Dignāga's Investigation of the Percept

In Dignāga’s Investigation of the Percept, an esteemed group of authors presents its results of a four-year collaborative research project on a short but philosophically dense text by Dignāga (c. 480 – c. 540 CE), the Ālambanaparīkṣā—the Investigation of the Cognitive Object/Point of Reference. In it Dignāga, who is generally considered to be the founder of the first Buddhist system of logic and epistemology, attempts to show the impossibility of an external material object functioning as a cognitive “point of reference” (ālambana). The ālambana is rather taken, in line with Yogācāra, as emerging—simultaneously with cognition as its internal form or content—from equally internal predispositions or mental seeds. In view of this, the authors’ translation of ālambana as percept is not as infelicitous as it first seems. Dignāga’s argument for such an idealist interpretation of ālambana profits mainly from the restrictive ontology of early Buddhism that posits only indivisible particles as substantially real and able to serve a function, whereas a collection of particles cannot do so, existing as they do only nominally, as something designated. Now as a percept for Dignāga must be able to cause cognition, and to transmit to it a representation of itself, it can be neither a particle—which could theoretically cause cognition but could not transmit a representation of itself—nor the contours of collections, which may resemble cognitive content but, given their mere nominal existence, cannot cause anything. A third widespread position attributed to philosophers such as Vāgbhaṭa, Uddyotakara, and Kumārila, namely that collected forms of particles have causes that produce cognitions with the appearance of collections, is ruled out on the basis that even in a collected state particles have no additional factor able to transmit distinct contours to cognition.

In the present publication, this central Yogācāra tenet—and that includes Dignāga’s refutation of external matter—is extensively elaborated on the basis of fine translations that balance philological accuracy with good readability. They include, besides Vinītadeva’s (c. 645 - c. 715) and Dignāga’s own commentaries, the detailed explanations of the Tibetan Dge lugs pa masters Gung thang Dkon mchog bstan pa’i sgron me (1762-1823) and Ngag dbang bstan dar (1759-1840). All translations are helpfully introduced and based on exemplary editions of the Tibetan texts, the Ālambanaparīkṣā and its Indian commentaries not having come down to us in their original Sanskrit.

Despite Dignāga’s Investigation of the Percept’s value as a readable and reasonably accurate translation, there are a couple of methodological concerns. In their introductory comment on methodology (xxviii – xxx), the authors do not regret the absence of the Sanskrit original and complain “scholars often speak of ‘Sanskrit originals’ as if a newly discovered Sanskrit manuscript is necessarily earlier—and thus more authoritative—than translations into other languages … If such a Sanskrit version of the Investigation of the Percept were to be discovered, it would be one more part of the corpus, and it would need to be edited in light of Tibetan and Chinese translations as well as commentaries” (xxviii – xxix). It is true that Sanskrit manuscripts are sometimes later than the extant Tibetan translations of them, and sometimes defective due to scribal or other transmission errors, but the point is not only that a critical comparison of even a later version of the Sanskrit with the Tibetan often allows us to correct the latter, but also that it is possible to reconstruct, at least in some passages, a text-critical “archetype,” which brings us closer to the time of the Tibetan text’s translation—itself a translation that was repeatedly copied and underwent change over the centuries until the available redactions of the Bstan ‘gyur. In view of this, it also makes sense to reconstruct central philosophical terms and phrases through comparison with texts, of which the Indian original is still available, as, for example, the Viṃśatikā by Vasubandhu, which is doctrinally close to the present text. Having followed these principles, Erich Frauwallner’s German translation of the Ālambanaparīkṣā remains the best available rendering into a Western language.

To give an example, throughout the publication (38, 41, etc.), the causal clause in the first verse of Dignāga’s root text is rendered as “because they do not appear to cognition.” “They” refers to particles, which are excluded from possibly being an ālambana (percept). The Tibetan of the causal clause (der mi snang ba’i phyir) has no noun equivalent for “cognition,” but it is obviously der that is being rendered by “to cognition.” The suffix –r in der, however, is not the dative/locative marker here, but a modal construction: “because [cognition] does not appear as them,” i.e., particles (the plural particle is not necessary in Tibetan). In line with Vasubandhu’s Viṃśatikā (line 17a), where der snang rnam rig is the Tibetan equivalent of the Sanskrit possessive construction tadābhāsā vijñaptiḥ, one can safely take der mi snang ba’i phyir along similar lines and translate “because it (i.e., the cognition) does not possess their (i.e., the particles’) image.” Douglas Duckworth and the other translators seem to have been aware of this possibility, as they write “there is an alternative reading of this line: “cognition arises as its appearance.”” It is not clear whether this refers to Dignāga’s causal clause under discussion here, but if it does, it should be corrected to: “because it (i.e., cognition) does not arise as their appearances” (38n3). The remark, which follows in the same footnote, namely that “this is consistent with Dharmakīrti’s understanding of perception,” while the authors’ choice “reflects the structure of Dignāga’s argument more closely,” cannot be shared by the reviewer, since my—and Frauwallner’s—translation accords with not only Dharmakīrti’s but also Vasubandhu’s understanding.

There are also a few issues with peculiar English translations of philosophical terms: Tibetan rdul phran rnam pa (Sanskrit: paramāṇvākāra), for example, is rendered as “features of minute particles” (43). Apart from “form” or “shape” (German “Gestalt”) being the standard translation of Tibetan rnam pa (Sanskrit ākāra), it is also the context here which requires “form”: minute particles do not transmit their always identical round shape to cognition, since otherwise, all of our visual percepts would be round. It should be also noted that the expression “Cittamātras” for followers of cittamātra (“mind-only”) throughout the monograph is not attested in any known Indian text and grammatically awkward. I suggest using Vijñānavādin(s) instead. Finally the correct German spelling of “ür-text” is “Urtext” (xxix).

These few minor points hardly count in view of an overall outstanding contribution to the study of Dignāga’s Yogācāra philosophy and its Tibetan reception in the Dge lugs pa school. The present monograph will be a starting point for any future study of the Tibetan reception of the Investigation of the Percept.

Klaus-Dieter Mathes is department head of South Asian, Tibetan, and Buddhist studies at the University of Vienna.

Klaus-Dieter Mathes

Date Of Review:

July 25, 2017