In his Pramāṇasamuccaya, Dignāga claims that there are two means of knowledge (pramāṇas): perception (pratyakṣa) and inference (anumāna). Perception has as its object the particular (svalakṣana); inference has as its object the universal (sāmānyalakṣana). The key distinguishing feature between the two is that “perception is free from kalpanā (imagination, conceptual construction)” (Pramāṇasamuccaya, 1.3a). This is an exclusive dichotomy: any means of knowledge must be either perception or inference, but not both.
The Madhyamaka philosopher, Candrakīrti, offers the following argument against Dignāga: “… if it is said that there are two means of knowledge (pramāṇas) through adherence to two characteristics – particular and universal, then that characterized thing, of which there are two characterizing marks (i.e., particular and universal), does that exist, or on the other hand, does it not exist? If it exists, then there is another third object of knowledge (prameya) than those two, so how are there two means? On the other hand, if that which is characterized does not exist, then the characterization is also without a basis, so how could there be two means of knowledge?” (Prasannapāda, p. 20, lines 20-23).
The Cārvāka skeptic Jayarāśi gives what I call “The Impossibility of Considering Duality Argument.” Jayarāśi begins with the notion that perception apprehends itself and inference apprehends itself, but neither can apprehend the other according to Dignāga. Jayarāśi concludes, “Thus, talking or thinking about the number [of means of knowledge] being two is impossible” (Tattvopaplavasiṃha, 3.3a). Eli Franco spells out the presupposition that makes the argument work:
In order to determine the number of means of valid cognition, one has to have them all as the object of one and the same cognition. However, according to the Buddhists, a cognition is not apprehended by another cognition, but only by itself. Nor is there an ātman which could coordinate the different cognitions. Thus, one may perceive perception by perception, and inference by inference, but never both at the same time. Consequently, whatever the number of means of valid cognition may be, there is no way of knowing it. (Franco 1994, 430 n. 184)
Thus, Jayarāśi’s argument rules out the possibility of even considering the Buddhist thesis that there are two means of knowledge. He ends his chapter on Buddhist epistemology with the following: “And when this (i.e., there being two means of knowledge) is not possible, saying ‘There are only two means of knowledge’ is the gesticulation of a fool” (Tattvopaplavasiṃha, 3.3a).
Are Candrakīrti and Jayarāśi giving roughly the same argument against Dignāga?
In Part Two I will ask: Do Candrakīrti and Jayarāśi raise serious problems for Dignāga’s epistemology?
Candrakīrti. (1960). Prasannapadā. Ed. P. L. Vaidya. Madhyamakśāstra of Nāgārjuna (Mūlamadhyamakakārikās) with the Commentary Prasannapadā by Candrakīrti. Dharbanga: Mithila Institute.
Dignāga. (2005). Pramāṇasamuccaya, Chapter One. Ed. Ernst Steinkellner.
Franco, Eli. (1994). Perception, Knowledge and Disbelief: A Study of Jayarāśi’s Scepticism,Second Edition. Delhi: Motilal Banarsidass.
Jayarāśi Bhaṭṭa. (1994). Tattvopaplavasiṃha. Ed. and Trans. Eli Franco. In: Perception, Knowledge and Disbelief: A Study of Jayarāśi’s Scepticism, Second Edition. Delhi: Motilal Banarsidass.