Saṃvṛti and Paramārtha in Mādhyamika and Advaita Vedānta

“Saṃvṛti and Paramārtha in Mādhyamika and Advaita Vedānta” — Murti, T. R. V. In M. Sprung (ed.), The Problem of Two Truths in Buddhism and Vedānta, D. Reidel, Dordrecht 1973, pp. 9–26. (The copyright line in the working copy gives the running short title, “Two Truths in Buddhism and Vedanta”; the full title is as carried on Sprung.)

Murti’s own statement of his position, eighteen years after The Central Philosophy of Buddhism (1955), which this wiki cites at n. 1 as the ground of its opening section. Elsewhere in the wiki he appears at second hand — through della-santina-madhyamaka-western-1986, hayes-nagarjunas-appeal-1994, galloway-logical-issues-1989, gokhale-madhyamika-dialectics-1992, Wood and Westerhoff — while Matrix D tabulates him as one of the nineteen framework-absent readings and makes him the exhibit of the Kantian overlay. The paper substantially complicates that characterisation.

Provenance. raw/ holds the working markdown copy only; there is no PDF. The text is clean and continuous with no OCR gaps in the body prose, but the notes on pp. 24–26 are garbled — note numbers, kārikā numbers and the Sanskrit verses they carry have come apart from one another in several places. Verse references below are taken from the body prose, or from notes whose Sanskrit could be matched to the verse quoted. Any kārikā number that will bear weight in this wiki should be checked against the printed volume.

Thesis / main argument

The paper is a comparison written against a consensus. Radhakrishnan, Stcherbatsky and Dasgupta had held that the difference between śūnyatā and Brahman is superficial and largely verbal; Murti holds the opposite, that “in spite of superficial similarities in form and terminology, the differences between them are deep and pervasive” (p. 9). He grants a long list of agreements first — both traditions hold the absolute to be transcendent to thought and devoid of empirical determinations, both make it immanent as the reality of appearance, both reach it in a non-discursive intuition, both derive appearance from a beginningless non-empirical ignorance (མ་རིག་པ་), both are forms of non-dualism that reject duality dialectically rather than establishing the absolute by positive argument, and both are therefore committed to two truths and to a provisional / definitive sorting of scripture. Having granted all this, he spends the rest of this wiki prising the two apart.

The wedge is a single distinction, stated at pp. 13–14 and restated in the conclusion: Madhyamaka is advaya, Vedānta is advaita. Non-duality (གཉིས་མེད་) as Nāgārjuna means it is freedom from the duality of the extremes — knowledge purified of conceptual distinctions, an epistemological programme whose whole concern is the correction of the knowing faculty. Advaita is knowledge of a differenceless entity, an ontological programme whose concern is the thing known. From this one difference Murti derives the rest: Madhyamaka has no cosmology, no theory of the ownership or object of ignorance, no doctrines of god, soul or matter, because for it the ontologising of the categories of thought is “the very worst sin one could commit” (p. 14). Vedānta has all of these, and a theory of māyā besides.

The conclusion turns from Vedānta to Frederick Streng, and it is here that this wiki does its real work for this project. Murti reads emptiness and the two truths as “a kind of Absolutism, not Nihilism” (p. 22) and refuses Streng’s reading, on which the ultimate is itself a species of the conventional that collapses once it has finished its work. His argument against Streng is soteriological before it is metaphysical: if nothing is left over as Real after everything has been declared relative, the cure has destroyed the patient. That is the sentence this wiki of should quote.

Key claims

• Madhyamaka begins with the world-illusion, Vedānta with an empirical one (pp. 11–13). Vedānta analyses the rope-snake, builds from it the distinctions real / false / utterly unreal (sat / mithyā / asat, with the false as anirvacanīya, neither), and extends the analysis analogically to the world. Murti holds the Mādhyamika does nothing of the kind: he gives illusion-examples (mirage, city of gandharvas, ball of foam, hair-ringlets) but nowhere analyses an empirical illusion to show what its illusoriness consists in, nowhere criticises rival theories of illusion, and nowhere distinguishes the false from the utterly unreal. He goes directly to the world-illusion, which is presented to him as the interminable conflict of philosophical viewpoints — “what has been called by Kant the Transcendental illusion” (p. 10).
• The catuṣkoṭi is a claim to exhaustiveness (pp. 14–15). The Mādhyamika must reject every alternative, actual and possible, so he schematises all philosophical views into “the four-cornered conspectus of negation”. Murti compares the ambition to Kant’s critique and Hegel’s dialectic, and is explicit that what is interesting is not the “topical soundness of the schematism” but the daring of the claim to universality — without it the criticism would not be a dialectic of all philosophy.
• The negation-principle is mutual dependence (pp. 15–16). Nothing analysed survives as a thing in itself; each relatum is what it is in relation to others, indefinitely. Cause and effect, substance and attribute, whole and part, self and states, fire and fuel, agent and act, definition and definiendum, all fall to the same argument: they are established neither as identical nor as different, neither together nor separately. Murti quotes Bradley twice as an exact parallel and notes the divergence — Hegel welcomes mutual dependence as the mark of the Real, the Mādhyamika takes it as the lack of inner essence.
• Reason itself is the conventional (pp. 16–18). Buddhi is equated with saṃvṛti and condemned as the principle of falsity; the ultimate (དོན་དམ་པ་) is “the utter absence of the function of Reason”. Murti reproduces Candrakīrti’s three definitions of the conventional (ཀུན་རྫོབ་) from the Prasannapadā — that which covers over, and so is identical with ignorance; mutual dependence, and so relativity; and saṃketa, convention or common usage (འཇིག་རྟེན་གྱི་ཐ་སྙད་) — and adds the loka / aloka division from Madhyamakāvatāra 6.24–26, aligning the second with the Vedāntic prātibhāsika.
• The two truths are two manners of apprehension, not two sets of objects (p. 19). “There are not two different spheres or sets of objects to which these apply… The difference is in our manner of looking at things.” He cites MA 6.23 for it. This is the subject-side reading of the two truths, stated flatly, by a Kantian absolutist.
• The conventional is means, the ultimate is end (p. 19). Saṃvṛti is upāyabhūta and paramārtha is upeyabhūta; the conventional is “the ladder or the jumping board”. The ultimate can only be reached negatively, by removing what thought has beginninglessly ascribed. The Four Noble Truths sort accordingly: cessation is the ultimate, the other three including the path are conventional, as are the four fruits, and the Prajñāpāramitā extends this even to Buddhahood.
• Ignorance is itself unreal for the Mādhyamika and quasi-positive for the Vedāntin (pp. 18–19). For Madhyamaka not only difference but identity is ignorance; conceptualisation as such, not any particular concept, is what obscures — buddhiḥ saṃvṛtir ity ucyate. And ignorance is māyā: were it real its products would be real and nothing could be negated. For Vedānta ignorance is the creative power of Brahman, beginningless, quasi-positive, with an epistemic status and a cosmology attached.
• Relation: Madhyamaka rejects both terms, Vedānta keeps one (pp. 20–21). The Vedāntin holds the two terms of a relation to be of different orders — the higher not exhausted by the relation and able to stand without it, the lower entirely exhausted within it. Particulars are false in their own nature (svarūpato mithyā), the substance false only in its seeming relation (saṃsargato mithyā). The Mādhyamika allows no such asymmetry.
• Language has no revelatory function (p. 21). “For the Madhyamika language is a conventional set of symbols which has no reference outside this convention to any referent, to fact or reality. The meaning of a term or concept is another term or concept.” Vedānta by contrast can indicate Brahman metaphorically (lakṣyārtha), knows Brahman only through the Upaniṣads, and may in the last analysis identify vāk with Brahman.
• Emptiness is freedom, and must not become a further view (pp. 21–22). The root of suffering is indulging in views (ལྟ་བ་) and imagination; freedom is the cessation of conceptual construction (རྣམ་པར་རྟོག་པ་); emptiness is the antidote taught for a soteriological purpose. But “we should not take the cure of the disease as one more reality. Both the disease and its cure are equally within phenomena”, and we are warned not to make emptiness another view.
• Against Streng, four moves (pp. 22–23). (i) “No views about reality” is not “a no-reality view” — that would be nihilism, and Nāgārjuna and Candrakīrti both protest against reading emptiness as abhāvārtha. (ii) Reality is accepted and characterised at MMK 18:9 as non-relative, quiescent, non-discursive, non-dual, and its presence can be indicated by an ascribed mark (samāropāt). (iii) A regress: is “all things are dependently arisen” itself a relative statement or a perception of truth? If a statement, its relativity rests neither on itself nor on its objects, and any higher statement grounding it starts an infinite regress; so it is more consistent to take it as a perception, not a logical statement. (iv) The headache argument: if nothing survives the rejection of everything as relative, the soteriological purpose cannot be secured — the cure removes the sufferer along with the complaint.
• Reality is self-existent (p. 16). “Tattva, or the Real, is something in itself, self-evident and self-existent.” Also: only the absolute as the unconditioned is real, and therefore it cannot be conceived as existence, non-existence, both or neither.

Methodology

Comparative philosophy of a particular mid-century kind: two systems set side by side on a fixed grid of questions — the status of appearance, the analysis of illusion, the theory of relation, the function of language, the ideal of freedom — with Western philosophers used as calibration throughout. Kant supplies the transcendental illusion and the model of a critique claiming exhaustiveness; Hegel supplies the contrast case on mutual dependence; Bradley supplies the parallel on relations, quoted twice from Appearance and Reality.

The primary material is real and it is Sanskrit. Murti works from the MMK, the Prasannapadā, the Madhyamakāvatāra, Prajñākaramati’s Bodhicaryāvatārapañjikā on Bodhicaryāvatāra 9, the Śikṣāsamuccaya, Atiśa’s Satyadvayāvatāra as quoted in the BCAP, and Śaṅkara’s Brahmasūtrabhāṣya, with the verses given in Sanskrit in the notes.

The framework status is the interesting thing, and it is not what the wiki has been assuming. The two truths are here in full, including the upāya / upeya structure, the ladder image, the sorting of the Four Noble Truths, Candrakīrti’s three definitions of the conventional, the loka / aloka division, and the subject-side gloss of MA 6.23. The provisional / definitive distinction is named at p. 9 as nītārtha and neyārtha, though it is not worked. The soteriological purpose is stated as such, and the warning against emptiness becoming a view is given in Nāgārjuna’s own terms. What is absent is the Three Turnings, any Tibetan material, and any use of the framework to sort the Madhyamaka corpus. What is added is the Absolute.

Notable quotes

• “I have interpreted sunyata and the doctrine of Two Truths as a kind of Absolutism, not Nihilism.” (p. 22)

Connections

• Della Santina / della-santina-madhyamaka-western-1986 — the anchor for the anti-Murti case, and testable against this paper rather than against a report. Della Santina’s three objections are the etymology of paramārtha and saṃvṛti, the pedagogical rather than ontological status of the division, and the unmappability of Kant’s ontological dualism onto a non-ontological Madhyamaka. The second objection is the one this paper is hardest for: Murti has the pedagogical account, in full, and is an absolutist anyway.
• Hayes / hayes-nagarjunas-appeal-1994 — .2 is the sharpest form of the Advaita-reduction charge; this paper is its anticipatory refusal, and note 5 above records why the refusal fails.
• Galloway / galloway-logical-issues-1989 — dismantles two of Murti’s examples of an allegedly suspended excluded middle (the integer between 3 and 4; the rope-snake). Those come from the 1955 book, not from here, but the rope-snake reappears in this paper as the Vedāntic case Madhyamaka does not argue from, which is a different use of it.
• Gokhale / gokhale-madhyamika-dialectics-1992 — Murti is the only secondary source in Gokhale’s 1992 bibliography and the absolutist reading he sets up and refuses. With Galloway and Hayes that makes three framework-external voices against Murti, on three different grounds (the Hegelian-absolutist reading refused; the excluded middle; the Advaita reduction).
• Sprung / sprung-lucid-exposition-1979 — editor of this volume, collaborator on the Lucid Exposition, and the deflationary terminus of the same account of language. See note 6.
• Wood / wood-nagarjunian-disputations-1994 — Wood’s stated project is to rehabilitate the early European nihilist reading against the post-Murti consensus. This is what he is arguing against, in Murti’s own words. The two are exact opposites on one axis and identical on another: both take the negation to terminate in something statable, and disagree only about whether what it terminates in is a Real or a nothing.
• Streng (no wiki page) — the named opponent of the conclusion; Emptiness: A Study in Religious Meaning (Abingdon 1967). Matrix D carries him as the Wittgensteinian-positivist overlay. This is the first case in the wiki of two Matrix D rows arguing with each other directly.
• Candrakīrti / Prasannapadā / Madhyamakāvatāra — the three definitions of the conventional (PsP p. 492), the loka / aloka division (MA 6.24–26), and the subject-side gloss (MA 6.23) all come from him.
• Radhakrishnan, Stcherbatsky, Dasgupta, Bradley, K. C. Bhattacharyya (no wiki pages) — the comparative cast. Bradley is quoted twice from Appearance and Reality, pp. 21 and 28, as reaching the Madhyamaka conclusion about relations by an independent route.