________________
148
INDIAN LOGIC
not two letters happening to be a seat of the same universal."15 This is the most crucial piece of argumentation and is further elucidated as follows : "Even when a letter is pronounced slowly in one case fast in another, slowness and fastness do not belong to this letter itself but to the sound that makes this letter manifest. Really, even granting that a common universal resides in two cases of the pronunciation of a letter, one cannot say that slowness and fastness belong to this letter itself but must allow that they are due to some accidental adjunct; so why not grant that the very wordness in question is due to a similar adjunct ? Moreover, a consonant sounds different when pronounced in association with a different vowel; hence if a vowel soạnds different that too must be due to its association with some accidental adjunct, so that, for example, it is not proper to speak of the vowel 'a' having eighteen types.''16 Thus it is concluded that a word cannot possibly yield a meaning (that is, cannot perform its sole function), unless it is something eternal.'? The point that a letter is something eternal is explained once more. This time it is argued that a thing perishes either when its component parts perish or when its locus perishes, but neither is possible in the case of a letter; that a letter is not made up of parts is proved on the ground that it is pronounced not partwise but either not at all or in full; to this is added that a letter requires no locus and even if the sky is supposed to be its locus this locus can never perish.18 Then it is pointed out that the usage is to the effect 'such and such a word is pronounced so many and so many a time', a usage that makes sense only in case one and the same word is pronounced on several occasions."' Lastly, it is submitted that to view a word as the same on the different occasions when it is pronounced is an indubitable case of recognition, it being maintained by the Naiyāyika himself (as against the Buddhist) that recognition is a perfectly legitimate case of valid cognition.20 In this connection it is said that a word is made manifest within a moment by a fleeting air-mass just as things are made manifest within a moment by the flash of lightning appearing in a pitch-dark rainy night.21 This immediately leads to the discussion as to how an eternally existing word is made manifest by an appropriate sound on this occasion or that.
The opponent objects : "On your showing, a sound supposed to make a word manifest is of the form of air.22 Now if such a sound