________________
THEORY OF KNOWLEDGE
111
be a mental svalaksana; but in most cases while contrasting a physical svalaksaņa to something else what Dharmakirti has in mind is a different thing altogether. The reason is that for all practical purposes Dharmakiti understands by svalaksana a physical svalakṣaṇa and contrasts it with sāmānyalakṣaṇa which is another crucial concept of his logic 3. If svalaksana alone constitutes real reality then the conclusion is automatic that a sāmānyalaksana lacks real reality14. But what is sāmānyalaksana ? By sāmānyalaksana, Dharmakīrti understands an abstract generic feature which real things are found to exhibit now here now there, and he denies real reality to it not because real things do not really exhibit it but simply because it is not itself a real thing - which is a truism. As a matter of fact, Dharmakirti's own treatment of perception and inference - the only two means of valid cognition recognised by him - goes to make clear as to how vital a role is played by sāmānyalaksaņa in each. Thus perception is here identified with the bare sensory experience which an object produces in the cogniser concerned, but it is at once admitted that perception thus understood serves no practial purpose unless followed by the attribution of an abstract generic feature - a sāmānyalaksana - to the object perceived. Similarly, inference is an impossibility unless the relation of invariable concomitance is observed to obtain between the probans and the probandum concerned, but this relation obtains not between a probans and a probandum conceived as two particular things but between them as possessed of this or that abstract generic feature - this or that sāmānyalaksana. Yet Dharmakirti feels that there is nothing anomalous about his emphatic denial of real reality to a sāmānyalakṣaṇa, and there are two reasons for this. First, Dharmakīrti finds it rather easy to point out loopholes in the concept of sāmānya as upheld by the philosophers belonging to the Nyāya-Vaiseșika and Mīmāmsā schools, and this misleads him into thinking that all talk about an abstract generic feature really characterising a real thing' must be erroneous. Secondly, Dharmakirti feels, mistakenly of course, that there results nothing incongruous in case an abstract generic feature is conceived negatively rather than positively; e.g. on his view it would be erroneous to suppose that all cows share in common the positive feature 'cowhood' but not at all erroneous to suppose that they share in common the negative feature 'absence of non-cows', a misconceived view.
Some details of Dharmakirti's treatment of perception (pratyaksa) and non-perceptual cognition (anumāna) are also noteworthy. Thus he
