Against arthāpatti as only technically distinguished from inference (Śālikanātha)

In contrast to his willingness to play down the differences with his Prābhākara opponents, Śālikanātha is quite straightforward in denying the understanding of arthāpatti, which he attributes to an anonymous opponent, and is clearly influenced by the Ślokavārttika’s treatment of the issue.
According to this opponent, the absence from home is the trigger insofar as it is itself thrown into doubt. Śālikanātha starts by asking how could this impossibility be conceived and comes with two possible options:

1. It is impossible insofar as the absence of the one is invariably connected with the absence of the other.

2. It is impossible insofar as the absence from home is impossible as long as one does not postulate the presence of Caitra outside.

The second option is easy to defeat, since it is not the absence from home which does not make sense, but rather only the being alive of Caitra. Why should one in fact doubt the absence from home once one has seen that Caitra is not there?

The first option deserves, by contrast, a longer treatment. Śālikanātha argues that it is tantamount to an inference based on negative concomitance only (kevalavyatirekin anumāna). This requires a short explanation: According to the Indian theory of inference, the inference is valid if sādhya and hetu don’t just happen to co-occur by chance, but are rather linked by an invariable concomitance. This is checked through the co-occurrence of the same inferential reason in similar instances (called sapakṣa) and its absence in dissimilar instances (called vipakṣa).

A valid inference should have both a positive concomitance (called anvaya) and a negative one (called vyatireka). However, epistemologists have discussed also the deviant case of inferences which seem to have only an anvaya, because there is no vipakṣa (e.g. “Everything is a product” for the Diṅnāga-Dharmakīrti school) or only a vyatireka, because there is no sapakṣa (e.g., “Nothing is eternal, because permanence is nowhere to be found”, again for the Diṅnāga-Dharmakīrti school).

Śālikanātha contends that the latter type does not work, because in order to establish the absence of the inferential reasons from all the dissimilar instances, one should be able to check them one by one, which is impossible. By contrast, the negative concomitance can be established only on the basis of a previously established positive concomitance, just as it happens in the case of the concomitance of fire and smoke.

The next step is even more interesting, since Śālikanātha suggests that the negative concomitance is established on the basis of the positive one exactly through arthāpatti. In fact, it is through arthāpatti that we know that, given that whenever there is the inferential reason there is also the thing to be inferred. Then, given that the absence of the thing to be inferred could not be possible otherwise, one concludes that also the inferential reason must be absent.

Therefore, the solution proposed by the opponent does not work since it leads to a kevalavyatirekin inference which is, in turn, parasitical on arthāpatti, since one first needs to ascertain the positive concomitance and then use arthāpatti to come to the negative concomitance. Once one has known the negative concomitance, as observed by Kumārila, one might well perform an inference on the basis of the established concomitance, but one would only end up knowing something already known, namely that Caitra is out (see ŚV arthāpatti, especially v. 67 with Uṃveka’s and Pārthasārathi’s commentaries thereon and Sucarita’s commentary on v. 19).

As for the “connection with an outer place” (bahirdeśasambandha), this is not further specified, so that we don’t know whether it just means “connection with any place other than his house” or “connection with a specific place outside his house”.

In favour of the latter option come two considerations:

1. Uṃveka discusses once the type of invariable concomitance one would need to be able to establish in order to make the arthāpatti a case of inference and refers to the fact that one should be on the door’s threshold and see at the same time Caitra’s absence from home and his presence in the garden (see the translation and discussion of Uṃveka’s crucial commentary on ŚV arthāpatti 34 in Freschi and Ollett’s translation).
2. If “outside of house” just meant “not in the house”, then Śālikanātha’s point about having to check all dissimilar instances would not make sense, since one would just need to check the single dissimilar instance, namely Caitra’s home. Therefore, it must mean “connected with a specific place outside of home”.