A Dictionary of Tocharian B.

• Author: Winter, Werner
• Position: Book Review

By DOUGLAS Q. ADAMS. Leiden Studies in Indo-European, vol. 10. Amsterdam: RODOPI, 1999. Pp. xxxiv + 830. $190.

In the book under review, the author sets out to do two things: he wants to present the vocabulary of Tocharian B as found in the surviving documents, including information about paradigmatic properties of the forms enumerated and comments on the meaning of the lexical entries (supported by evidence from bilingual material and/or chunks of text quoted and translated), as well as an inventory of etymological proposals made in the past, supplemented by novel interpretations of his own. I have discussed the first component in some detail elsewhere (Central Asiatic Journal 45 [2001]: 129-38), pointing out, inter alia, the embarrassingly high number of technical and substantial errors on the part of Adams, which greatly reduces the value of the Dictionary as a reference work; in addition, mention has to be made of the author's failure to pay systematic attention to dialectal variation and to differences between forms in prose and metrical contexts, respectively, which leads to inadequacies in the establishment of lemmas and of normalized forms within paradigms listed. As my review in CAJ provides ample documentation, I can largely disregard the descriptive component of Adams's work in the comments I want to make here.

Prior to this volume, three book-length publications intending to serve as etymological dictionaries of the Tocharian languages had appeared in print, viz., Van Windekens 1941, Van Windekens 1976, and Hilmarsson 1996. Of these, the first was based on incomplete and by now entirely outdated material; the second suffers greatly from its author's lack of philological competence in Tocharian and an almost unlimited willingness to accept all kinds of semantic changes. The third book was prepared by its author during the last months before his untimely death and published posthumously; it remained a torso but retains its value thanks to Jorundur Hilmarsson's astute insights into Indo-European matters in general and his ability to use Tocharian texts and to analyze them competently.

A further dictionary, while primarily intended as a full inventory of all Tocharian A forms then known (Poucha 1955 [omitted from Adams's list of references]), contains a fairly long list of etymological proposals. These, however, are of limited value, as Poucha did not have access to more than a small portion of the Tocharian B lexicon. (As the book was completed in 1947, not even Sieg-Siegling 1949 could be used.) The same difficulty prevailed in the case of Holger Pedersen's work; in his book of 1941, still impressive as a study of aspects of the Tocharian languages in an Indo-European context, Pedersen had very little to go on in trying to come up with persuasive conclusions about the lexicon of the immediate ancestor of the historically attested languages Tocharian A and B, since for the latter he lacked the information contained in Sieg-Siegling 1949 and the publications to follow. The otherwise extremely useful handbooks written by scholars of the Berlin-Gottingen tradition (Sieg-Siegling-Schulze 1931, Sieg-Siegling 1949, Krause 1952, Krause-Thomas 1960, Thomas-Krause 1964) contributed only marginally to an etymological assessment of vocabulary items; likewise Pinault 1989, while rich in etymological remarks, was not intended as an etymological dictionary.

By far the greatest importance for an understanding of the etymological background of Tocharian lexical items attaches to numerous articles on individual problems by various scholars, both specialists in Tocharian studies and otherwise active Indo-Europeanists making occasional forays into the Tocharian field. Papers that came out before the mid-1970s have largely been covered in Van Windekens 1976; these and later publications can be expected to have been evaluated by Adams--by and large, that is the case, although the list of references contains a number of surprising omissions, e.g., Couvreur 1955-56; Ivanov 1959; Klingenschmitt 1994; Kolver 1965; Lubotsky 1994; Pinault 1989; K. T. Schmidt 1986; Thomas 1952; Winter 1952, 1955, 1961, 1991, 1992, 1994, etc. (Some of these titles are cited in the book itself, but not in the bibliography.)

As in work with other languages, etymological evaluation of Tocharian data has to go through the following series of steps. (1) Forms to be interpreted must be understood with respect to their synchronic status within the language under consideration; this means phonological and morphological properties have to be fully under control. (2) Still on the intralanguage level, internal reconstruction may make it possible to identify innovations or to undo the effects of paradigmatic leveling, thus creating a better input for diachronic analysis. (3) Next, the task will be comparison between different languages, with the languages suspected to be most closely related being the first to be studied. (4) Comparison establishing recurrent regularities will lead to the positing of historically underlying forms assignable to the ancestral language of the subgroup studied. (5) The construct thus arrived at will be compared with similarly obtained constructs for other subgroups more distantly related to the subgroup investigated first. (6) Again, comparison now within a somewhat larger group leads to the positing of historically underlying forms pertaining to a stratum of greater time depth. (7) Comparison and reconstruction will be repeated on even larger groupings of languages thought to be related, as long as there are sufficient data for the necessary hypotheses.

The scenario proposed here is of course idealized and more complete than what can normally be achieved in one's comparative-reconstructional work; still, it may be interesting to see how well Adams's analyses and findings shape up. My coverage will necessarily be far from complete; as one task of a review is to provide material for a possible revision of the claims made by an author, there will be a certain highlighting of inadequacies. To keep my comments reasonably brief, I will attempt to select matters that seriously affect the author's results. In my argumentation, I will refer to the steps just enumerated.

(1) (i) TB a (under accent a) is viewed as...