Part 5 - Mechanism
         of anumAna Pt. 2  
         VP discusses how the processes of inferential knowledge take  place. In the smoke-fire example, there is a latent impression in the mind  formed by generalization from the particular observation that, when smoke is present,  there is fire. This latent impression is in unmanifest form. However, when I  have a perceptual knowledge of smoke on a distant hill, that latent impression  manifests in the form of a vyApti, providing the concomitant relation between  smoke and fire. Hence, VP rules out the naiyAyika’s position that recollection  arises from its prior non-existent state. VP says that the latent impression  relating to vyApti exists, and forms a basis for the recollection of the vyApti.  Similarly, it dismisses their position that recollection destroys the preexisting  latent impression.  
         In addition, VP says that the latent impression has to be  awakened in order to form a basis for recollection. If it is not awakened by  the perception of the hetu, an un-awakened latent impression cannot give rise  to inferential knowledge, since the vyApti has not materialized in the mind.  Hence, one can consider the awakening of the latent impression as an auxiliary  cause, since it forms a basis for recollection of the vyApti. Thus, VP says  that the inferential knowledge ‘the hill has fire’ arises as soon as I see  smoke on the distant hill, because this triggers the latent impression that  gives rise to the recollection of vyApti. Immediate recollection takes place  and the mind infers that there is a fire on the distant hill. There is no  reason to have a third factor that involves parAmarsha (reflection) as  discussed above, which is only a cumbersome addition not needed to arrive at  the inferential knowledge. In the example, the inferential knowledge is only  that there is a fire on the distant hill because I see smoke there. The smoke  and the hill are objects of perceptual knowledge. The smoke and the hill are  therefore objective knowledge based on their attributive content. The  inferential knowledge that there is fire is 
           not an objective knowledge with attributive content of fire. (This was  discussed earlier in relation to perception.)  
  
           The vyApti or invariable concomitance involves the coexistence of the sAdhya (the  thing that is inferred). In our fire example, it must be valid for all  situations where the existence of the hetu (in our example, smoke) is observed.  This concomitant relation between the two has been established by the  observation of both and without any exceptions, i.e. without observing smoke occurring  at anytime without fire. VP says that it does not matter whether this  coexistence of the two is observed once or many times, as long as no violation  of their coexistence is noted. What counts is observation of the coexistence  without any violation. Other philosophers say that the observations should be  more than once, and the more the better, in order to establish the universality  of the vyApti, without any violations. advaita and nyAya agree that one  observation is enough, since vyApti is both deductive and inductive as long as  no exceptions are observed.  
  
           Types of inference according to naiyAyika-s 
         Based on anvaya and vyatireka logic, the naiyAyika-s propose  three different types of vyApti-s or invariable concomitant relations. 
           a) anvaya-vyatireka (affirmative-negative) 
           b) kevala anvaya (purely affirmative) 
          c) kevala vyatireka (purely negative) 
           In the first case, the concomitant relation between the hetu and sAdhya are  related to each other both affirmatively and negatively. This is determined by  observation of their co-presence and co-absence. In the case of smoke and fire,  the positive concomitant relation is: ‘smoke is, fire is - as in the kitchen’.  The negative concomitant relation is: ‘smoke is not, fire is not, as on the  lake’. I.e. there is agreement in their presence (as positive or affirmative vyApti)  and also there is agreement in their absence (as negative vyApti). Both  establish the relation between smoke and fire.  
         Advaitins do not subscribe to the requirement for both; they  only subscribe to the affirmative aspect, not the negative. ‘When there is  smoke, there must be fire’ is the affirmative and is sufficient for inferential  knowledge. I see the smoke on the distant hill and, based on the anvaya vyApti  or affirmative invariable concomitance that whenever there is smoke there must  be fire, I can infer that there is fire on the hill.  
         naiyAyika-s say that even the negative  concomitance can give inferential knowledge. I see the smoke on the distant  hill. Now, applying the negative concomitance, we have to say: ‘if there is no  smoke, there must be no fire, as on the lake’. But, since there is smoke on the hill, there must be  fire. First of all (say the advaitins), this is a roundabout logic. Secondly,  the vyatireka or negative logic is faulty for many reasons. Firstly, if there  is no smoke on the lake, many things may not be there along with the absence of  fire. Hence, co-absence may not be generic to smoke in relation to fire. Hence,  the inter-relation between smoke and fire is not invariable for the concomitance  to work. Hence, Advaita rejects this requirement of negative concomitant  relation to arrive at inferential knowledge. For advaitins, the negative  concomitance discussed above comes under the pramANa of ‘postulation’ rather  than being a 
vyApti. In order to establish inferential knowledge between hetu and sAdhya,  all we need is a positive concomitant relation between the two.  
Purely affirmative concomitant relation is the second type, according to naiyAyika-s.  Purely affirmative concomitance involves sAdhya, the thing to be inferred to be  present everywhere or, to put it technically, it is not counter-positive to  non-existence. The counter-positive to non-existence in simple terms is ‘existence’,  as it is opposite to non-existence. They give, as an example of purely  affirmative concomitance: ‘the jar is nameable, because it is knowable’,  because ‘nameability’ (sAdhya or thing that is inferred) is everywhere, since  whatever is knowable is nameable. Since the absence of knowability and  nameability is nowhere to be observed, the knowledge of negative concomitance  is not possible. Hence, the naiyAyika-s argue that this is a case of pure  affirmative concomitance. The advaitins obviously reject this. For them, that  which is counter-positive to non-existence is existence itself, which is brahman,  brahman is non-dual and, by definition, cannot have any qualifications whatsoever.  There is no co-presence of anything else with Brahman. Hence, they do not  subscribe to kevala anvaya or purely affirmative concomitance.  
Similarly, there is also purely negative inference or kevala vyatireka,  according to naiyAyika-s; i.e. inference that is solely based on negative  invariable concomitance. The example they give is: ‘God is omnipresent, because  He is the creator’. The vyApti for this is the purely negative invariable  concomitance: ‘whoever is not omnipresent is not the creator’. No knowledge of  positive or affirmative invariable concomitance is possible in this case. If  there were, this would take the form of, for example: ‘He is omnipresent,  therefore He is the creator’. Such a statement is not possible, since  co-presence of ‘omniscience’ and ‘creatorship’ is nowhere to be observed. advaitins  reject the purely negative concomitance as a basis for inference, since such a  knowledge is not possible.  
Hence, advaitins reject both purely affirmative (kevala anvaya) and purely  negative (kevala vyatireka) invariable concomitant relationships between two  entities, hetu and sAdhya. They only subscribe to positive (not purely  positive) or anvaya vyApti (invariable concomitance). VP establishes this by  rejecting the naiyAyika-s position. For more detailed discussion of the above,  please refer to 'Methods of Knowledge' by Swami Satprakashananda.
 Proceed to the next essay.  |