Subschools of Mādhyamika
Tsong kha pa and his followers, the dGa’-ldan-pa (Gandenba) and then the dGe-lugs-pa school, reject this all-encompassing syncretism, arguing that Yogācāra and Mādhyamika are quite distinct schools, and that Tantra and the Tathāgatagarbha theories are to be interpreted along the lines of Nāgārjuna, and especially along the lines of the branch of Mādhyamika founded by Candrakīrti, the so-called *Prāsaṅgika or Thal-’gyur-ba (Talgyurwa) school. Whereas the *Svātantrika or Rang-rgyud-pa (Ranggyuba) philosophy of Indian Mādhyamika, attributed by Tibetans to Śāntarakṣita, Bhāvaviveka et al., had been dominant in the earlier diffusion of Buddhism in Tibet (starting in the eighth century) and had been the philosophy of the first traditions in the beginning of the second diffusion (from the eleventh century on), Tsong kha pa came squarely down on the side of the *Prāsaṅgika Mādhyamika, a school which owes much of its influence in Tibet to Pa tshab Nyi ma grags (Batsap Nyimadrak, b. 1054/55), who was the translator of Candrakīrti’s Madhyamakāvatāra (Introduction to the Middle Way) and other works. Note that the terms *Prāsaṅgika (thal ’gyur ba) and *Svātantrika (rang rgyud pa), although frequently used in the modern secondary literature, are not clearly attested in Indian texts. This terminology to distinguish between Mādhyamika schools is Tibetan in origin and is especially found in a genre of literature which had great importance in Tibet, the hierarchical presentation of the ‘systems of tenets’ (siddhānta; grub mtha’) of the four Indian Buddhist schools and their various subschools.
There were numerous theories among Tibetans of the eleventh to fifteenth centuries about what the difference between *Svātantrika and *Prāsaṅgika actually came down to – indeed, the matter took on such importance that the question of how to construe the difference became one of the key issues in Tsong kha pa’s thought, just as it was a key issue for his predecessors and for many of his successors. Some Tibetans had conceived of the difference between *Svātantrika and *Prāsaṅgika solely in terms of the logical forms which the two schools employed (namely, objectively valid reasons versus mere reductio ad absurdum); others focused on the fact that *Prāsaṅgika rejected means of valid cognition (pramāṇas) while *Svātantrika did not, or on the fact that *Prāsaṅgikas did not have any theses (pakṣa) whatsoever of their own, not even on conventional matters. (Compare ’Jam dbyangs bzhad pa’s third type of erroneous Mādhyamika.) Reacting to the views of these opponents, Tsong kha pa formulated the difference as turning, above all, on a difference in ontology, in particular on the acceptance or rejection of things being in some way established with/through their own intrinsic natures (rang bzhin; svabhāva). Svabhāva, a key notion that Nāgārjuna and his commentators argue against, figures throughout the Tibetan presentations of the systems of Buddhist schools in the grub mtha’ literature, but with each school adopting an ever subtler position on the question – it is only, according to Tsong kha pa, the *Prāsaṅgika branch of the Mādhyamika school which rejects it completely and unconditionally; the *Svātantrikas still hold on to a kind of rarefied version of svabhāva, albeit only on the level of conventional truth. The basic argumentation for this attribution turns on the *Svātantrikas’ own construal of the Yogācāra three-nature theory.
The question as to precisely what a Mādhyamika denies became a major issue in Tibet. A dominant trend in Tibetan Mādhyamika thought was that promulgated by Pa tshab’s disciple Zhang Thang sag pa (Shang tangsakba), Go ram pa bSod nams seng ge and others, a position which came to be known as ‘the Mādhyamika free from extremes’ (mtha’ bral dbu ma) and which was often depicted as ‘the view of neither being nor nonbeing’ (yod min med min gyi lta ba). This view advocated using Mādhyamika-style arguments to show the incoherence of any and all things that one might analyze, with the result that no predication of being, nonbeing, both being and nonbeing, or neither being nor nonbeing is rationally justified. We thus have, on this interpretation of the Mādhyamika, a literal and unqualified negation of all members of the ‘tetralemma’ (catuṣkoṭi), with the purpose of thoroughly overcoming the possibility of any conceptualizations and discursive thought. This ‘Mādhyamika free from extremes’, it should be remarked, is a pretty fair characterization of the usual Indian position, and corresponds quite well to the way in which certain important Western authors would characterize Indian Mādhyamika. (In this sense, S. Matsumoto (1990) is certainly right to stress just how remarkable and unique Tsong kha pa was in the history of Indo-Tibetan Mādhyamika in opposing the fundamental position of the ‘Mādhyamika free from extremes’.)
The question does, however, arise as to whether this version of the fourfold negation leads to a deviant logic, as negating nonbeing would seem to imply being, and negating the last lemma would seem to imply being or nonbeing. Tsong kha pa and his school avoid these types of unwelcome consequences by adding the modal qualifiers ‘truly’ (bden par) and ‘conventionally’ (tha snyad du) to the tetralemma, so that, for example, the first two negations end up being that things are not truly existent and not conventionally nonexistent. The tactic obviously avoids violations of logical laws. Equally, it allows the Mādhyamikas to make a clear separation between the conventional and the ultimate, so that they can argue that a thing like a pot exists (conventionally) and is established by means of valid cognition. We can thus easily hold the thesis that the pot exists, but where the Mādhyamika philosophy comes in is to show that it does not truly exist. Indeed, the advocates of the ‘Mādhyamika free from extremes’, as well as the gZhan-stong-pas, do have serious problems in accounting for conventional truths: to say, for example, that such truths merely exist for mistaken minds (blo ’khrul ba’i ngor yod pa) and are not established by means of valid cognition (pramāṇas) comes perilously close to just falling into the trap of subjectivism, where one says that things exist or have such-and-such properties when one believes that they do. (Quite a bit more than that is needed to explain how we can discover new things and how we can show that things which many people believe in at some particular time do not exist and have never existed even conventionally.) The issue of the acceptance or rejection of pramāṇas establishing conventional truths is developed at length from the fifteenth century onwards, starting with a critique of Tsong kha pa on this and other points by the Sa-skya-pa thinker sTag tshang lōtsa ba (Daktsang lodzawa, b. 1405).
It should be stressed that the logical simplicity and lucidity of Tsong kha pa’s version of Mādhyamika comes at a high exegetical and philosophical price: not only is he constrained to add words everywhere to Indian texts, but, what is perhaps more serious, he can well be accused of inventing a factitious notion, ‘true existence’, which he then refutes as incoherent and contradictory, leaving the ordinary world intact and even reified. This was a common criticism of Tsong kha pa. Arguably, as his critics maintained (be they advocates of the ‘Mādhyamika free from extremes’ or gZhan-stong-pas), Nāgārjuna and his commentators were showing the utter incoherence of ordinary things themselves, and were leaving nothing whatsoever unassailed and intact among these conventional entities.