________________
74
::
INDIAN LOGIC
he refuses to concede that this "absence' itself is an object of perception; thus he argues tbat the locus of an 'absence' is an object of perception but not this absence' itself just as the colour of a distant column of fire is an object of perception but not its touch. It is admitted that in the cited illustration the touch of fire is an object of inference, 5 but the implication is not that an 'absence' is cognized through inference. For a little later on it is argued that an 'absence cannot be cognized through inference because in this case it cannot be ascertained as to what acts as the locus-of-inference, what as the probans, what as the probandum, also because here it is impossible to formulate the needed relation of invariable concomitance (which formulation requires that at some stage an absence' must have been made an object of perception). Let us recall that somewhat similar was the Kumārilite's mode of arguing when he was out to demonstrate that a case of arthāpatti is not a case of inference. But this time the suggestion that an 'absence' is cognized through arthāpatti is rejected even without being given a moment's thought.? And yet what actually happens is that if x and y are such that they are either both cognized together or both not-cognized together then in case x is cognized but not v this cognition-of* implies the cognition-of-'absence of y'. This in essence is how the Buddhist puts the matter when he would submit that here 'absence of y' is cognized through an inference of the svabhāva-anumāna type where 'cognition of x' acts as probans; (svabhāva-anumāna essentially covers the same ground as the Mimāṁsā's arthåpatti and the Buddhist is also of the view that an 'absence' is no independent real, so that what is here inferred is not an 'absence' as such but an usage as to absence). In any case, we have thus seen how on the Kumārilite's showing the cognition of an 'absence' is not a case of perception and not a case of inference. In the meantime he has considered a case which to him appears to clinch the matter decisively. Thus when one observes a place without particularly noticing whether x is present there or not, one can later on rightly say that x was absent there at that earlier occasion, the Kumārilite argues that this later cognition of ‘absence of x' is obviously not had