Tsong kha pa and the Myth of the Given

My thesis is that Tsong kha pa’s distinction between the methodological approaches of Svātantrika-Mādhyamikas and Prāsaṅgika-Mādhyamikas has been widely misinterpreted.¹ As the distinction is usually construed, it is taken to turn on the different appraisals, by Svātantrikas and Prāsaṅgikas, of the epistemological and ontological costs and benefits of foundationalism. The standard interpretation of Tsong kha pa’s distinction ascribes to him the claim that Svātantrikas underestimate the ontological costs of foundationalism, in particular the cost of ontological commitment involved in accepting a form of the Given, as Svātantrikas seem to do in their use of svalakṣanas to claim authority for observation reports. A lemma in my argument here is that Sellars’s famous critique of foundationalism is misunderstood when taken as a total rejection of empiricist foundationalism. Sellars’s critique of the Given does not bear on the sort of foundationalism concerned with the authority of particular observation reports, such as are needed to block the inferential regress that lends significance to the metaphor of foundations. Sellars’s critique is concerned rather with the question of entitlement, which has to do, more fundamentally, with the kind of normativity associated with the very notion of ascribing justificational authority to any observation reports whatsoever. In light of this understanding of Sellars’s critique of the Given, I argue that Tsong kha pa’s Svātantrika-Prāsaṅgika distinction turns on what he sees as a divergence in the therapeutic efficacy of their respective methodologies, and that it does not imply a difference in the two schools’ ontological commitments.

For Tsong kha pa, the crux of the matter is that it is impossible for a Mādhyamika and a non-Mādhyamika to have commensurable perceptual apprehensions of the logical subject of an inference in any debate about intrinsic existence. Svātantrikas take one approach to dealing with this inevitable incommensurability, while Prāsaṅgikas take a different approach. The Prāsaṅgika approach involves the use of opponent acknowledged inferences (gzhan grags kyi rjes su dpag pa, or gzhan grags kyi sbyor ba), which, as I shall show, operate on the presupposition that even where total commensurability is impossible, partial commensurability, at the level of observation reports, is a necessary condition for the possibility of rational discourse between two individuals (or between opposing perspectives within a single individual’s divided consciousness). So, Tsong kha pa’s account of opponent-acknowledged inferences, as worked up in the final section of my paper, is relevant, more broadly, to the contemporary discussion of the foundations of rationality as evinced in the work of such figures as Habermas and Brandom.

The doxographic distinction between Svātantrika-Mādhyamikas and Prāsaṅgika-Mādhyamikas is a topic that is of central importance for Tibetan Buddhist philosophers, with differences between the main “schools” or “sects” of Tibetan Buddhism finding expression in distinct interpretations of how to understand the difference between the two sub-schools of Madhyamaka philosophy. The basic difference between the two sub-schools arises from a procedural difference classically delineated in the works of Bhāvaviveka and Candrakīrti. Svātantrikas follow Bhāvaviveka in countenancing the use, by Mādhyamikas in debate with non-Mādhyamikas, of certain types of inferences which require both parties to the debate to agree as to the mode of presentation or ontological status of the parts of the inference (subject, predicate, reason, and so forth). Prāsaṅgikas, on the other hand, following Candrakīrti, insist that Mādhyamikas in debate with non-Mādhyamikas cannot use such inferences because it is precisely the mode of presentation and ontological status of the subject of debate that is in dispute for Mādhyamikas and non-Mādhyamikas. In the Tibetan context, there are, speaking generally, two basic types of interpretation of the significance of this procedural disagreement between Svātantrikas and Prāsaṅgikas. On one interpretation, the difference between Svātantrikas and Prāsaṅgikas is thought to involve no significant consequences for the way in which...