A History of Indo-European Verb Morphology.

• Author: Klein, Jared S.

My criticism of this book begins with its title. This is not a "history" but a "prehistory," and therein lies the difficulty. Quoting from an earlier book of his on PIE noun inflection, Shields states that "there are very few unique solutions to the problems posed by historical and comparative linguistics" (p. 113). Yet the fact remains that linguistic states attested in the historical record have prehistories, i.e., some particular set of prehistoric events occurred, leading to their existence; others did not. Shields' frequent statement that a given sequence of events "could" or "may" have occurred can in a trivial sense not be denied, inasmuch as the term "could" excludes only absolute impossibility. The question is rather, how likely are the proposals to represent prehistoric truth? It seems to this reviewer that there are certain methodological prerequisites for success in building prehistoric scenarios for linguistic events. While these do not assure that a given proposal is likely to capture prehistoric reality, they at least help the scholarly consumer of such proposals feel that he is being securely led in a reliable manner, such that the "could" referred to above lies more toward the probabilistic pole of "likely" than "unlikely." These prerequisites include a complete command of the data in its philological aspect. After all, our raw materials come from ancient texts which must be evaluated for their trustworthiness as witnesses, interpreted, and chronologically arranged. There must be a willingness to open up the text and "see for oneself," so that one does not have to depend on the all-too-often unreliable statements of others. Next, there must be an inclination to see prehistorical events bidirectionally, that is, to be as concerned with the derivation A [less than] *B as with the development *B [greater than] A. Great leaps should be avoided and small steps encouraged. Moreover, minimal assumptions are to be preferred to maximal ones. Prehistoric events are essentially beyond our control. The greater the hypothetical leap, the less our ability to verify it. When such leaps are heaped up one upon the other, we have merely multiplied improbabilities. Superimposed on all the above should be common sense and good judgment - intangible characteristics which are, however, indispensable if one's proposals are to be believed by one's colleagues. Finally, one's proposals for prehistoric developments should be consonant with linguistic events occurring in the historical record.

By most of the preceding standards (for the last, cf. n. 3) Shields receives low grades. First, he gives no evidence of philological expertise, of ever having looked at a textual problem first hand. Evidence of this pervades the entire book. Thus, in his discussion of various deictic particles, so crucial to his reconstructions, he cites with...