Tattvasaṃgraha 及び Tattvasaṃgrahapañjikā 第18章「推理の考察(Anumānaparīkṣā)」和訳と訳注 (2)=An Annotated Translation of the 18th Chapter (Anumānaparīkṣā) of the Tattvasaṃgraha and Pañjikā thereon (2)
This article is an annotated Japanese translation of the 18th chapter (anumānaparīks.ā, vv. 1361-1485) of the Tattvasaṃ graha (=TS) and the Tattvasaṃ grahapañjikā (=TSP), which follows my previous paper (Shiga[2007]). The main subjects of the portion which I translate here (vv. 1418-1454) are as follows: (a) problems concerning three kinds of logical reason, (b) the five-membered logical formulation asserted by the Naiyāyika and (c) the inference theory of the Mīmāṃ saka (Kumārila). Regarding the subject (a), various issues have already been pointed out and examined by Dharmakīrti. An opponent raises an objection: there are logical reasons in mundane inference which are not classified into any of the three kinds (an essence, an effect, noncognition). The Buddhists claim that various logical reasons may be reduced to any of the three kinds as long as the inference in question is valid, because there must be a natural connection, i.e., tādātmya or tadutpatti, between the logical reason and what is to be proved. It is to be noted here that ‘a mirror image’ (pratibimba) is taken as an example of logical reason proving a real image which cannot be classified into the three kinds. Kamalaśīla cites the passages from the Abhidharmakośabhāṣya as the opponent’s view. In the subject (b), the point of dispute is whether the first, fourth and fifth members, i.e., ‘thesis’, ‘application’ and ‘conclusion’ are necessary or not. The important point to note here is that the views of Bhāvivikta (and so on) (TSP 514,9-16 on TS 1437-1438) and Aviddhakarṇa (TSP 516,10f on TS 1440), who flourished between Dignāga and Dharmakīrti and whose writings are not existent, are introduced with their names. It is also noteworthy that Kamalaśīla quotes some passages which have not been found in PS/PSV or NMukh etc. under the name of Dignāga (TSP 514,8f on TS 1437-1438 and TSP 515,14f on TS 1439). In the subject (c), the following two kinds of inference theorized by the Mīmāṃ saka are introduced and criticized: (1) the inference whose connection is observed on the basis of direct perception (the particularity) and (2) the inference whose connection is observed on the basis of commonality. Starting from Sabarasvāmin’s definition of the two kinds ´ of inference, Kumārila develops his own argument. The Buddhists reply that inference is limited to the second one, because, from the beginning, the object of inference is not the particularity but the commonality. As is seen in the citation of the views of opponents in TS/TSP, Sāntarakṣita and Kamalaśīla quote their views quite faithfully in the original form and criticize theṃ This style is one of the distinguishing features of TS/TSP. Through this translation, I wish to examine and analyze comprehensively various aspects of the polemics in the world of Indian logic at that time, shedding light on each thought and the background of the Indian logicians who appear in this text as opponents, as well as tracing the descent of the Buddhist logic back to Dignāga and Dharmakīrti.
