________________
248 Philosophy of the Rāmānuja School of Thought [ch. corroborated by another testimony, and that by another, and this will lead us to infinite regress.
In repudiating other views Meghanādāri points out that if validity is admitted as belonging to the collocative causes of knowledge (involving the self, the senses, and the object), then even the object would have to be regarded as a pramāņa, and there would be no prameya or object left. Again, if affirmation is regarded as being of the nature of awareness, then even memory-knowledge has to be regarded as valid, since it is of the nature of awareness. Further, if affirmation of validity be of the nature of power, then such power, being non-sensible, has to be manifested by some other means of knowledge. If, again, validity is supposed to be produced by the causes of knowledge, then the dictum of the self-manifestation of validity would have to be given up. Uncontradicted behaviour also cannot be regarded as a definition of validity, for in that case even memory has to be regarded as valid by itself. It cannot also be defined as merely knowledge as such, for knowledge, not being able to turn back on itself to apprehend its own validity, would have to depend on something else, and that would imply the affirmation of validity through extraneous reference (paratah-prāmānya). Again in those cases where the cause of error is known, the cognition, though known as erroneous, irresistibly manifests itself to us (e.g. the movement of the sun). The assumption that all knowledge is associated with its validity is inapplicable to such cases. If, again, it is held that, whenever a later cognition rejects the former, we have a clear case as to how the invalidity of the previous cognition is demolished by the valid knowledge of a later moment; it may be urged that, when the generic knowledge of an object is replaced by a cognition of details, we have a case when one cognition replaces another, though it does not involve any criticism of the former knowledge.
In the Bhāțța view, where it is supposed that when the object attains its specific cognized character its knowledge as an internal operation is inferred, both validity and invalidity ought to depend upon the objects. If, however, it is urged that the notion of validity shows itself in the faultless character of the instruments and condition of cognition, that would also imply the notion of validity as of extraneous origin. In the Prābhākara view, where knowledge is supposed to reveal the knower, the object and knowledge in one
