Madhyamaka Wiki

Three Tibetan Positions on Tathāgatagarbha and the Four Extremes

9 min read

The claim

The standard Tibetan doxographic framing of tathāgatagarbha sets zhentong (truly existent buddha-nature, Jonang) against rangtong (universal emptiness without exception, Sakya/Geluk). The wiki’s primary-grounded sources show this binary is too coarse. A third position is consistently articulated at MA 11.34 across three lineages — Sakya (Gorampa), Nyingma (Mipham), and contemporary Khyentse-line oral transmission (Dzongsar Khyentse) — and it resists collapse to either pole.

The three-position structure:

  1. Truly existent zhentong. Dolpopa (via taranatha-essence-other-emptiness-2007): the pariniṣpanna / tathāgatagarbha is empty only of “other” (conventional adventitious phenomena), not empty of itself. The ultimate is permanent, truly existent, and possesses svabhāva. Tāranātha calls it the Great Middle Way and reads Buddhapālita, Bhāviveka, Candrakīrti, et al. as “actually” zhentong on careful reading.

  2. Freedom-from-extremes tathāgatagarbha. Gorampa at MA 11.34 (gorampa-removal-wrong-views); Mipham in Word of Chandra MA 11.34 and Supplementary Discussion (mipham-introduction-middle-way-2002); Dzongsar Khyentse (dzongsar-khyentse-madhyamakavatara-2003). Tathāgatagarbha is real in a sense that exceeds a deflationary med dgag of bden grub, anchored in the Uttaratantra nine-examples framework, but is beyond all four extremes and therefore cannot be described as “truly existent.” Dzongsar Khyentse’s articulation is the sharpest: tathāgatagarbha must be beyond all four extremes or it collapses into a non-Buddhist ātman.

  3. Deflationary med dgag treating tathāgatagarbha readings as neyārtha. Tsongkhapa: tathāgatagarbha-talk is provisional (neyārtha) and reduces on careful reading to the standard med dgag of bden grub. Ninth Karmapa (karmapa-feast-fortunate-1578) classifies tathāgatagarbha and zhentong as provisional teachings aimed at śrāvaka anxieties about emptiness.

The argument is that position (2) is genuinely distinct from (1) and (3), is cross-lineage attested, and provides one of the wiki’s strongest underused data points for the chapter of outline.

Evidence for

Position (1): Truly existent zhentong

Now primary-grounded at dolpopa-mountain-doctrine-1333 (previously held only through taranatha-essence-other-emptiness-2007’s citations). The position:

  • The pariniṣpanna / tathāgatagarbha is permanent, unchanging, and truly existent. It is empty of the paratantra and parikalpita, not empty of itself — “if the matrix-of-one-gone-thus were self-empty, it would not exist at all” (MD 213).
  • The standard universal-emptiness reading is partial; it correctly identifies the conventional as rang stong but mistakes the ultimate’s mode of emptiness.
  • Buddha-nature is the basis of the path; awakening is the manifestation of what is already there (an effect of separation), not the production of a new condition (MD 55).
  • Affirming negative (paryudāsapratiṣedha, ma yin dgag) rather than non-affirming (prasajyapratiṣedha, med dgag) is the proper mode of negation for the ultimate (MD 470).

Tāranātha’s hermeneutical move — reading the Indian Madhyamaka commentators as “actually” zhentong — is recorded as a flag rather than endorsed; Dolpopa’s own version is milder (Nāgārjuna and his successors assert the matrix in their final thought, MD 102–106).

Complication from the primary text (the Position-1/Position-2 boundary is not as clean as the binary suggests). Dolpopa himself reaches for Position-2 vocabulary: the ultimate is “the middle devoid of extremes… not exhausted as a mere non-affirming negative… a third category” beyond existence and non-existence (MD 309.7), citing Bodhibhadra and the Questions of Kāśyapa. He thus wants both “truly existent / permanent” (Position 1) and “beyond the four extremes” (Position 2). This is developed as an objection below; it does not dissolve the three-position structure but shows that the line between Position 1 and Position 2 runs through Dolpopa’s own text, not only between Dolpopa and Gorampa/Mipham.

Position (2): Freedom-from-extremes tathāgatagarbha

Gorampa at MA 11.34 (gorampa-removal-wrong-views). Gorampa’s positive tathāgatagarbha passage is anchored in the Uttaratantra nine-examples framework. The tathāgatagarbha is genuinely there (in a sense that exceeds Tsongkhapa’s neyārtha reduction); it is not a deflationary med dgag. But it is also not truly existent — it is beyond all four extremes.

The position fits Gorampa’s broader catuṣkoṭi commitment (catushkoti-must-negate-all-four-extremes). The med dgag of bden grub is the negation of only the first extreme; a real ultimate must transcend all four. Tathāgatagarbha read this way is the real ultimate as freedom from all four extremes — neither existent (Dolpopa) nor non-existent (Tsongkhapa-deflated) nor both nor neither.

**Mipham at MA 11.34 and Supplementary Discussion ** (mipham-introduction-middle-way-2002). Mipham identifies dharmadhātu with tathāgatagarbha as the ultimate-in-itself. The structural derivation from MA 6.29 (rnam grangs ma yin pa’i don dam / rnam grangs pa’i don dam) is the same one Gorampa uses 430 years earlier. Mipham agrees with the Uttaratantra anchoring; he also agrees that the Geluk med dgag of bden grub, treated as the real ultimate, fails to transcend the four extremes. Independent Nyingma articulation of the Sakya position — across two lineages and four centuries.

Dzongsar Khyentse (dzongsar-khyentse-madhyamakavatara-2003). The sharpest formulation: “buddha-nature must be beyond the four extremes.” Otherwise it collapses into a non-Buddhist ātman. Dzongsar Khyentse arrives at this from Gorampa’s sa bcad (his structural authority for the MA teachings is Gorampa, not Mipham) and transmits it to a contemporary English-speaking Western audience. The fact that a Nyingma-affiliated master selects the Sakya commentary as structural authority on tathāgatagarbha is itself evidence of cross-school stability of position (2).

Position (3): Deflationary med dgag-only

Tsongkhapa: tathāgatagarbha readings are neyārtha (provisional). On careful reading they reduce to the standard med dgag of bden grub. The Uttaratantra is a teaching for those who would otherwise fall into nihilism; it is not an alternative ultimate. (tsongkhapa-illuminating-intent-1418 and Tsongkhapa’s Drang nges legs bshad snying po per Hopkins’s annotations to taranatha-essence-other-emptiness-2007.)

Ninth Karmapa (karmapa-feast-fortunate-1578) classifies tathāgatagarbha and zhentong readings as provisional teachings aimed at śrāvaka anxieties about emptiness. Position is closer to Tsongkhapa than to Gorampa or Mipham on this specific doctrinal point — though the Karmapa disagrees with Tsongkhapa elsewhere (object of negation, conventional truth).

The cross-position convergence on excluding (1)

Sakya (Gorampa, position 2 + Drakpa Gyaltsen’s twelfth-century Sakya rejection of zhentong-shaped positions), Geluk (Tsongkhapa, position 3), Karmapa (position 3 on tathāgatagarbha), Nyingma (Mipham, position 2), Khyentse-line oral (position 2), and Geluk doxography (Thuken’s external structural reduction at thuken-crystal-mirror-1802) all converge on excluding position (1) from Madhyamaka. The convergence is across positions (2) and (3): two structurally different responses to tathāgatagarbha that nonetheless agree on the boundary verdict.

Uttaratantra anchoring

The freedom-from-extremes tathāgatagarbha is anchored in the Uttaratantra’s nine-examples framework — the tathāgatagarbha present in sentient beings as gold concealed in earth, a buddha-image wrapped in tattered cloth, and so on. Position (2) reads this as a genuine doctrinal claim about the buddha-element rather than as provisional rhetoric. Position (3) treats the same passages as neyārtha. Position (1) reads them as proof of a truly existent ultimate. The same Indian text supports three positions; what differentiates the three is the hermeneutical framework brought to it.

Evidence against / objections

  • Position (3) reduction. Tsongkhapa’s case for collapsing position (2) to position (3) is that the “freedom from four extremes” formulation is just a med dgag with extra rhetoric: a non-affirming negation of the four extremes is a med dgag. If this reduction succeeds, position (2) is not actually distinct from position (3). The Gorampa/Mipham response is that the conceptualisation of med dgag is itself a graspable position on the existence-extreme (see grasping-emptiness-is-itself-a-form-of-nihilism), and a real freedom from four extremes therefore cannot be a med dgag in the Tsongkhapa-codified sense.
  • Position (1) absorption. Tāranātha’s response is that position (2) is structurally zhentong once the Indian textual base (the late Maitreya works, the Tathāgatagarbha sūtras, the Uttaratantra) is followed through. The freedom-from-extremes language is a Sakya/Nyingma rhetorical mask over what is really a zhentong commitment. The Sakya/Nyingma response is Drakpa Gyaltsen’s twelfth-century pre-Dolpopa Sakya rejection of zhentong-shaped positions: position (2) is structurally distinguishable from zhentong because it does not commit to the truly-existent-ultimate move.
  • Dolpopa’s own freedom-from-extremes language blurs the (1)/(2) boundary (now primary-grounded at dolpopa-mountain-doctrine-1333). The cleanest way to keep positions (1) and (2) distinct is: Position 1 asserts “truly existent / permanent,” Position 2 refuses it as a catuṣkoṭi violation. But the Mountain Doctrine shows Dolpopa also using the Position-2 idiom — “a third category… free from existence and non-existence,” “not exhausted as a mere non-affirming negative” (MD 309.7, 470) — alongside “truly existent / permanent.” Two readings of this datum: (a) Dolpopa is inconsistent (he wants the existence-extreme and freedom-from-extremes at once), which is grist for the Gorampa/Thuken reductions; or (b) the real difference between (1) and (2) is finer than “asserts true existence vs denies it” — it is whether the freedom-from-extremes is compatible with a permanent self-arisen wisdom (Dolpopa: yes) or requires its negation (Gorampa, Mipham: yes). Either way, position (2)‘s distinctness from (1) cannot rest solely on the presence/absence of freedom-from-extremes language, because Dolpopa has it too. The load-bearing distinction is the permanence / true-existence commitment, not the four-extremes rhetoric. Flag this; it sharpens rather than weakens the three-position claim, but it relocates where the line is drawn.
  • Genuinely Indian or Tibetan codification? The three-position structure is articulated in Tibetan polemics. The Indian textual base does not pre-codify it. A sceptic could argue that position (2) is a Tibetan philosophical achievement that the Indian sources neither anticipated nor required — i.e. that it is a coherent position but not an Indian-attested position. The framework-necessity thesis is partly insulated against this objection: the framework is Mahāyāna, not specifically Indian-Madhyamaka, and the Uttaratantra is Indian.
  • Heterogeneity within position (2). Gorampa, Mipham, and Dzongsar Khyentse may agree on the form of the position (freedom from four extremes) but differ on its content. Gorampa is rigorously catuṣkoṭi-driven; Mipham draws more heavily on Yogācāra-Madhyamaka two-step methodology; Dzongsar Khyentse is contemporary oral. Whether the three articulations are the same position or three positions that share a structural shape is a further question.

Linked pages

Graph View

Backlinks