Definition
A non-affirming negation (prasajyapratiṣedha, Tib. med dgag) is a negation that simply removes its object without implying any positive remainder. In contrast, an affirming negation (paryudāsapratiṣedha, Tib. ma yin dgag) removes one thing while affirming another. For example, “the pot is not blue” (affirming: implies it is some other colour) vs. “there is no pot” (non-affirming: nothing further is implied).
In Madhyamaka, emptiness is classified as a non-affirming negation: it simply negates intrinsic nature (svabhāva) without affirming any positive ultimate reality in its place.
Interpretations
Tsongkhapa (in his own words): In Illuminating the Intent, Tsongkhapa insists that ultimate truth IS an object of knowledge and is obtained by rational cognition of suchness — but it is “not established through its own essence.” Being obtained by meditative equipoise does not make something truly existent. He explicitly distinguishes two senses of “ultimate” to avoid the trap: the rational cognition characterised as “ultimate” does establish phenomena, but existence through a thing’s own objective mode of being does not. Tsongkhapa is aware of the objection that his approach reifies emptiness, and his two-senses distinction is designed to pre-empt it. (From tsongkhapa-illuminating-intent-1418, Ch 9 and Ch 11)
Tsongkhapa’s prasajya / paryudāsa doctrine for the MMK (see tsongkhapa-ocean-of-reasoning-1408 Ch 1): the earlier (1407–08) and structurally more developed Ocean treatment supplies a verse-by-verse account of non-affirming negation as the form Madhyamaka negation takes. Several moves stand out. First, all four negations in MMK 1.1 (“not from itself, not from another, not from both, not without a cause”) are external (prasajya) negations: none projects an alternative. Second, Tsongkhapa engages the prasajya / paryudāsa distinction against Bhāviveka’s Tarkajvālā definition (Tarkajvālā 59b: “the external negation is just a mere elimination of the entity of a thing but does not establish anything else of that kind”). Third, he addresses two opponents at once — those who claim prasajya-negation cannot be the conclusion of an argument (it can, he answers, citing Vigrahavyāvartanī v. 26), and those who claim that establishing essencelessness by analytic cognition would render essencelessness itself truly existent (analytic cognition merely eliminates the object of negation; existence-of-essencelessness is not implicitly projected). Fourth, the “fat man who does not eat during the day” example separates prasajya from implicit-projection cases — “the sprout is without essence” carries no express implication. Fifth, autonomy is explicitly rejected (“Since it makes no sense for Mādhyamikas to propound autonomous arguments…”). Sixth, the Buddhapālita–Bhāviveka exchange is reconstructed in granular detail, with Tsongkhapa defending Buddhapālita’s compressed reductio form against Bhāviveka’s charge that it lacks premises and example. This is the most extensive scholastic treatment of the debate in the sources read so far, and the primary basis for Tsongkhapa’s claim that prasajya-negation is a thesis-establishing procedure (bearing on the no-thesis question; cf. vv-29-three-readings).
A qualifier-nuance from Ocean Homage .1.1.2 (Sarnath ed. p. 30): Tsongkhapa attributes the verbal-qualifier requirement (“ultimately,” “really,” “truly”) to the Svātantrika mādhyamikas and disavows it for Prāsaṅgika: “In the system of Buddhapālita and Candrakīrti, if any of the former or the latter modifying phrases are applied there is no need for a second modifying phrase; so the former and latter phrases are similar. Despite this they do not maintain that the modifying phrase ultimately is not applied to the object of negation. At the same time, they do not maintain that such things as arising cannot be negated without applying the modifying phrases in these two sets.” This complicates the common shorthand that makes Tsongkhapa a straightforward qualifier-requirer. What is non-negotiable for him is the substantive mode-of-existence distinction (mere-existence versus essential-existence), not the verbal qualifier, which for Prāsaṅgikas is optional but available. His disagreement with Mipham, Gorampa, and the Ninth Karmapa sits at the mode-of-existence level, not the verbal one.
Tsongkhapa’s three reductios for med dgag (Ch 12, on MA 6.34–36): In a structurally innovative reading he acknowledges as unique to him — his teacher Rendawa and his colleague Lochen Kyabchok Palsang group these stanzas differently — Tsongkhapa reads MA 6.34–36 as three unwanted consequences against any opponent (including the Svātantrika) who concedes that things exist through their intrinsic characteristic even on the conventional level. (1) Annihilation by equipoise: if a sprout existed by intrinsic characteristic, the ārya’s meditative equipoise, which does not perceive the sprout, would destroy it, making the wisdom realising emptiness “a hammer to a vase.” (2) Conventional truths withstanding ultimate analysis: if the sprout existed by intrinsic characteristic, ultimate analysis would find it, collapsing the distinction between ultimate and conventional analysis. (3) Ultimate arising not negated: once intrinsic characteristic is conceded on the conventional, ultimate arising slips back in by the back door, since “intrinsic characteristic” just is what ultimate arising means. The argument is Tsongkhapa’s central defence of med dgag as the necessary form of Madhyamaka negation: any negation weaker than non-implicative leaves intrinsic existence standing under one of these three guises. He names Bhāviveka’s brand of Svātantrika as the target. The Heap of Jewels Sutra is invoked: “the middle path perfectly discerning reality does not make phenomena to be empty. Phenomena are themselves empty.” (From tsongkhapa-illuminating-intent-1418, Ch 12)
Tsongkhapa (via Gorampa’s critique): Gorampa charges that Tsongkhapa treats emptiness as a non-affirming negation that is the proper object of inferential cognition — making emptiness a specific, identifiable cognitive content, a conceptually accessible absence. This grasps at emptiness as an object. (From gorampa-distinguishing-views-1469)
Gorampa’s critique: Gorampa’s central claim is not that med dgag is the wrong form of negation but that Tsongkhapa, having correctly used med dgag to negate bden grub, then commits two further errors: (a) treating that med dgag as itself the real ultimate, and (b) refusing to negate the conceptual apprehension of that med dgag. Gorampa’s own formulation: “to accept that, once one has negated truth, the conceptualisation of emptiness should not itself be negated is to err on one of the most important points of the textual tradition of the Madhyamaka.” The point is structural. The med dgag of bden grub is only one of the four koṭis; eliminating it leaves three more standing, and grasping at it as the ultimate is itself an apprehension of the existence-koṭi at the level of the apprehending mind. The real ultimate lies beyond all four koṭis simultaneously, accessible only through āryan equipoise that withdraws even the appellation “ultimate.” For Gorampa this is not anti-med dgag but anti-reified med dgag: the negation must remain a stepping-stone (the quasi-ultimate) rather than become a destination. This stands in tension with the Geluk position that med dgag is the proper form of Madhyamaka negation and that ultimate truth is a genuine object of knowledge; the two positions disagree, and the disagreement is real. (From gorampa-distinguishing-views-1469 )
Gorampa’s rejection of double negation: The negation of existence does not entail the acceptance of non-existence. Gorampa’s position is yod min med min — “neither existent nor non-existent.” The negation of the first koṭi does not produce the second; all four koṭis of the catuṣkoṭi are negated. (From gorampa-distinguishing-views-1469)
Dolpopa — the ultimate is an affirming negative (see dolpopa-mountain-doctrine-1333 470): Against the shared Tsongkhapa/Gorampa commitment that the ultimate is a non-affirming negation, Dolpopa identifies the ultimate nine times in the Mountain Doctrine as an affirming negation (ma yin dgag, མ་ཡིན་དགག་, paryudāsapratiṣedha). It excludes conventionalities, so it is a negative; but it is “not a mere negative, or non-affirming negative, in that it is self-arisen pristine wisdom (རང་བྱུང་ཡེ་ཤེས་) endowed with buddha-qualities of body, speech, and mind” (Hopkins 2006, p. 470; introduction, p. 23). His formula: “the negative term ‘does not possess’ is an affirming negative because those types of conceptual consciousnesses do not exist [in the mode of subsistence], and it possesses naturally luminous pristine wisdom, devoid of them” (Hopkins 2006, p. 470). This is the sharpest single formulation of the zhentong break from rangtong on the form of negation. Tsongkhapa and Gorampa disagree about almost everything else but agree the ultimate is a med dgag; Dolpopa denies precisely this. There is a complication: the same text also calls the ultimate “the middle devoid of extremes… a third category… not exhausted as a mere non-affirming negative” (Hopkins 2006, p. 335). Dolpopa wants the ultimate to be neither a med dgag nor reducible to the first koṭi, which partly converges his language with Gorampa’s freedom-from-extremes register even as their verdicts diverge — he keeps “truly existent / permanent,” Gorampa rejects it. The two are recorded as distinct, not merged; see tathagatagarbha-and-four-extremes.
Galloway — the Madhyamaka prasajya is a different negation from everyone else’s (see galloway-logical-issues-1989 pp. 6–9). The framework-external formulation, from set theory rather than from the dgag bya literature, and the sharpest statement in the wiki of what “non-affirming” costs. In the grammarians’ and the non-Madhyamaka schools’ usage, prasajya-pratiṣedha negates the predicate and paryudāsa-pratiṣedha complements the term; both leave a statement with content (“a brāhmaṇa is not here” is weaker than “a non-brāhmaṇa is here”, but it is still information). Galloway’s claim is that for the Mādhyamika any negation carrying an implication is a paryudāsa whatever its syntax, and that the properly Madhyamaka negation yields a zero statement — applied to A(x) it licenses neither A(x) nor ¬A(x), “because it allows us to conclude nothing whatever.” Term and predicate are both to be negated prasajyavat: the whole statement, subject included, is voided. His textual base is Bhāviveka’s Prajñāpradīpa on MMK 1.1, where “things do not arise” is rejected as annihilationist and “things arise, just not from themselves” as eternalist, and only “things do not (arise from themselves)” is accepted, with no implication drawn; and the Tibetan translators’ non-literal med par dgag pa / ma yin par dgag pa, coined on two verbs of being whose distinction of implication neither Sanskrit nor English marks. Two consequences the Tibetan literature also reaches by other routes: the negation of the bhāva does not support abhāva, which is “itself also explicitly repudiated”; and the absence of a pratijñā at VV 29 just is the zero statement (n. 5, after Ruegg 1983 — see No-thesis). Set beside Tsongkhapa this is an instructive contrast rather than an agreement: both hold the negation non-implicative, but Tsongkhapa holds that it can be the conclusion of an argument and establishes the mere elimination of essence, where Galloway holds that it establishes nothing at all. Gorampa’s yod min med min is closer to the zero statement than Tsongkhapa’s med dgag is.
Haribhadra and Ratnākaraśānti divide over a single sūtra phrase (see astasahasrika-84000-2026, Introduction and nn. 60, 95). The clearest Indian instance in the wiki of the distinction being applied, by named opponents, to one scriptural sentence. The Aṣṭasāhasrikā Prajñāpāramitā’s most-cited line is Subhūti’s “that mindset is free of a mindset” (acitta, 1.7). Haribhadra, as a Mādhyamika, reads the privative as a non-affirming negation: such a mind appears but does not ultimately exist. Ratnākaraśānti, as a Yogācārin, reads it as an implicative negation (མ་ཡིན་པར་དགག་པ་): it is a mind insofar as it is aware, and free of any mind insofar as it is free of grasping. The Sanskrit privative a- carries both, and the sūtra does not disambiguate.
Two things follow. First, the med dgag / ma yin dgag fault line is not a Tibetan scholastic development applied retrospectively to Indian materials; it is already the operative division between the two great Indian commentators on the founding sūtra, and it separates them along exactly the Madhyamaka / Yogācāra line. Second, this bears directly on Galloway’s claim above that the Tibetan pair med par dgag pa / ma yin par dgag pa is a non-literal coinage marking a distinction the Sanskrit does not. The translation record supports him in part and complicates him in part: the Chinese tradition rendered these privatives systematically as “non-x” and the Tibetan systematically as “x-less”, which is the paryudāsa and prasajya readings distributed by translation tradition rather than by argument. But n. 60 records Tibetan translators going the other way when a commentator pushed them — Ngok Lotsāwa renders acittam as sems ma yin pa (“not a mind”) in Ārya Vimuktisena’s Vṛtti (Toh 3787, F.45.a.4). So the Tibetan choice was a construal made under commentarial pressure, not an artefact of the language.
Tsongkhapa on prasajya negation and the excluded middle (via Jinpa): In jinpa-tsongkhapa-qualms-1999, Jinpa shows Tsongkhapa arguing that even the non-affirming negation (prasajya) must presuppose the law of excluded middle to be effective. He cites Vigrahavyāvartanī 26b: “If the absence of intrinsic being is reversed, intrinsic being becomes established.” Without subscribing to fundamental principles of logic, one is left only in a state of indecision. The argument is aimed at the “no-thesis” proponents who reject all logical argumentation. For Tsongkhapa, the ultimate insight into the middle way must be an active state of “knowing,” not a withdrawal of cognitive activity; otherwise it is indistinguishable from Hva-shang’s no-thought meditation.
Textual loci
- MMK 13:18 — emptiness as antidote to all views, not itself a view
- MMK 22:11 — “empty,” “non-empty,” “both,” and “neither” should not be stated
- Aṣṭasāhasrikā Prajñāpāramitā 1.7 — “that mindset is free of a mindset” (acitta), the phrase over which Haribhadra and Ratnākaraśānti divide med dgag against ma yin dgag
- dolpopa-mountain-doctrine-1333 470 — the zhentong counter-case: the ultimate as affirming negative (ma yin dgag), nine-fold, against the rangtong med dgag
- Candrakīrti’s Prasannapadā on the distinction between prasajya and paryudāsa negation
- Tsongkhapa’s Ocean of Reasoning — systematic treatment of emptiness as med dgag (see tsongkhapa-ocean-of-reasoning-1408 Ch 1: the prasajya / paryudāsa distinction; the four-negation external-reading of MMK 1.1; the Buddhapālita-defence)