Variation and Change in Tocharian B.

• Author: Hitch, Doug
• Position: Book review

Variation and Change in Tocharian B. Leiden Studies in Indo-European, vol. 15. By MICHAEL PEYROT. Amsterdam: RODOPI, 2008. Pp. 262. $70.

If it is not already, this book will become an essential volume on the desk of researchers of Tocharian B. The author has examined every document accessible to him in painstaking detail, taking into consideration paleography, phonology, morphology, syntax, find site, and content. Probably just a handful or fewer scholars have such an in-depth familiarity with the corpus. When he makes a point, he does so drawing on all the relevant forms. His arguments are solid and convincing. There is little speculation and where there is, he recognizes it. The biggest criticism may be that the author is too modest and tentative when making very good points. This is not merely "a plea for Stumpf's theory" (p. 11), but a mountain of solid evidence and argumentation, and a significant contribution to the field.

This book may represent a turning point in Tocharian B studies for several reasons. It establishes beyond doubt that the linguistic variation uncovered by Werner Winter in a seminal article in this journal more than fifty years ago reflects temporal rather than geographic dialects. It sets up a schema through which the language in a particular B document can receive a relative date within the corpus. It puts historical linguistic analyses of B on much firmer footing. It is also the first large grammatical study in English, which will make this language more accessible to a wider audience.

Tocharian B dialect studies took a big step forward in 1955 when Werner Winter identified three geographic dialects: Western from Min-Oy Qizil, Central from Sorcuq, and Eastern from Murtuq/Singim. In 1976, in his Habilitationsschrift, Peter Stumpf argued that these were temporal dialects with the oldest from Min-Oy Qizil, the middle from Sorcuq, and the youngest from Murtuq/Singim. Stumpf's theory did not gain much notice and the received view seems to have remained that of Winter. Perhaps this is partly because Stumpf died in 1977 and could not promote his idea, and partly because his work was not published until 1990. Peyrot has been able to evaluate and expand on Stumpf's ideas using a much larger corpus, and taking advantage of more than a quarter century of progress in the field.

I will give just one random example of how this improved understanding of dialect chronology aids our understanding of the language. The third plural active ending of the oy-optative-imperfect has a long form -oyem and a short form -om. While Krause reasonably suggested that the longer ending was more original and shortened over time, Peyrot proves that the reverse is true, Archaic documents have only the short form and late documents have only the long form. This observation allows him to conclude, obviously correctly, that the -oy- formant was reintroduced by analogy, as only the third plural was without it (pp. 142-44).

After thirty pages of useful information under the headings "0. introduction," "1. aim," and "2. method," Peyrot, in "3. variant forms," launches into a series of technical linguistic essays on the attested variations. Under "phonology" he discusses thirty-seven points, "morphology" fifty-three, "syntax" two, "lexicon" nineteen, and "writing and spelling" five, for an impressive one hundred six distinct topics. In almost every case, Peyrot is able to show where the variation, or language change, fits into the overall chronological pattern. Where the evidence is not sufficient to draw conclusions, he states this.

After giving "criteria for text types" (pp. 180-82), which allow classification of a text as archaic, classical, or late, the author gives a "chart of variation sets" (pp. 183-85), which shows how each of the one hundred six variations intersects with the chronological classification. It is easy to imagine researchers pinning a copy of this chart to the wall for handy reference when working through a...