On the chronology of sound changes in Tocharian. Volume I: From Proto-Indo-European to Proto-Tocharian by Don Ringe, Jr. (review)

REVIEWS On the chronology of sound changes in Tocharian. Volume I: From Proto-IndoEuropean to Proto-Tocharian. By Don Ringe, Jr. (American Oriental series, 80.) New Haven, CT: American Oriental Society, 1996. Pp. xxv, 203. Reviewed by Douglas Q. Adams, University ofIdaho Ringe's book is an important new survey of Tocharian historical phonology, a survey that is at the same time both ambitious and conservative. It is ambitious in that he is trying not only to explicate the phonological changes that created Proto-Tocharian from Proto-Indo-European but to put them all in their correct relative order. He really wants to write a history of Tocharian rather than assemble a list of rules that will tum Proto-Indo-European into Proto-Tocharian. (A proposed companion book would trace the history of the Tocharian languages from Proto-Tocharian down to the attested Tocharian A and B.) His work is conservative in several ways. He starts from a reconstructed PIE that is in all essentials that of Mayrhofer (Cowgill & Mayrhofer 1986) with the proviso that there were three laryngeals which he tentatively reconstructs as *x, *x, and *xw. He specifically rejects the 'glottalic' theory, but it must be admitted that Tocharian, in whose attested languages all three series ofPIE obstruents have merged as unaspirated voiceless stops, offers little succor to any particular theory of the phonetics of PIE obstruents. He is also conservative in staking out a strict neogrammarian approach with regard to sound change. At the same time he recognizes the methodological paradox that 'we cannot propose reliable Tocharian etymologies until we have discovered the sound laws, yet we can only discover the Tocharian sound laws by the analysis of reliable etymologies!' (xvi-xvii). He seeks to resolve the paradox by relying on inherent plausibility, trying to build his analysis on Tocharian words which agree closely in both meaning and shape with solidly reconstructed PIE words and which are unlikely to have undergone analogical reshaping. He seeks fewer (but better) etymologies on which to build more reliable sound laws. Inevitably not everyone will agree that R has always hit the mark in choosing which etymologies to rely on: Tocharian B kenmer 'excrement' is probably a ghostword (Hilmarsson, 1996: 130), Tocharian A s'ärme does not mean 'summer-heat' but 'winter' (Schmidt 1994:280-1). But it is unlikely that anyone else would have done any better. Considerably more problematic than choosing which etymological connections to start from is the need to work with words which have not been (potentially) subject to analogical reshaping. This is a hard requirement for an Indo-Europeanist since all PIE nouns, adjectives, and verbs come with elaborate paradigms which can cause endless restructuring. In addition, words may be in semantic sets (e.g. kinship terms, numerals) which may come to influence the shape (of some) of their members. For example, Tocharian B mScer, Tocharian A mâcar, 'mother,' certainly a word whose ultimate etymological connections Eire obvious and unassailable, may be taken as giving strong evidence that PIE *-Sgave Proto-Tocharian *-ä- (R's *-a-) or dismissed as an analogical rebuilding of the expected *mocer on the basis of the word for 'father,' Tocharian B pâcer, Tocharian A pScar. R's desire to take the straight, high road of linguistic reconstruction is both understandable and laudable. However, in reality the road is often winding and even marshy. Linguistic reconstruction remains at least as much an art as a science. The situation the Tocharianist faces in reconstructing the history of the stock can be illustrated with a couple of examples (examples which could be easily multiplied). The first example concerns the fate of PIE *-a- (*-ex-). The debate is nicely dichotomized: There are those who believe it became Tocharian -S- (R's Proto-Tocharian *-«-) and those who think it became Tocharian -o- (R's Proto-Tocharian *-o-). R is in the latter camp, I in the former. It is illustrative of the Tocharianist's situation to look briefly at the evidence for or against to see why such a divergence of opinion exists and persists. R adduces five examples of PIE *-Sgiving Proto-Tocharian *-o- (97-98): *bhrexter 'brother' (TochA pracar, TochB procer), *bhSghu- 'arm...