This long harangue of our opponent, however, is but a vain attempt to strike the sky with his fist; for (_a._) we mean by our phrase "self-proved" that while right knowledge is produced by the instrumental causes of knowledge, it is not produced by any other cause (as "defect," &c.) The following is our argument as drawn out in full:--Right knowledge is not produced by any other instrumental causes than those of knowledge, while, at the same time, it is produced by these, because it is not the site of wrongness of knowledge,--just like a jar.[294] Nor can Udayana's[295] argument be brought forward as establishing the dependence of authoritativeness on something external, for it is swallowed up by the dragon of the equally potent contradictory argument. "Right knowledge is not produced by any cause which is other than the causes of knowledge and is also other than 'defect,'[296] from the very fact of its being knowledge--like wrong knowledge." Again, since right knowledge can arise from the causes of knowledge _per se_, it would be a needless complexity to suppose that anything else is a cause, whether you call it a _guṇa_ or the absence of a "defect" (_dosha_).[297] "But surely if the presence of a defect is the cause of wrong knowledge, it is difficult to deny that its absence must be a cause of right knowledge?" We meet this, however, by maintaining that the absence of defect is only an indirect and remote cause, as it only acts negatively by preventing wrong knowledge. As it has been said-- "Therefore we reasonably conclude from the presence of _guṇas_ the absence of 'defects,'[298] from their absence the non-existence of the two kinds of non-authoritativeness,[299] and from this the general conclusion."[300] (_b._) We maintain that the recognition of right knowledge is produced by the same causes only which make us perceive the first knowledge[301] [_sc._ the eye, mind, &c.] Nor can you object that this view is precluded, because it would imply that there could be no such thing as doubt; for we answer that doubt arises in cases where, although all the causes which produce knowledge are present, there is also the simultaneous presence of some opposing cause, as a "defect," &c. As for your argument [O Naiyáyika! given _supra_, in p. 198, lines 17-24], I ask, Is your own argument an authoritative proof by itself or not? If it is, it proves too much [for it would properly apply to itself and lead us to infer its own dependence on external proof, whereas you hold it to be independent of such]; and if it is not, we should have a case of _regressus in infinitum_, for it will want some other proof to confirm its authoritativeness, and this too in its turn will want some fresh proof, and so on for ever.