 'to set moving' (whose correct meaning is 'to go towards smth.') is compared with IE *H2ep- / *H2ebh- 'to move quickly, to run, to flow; (flowing or running) water, river, stream, current', but the quoted verbal meaning is unattested in IE, while the 'water' semantics has nothing to do with the Hurrian meaning. Another group consists of connections with highly unlikely semantic shifts. E.g., Hurr. na(w) (p. 121) 'to graze, feed' is compared with IE neuo- 'new'. In a third type of cases, the authors arbitrarily segment IE roots into a monoconsonantal nucleus and a "root extension". E.g., Hurr. asuhi 'pine-tree' is compared with local IE *osp- 'aspen, poplar' which is analyzed as *os(-p-) by F&B. In the fourth group I would place those etymologies whose proposed IE cognates are isolated within the IE family, i.e. attested in one subbranch only.4 E.g., Hurr. keb=an (p. 111) 'to bring' (whose correct translation is rather 'to send') is compared with Germanic *gebanan 'to give' (maybe related to Lat. ha-beo 'to hold, to have', OIr. gaibid 'to take'); Hurr. mane, =ma/=me '3rd p. sg. pronoun' is compared with some dubious Celtic m-forms with the meaning 'here'. A lot
4 See Ringe 1999 on the theoretical easiness to find isolated comparanda for any CVC-root when many languages are compared simultaneously.
of etymologies consist of forms with a vague general semantics like 'to destroy' etc.
Nevertheless, F&B do quote a few comparisons that could make an unexperienced reader believe that Hurrian is indeed somewhat related to Proto-IE: e.g., Hurr. ass 'to be seated' ~ IE *es- 'to sit'.
As I have already mentioned above, relationship between languages must be proven by root etymologies which belong to the basic vocabulary. The core of the basic vocabulary is the so-called Swadesh 100-word-list. The Swadesh list is not homogenous, but its entries possess different degrees of stability. In particular, it is possible to single out the 50 most stable words, and this shortened 50-wordlist is a very suitable tool for a quick language relationship check (see Staros-
tin G. 2010 for detail). E.g., there are 23 matches between Ancient Greek and Old Indic in the 50-wordlist (ooreov ~ asth- 'bone', kuwv ~ svan- 'dog' and so forth). Correspondingly, the Proto-IE ~ Proto-Uralic comparison yields from 12 to 14 matches on the same list (*ued— *wete 'water', *l/nomen ~ *l/nime 'name' and so on), depending on acceptance/rejection of a couple questionable etymologies.5
Let us now look at the Hurrian 50-item wordlist compiled by the reviewer (for the general principles of the compilation process now see Kassian et al. 2010). Index figures that accompany the English entries designate the relative index of stability as calculated by S. Starostin (Старостин 2007a; Starostin G. 2010; 'wei' is the most stable word, 'night50' — the least stable one).
Meaning Hurrian word Textual comments
ashes38 salmi Ugar.C. (Huehnergard 1987/2008: 378). Cf. the verbal form salm=ul=ust=ed=a (Msk.) 'to burn(?)' (Andre-Salvini & Salvini 1998: 15).
bird33 eradi Ugar.C., Bogh. (Wegner 2007; Haas & Wegner 2004: 341; Andre-Salvini & Salvini 1998).
black48 timeri Bo.Bil. (Catsanicos 1996: 227). Attested in the collocation timeri ese = Hitt. tankui tegan 'the Underworld', lit. 'black earth' (not 'dark earth', since Hitt. tankui = Sum. GIGGI/GE6 = Akkad. salmu mean 'black').
blood20 zurgi Msk., Bogh. (GLH: 309; Wegner 2007).
bone34 —
dieis — It is possible that ull=ul (Bogh.) means 'to die', an intransitive(?) stem of ull- 'zerstören'. Cf. GLH: 279; Salvini & Wegner 2004 (ChS 1/6): 42, 184.
dogi6 erwi, erbi Bo.Bil., Bogh. Ugar.C. (Catsanicos 1996; GLH: 83; Andre-Salvini & Salvini 1998)
drink15 —
dry24 — Cf. the participle sib=a (Bo.Bil.) which theoretically may mean 'dried-up, dried-out', Catsanicos 1996: 275; Neu 1996 (StBoT 32): 316-319.
ear32 nui, nuwi (probably also nuhi) Bo.Bil. Bogh. (Catsanicos 1996: 273; Wegner 1995: 121-2). The Hurr. word leli- means 'a k. of beverage (vel sim.)' (not 'ear'!) and does not seem inherited in view of the initial l-.
eat25 ul ul=an=o=m '(the dog) began to eat (the bread)' (Bo.Bil., Catsanicos 1996: 216). The Hurr. verb refers to a dog and could therefore mean 'to devour' rather than the neutral 'to eat' (of humans). But since it is translated by the basic Hitt. verb ed- 'to eat (of humans)' rather than by Hitt. karap- 'to eat (of animals)', I tentatively consider ul- to be the default Hurr. verb 'to eat'.
egg47 —
eye4 si Bo.Bil., Bogh. Attested with the "article": st=na (Catsanicos 1996: 270; Wegner 1995: 124-5). The previous candidate for 'eye' pasi- actually denotes 'mouth'. The noun furi (<fur- 'to see, to look') rather means something like 'look', not 'eye' per se (Wegner 2007).
fire7 tari Bo.Bil., Mit., Bogh., Ugar.C. (Catsanicos 1996: 231; GLH: 257; Huehnergard 1987/2008: 378; Andre-Salvini & Salvini 1998).
foot43 uri, ur=ni 'foot, leg, thigh' Bo.Bil., Bogh., Ugar.C. (Wegner 2007; Huehnergard 1987/2008: 379; Andre-Salvini & Salvini 1998). The similar noun ugri- (Bogh.) is attested in the meaning 'leg of a table' (GLH: 277).
hair27 —
5 Calculated by G. Starostin, pers. comm., as part of our ongoing research on the Preliminary Lexicostatistical Tree of the world's languages (within the "Evolution of Human Language" project, supported by the Santa Fe Institute)
Meaning Hurrian word Textual comments
handu suni, summi suni — Bo.Bil. (Wegner 2007: 219; Catsanicos 1996: 200), summi — Bogh., Ugar.C. (GLH: 242). The stem summi might be derived from suni with the nominal suffix -me (on which see Wegner 2007: 59), if mm < nm.
head49 pahi Bogh., Mit., Mari, Ugar.C., Ugar.A. (GLH: 192; Andre-Salvini & Salvini 1998).
hear45 has Tisatal, Bo.Bil., Bogh., Mit., Ugar. (Wegner 2007; GLH: 95; Catsanicos 1996: 267). Another verb — salh (Mari, Bo.Bil., Bogh.) — seems to have the meaning 'to listen (to), hearken' (Wegner 2007; Catsanicos 1996: 267).
hearti4 tisa Bo.Bil., Bogh., Mit., Ugar.C. (GLH: 266; Catsanicos 1996: 199).
horn44 —
I3 is= [abs., erg.], so= / su= [obl.] Wegner 2007: 82.
kill42 — Cf. uw 'to slaughter (ox etc.)', sur 'to slaughter (small cattle)' (Catsanicos 1996: 234).
leaf41 —
lousei7 aphe Ugar.C. (Andre-Salvini & Salvini 1998; Andre-Salvini & Salvini 1999).
meat46 uzi Bogh., Ugar.C. (GLH: 291).
moonis kusuh Name of the Moon god. Mari, Emar, Nuzi, Msk., Ugar., Bogh. (GLH: 156).
mouth31 pasi, fasi Mari, Bogh., Ugar.C., Ugar.A. (Wegner 2007; GLH: 197; Andre-Salvini & Salvini 1998).
nail19 —
name10 —
new23 suhe Ugar.C. (Andre-Salvini & Salvini 1998; Huehnergard 1987/2008: 378).
night50 —
nose29 punhi, puhhi Bogh. (punhi), Ugar.C. (puhhi), see Wegner 1995: 123-4; Andre-Salvini & Salvini 1998.
notsö =u (also =wa, =ud, =wur) The basic Hurr. negation morpheme is u which is attested as =u in ergative indicative forms of the 1st & 2nd p., =ud in ergative indicative forms of the 3rd p. (Bo.Bil., probably Ugar.C.), =wur in the stem mann- 'to be' (Bo.Bil. only), =wa in non-indicative (desiderative) forms of the 1st & 3rd p. Although in the 3rd p. u-based morphemes began to be superseded by the old prohibitive particle ma already in Bo.Bil., Bo.Bil. still retains ergative forms in =ud. The antiquity of the morpheme u is proved by Urartian data, where ui and =ure are negative exponents in transitive indicative forms (see Yakubovich forthc.).
not30 =kkV =kkV is the basic Hurr. negative exponent in non-ergative (i.e. intrans. & antipass.) indicative forms. This morpheme seems lost in the Urartian branch.
one21 sukko Bo.Bil., Mit., Bogh. (Wegner 2007: 81).
rain39 isena Ugar.C. (Andre-Salvini & Salvini 1998).
smoke36 hiuri Ugar.C. (Andre-Salvini & Salvini 1998). The well attested form hubr=ushi 'incense burner' may be derived from the same root, although the vowel alternation is not clear.
star40 — The Hurr. word zuzuhe (Ugar.C.) means something like 'container, box, vessel' rather than 'star' (see Huehnergard 1987/2008: 26, 54, 377) and seems to be a variant of Hurr. zizzuhi 'cruche'.
stone9 —
sun35 simigi Name of the Sun god. Msk., Mit., Ugar., Bogh. (GLH: 232).
tail26 —
thou5 fe= Wegner 2007: 82.
tongues irde Bogh. (Wegner 1995: 117-8).
tooth22 sini=ni Ugar.C. (Andre-Salvini & Salvini 1998: 16-17).
tree37 tali 'tree; wood' Bo.Bil., Msk., Ugar.C., Bogh. (GLH: 253; Catsanicos 1996).
two2 sin(i) Mari, Mit., Bogh., Ugar.A. (Wegner 2007: 81; GLH: 234).
Meaning Hurrian word Textual comments
water28 siwe [archaic?], siye Mari (siwe), Bo.Bil., Ugar.A., Ugar.C. (siye), Bogh. (siye, siwe) (GLH: 230; Wegner 2007; Catsanicos 1996).
we1 sa(=tti=lla) [abs.], sie(=s=) [erg.], sa(=s=) [obl.] Wegner 2007: 83.
what12 — Pace Хачикян 2010b: 138, the Hurrian interrogative stem iya (Bo.Bil., Mit.) is not attested in the meaning 'what?'.
who6 abe, aw= Bo.Bil., Mit. (Wegner 2007: 85; Хачикян 2010b: 138).
Only six items on this list are etymologized by F&B. Five of these etymologies are rather weak and feeble:
1) is- [abs., erg.], so-/su- [obl.] 'I' (p. 116) ~ IE *eg- / *egh- 'I'. Phonetically and morphologically unconvincing.
2) nui 'ear' (p. 123) ~ IE *neuks- 'to smell (trans.)', attested in Slav. and Germ. only.6
3) tali 'tree; wood' (p. 130-1) ~ IE *dhal- 'to bloom; to be leafy' as reconstructed in Pok. 234. It is very likely, however, that this reconstruction covers a whole number of etymologically unrelated roots with different vocalism and meanings in several IE subbranches. Celt. *dol-V- 'leaf' (Matasovic 2009: 102 f.), Celt. *dal-n- 'come into being, turn into something' (Matasovic 2009: 88 f.). Germ. *deljaz 'dill', OHG toldo 'foliage' (Orel 2003: 70). Greek 8dAAw, whose primary meaning may be something like 'to spring forth with or from moisture' (Lowenstam 1979). Arm. dalar 'green, fresh' (Martirosyan 2010: 231). Alb. dal 'to go out' (Orel 1998: 54). It is possible that some of these roots could eventually be etymologically united, but I am not sure that a uniform root *dhVl- with the general semantics of 'plant' is truly reconstructible for the Proto-IE level.
4) timeri 'black' (p. 133-4) ~ IE dhem- 'dark', attested in Celtic (Middle Irish) and Germanic only, see Matasovic 2009: 95 (F&B quote MIr. deim 'dim, dark' as OIr. dem 'black, dark'). Can hardly be projected onto the Proto-IE level.
5) zurgi 'blood' (p. 143) ~ IE *dhor- 'to gush forth, to burst forth, to spurt'. Semantically unconvincing.
The sixth etymology could be accepted as an additional comparison between languages whose genetic relationship has already been established:
6) siwe, siye 'water' (p. 128-9). Several roots of the shape sVu- with a general semantics of 'moisture' can indeed be found within IE languages
6 The reconstruction of the variant *neus- for some Germanic forms is probably unnecessary.
(although F&B merge all of the IE forms in one unified *sew-/*sow-/*su-): IE *suue- 'to rain' (Toch., Greek, Alb.)7 ~ IE *srn- 'juice; sea' (Balt. & Germ. only) ~ IE *sau- (~ -o-) 'juice; to squeeze the juice' (Toch., In.-Ir., Germ., Celt.), see Piet.dbf.
Not a single entry on the Hurrian 50-item wordlist has a direct semantic match with the corresponding entry for Proto-IE. Therefore I must conclude that the Hur-rian-IE root comparison also fails.
4. Conclusions. Fournet and Bomhard have not managed to demonstrate the relationship between Hurrian and IE. I suppose that it should be clear to everybody who is familiar with Hurro-Urartian and IE languages that these two families are genetically unrelated. The philological, esp. Hurritological level of the book under review is also not particularly high.
The genealogical attribution of Hurro-Urartian remains an open question. The most natural assumption, in view of the geographical distribution and typological similarity, would be to include HU into the East Caucasian (Nakh-Dagestanian) stock of the North Caucasian linguistic family. This was originally proposed as early as the second half of the 19th c.; more recently, this idea was further developed by some Soviet authors — J. Braun, G. Klimov, Ju. Deseriev and I. Diakonoff (see, e.g., Diakonoff 1971: 161 ff.; Дьяконов 1978), after which the monograph Hurro-Urartian as an Eastern Caucasian Language by Diakonoff and S. Starostin appeared (Diakonoff & Starostin 1986). In the last few decades, however, there has been some very serious progress in the North Caucasian linguistics as compared to the 1980s. The fundamental reconstruction of the Proto-North Caucasian language by S. Nikolaev and S. Starostin was completed and published as NCED (1994). Comparable progress was made in the reconstructions of the Proto-Yeniseian8
7 Or IE *suH-/sHu-, if the Hitt. stems suhha-, ishuwa- 'to throw, strew, pour out' are indeed related.
8 See Старостин 1982/2007 and Yenet.dbf, based on Старостин 1995 and Werner VWJS with additions and corrections.
and Proto-Sino-Tibetan9 languages. Finally, the theory of a Sino-Caucasian linguistic macrofamily, which unites North Caucasian, Basque, Yeniseian, Bu-rushaski and Sino-Tibetan families, has partially been substantiated.10
It is currently obvious that HU can hardly be a member of the East Caucasian stock. On the other hand, there is a number of rather striking isoglosses between HU and Sino-Caucasian.11 Thus, it is very likely that HU is an extinct member of the Sino-Caucasian macrofamily.12 It is interesting that, lexi-costatistically, HU seems to be closer to the Yeniseian family rather than to the North Caucasian or Sino-Tibetan ones. A similar situation is observed with another Asia Minor extinct language — Hattic, which also possesses some exclusive isoglosses with Proto-Yeniseian (see Kassian forthc.), although there is no specific visible relationship between HU and Hattic.
In any case, further etymological studies of HU will be possible only after a Hurro-Urartian dictionary has been compiled and a Hurro-Urartian reconstruction (based on such a dictionary) is accomplished. As of
9 See Stibet.dbf, based on Peiros & Starostin 1996, but seriously improved.
10 For the comparative phonetics of the Sino-Caucasian macrofamily see S. Starostin's SCC (this work was not finished and therefore remains unpublished). The highly preliminary Sino-Caucasian etymological dictionary by S. Starostin is available as Sccet.dbf. Some other papers by the same author, dedicated to the Sino-Caucasian problem, can be found in Старостин 2007 (both in Russian and English).
11 Cf., e.g., the following entries of the 50-wordlist (SCauc. forms are quoted after the Tower of Babel project databases):
Hurr. is- [abs., erg.], so-/su- [obl.] 'I' (1st p. sg. pronoun) ~ SCauc. *zV 'I' > NCauc. *zo, erg. *?ez(V), gen. *?iz(V), obl. *za- 'I', Yen. *?aj 'I', Burush. *ja 'I' (see the discussion in Caucet.dbf sub *zo about the distribution of stem variants within the paradigm).
Hurr. fe- (< *we-) 'thou' (2nd p. sg. pronoun) ~ SCauc. *wV 'thou' > NCauc. *uo 'thou' (the direct stem), Yen. *2aw (/*?u) 'thou', Burush. *u-n 'thou'.
Hurr. time=ri 'black' ~ SCauc. *rVmV 'black, dark' > STib. *rVm 'dark' ~ Yen. *tum- 'black'.
Hurr. zur=gi 'blood' ~ SCauc. *catwV > NCauc. *çatwV 'vital force (vel sim.)', Yen. *sur 'blood', Basque *i-serdi 'sweat; sap'
Hurr. aphe 'louse' ~ SCauc. *?amkV 'louse, small insect' > NCauc. *2ëmkV 'a k. of insect, vermin, worm', STib. *moy / *mok (~ -Й-) 'midge, moth', Yen. *2Ske (~ x-) 'louse'. The comparison seems very likely, although the Hurrian denasalization mk > pk requires additional examples. Note that the cluster mK seems synchronically unattested in proper Hurrian words. An interesting parallel is the name of the Sumerian goddess Ddamkina which can sometimes be spelled as Dtap-ki-in-na in Hurrian contexts (GLH: 70; van Gessel 1998 1: 438). An assumption of further fricativization pk > ph also looks natural, since the cluster pk is probably unattested in proper Hurrian words.
12 At the beginning of the 2000s, S. Starostin himself tended to lean towards the same conclusion.
today, we do not even possess synchronic Hurrian and Urartian thesauri.
Literature
André-Salvini & Salvini 1998 — B. Andre-Salvini & M. Salvini. Un nouveau vocabulaire trilingue sumérien-akkadien-hour-rite de Ras Shamra // Studies in the Culture and Civilization of Nuzi and the Hurrians 9 (1998). P. 3-31.
André-Salvini & Salvini 1999 — B. Andre-Salvini & M. Salvini. La colonne I du vocabulaire Sa trilingue RS 94-2939 // Studi Micenei ed Egeo-Anatolici 41/1 (1999). P. 145-148.
CAD — The Assyrian Dictionary of the Oriental Institute of the University of Chicago. Chicago, 1956 — .
Campbell & Poser 2008 — L. Campbell & W. J. Poser. Language Classification: History and Method. Cambridge University Press, 2008.
Catsanicos 1996 — J. Catsanicos. L'apport de la bilingue de Hattusa à la lexicologie hourrite // Amurru 1: Mari, Ébla et les hourrites. Paris, 1996. P. 197-296.
Caucet.dbf — North Caucasian etymological database by S. L. Niko-layev & S. A. Starostin (= NCED) // Available online at Tower of Babel Project.
CDA — J. Black, A. George, N. Postgate. A Concise Dictionary of Akkadian. 2nd ed. Wiesbaden, 2000.
Diakonoff 1971 — I. M. Diakonoff. Hurrisch und Urartäisch (Münchener Studien zur Sprachwissenschaft, Beiheft, N.F., 6). München, 1971.
Diakonoff & Starostin 1986 — I. M. Diakonoff, S. A. Starostin. Hurro-Urartian as an Eastern Caucasian Language (Münchener Studien zur Sprachwissenschaft, Beiheft, N.F., 12). München, 1986.
ePSD — Electronic Pennsylvania Sumerian Dictionary Project (http://psd.museum.upenn.edu/epsd/index.html).
GLH — E. Laroche. Glossaire de la langue hourrite. Paris, 1980.
Haas & Wegner 2004 — V. Haas & I. Wegner. Das Gegenwortpaar 'wahr' und 'falsch' im Hurritischen // D. Groddek & S. Rößle (eds). Sarnikzel. Hethitologische Studien zum Gedenken an Emil Orgetorix Forrer. Dresden, 2004. P. 339-344.
Helimski 1987/2000 — E. Helimski. A "new approach" to Nostratic comparison // Е. А. Хелимский. Компаративистика, уралистика: Лекции и статьи. Moscow, 2000. P. 476480 [First publ.: Journal of the American Oriental Society 107/1 (1987). P. 97-100.]
Huehnergard 1987/2008 — J. Huehnergard. Ugaritic Vocabulary in Syllabic Transcription (HSS 32). 1st ed.: Atlanta, 1987; 2nd repr. ed.: Winona Lake, 2008, with addenda on pp. 375-406.
Kassian forthc. — A. Kassian. Hattic as a Sino-Caucasian language // Ugarit-Forschungen. Forthcoming.
Kassian et al. 2010 — A. Kassian, G. Starostin, A. Dybo, V. C hernov. The Swadesh wordlist. An attempt at semantic specification // Journal of Language Relationship, No. 4 (2010) [the present volume]. P. 46-89.
Lowenstam 1979 — S. Lowenstam. The meaning of IE *dhal- // Transactions of the American Philological Association 109 (1979). P. 125-135.
Martirosyan 2010 — H. K. Martirosyan. Etymological dictionary of the Armenian inherited lexicon. Brill, 2010.
Matasovic 2009 — R. Matasovic Etymological Dictionary of Proto-Celtic. Brill, 2009.
NCED — S. L. Nikolayev, S. A. Starostin. A North Caucasian Etymological Dictionary. Moscow, 1994 [reprinted: 3 vols. Ann Arbor: Caravan Books, 2007.]. Available online as Caucet.dbf.
Neu 1988 — E. Neu. Das Hurritische. Ein altorientalische Sprache in neuem Licht. (AAWLM 1988 no. 3). Mainz/Stuttgart, 1988.
Neu 1996 — E. Neu. Das hurritische Epos der Freilassung I: Untersuchungen zu einem hurritisch-hethitischen Textensemble aus Hat-tusa (Studien zu den Bogazköy Texten 32). Wiesbaden, 1996.
Orel 1998 — Vl. Orel. Albanian etymological dictionary. Brill, 1998.
Orel 2003 — Vl. Orel. A handbook of Germanic etymology. Brill, 2003.
Peiros & Starostin 1996 — I. I. Peiros, S. A. Starostin. A Comparative Vocabulary of Five Sino-Tibetan Languages. 6 vols. Melbourne, 1996.
Piet.dbf — Indo-European etymological database by Sergei Nikolayev // Available online at Tower of Babel Project.
Pok. — J. Pokorny. Indogermanisches etymologisches Wörterbuch. Bern, 1959.
Ringe 1999 — D. Ringe. How hard is it to match CVC-roots? // Transactions of the Philological Society 97/2 (1999). P. 213-244.
Salvini & Wegner 2004 — M. Salvini & I. Wegner. Die mythologischen Texte (ChS 1/6). Roma, 2004.
SCC — S. A. Starostin. Sino-Caucasian. Unfinished MS, the middle of the 2000s. Available online at Tower of Babel Project.
Sccet.dbf — Sino-Caucasian etymological database by S. A. Starostin // Available online at Tower of Babel Project.
Starostin G. 2010 — G. S. Starostin. Preliminary lexicostatistics as a basis for language classification: A new approach // Journal of Language Relationship, No. 3 (2010). P. 79-116.
Stibet.dbf — Sino-Tibetan etymological database by S. A. Starostin (= Peiros & Starostin 1996, but with serious improvement) // Available online at Tower of Babel Project.
Tower of Babel project — Etymological database project "The Tower of Babel", headed by S. A. Starostin. In affiliation with the "Evolution of Human Languages" (EHL) Project at the Santa Fe Institute. Available online: http://starling.rinet.ru/.
van Gessel 1998 — B. van Gessel. The Onomasticon of the Hittite Pantheon (HbOr I/33). Leiden, 1998-2001.
Wegner 1995 — I. Wegner. Die hurritischen Körperteilbezeichnungen // Zeitschrift für Assyriologie 85 (1995). P. 116-126.
Wegner 2007 — I. Wegner. Hurritisch. Eine Einführung. 2nd rev. ed. Wiesbaden, 2007.
Werner VWJS — H. Werner. Vergleichendes Wörterbuch der Jenis-sej-Sprachen. 3 vols. Wiesbaden, 2002.
Wilhelm 1988 — G. Wilhelm. Gedanken zur Frühgeschichte der Hurriter und zum hurritisch-urartäischen Sprachvergleich //
V. Haas (ed.). Hurriter und Hurritisch (Xenia, 21). Konstanz, 1988. P. 43-67.
Wilhelm 1992 — G. Wilhelm. Hurritische Lexikographie und Grammatik. Die hurritisch-hethitische Biingue aus Bogazköy // Orientalia NS 61 (1992). P. 122-141.
Wilhelm 1993 — G. Wilhelm. Zur Grammatik und Lexicon des Hurritischen // Zeitschrift für Assyriologie 83 (1993). P. 99-118.
Wilhelm 2004 — G. Wilhelm. Hurrian // R. D. Woodard (ed.). The Cambridge Encyclopedia of the World's Ancient Languages. Cambridge: Cambridge University Press, 2004. P. 95-118.
Yakubovich forthc. — I. Yakubovich. Morphological Negation in Urartian // Aramazd. Armenian Journal of Near Eastern Studies. Forthcoming.
Yenet.dbf — Yenisseian etymological database by S. A. Starostin (= Старостин 1995 & Werner VWJS with additions and corrections) // Available online at Tower of Babel Project.
Бурлак & Старостин 2005 — С. А. Бурлак, С. А. Старостин. Сравнительно-историческое языкознание. 2-е изд. М., 2005.
Дьяконов 1978 — И. М. Дьяконов. Хуррито-урартский и вос-точнокавказские языки // Древний Восток 3. Ереван, 1978. С. 25-38.
Старостин 1982/2007 — С. А. Старостин. Праенисейская реконструкция и внешние связи енисейских языков // Старостин 2007. С. 147-246. [First publ.: Кетский сборник. Л., 1982. С. 144-237]
Старостин 1995 — С. А. Старостин. Сравнительный словарь енисейских языков // Кетский сборник (Studia Ketica). Вып. 4. М., 1995. С. 176-315.
Старостин 2007 — С. А. Старостин. Труды по языкознанию [Works in Linguistics]. М., 2007.
Старостин 2007a — С. А. Старостин. Определение устойчивости базисной лексики [Defining the Stability of Basic Lexicon] // Старостин 2007. С. 827-839.
Хачикян 2010a — М. Л. Хачикян. Хуррито-урартские языки // Языки мира. Древние реликтовые языки Передней Азии. М.: РАН, Институт языкознания, 2010. С. 118-125.
Хачикян 2010b — М. Л. Хачикян. Хурритский язык // Языки мира. Древние реликтовые языки Передней Азии. М.: РАН, Институт языкознания, 2010. С. 126-148.
Хелимский 1989/2000 — Е. А. Хелимский. К оценке надежности индоевропейско-семитских лексических сопоставлений // Е. А. Хелимский. Компаративистика, уралистика: Лекции и статьи. М., 2000. P. 481-486 [First publ.: Палеобал-канистика и античность. М., 1989. P. 13-20.]
