________________
XXVII] Pramāṇas (ways of valid knowledge) 161 nor the object can be called the instrumental causes of knowledge, though they may in some sense be admitted as causes, and they do not cause an entity to be an object of knowledge either. Our knowledge does not in any way modify an object of knowledge, but an entity becomes known when knowledge of it is produced. Truth, by which is understood exact agreement of knowledge with its object, belongs properly to knowledge alone (jñānasyaiva mukhyato yāthārthyam). The instruments of knowledge can be called true (yathārtha) only in an indirect manner, on the ground of their producing true knowledge (yathārtha-jñāna-janaka yathārtha)". But yet the definition properly applies to the instruments as well, since they are also yathārtha in the sense that they are also directed to the object, just as knowledge of it is. So far as they are directed towards the right object of which we have right knowledge, their scope of activity is in agreement with the scope or extent of the object of knowledge. So it is clear that pramāna is twofold: pramāņa as true knowledge (kevala pramāna) and pramāna as instrument (sādhana) of knowledge (anu pramāna). This kevala pramāna is again twofold, as consciousness (caitanya) and as states (vrtti). This consciousness is described by Jaya-tīrtha as superior, middling and inferior (uttama-madhyamādhama), as right, mixed, and wrong; the vrtti is also threefold, as perception, inference, and scriptures (āgama). The anu pramāna also is threefold, as perception, inference and scriptures. A question arises, whether the term pramāņa could be applied to any right knowledge which happens to be right only by accident (kākatālīya) and not attained by the proper process of right knowledge. Thus, for example, by a mere guess one might say that there are five shillings in one's friend's pocket, and this knowledge might really agree with the fact that one's friend has five shillings in his pocket; but, though this knowledge is right, it cannot be called pramāņa; for this is not due to the speaker's own certain knowledge, since he had only guessed, which is only a form of doubt (vaktur jñānasya samśayatvena aprasangāt). This also applies to the case where one makes an inference on the basis of a misperceived hetu, e.g., the inference of fire from steam or vapour mistaken for smoke.
The value of this definition of pramāņa as agreement with objects of knowledge (yathārtha) is to be found in the fact that it 1 Ibid.
* Ibid. p. 250.
