What this matrix compares
If the object-of-negation matrix asks what does Madhyamaka deny?, this matrix asks what survives the denial? — i.e. what is the status of conventional truth (saṃvṛtisatya / ཀུན་རྫོབ་བདེན་པ་, kun rdzob bden pa) once the negation has done its work?
The page now carries two matrices:
- The Indian matrix — the three founding Indian Mādhyamikas whose handling of saṃvṛti the Tibetan debate inherits: Bhāviveka (who takes the Sautrāntika view as conventional truth), Śāntarakṣita (who takes Yogācāra/Cittamātra as conventional truth), and Candrakīrti (the Prāsaṅgika, whose conventional is hardest to state because he declines a grounded one). They are laid out against five axes adapted from Ruegg’s six criteria, which “all converge on the status of saṃvṛti”: (i) what conventional truth is / its basis; (ii) whether self-characteristic (svalakṣaṇa) is admitted on the conventional level; (iii) whether external objects are real conventionally; (iv) what warrant / grounding conventional truth has (philosophical pramāṇa-grounding vs everyday-practice grounding only); (v) whether the veridical/distorted (tathya/mithyā) distinction is objective, and whether the paramārthatas qualifier is used. A final primary-text column records the source on which each row rests.
- The Tibetan matrix — the six Tibetan thinkers who address the question with a developed primary-text position, one per major lineage, diverging along five axes: (i) what makes something conventionally true; (ii) what kind of warrant; (iii) whether conventional truth is internally graded; (iv) what the conventional/ultimate relationship is; (v) whether the buddha-side perception of conventionalities counts as conventional truth at all. A final primary-text column records the source on which each row rests.
Both matrices are constructive — they lay the readings side by side so the divergence is structurally visible. The wiki’s adjudicative arguments live elsewhere (grasping-emptiness-is-itself-a-form-of-nihilism for the Gorampa-Tsongkhapa contrast at MA 6.23; zhentong-is-refined-cittamatra-not-madhyamaka for the Sakya/Geluk verdict on Dolpopa’s three-natures conventional; madhyamaka-is-method-not-standalone-system for the structural reading the Indian matrix makes visible).
The Indian matrix
The three Indian founders of the methodological streams the Tibetans call Svātantrika and Prāsaṅgika diverge most sharply not on the ultimate (where all three reach niḥsvabhāva) but on saṃvṛti. The two Svātantrikas each import a positively-grounded conventional ontology from a lower tenet system — Bhāviveka the Sautrāntika’s external realism, Śāntarakṣita the Yogācāra’s mind-only — whereas Candrakīrti the Prāsaṅgika declines any such grounding and keeps conventional truth “in accordance with the world” alone.
| Thinker | What conventional truth is | Svalakṣaṇa conventionally? | External objects real conventionally? | Warrant / grounding | Tathya/mithyā + paramārthatas qualifier | Primary text |
|---|---|---|---|---|---|---|
| Bhāviveka (Sautrāntika-Svātantrika) | The world’s appearances as the Sautrāntika construes them — externally real entities bearing their own characteristics; a momentary, “cinematic” account of process (the goer as “a collection of conditioned factors originating continuously in another place”, Prajñāpradīpa Ch 2 on MMK 2.22c, ames-prajnapradipa-ch2-1995) | Yes — self-characteristic (svalakṣaṇa) is admitted at the level of saṃvṛti (one of Ruegg’s six criteria; ames-prajnapradipa-ch2-1995) | Yes — external objects are real on the conventional level (the Sautrāntika construal; Bhāviveka) | Robust: terms are commonly established (ubhayasiddha) between disputants; conventional reality supports autonomous inference | Objective veridical/distorted distinction (real elephants vs mirror-elephants) — which Candrakīrti rejects as objective; the paramārthatas (“ultimately”) qualifier is affixed to negations | Prajñāpradīpa Ch 2 (MMK 2.22c; the paramārthatas qualifier) — ames-prajnapradipa-ch2-1995 |
| Śāntarakṣita (Yogācāra-Svātantrika) | Mere appearance (snang ba tsam) that is the mind’s own projection — external objects refuted, retained as mind-only; causally efficient, “satisfactory only when unexamined, subject to birth and destruction” (Madhyamakālaṅkāra v. 64) | Yes, but mind-only — reflexive awareness (རང་རིག་, svasaṃvedana) is “the sine qua non of valid cognition on the conventional level”; the causally-efficient thing is the sole authentic object of valid cognition (MA General Introduction, raw pp. 102–103) | No — external objects do not exist; phenomena arise “by the power of the mind” (MA v. 92) | Robust, Dharmakīrtian conventional pramāṇa — “when I investigate outer phenomena, I take the Sautrāntika as my starting point” (Dharmakīrti, cited by Mipham); the two valid reasonings held without contradiction | Sautrāntika veridical/mistaken (yang dag / yang dag min) admitted conventionally; the principal grading sits on the ultimate side — approximate (rnam grangs pa’i don dam) vs actual (rnam grangs ma yin pa’i don dam) | Madhyamakālaṅkāra vv. 64, 91–93 + the five propositions (svasaṃvedana conventional, Gen. Intro. pp. 102–103) — shantarakshita-madhyamakalankara / Madhyamakālaṅkāra |
| Candrakīrti (Prāsaṅgika) | Mere worldly consensus — things “accepted just as they are known in the world” (‘jig rten gyi tha snyad), unanalysed; saṃvṛti glossed both as “utterly obscured” by ignorance and as social convention (sprung-lucid-exposition-1979 pp. 230–232); “the chariot… accepted in accordance with the world” (MA 6.160) | No — negates all six synonyms of intrinsic existence even conventionally; no svalakṣaṇa at any level (the load-bearing contrast with Bhāviveka; Prāsaṅgika-Svātantrika) | Accepted “in accordance with the world” — neither reduced to mind nor given external-realist grounding; he supplies no positive analysis of them | Everyday-practice grounding only, not philosophical grounding (Phya pa’s Q5; westerhoff-candrakirti-2024); the relative is the object of mistaken cognition (Prasannapadā §X) — no objective pramāṇa-grounding | Veridical/distorted drawn only from the worldly perspective, never objectively (Tsongkhapa’s reading of Candrakīrti); the paramārthatas qualifier is rejected as meaningless (sections IX–X; candrakirti-prasannapada-mmk1-excerpt) | Madhyamakāvatāra 6.160 (and 6.23–28); Prasannapadā on MMK 1.1 §IX–X — candrakirti-prasannapada-mmk1-excerpt / sprung-lucid-exposition-1979 |
What the Indian matrix shows
One axis does all the work: whether the conventional is grounded from outside. Bhāviveka and Śāntarakṣita differ on which lower system supplies the conventional inventory (Sautrāntika external realism vs Yogācāra mind-only) but agree on the deep structural point — the conventional is a positively-grounded ontology, admitting svalakṣaṇa, warranted by a robust pramāṇa, with an objective internal grading, and compatible with the paramārthatas qualifier. Candrakīrti declines every one of these: no svalakṣaṇa even conventionally, no philosophical grounding, no objective veridical/distorted distinction, no qualifier. This is the Indian-level exhibit of madhyamaka-is-method-not-standalone-system: the Svātantrika conventional layer is a received input swapped between Sautrāntika and Yogācāra registers without touching the ultimate analysis, while the Prāsaṅgika declines a conventional ontology by design. It is also why the Prāsaṅgika conventional is “harder to assert” — it is deliberately deflationary: a use-level convention, not a content-level ontology. (Modern gloss: Westerhoff reads Candrakīrti’s conventional as transactional-pragmatic — “those truths which are pragmatically successful within this framework” — with “semantic insulation” between the truths and no rock-bottom layer, westerhoff-candrakirti-2024. Garfield supplies the analytic-idiom version of the same deflationary conventional, framework-internally: in garfield-fundamental-wisdom-1995 the via media on causation is reached “by taking conventions as the foundation of ontology, hence rejecting the very enterprise of a philosophical search for the ontological foundations of convention,” and “to exist is to exist conventionally, dependently” — conventional truth held “in accordance with the world” with no privileged level and no positive Madhyamaka inventory. Like Westerhoff, Garfield is a modern gloss on the Candrakīrti row, not a matrix entry — he reads the Prāsaṅgika conventional rather than supplying a rival one.)
The “buddha cognises conventionalities?” axis is omitted here. It is poorly attested for the Indian founders (Bhāviveka especially) and is better left to the Tibetan matrix, where the primary texts speak to it directly. Śāntarakṣita/Mipham do speak to it via the equipoise/post-meditation structure (no assertions in equipoise; the Ārya functions within thought-and-word in post-meditation), but Bhāviveka and Candrakīrti do not give a developed same-kind answer, so the column would be one-third empty.
The Tibetan matrix
The six Tibetan thinkers below inherit the Indian saṃvṛti problem and re-pose it on their own axes. (Note the structural rhyme: Gorampa’s snang tsam conventional descends from the same deflationary impulse as Candrakīrti’s; Tsongkhapa’s robust pramāṇa-warranted conventional rhymes with the Svātantrika grounding; Mipham, following Śāntarakṣita, cross-cuts — a validly-cognised warrant on the grounded side but a “mere appearance” account of what conventional truth is; Dolpopa’s three-natures conventional is Yogācāra-import — see zhentong-is-refined-cittamatra-not-madhyamaka.)
| Thinker | What is conventional truth? | Warrant | Internal grading | Relation to ultimate | Buddha cognises conventionalities? | Primary text |
|---|---|---|---|---|---|---|
| Tsongkhapa | Phenomena posited through mere conceptualisation in dependence on bases of designation; not mere illusion — vases exist, rope-snakes do not. Three criteria (now primary-grounded on MMK directly at tsongkhapa-ocean-of-reasoning-1408 Homage .1.1.2, with the parallel MA-anchored statement at tsongkhapa-illuminating-intent-1418 Ch 11): (a) acknowledged within unanalysed conventional cognition (tha snyad pa’i shes pa = Candrakīrti’s ma dpyad pa’i shes pa), (b) not invalidated by other conventional valid cognition, (c) not invalidated by ultimate analysis. Ocean Homage .1.1.2 also supplies the load-bearing primary statement that conventional things are “merely posited by nominal convention” and “established by authoritative cognition” without contradiction — the “merely” precludes inherent existence but not pramāṇa | Full pramāṇa-warrant; valid worldly cognition delivers conventional truth — and “deferring to the world” appeals to the unanalysed perspective shared by reflective philosophers and ordinary persons alike, not the lowest common denominator of cowherd-cognition | Veridical (unimpaired senses) vs distorted (impaired senses), but only from the worldly perspective; from the ārya perspective both are equally mistaken (MA 6.24–25, Ch 11) | Two natures of each phenomenon (tsongkhapa-ocean-of-reasoning-1408 Ch 24 p. 406): “each phenomenon has two natures, an ultimate and a conventional,” found by two distinct cognitive processes — explicitly NOT a single nature seen from two perspectives. At MA 6.23 (tsongkhapa-illuminating-intent-1418) the same point is articulated as “single nature, distinct conceptual identities.” The 1407–08 Ocean statement is the sharper anti-subject-side formulation; both texts together place Tsongkhapa on the object-side of the subject/object two-truths divide — the lone object-side voice against the four-voice subject-side coalition (Gorampa, Karmapa, Mipham, Dzongsar Khyentse) — note that this counts texts/thinkers; this wiki’s folds Dzongsar Khyentse under Gorampa (his 2003 commentary is arranged on Gorampa’s sa bcad and transmits Gorampa’s reading), so on a voice-count records a three-voice subject-side coalition (Gorampa[+Khyentse], Karmapa, Mipham) against Tsongkhapa | Yes — at buddhahood, equipoise and post-equipoise no longer alternate; a single gnosis with two modes (ji lta ba and ji snyed pa) cognises both truths within one moment, and the dualistic appearances of conventional truth that appear in the buddha’s perception are not erroneous (Ch 25, on MA 11.10–17) | Ocean of Reasoning Ch 24 (two natures) — tsongkhapa-ocean-of-reasoning-1408; Illuminating the Intent Ch 11 (the three criteria) — tsongkhapa-illuminating-intent-1418 |
| Gorampa | Mere appearance (snang tsam) exhausted by deluded perception; conventional truth obscures both subject and object (MA 6.28) | Deluded-perception-bound; conventional cognition is the cognition of saṃvṛti’s misleading character | Two-level ultimate (quasi-ultimate rnam grangs pa’i don dam and real ultimate mtshan nyid pa’i don dam) — the quasi-ultimate is “compared to the stainless equipoise of āryans, only a conventional truth” (.1.2.1–2). Worldly conventional itself is not sub-graded | Not single-nature-distinct-identity; the two truths are not collapsible by qualifier (MA 6.23 five-step reductio) — the single-nature reading collapses the truths into one another | Buddhas have no saṃvṛti cognition in their own continua; conventionalities appear to others via the buddha-side compassionate display | Distinguishing the Views — gorampa-distinguishing-views-1469; Removal of Wrong Views (MA 6.28; the MA 6.23 reductio) — gorampa-removal-wrong-views |
| Mipham | Mere appearance that is nonetheless real at its own level — what undeceiving conventional cognition (tha snyad pa’i tshad ma) finds, following Śāntarakṣita’s Yogācāra-Madhyamaka synthesis. The difference between the two truths is epistemic, not ontological: “this double identity is the conclusion of correct reasoning; it has no reality on the level of being” (MA 6.23; Padmakara fn. 95). Neither Gorampa’s pure deluded-perception snang tsam nor Tsongkhapa’s bden grub-qualified posit (mipham-introduction-middle-way-2002) | Conventional valid cognition of the Dharmakīrtian kind inherited from Śāntarakṣita; reflexive awareness (svasaṃvedana) retained on the conventional level; robust at its own level — on the warrant axis Mipham sits with the grounded side, not with Gorampa | Correct vs mistaken conventional (validly vs non-validly cognised) on the worldly side; the load-bearing grading is on the ultimate — the two ultimates: approximate (rnam grangs pa’i don dam, a non-affirming negation / med dgag) and actual (rnam grangs ma yin pa’i don dam, beyond all four extremes) | The two truths are indivisible; “single nature, distinct identities” is a conclusion of reasoning with “no reality on the level of being”, not an object-side ontological claim — placing Mipham on the subject-side coalition (with Gorampa, the Karmapa, Dzongsar Khyentse) against object-side Tsongkhapa. The actual ultimate is identified with the dharmadhātu / tathāgatagarbha | Yes, but non-dually. At buddhahood the dualistic mind and mental factors cease (“the mind is stopped”, MA buddhahood ch. v. 17), yet non-dual primordial wisdom (ye shes) “perceives all things in their multiplicity” (ji snyed pa) — a “wisdom perception” of one taste, not the dualistic perception of ordinary beings. Contrast Tsongkhapa: the dualistic appearances do not remain. (Buddha-specific answer developed at mipham-introduction-middle-way-2002, MA vv. 11–17 — full treatment on the source page; the equipoise/post-meditation answer at shantarakshita-madhyamakalankara covers the ārya, not the buddha.) | The Word of Chandra (MA 6.23) — mipham-introduction-middle-way-2002; Madhyamakālaṅkāra commentary, the two ultimates — shantarakshita-madhyamakalankara |
| Ninth Karmapa | At no analysis, worldly conventions are accepted; at slight analysis, phenomena seen as empty of inherent nature; at thorough analysis, even emptiness transcended | Three-stage pramāṇa — each stage has its own warrant; “no-analysis” warrant is not full pramāṇa but is the level at which conventional transactions operate | Three stages of analysis — the grading is on the analysis, not on the truth; conventional truth is the level where no analysis is in operation | Three stages mean conventional and ultimate are progressively reached, not divided by single-nature/distinct-identity distinction | Self-awareness does not exist even conventionally (agreeing with Tsongkhapa, against Gorampa and Shakya Chokden); buddha-nature is provisional and conventional, not ultimate (against Dolpopa) | Feast for the Fortunate — karmapa-feast-fortunate-1578 |
| Dolpopa | Imputational and other-powered natures (the first two of the three natures); empty of both their own entities and others’ entities (“non-existent emptiness” + “existent emptiness”) | Conventional warrant is Yogācāra-shaped — three-natures analysis structures the conventional; paratantra is the locus of dependent functioning | Two-tier conventional: imputational (non-existent) and other-powered (existent). The three natures are read as a graded conventional; the thoroughly established (tathāgatagarbha) is the ultimate, never conventional | Conventional and ultimate are categorially separated: conventional is self-empty; ultimate is other-empty (empty of conventionalities, not of itself) | Buddha-side primordial wisdom cognises the thoroughly established; conventionalities are cognised only insofar as they are empty of their own entity — not as truths in their own right | Mountain Doctrine — dolpopa-mountain-doctrine-1333 |
| Gendün Chöpel | Fictions (Adornment ¶15: “regarding these conventional phenomena that have the nature of fictions, it comes down to the fact that there is nothing suitable other than a mere decision by this mind of ours, which is itself a source of fictions”) that the unenlightened mind has no choice but to act on (¶19, ¶51); not mere appearances of something but mere mind-decisions full stop | No pramāṇa-warrant for conventional truth in any robust sense (the 21 refrain-verses ¶56–¶76 each end in tha snyad tshad grub ‘jog la blo ma bde, “I am uncomfortable about positing conventional validity”). GC explicitly rejects (¶100) the no-tenets / tenets distinction as a basis for sub-grading the conventional | The assertions-for-oneself / assertions-for-others distinction (¶51, ¶80–¶84) supplies a context sub-grading but not a content sub-grading. Necessary involuntary conventions (being burned by fire, cooled by water) are assertions-for-others; the yogin in equipoise makes none of these assertions in her own system | Strict incommensurability / chasm-reading (¶196): “Emptiness completely contradicts the world … there is no commonality whatsoever between the way things are perceived by the ignorant mind and the way they are perceived by the enlightened mind.” The Geluk doctrine of “the compatibility of the two truths” is a gross error | No in any way meaningful to the unenlightened (¶22–¶28, ¶90–¶93). The buddha-fields, the inconceivable powers, the tantric iconography are calibrated to the audience’s predilections (¶23); the actual nature of buddha-cognition is literally inconceivable — “this secret of yours cannot be told” (citing MA XII.39d at ¶23); the Gorampa-style “compassionate display” answer is unstable on GC’s reading because it preserves a continuity of cognitive content between unenlightened and enlightened sides that the chasm-reading denies | Adornment for Nāgārjuna’s Thought (Klu sgrub dgongs rgyan) — lopez-madmans-middle-way-2006 |
What the Tibetan matrix shows
Two divergent shapes for “what makes conventional truth a truth.” Tsongkhapa’s mere conceptualisation in dependence on bases of designation makes conventional truth robustly real at its level — the rope-snake / vase asymmetry shows what robustness amounts to. Gorampa’s mere appearance exhausted by deluded perception makes conventional truth an epistemic-deception object: it is not nothing, but it is not warranted by pramāṇa; it is the object that saṃvṛti-cognition reaches because saṃvṛti-cognition is what it is. The Karmapa’s three-stage answer is structurally distinct from both: conventional truth is the level at which no analysis is operating, with each subsequent analytic stage progressively dissolving it. Dolpopa’s three-natures answer is structurally Yogācāra: conventionality is the imputational + other-powered territory, with the thoroughly established categorially distinct.
Tsongkhapa’s three criteria are now primary-grounded. Until tsongkhapa-illuminating-intent-1418 was summarised, Tsongkhapa’s “robust conventional” was characterised on the matrix only abstractly. Ch 11 supplies the operational test: a fact is conventionally existent iff (a) it is acknowledged within unanalysed conventional cognition; (b) it is not invalidated by other conventional valid cognition; (c) it is not invalidated by analysis probing the ultimate nature of reality. This makes the rope-snake / vase asymmetry diagnostic: rope-snakes fail criterion (b); intrinsically existent vases fail criterion (c); ordinary vases pass all three. The three criteria are also the primary-text basis for distinguishing Tsongkhapa’s “deferring to the world” from a defeatist or relativist appeal — the relevant cognition is the unanalysed perspective shared by everyone, not the lowest worldly common denominator.
The qualifier question reappears at the conventional column. Tsongkhapa’s qualifier-tradition has its conventional-side correlate: bden grub is what the negation reaches, and conventional phenomena survive without a qualifier. The other three thinkers do not need a bden grub qualifier on the ultimate side because their conventional-side handling is already non-robust (Gorampa) / progressive (Karmapa) / structurally Yogācāra (Dolpopa). The qualifier-tradition and the robust-conventional are reciprocal commitments.
The Karmapa is the structural outlier. On the ultimate-side qualifier question (see object-of-negation) the Karmapa converges with Gorampa against Tsongkhapa. On the conventional-side question he diverges from Gorampa: self-awareness does not exist even conventionally (agreeing with Tsongkhapa); buddha-nature is provisional and conventional (against Dolpopa). The Karmapa’s three-stage structure is not reducible to either Geluk or Sakya — it is a Kagyü innovation that handles the conventional through progressive analysis rather than through grading-internal-to-the-truth.
Mipham cross-cuts the matrix rather than occupying one corner of it. This is why he was initially held off the matrix (see the resolved note under Open questions). On what conventional truth is his “mere appearance, yet real at its own level” reads close to Gorampa’s deflationary snang tsam; but on warrant his Dharmakīrtian conventional pramāṇa (inherited from Śāntarakṣita, svasaṃvedana retained) sits with the grounded side, near Tsongkhapa and the Svātantrikas. His relation-to-ultimate answer — the indivisibility of the two truths and the two-ultimates structure (approximate med dgag / actual beyond the four extremes) — is sui generis, shared in shape with Gorampa’s quasi-ultimate/real-ultimate but derived through the Yogācāra-Madhyamaka synthesis rather than Gorampa’s prasaṅga-only route. The “epistemic, not ontological” reading of the two-truths relation places him with the subject-side coalition against object-side Tsongkhapa (cf. object-of-negation). So Mipham does not add a seventh structural type; he is the diagnostic case showing the five axes are genuinely independent — a thinker can take the grounded warrant and the deflationary account of appearance at once.
Dolpopa’s category-shape is Yogācāra-import (now primary-grounded at dolpopa-mountain-doctrine-1333). Three of the five diagnostic axes (warrant, grading, relation-to-ultimate) classify Dolpopa as structurally closer to Yogācāra than to the other three Mādhyamikas. This is the structural basis for Red mda’ ba’s “refined Cittamātra” reduction (cf. zhentong-is-refined-cittamatra-not-madhyamaka). The Mountain Doctrine confirms the two-tier conventional directly: self-empty conventionalities “do not appear to wisdom of reality, since they do not exist in the mode of subsistence” (MD 209, 527), and “the spheres of wisdom and consciousness are separate” (MD 527). The matrix therefore shows the structural argument visually: Dolpopa’s conventional-side handling is incommensurable with the rangtong cohort’s, not merely contrary to it.
Buddha-cognition of conventionalities is a cleanly six-way split. Tsongkhapa: yes, in post-equipoise — the dualistic appearances appear and are non-erroneous. Mipham: yes, but non-dually — the dualistic mind ceases while non-dual ye shes cognises multiplicity (ji snyed pa) as a “wisdom perception”, so dualistic appearance does not survive (the sharp contrast with Tsongkhapa). Gorampa: no in their own continua, yes via compassionate display. Karmapa: refuses the question of buddha-side conventionality (buddha-nature is provisional, conventionality is the no-analysis level). Dolpopa: buddha-side wisdom cognises only the thoroughly established; conventionalities are cognised only in their structural emptiness. Gendün Chöpel: no in any way meaningful to the unenlightened — the chasm-reading destabilises even Gorampa’s “compassionate display” answer because it denies any continuity of cognitive content between the unenlightened and enlightened sides. No two of the six answers coincide. This is the sharpest single column in the matrix, and GC’s chasm-reading sits at the maximally apophatic pole.
GC sharpens the conventional-side question by rejecting both the warrant and the grading. Tsongkhapa’s robust-conventional and Gorampa’s mere-appearance both still presuppose some conventional sub-structure — Tsongkhapa’s three criteria, Gorampa’s deluded-perception register. The Karmapa’s three stages and Dolpopa’s three natures are sub-gradings of the conventional in a different mode. GC refuses all four substructures: there is no pramāṇa-warrant, no progressive analysis, no Yogācāra hierarchy, and no veridical / distorted sub-grading. What remains is the context-relative distinction between assertions-for-oneself and assertions-for-others — a use distinction, not a content distinction. This is the conventional-side correlate of GC’s chasm-reading: just as the enlightened side cannot be reached by any continuation of conventional cognition, the conventional side cannot be sub-graded by any extension of unenlightened cognition.
Cross-school witness count (Tibetan matrix)
For each diagnostic feature: how many of the six Tibetan thinkers hold it?
- Conventional truth as robust posit-by-designation, pramāṇa-warranted: 1/6 (Tsongkhapa) — Mipham shares the grounded warrant but not the object-side posit-by-designation
- Conventional truth as mere appearance / deluded-perception-bound: 1/6 (Gorampa) — the Karmapa’s “no-analysis level” is structurally adjacent, and Mipham’s “mere appearance real at its own level” is adjacent on the what-is-it axis only
- Conventional truth structured by Yogācāra three-natures: 1/6 (Dolpopa)
- Conventional truth as mere fictions with no warrant or sub-grading: 1/6 (Gendün Chöpel)
- Two-level ultimate folding the quasi-ultimate into conventional: 2/6 (Gorampa; Mipham’s approximate/actual two-ultimates is the closest structural twin, derived independently) — the Karmapa’s three-stage structure is a looser analogue
- Buddha cognises conventionalities in own continuum: 2/6 (Tsongkhapa; Mipham) — but on different terms: Tsongkhapa retains non-erroneous dualistic appearance, whereas Mipham’s buddha cognises multiplicity (ji snyed pa) through non-dual ye shes with the dualistic mind ceased. 1/6 if the criterion is restricted to dualistic cognition (Tsongkhapa alone)
The near-uniform one-each pattern across these axes shows that the conventional-side question is more divergent than the object-of-negation question. On the dgag bya matrix, Tsongkhapa is the 5-vs-1 outlier; on the conventional-truth matrix, each thinker holds a structurally distinct position. GC at the maximally-apophatic pole and Mipham as the axis-independent cross-cut both sharpen the structural diversity rather than expanding into sub-clusters of existing positions.
Linked pages
- Two Truths — the structural frame
- Svabhāva — what the qualifier negates; conventional-side qualifier reciprocity
- Rangtong-Zhentong — Dolpopa’s row’s structural source
- dolpopa-mountain-doctrine-1333 — primary source for Dolpopa’s row (the three-natures two-tier conventional)
- object-of-negation — sister matrix; the qualifier-question and conventional-question are reciprocal
- grasping-emptiness-is-itself-a-form-of-nihilism — adjudicative argument on Tsongkhapa-vs-Gorampa
- zhentong-is-refined-cittamatra-not-madhyamaka — adjudicative argument on Dolpopa’s row
- framework-internal-debate-is-productive — meta-argument the matrix exhibits
- madhyamaka-is-method-not-standalone-system — the conventional-side claim that the Prāsaṅgika declines a positively-grounded conventional ontology, which the Indian matrix makes visible (Bhāviveka/Śāntarakṣita supply a grounded conventional, Candrakīrti declines one); object-side Tsongkhapa is the hard case for it on the Tibetan matrix
- Indian matrix scholars: Bhāviveka (Sautrāntika conventional), Candrakīrti (the deflationary Prāsaṅgika conventional), Śāntarakṣita (Yogācāra conventional)
- Indian matrix sources: ames-prajnapradipa-ch2-1995 (Bhāviveka’s svalakṣaṇa + cinematic motion), candrakirti-prasannapada-mmk1-excerpt / sprung-lucid-exposition-1979 / westerhoff-candrakirti-2024 (Candrakīrti), Madhyamakālaṅkāra / shantarakshita-madhyamakalankara (Śāntarakṣita)
- sprung-lucid-exposition-1979 — Western-academic deflationary contrast at MMK 24
- mipham-introduction-middle-way-2002 — Mipham Tibetan-matrix row
- lopez-madmans-middle-way-2006 — primary source for the GC row