Agents and Actions in Classical Tibetan

Abstract

The translations and studies contained here are the result of several years' intermittent work by the authors, together and separately, on the indigenous grammatical scholarship of Tibet. The translations in Chapter II were drafted and polished in reading sessions with Prof. KATSURA Shöryü in the Department of Indian Philosophy, Hiroshima University, during the years 1984-1985. Three papers on this and related material have been presented orally to the scholarly community: one by TILLEMANS at the International Association of Tibetan Studies Conference in Munich 1985 (appeared as TILLEMANS 1988); another by HERFORTH at the International Conference on Sino-Tibetan Languages and Linguistics in Bangkok 1985 and a third by HERFORTH at the National Meeting of the Linguistic Society of America in Seattle 1986. In 1987 Prof. KATSURA authored a research report to the Japanese Ministry of Education which contains an annotated Japanese translation with Tibetan text of the treatise of dByans can dga' ba'i bio gros to be found in our Chapter IL Prof. KATSURA'S report includes a systematic presentation of the topical outlines (sa bead) to Si tu Pan chen's commentary on the rTags kyi 'jug pa. The present contribution is divided into three chapters. In the first essay, TILLEMANS presents the principal notions and currents of thought in Tibetan grammarians' treatment of verbal forms, actions and agents in their language. Chapter II, a joint effort on the part of the two authors, contains translations and editions of the texts by dByans can dga' ba'i bio gros and Si tu Pan chen Chos kyi 'byun gnas which constitute the starting-point for our own analyses. In Chapter III, HERFORTH examines the notions transitivity and voice in Classical Tibetan, incorporating some of the insights of the native grammarians and interpreting them in the vocabulary of late 20th-century typological linguistics. The problems treated here are, by common consent both East and West, among the most recalcitrant in the study of Tibetan grammar. The reader need only refer to the title of dByans can dga' ba'i bio gros treatise, "The difficult points of the rTags kyi 'jug pa"y to realize that the Tibetans themselves consider this area a formidable challenge. R. A MILLER speaks of an "impenetrable tangle" created by earlier scholars; NISHIDA Tatsuo, while recognizing the importance of the perspective provided by indigenous Tibetan grammatical categories, cautions that they are "virtually unfathomable for the uninitiated"/ Needless to say, we approach these problems with some diffidence, but encouraged in the belief that progress toward solutions can best be made through cooperation between Tibetanists and linguists. It is probably fair to say that grosso modo the motivating problem of this study was thlt of the applicability of the active-passive distinction to Tibetan and the connection between this distinction and the indigenous grammarians' classifications in terms of "self (bdag) and "other" (gzan). Certain contemporary writers, such as M. HAHN, have spoken of the voice-neutrality of Tibetan verbs. Speaking of verbal forms ending in pa/ba, he writes : "Sie kann gemäss der impersonalen Nature des tibetischen Verbs, das keine Unterscheidung von Aktiv und Passiv kennt (...) aktivisch — 'ein Sehender' — und passivisch — 'einer, der gesehen wird' — interpretiert werden." (HAHN 1985, p. 28) And further on (p. 58) we find: "Es ist zu betonen, dass es im Tibetischen bei ein und demselben Verb keinen Genuswechsel gibt, selbst wenn man diesen in der Übersetzung gelegentlich aus stilistischen Gründen vornehmen wird. So lautet z.B. der Satz rgyalpos dgra bo gsod do in genauer Wiedergabe 'Es findet ein Feind-Töten statt durch den König (als den Urheber der Verbalhandlung).', die man dann mit gleicher Berechtigung in 'Der König tötet den Feind.' und 'Der Feind wird vom König getötet.' umformen kann." Other writers, such as BACOT and REGAMEY have invoked the traditional Tibetan grammatical notions of bdag and gzan in order to justify a position which to quite a degree seems to involve the same voice-neutrality. Our problem thus took on several aspects: How did the Tibetan grammarians themselves view bdag and gzan? How well did BACOT understand the theories of these grammarians? Given a revised understanding of bdag and gzan, do the Tibetan grammarians use these notions to make a distinction of some sort between active and passive? The first two questions are treated in Chapters I and II. To answer the third question it was advantageous to see the Tibetan language in its proper place as an ergative language and thus reformulate things in an alternative set of terms which are arguably closer to bdag and gzan — agent-prominence and patientprominence. A result of our investigations, however, is that it seems clear that the complete active-passive neutrality of which HAHN speaks is not borne out by Tibetan grammarians' analyses of their language. In effect, to take HAHN'S example where the verb gsod pa is in the present tense, the rendering 'Der König tötet den Feind' is preferable, and not for stylistic reasons, but rather because the present tense form gsod pa is indeed exclusively agent-prominent, or bdag, just as the grammarians unanimously say it is. Had the verb been gsad pa or gsad par bya instead of gsod pa, a passive rendering would have been more in keeping with the fact that future forms seem to be patient-prominent, or gzan, according to the grammarians. Nor is there probably much reason to suppose with HAHN and REGAMEY that the voice-neutral nominalized version, 'Es findet ein Feind-Töten statt durch den König (als den Urheber der Verbalhandlung)', is somehow or another the most accurate rendering. At least in the case of REGAMEY, a good part of the motivation for such a position seems to have come from the ill-fated theories of BACOT on bdag and gzan. However, the general problem of the noun/verb distinction in Tibetan cannot be taken up in any detail here.iö Finally, two cautionary notes are in order, one for the specialist in Sanskrit grammar, one for the philosopher studying Madhyamaka Buddhist argumentation. We have occasionally supplied Sanskrit terms when it seemed clear that the Tibetans were borrowing or alluding to Sanskrit notions. Strictly speaking, it is somewhat odd to speak of equivalences at all in that the texts we are dealing with are indigenous Tibetan works and not translations from Sanskrit. Nonetheless, in certain cases of patent concept-borrowing, we no doubt can give the original Sanskrit notions, the best example being the triad bya byed las gsum (kriyâ, kartr, karman; "action, agent, object"), a bona fide Indian-based set of terms which Tibetan writers (such as dByans can dga' ba'i bio gros) associated with bdag and gzan and then interpreted for their own purposes. But, as is the case for most Tibetan grammatical terms, they have to be understood primarily in their Tibetan contexts. To take another example, byed pa po la yodpa'i bya ba ("action pertaining to the agent") would, prima facie at least, seem to be the equivalent of kartrsthakriya(ka)y a term to be found in e.g. Patanjali's Mahäbhäsya where it refers to roots such as GAM ("to go") whose action is found in the agent. {PAC ["to cook"] on the other hand, is an example of a root which is karmasthakriya( ka)9 because the action functions in the object in the sense that it produces a change there. See e.g. ABHYANKAR 1977 p. 109, 112.) If, however, we look at how the Tibetan term is actually used, then inspite of the fact that byed pa po la yodpa'i bya ba seems to be a direct translation of kartrsthakriya(ka), the Tibetan term is used very differently: firstly, it is not applied to roots (as in Sanskrit), but is related primarily to certain specific tenses; secondly, in Tibetan the term byed pa po la yodpa'i bya ba refers to the effort or exertion (rtsol ba) of an agent, rather than a type of action (such as "going") which is found functioning in the agent; thirdly, the Tibetan verbs like "to go" ('gro ba) are not classified as byed pa po la yod paï bya ba. Another point worth bearing in mind when one takes up the problem of finding Sanskrit equivalences is that the Tibetan grammatical vocabulary seems to be more limited than that of Sanskrit, so that the Tibetan terms are often highly ambiguous, doing double or triple duty for possible Sanskrit notions. In short, the relationship between Sanskrit Vyäkarana and Tibetan grammar is generally complex and very indirect. No doubt it needs to be studied, but at least at the present state of the research on bdag and gzan and related notions, the enterprise of searching for influences from Sanskrit is by and large inconclusive.