In commenting on Śabara's commentary on the Mīmāṃsāsūtra, Kumārila has to explain the relationship between the eternal Vedic texts which have no origin, and those texts known as smṛti (literally "remembered" texts, in contrast to śruti, those which are "heard") which are authored by human beings such as Manu but contain moral insight. This discussion starts at MS 1.3.1, where the first (incorrect) position is presented that people should ignore anything which isn't Vedic. This would mean rejecting the smṛtis. This position is untenable for a number of reasons, and Kumārila discusses why it is we should think that smṛtis are in fact based on lost Vedic texts.
This discussion is quite interesting, epistemologically, since it involves the epistemic instrument of postulation (arthāpatti). Kei Kataoka has discussed it in "Manu and the Buddha for Kumārila and Dharmakīrti." Kumārila argues that of the five possible explanations for Manu's writing his smṛti, the best one, for reasons of such principles as theoretical simplicity, is that he is working from a lost Vedic text. On the other hand, we can explain the Buddha's texts in other ways, such as being motivated by greed, or being himself deluded.
|Gautama Buddha, Moses, and Jesus.|
Why bring in Buddhist doctrine here? Well, if there is going to be room for non-Vedic moral authority, this might mean that even the Buddhist texts could have some moral authority. Perhaps there is a common Vedic ground even for the Buddha's teachings, where he happens on truth? For somewhat analogous reasoning in a more recent context, we might look to Vatican II, which set out a declaration that says
The Catholic Church rejects nothing of what is true and holy in these religions. She has a high regard for the manner of life and conduct, the precepts and doctrines which, although differing in many ways from her own teaching, nevertheless often reflect a ray of that truth which enlightens all menOf course, for the Catholic Church, divine revelation is limited to the written scriptures (for them, authored by God). But we can see some broadly similar motivation in Vatican II and the discussion in the MS. It seems like there are other texts and teachings which manage to get things "right," morally speaking. Thus we need some explanation for them. So rather than taking the extreme first position that we should reject anything non-Vedic, we might take up the view that we should accept anything which does not contradict the Vedas. (This is one interpretation of MS 1.3.5.)
But this option is too liberal for Kumārila. After all, even in the case of Vedic texts, which are accepted as teaching dharma, merely understanding the Veda is insufficient for them to be morally fruitful. They must be learned with a teacher, they must be learned by the right group of people (e.g., not a śūdra), and so on. Further, there is an additional problem for human authored texts which is not found for the Vedas: we have to evaluate the moral conduct of the author of these non-Vedic texts. For instance, the Buddha goes beyond his caste boundaries as a kśatriya in teaching. So, Kumārila argues, we should look to the writings of genuinely good people.
It is at this point that a counterexample is raised by the pūrvapakṣin. Wait a second, they say--even if we look at supposedly "good" people, they do bad things. There's a long list of examples here, including Prajāpati, Indra, Dhṛtarāśtra, and Arjuna. On top of that, there is the problem that people in different parts of the world have different practices. In some places, Brahmin women drink alcoholic drinks, and others they do not. And, further, it seems there's a vicious circle if we are trying to identify moral teaching based on the character of the author, since we also need to have moral teaching in order to evaluate the author's character! Jonardon Ganeri (2004) has pointed out that this seems analogous to the dilemma in Euthyphro (222).
This is just a very brief summary of some of the arguments raised by the pūrvapakṣin in this section, doubting the role of moral role models in knowing dharma. In his reply, Kumārila distinguishes between the authority of the Vedas and the authority of good people. He argues that we call good people "good" because they act in ways that align with the express commands of the Vedas. The explicit injunctions of the Vedas are always the strongest guide to dharma. There is no vicious circle because one always start with these explicit statements. If someone who routinely follows the Vedas does some action which is not in direct contradiction to the Vedas, then we can consider that action good. Kumārila compares the relationship between knowers of the Vedas and the Vedas themselves to the relationship between things which emerge from salt mines and the salt mine. The salt mine makes things salty, and the Vedas analogously make things (=people who know it correctly) Vedic.
Kumārila deals with the problem of local customs by emphasizing the universal nature of Vedic injunctions, and arguing that the particular restrictions about alcohol in the Vedas are in fact consistent with the regional variations. Since Brahmin women are not actually commanded not to drink alcohol (of a particular kind), the ones who refrain are not doing anything wrong, and the ones who engage aren't either. And as for the textual examples of good people doing bad things, there is a range of possible strategies (some of which might also be interestingly compared to hermeneutic strategies for apparently bad actions in the Hebrew Bible). We can say that the text itself sometimes condemns actions, which means it wouldn't be a problem as a moral guide. Or, perhaps the text admits of a figurative meaning or otherwise different meaning which makes the apparent badness of the action dissolve.
But back to the Buddhists. Kumārila argues that since the few Vedic truths in Buddhist teachings are so mixed up in bad reasoning (he gives examples of some of these), these truths are "like milk put in the skin of a dog" (śvaṭṭi nikṣiptakṣīravad). That is, they are polluted and not to be drunk. Of course, Dharmakīrti will argue otherwise, using the structure of arthāpatti against Kumārila (see Kataoka 2011 on this). He thinks that the Buddha's teachings can't be explained unless we accept that the Buddha had some direct perception of dharma. In contrast, we can explain the Vedas otherwise.
While much of my personal interest in Mīmāṃsā is in his linguistic philosophy, it's important to remember that for Kumārila (and indeed, Mīmāṃsakas in general), philosophy of language is never an abstract concern divorced from dharma. And while I would be hesitant to assign large sections of the Tantravārttika in an introductory philosophy class, I do think that this portion of the TV is a nice rejoinder to the oft-repeated worry that Indian philosophy lacks anything like "ethics." Of course, the concerns are deeply Vedic and we cannot entirely abstract from the ritual context, as ritual is a moral instrument. But Kumārila is grappling with problems that any philosopher of religion today would recognize.
Ganeri, Jonardon. "The Ritual Roots of Moral Reason" Kevin Schilbrack (ed), Thinking Through Rituals: Philosophical Perspectives. New York: Routledge, 2004, pp. 213-230.
Kataoka, Kei "Manu and the Buddha for Kumārila and Dharmakīrti," Helmut Krasser, Horst Lasic, Eli Franco, Birgit Kellner (eds), Religion and Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International Dharmakīrti Conference. Vienna, August 23–27, 2005. Wien 2011, pp. 255–269