Recognizing Reality: Dharmakīrti's Philosophy and Its Tibetan Interpretations

Terminology note

Dreyfus uses several English terms that differ from our glossary-preferred forms. Throughout this page (as in the rest of the wiki) we use the preferred terms in our prose; direct quotations retain Dreyfus’s wording. Mappings: self-awareness (Dreyfus: “self-cognition” / “apperception”, རང་རིག་); investigative thought (Dreyfus: “correct assumption”, ཡིད་དཔྱོད་); appearance-without-ascertainment (Dreyfus: “inattentive cognition”, སྣང་ལ་མ་ངེས་པ་); conceptual cognition (Dreyfus: “conceptual thought” / “conception”, རྟོག་པ་).

Main argument

Dreyfus reads Dharmakīrti’s epistemological system through the lens of a deep, philosophically substantive disagreement between its later Tibetan interpreters: the antirealist Sa-gya (ས་སྐྱ་) tradition centred on Sakya Paṇḍita’s Tshad ma rigs gter (the Treasure, i.e. pramanayuktanidhi) and developed by gorampa and sakya-chok-den; and the moderately realist Ge-luk (དགེ་ལུགས་) tradition of gyel-tsap, kay-drup, and Ge-dün-drup, descended from the “new epistemology” (ཚད་མ་གསར་མ་) of ngok-lotsawa and cha-ba. His central thesis is that the problem of universals is the load-bearing axis around which Dharmakīrti’s ontology, philosophy of language, and epistemology all turn — and that neither Tibetan interpretation succeeds in fully resolving the tensions internal to Dharmakīrti’s system (p. 14).

Dharmakīrti’s system rests on a strict dichotomy between specifically characterised phenomena (svalakṣaṇa, རང་མཚན་), which are real, momentary, causally efficient individuals, and generally characterised phenomena (sāmānyalakṣaṇa, སྤྱི་མཚན་), which are unreal, permanent, causally inefficient conceptual constructs (pp. 67–68). Perception apprehends only the former; inference apprehends only the latter (p. 68). This produces what Dreyfus identifies as the central tension of the whole tradition: if perception and inference have categorically different objects, how can they cooperate in the cognitive process at all? (pp. 274–75, 395–96).

Dreyfus argues that the Ge-luk response was revisionist: following cha-ba and Dharmottara, they reintroduced moderate realism — admitting universals as real properties dependent on their instances — and reinterpreted perception as cognitively active (eliminating superimpositions, ascertaining objects), thereby weakening the bare-sensing model in favour of a propositionally articulated perception (pp. 366–67, 443–44). The Sa-gya response was conservative: sakya-pandita retained Dharmakīrti’s bare-sensing model and his radical dichotomy, and bridged perception and conception through self-awareness (svasaṃvedana, རང་རིག་) — “sense consciousness is like the fool who sees; conception is like a blind skilful speaker; self-cognition is like a person with complete senses, who introduces one to the other” (Sa-paṇ quoted p. 397). Dreyfus judges Sa-paṇ closer to Dharmakīrti on the bare-sensing point, but closer to Dignāga (not Dharmakīrti) on apoha and the signifier/signified analysis (pp. 274–75, 446).

Chapter breakdown

The book is in two parts (“Book One” on ontology and language; “Book Two” on epistemology), totalling 27 chapters plus two introductions and a conclusion. Section labels below give the move Dreyfus makes, not just the topic.

Introduction I — methodological defence of treating Tibetan commentators as genuine philosophical partners; commentary as a mode of philosophising (pp. 3–7).

Introduction II — historical reconstruction: situates Sa-paṇ as the pivot of the Tibetan reception, dating the conflict of interpretations to the late-14th/15th century, with Yak-don (1348–1414) reviving interest in the Treasure against the dominant realist reading inherited from Cha-ba (pp. 23–28). Argues that the modern reified labels “Sa-gya” and “Ge-luk” obscure the more fluid 15th-century situation but are unavoidable for a topic where doctrinal lines really do follow sectarian ones (pp. 33–34).

Ch. 1 — Reads Dharmakīrti’s whole project as a refusal of Nyāya substance-ontology; distinguishes five senses of “realism” and restricts his usage to senses (2) [universals] and (5) [commonsense objects] (pp. 54–55).

Ch. 2 — Lays out Dharmakīrti’s three “binary equivalences” — thing/non-thing, impermanent/permanent, svalakṣaṇa/sāmānyalakṣaṇa — and the three identity conditions (spatial, temporal, entity-determination) that disqualify universals from reality (pp. 70–71). Notes Steinkellner’s identification of an evolution from the Pramāṇavārttika’s “inference from disintegration” to the later “inference from existence” (pp. 65–66).

Ch. 3 — Argues that Sa-paṇ’s notorious claim that “only specifically characterised phenomena are objects of comprehension” is not the extreme view Bo-dong and Gyel-tsap accuse it of (which would be self-refuting); rather, on Ngak-cho’s reading, Sa-paṇ means only that real things are primary objects of comprehension while conceptual constructs are merely derivative (pp. 78–80). Lo Ken-chen disagrees with this softening (p. 80).

Ch. 4 — Examines Dharmakīrti’s wavering between Sautrāntika and Yogācāra frameworks; argues this is not confusion but a strategy of “ascending scales of analysis” (pp. 59, 83ff).

Ch. 5 — Ge-luk thinkers (especially kay-drup) reinterpret svalakṣaṇa to include commonsense objects, anchoring epistemology in common sense rather than in atomic reduction (pp. 106ff).

Ch. 6 — Distinguishes nominalism from conceptualism; reads Dharmakīrti as a conceptualist (not nominalist), since universals on his view are mind-dependent constructs that nonetheless have some weak existence (pp. 133–34). Introduces the moderate realist’s predicament: any view that universals exist in their instances must invent an intermediary relation (Scotus’s “formal distinction”, Kumārila’s identity-in-difference) that is “less than real, more than conceptual” (pp. 136, 139–40).

Ch. 7 — Dharmakīrti’s arguments against real universals; resemblance theory.

Ch. 8 — Sa-gya antirealism on predication, distinguishers (ldog pa), and individuations.

Ch. 9 — Ge-luk Collected-Topics-style moderate realism; arguments for the reality of universals.

Ch. 10 — Historical survey of realism in the Buddhist tradition.

Ch. 11 — Introduces apoha; distinguishes Dignāga’s view (apoha as internal articulation of language) from Dharmakīrti’s (apoha as response to Hindu charges of arbitrariness, grounded in causal connection with experience).

Ch. 12 — Dharmakīrti on concept formation; the negative nature of conceptuality; concepts as useful fictions (pp. 133).

Ch. 13 — Negation; Śāntarakṣita’s transformation of apoha into objective elimination (don rang mtshan gyi gzhan sel) which Ge-luk takes as a real negative property (pp. 276–77).

Ch. 14 — Object universal (don spyi) and term universal (sgra spyi) in Ge-luk vs. Sa-gya formulations.

Ch. 15 — Argues the Sa-gya signifier/signified analysis (e.g. Gorampa’s distinction between conceived signified and direct signified, pp. 273–74) is closer to Dignāga than Dharmakīrti; Ge-luk thinkers, by contrast, hold that names succeed by capturing real properties — taking the “objective elimination” as a real common feature (Kay-drup, pp. 278–79).

Ch. 16 — Reconstructs Dharmakīrti’s twin characterisations of pramāṇa: nondeceptive cognition (avisaṃvādi) understood pragmatically, and revealing of a previously unknown object understood intentionally (pp. 289–90). Dharmottara’s “Gettier-like” example of mistaking a mirage for water but stumbling on water under a rock shows that practical success alone cannot define validity — a normative element is required (pp. 292–93).

Ch. 17 — Was Dharmakīrti a pragmatist? Argues no: novelty + intentionality + non-deceptiveness must be combined.

Ch. 18 — The deepest difficulty in Dharmakīrti’s system: how can inference be valid when its object (the universal) is unreal?

Ch. 19 — Representationalism vs. direct realism; aspects (ākāra, rnam pa).

Ch. 20 — Dharmakīrti’s definition of perception as free from conception and non-mistaken.

Ch. 21 — Dharmottara’s innovations: perception bears on two types of object (held vs. conceived), allowing some coordination with conception (pp. 359–63); seeds of the “new epistemology” (pp. 363–64) — including the unorthodox notion of a “determinate perception”.

Ch. 22 — cha-ba’s Tibetan new epistemology: introduces the distinction between perception (always non-mistaken) and valid perception (which determines its object by eliminating superimpositions, p. 366); introduces the categories of subsequent cognition (བཅད་ཤེས་) and appearance-without-ascertainment (Dreyfus: “inattentive cognition”, སྣང་ལ་མ་ངེས་པ་) to handle perceptions that fail novelty or fail to ascertain (pp. 365–66).

Ch. 23 — Cha-ba’s typology of four objects (appearing/held/object-of-application/conceived); Sākya Chok-den’s polemical argument that “appearing object” (snang yul) and “conceived object” are Tibetan creations without Indian source — though Dreyfus shows this overstates: Dharmottara does use “conceived object” (pp. 384–87).

Ch. 24 — sakya-pandita’s critique of the new epistemology: rejects investigative thought (yid dpyod, ཡིད་དཔྱོད་) as a fabricated category collapsible into doubt or inference; rejects “subsequent cognition” as a category since perception merely holds objects in flux; rejects appearance-without-ascertainment as a category since perception never determines anything — so the distinction between perceptions that induce certainty and those that don’t is a function of the perceiver’s conceptual state, not of perception itself (pp. 390–94). Rejects the explicit/implicit realisation distinction as a Tibetan invention (pp. 395). Sa-paṇ’s positive solution: self-awareness mediates between bare perception and blind conception (p. 397).

Ch. 25 — Perception and self-awareness (rang rig, svasaṃvedana); gorampa develops a representationalist reading of Dharmakīrti against Ge-luk direct realism.

Ch. 26 — Are external objects perceptible? Gorampa: only via the aspect that represents them, so external objects are hidden (lkog gyur); the Sautrāntika causal account commits one to this. Ge-luk attempts to combine the Sautrāntika causal theory with the equation of held and appearing objects — Sākya Chok-den shows this is incoherent (pp. 385–86).

Ch. 27 — Yogācāra in Dharmakīrti; True vs. False Aspectarians; Sākya Chok-den’s reading of Dharmakīrti as moving up an ascending scale culminating in False Aspectarian Madhyamaka (pp. 435–37). Dreyfus argues Dharmakīrti himself does not commit either way — his aim is epistemological rather than soteriological (pp. 437–38).

Conclusion — Philosophy as education of the mind. The problem of universals is a Wittgensteinian riddle, not a solvable problem; the value of studying Dharmakīrti’s tradition lies in the conceptual training it provides, which prepares the Tibetan student for the Madhyamaka critique of foundationalism — Candrakīrti’s rejection of Dignāga’s systematic epistemology in favour of common usage (pp. 451–55). Dzong-ka-ba’s Ge-luk reads Dharmakīrti’s logical and epistemological resources back into a Madhyamaka framework; Gorampa’s Sa-gya treats Dharmakīrti’s analyses as a propaedeutic whose ultimate function is to expose its own limits (pp. 454–56).

Key claims

1. The Sa-gya / Ge-luk split is grounded in the problem of universals, not just sectarian politics. Sa-gya antirealism follows Dharmakīrti literally; Ge-luk moderate realism follows Cha-ba’s “new epistemology” via Dharmottara (pp. 128, 134, 444).
2. Dharmakīrti’s three identity conditions for the real: spatial determinacy (deśaniyata), temporal determinacy (kālaniyata), entity-determinacy (ākāraniyata). Universals fail all three (pp. 70).
3. The “inference from existence” (sattvānumāna) supersedes the “inference from disintegration” (vināśitvānumāna) in Dharmakīrti’s later works. This is Steinkellner’s observation, used by Dreyfus to argue against treating Dharmakīrti’s Pramāṇavārttika, Pramāṇaviniścaya, and Vādanyāya as a single uniform doctrine (pp. 65–66).
4. Sa-paṇ holds that only specifically characterised phenomena are real objects of comprehension — a position that drew sustained criticism from Bo-dong, Gyel-tsap, and others as self-refuting, since the very statement is itself about an unreal universal (pp. 78–79).
5. The Tshad ma rigs gter Auto-Commentary contains discrepancies with the root verses, which Gyel-tsap exploits to argue that some passages are corruptions; Sa-gya commentators (Yak-don, Gorampa) defend the Auto-Commentary as authentic (pp. 26–27).
6. Cha-ba introduced both the seven-fold typology of mental states and the new pramāṇa-vocabulary (subsequent cognition, appearance-without-ascertainment, investigative thought, four-fold object typology) which Sa-paṇ explicitly rejects in the Treasure (pp. 389–94).
7. Sa-paṇ rejects investigative thought (Dreyfus: “correct assumption”, yid dpyod, ཡིད་དཔྱོད་) as a pseudo-category: it must collapse into doubt or inference. Dreyfus notes that later Ge-luk textbooks (Paṇ-chen Sö-nam-drak-pa, citing Sa-paṇ) astonishingly misread this refutation as an endorsement (pp. 389–90).
8. Self-awareness (rang rig, རང་རིག་) does the bridge work in Sa-paṇ’s system that “elimination of superimposition” does in Cha-ba’s: it ensures conceptions operate on the same objects given to perception (p. 397).
9. Dharmottara is the seed of the new epistemology, not Cha-ba. Cha-ba radicalises moves already present in Dharmottara — assigning perception a “conceived object” alongside its “held object”, attributing a more active role to perception (pp. 359–64).
10. Gorampa is a representationalist about external objects, holding that external objects are “hidden” (lkog gyur) and accessed only via their aspect; Ge-luk thinkers are direct realists (pp. 416–25).
11. Sākya Chok-den underwent a real philosophical evolution through three phases (his Defeater of 1474 belongs to the early Sa-paṇ-orthodox phase; the emptiness-of-other (gzhan stong) view enters only after his 1484 meeting with the Seventh Karmapa, c. 1489–1502) (pp. 28–29).
12. Sa-gya readings of apoha — especially the signifier/signified distinction in Gorampa and Sākya Chok-den — are closer to Dignāga than to Dharmakīrti, because both Dignāga and Sa-paṇ are primarily occupied with refuting realism (pp. 274–75).
13. The Ge-luk attempt to combine Sautrāntika causal theory with the equation of held and appearing objects is incoherent (Sākya Chok-den’s argument): the causal theory requires a time gap between object and perception, which contradicts the claim that the held object appears to perception (pp. 385–86).
14. Candrakīrti’s Madhyamaka critique of Buddhist epistemology is anti-foundationalist; Tibetan thinkers (especially Ge-luk) try to graft Dharmakīrtian categories onto a Prāsaṅgika frame and end up with an unstable synthesis (pp. 451–56).

Notable quotes

• “Because all things essentially abide in their own essence, they partake in the differentiation between [themselves and the other] similar and dissimilar things.” — Dharmakīrti, PV III, on the nature of svalakṣaṇa (p. 69).
• “It is completely logical to say that [the universal], which was present elsewhere and did not move from its own place, exists in what has its origin in a place other than it!” — Dharmakīrti’s mockery of Nyāya inherence (p. 70).
• “Whatever exists disintegrates as [for example] a jar. Sound now exists. Thus [is] a reason of [identical] essence [stated].” — Sa-paṇ glossing the inference from existence (p. 65).
• “[Dharmakīrti] did not speak [of valid cognition] as being of two [types] due to the fact that there are [in reality] two [types of] object of comprehension. He declared that there are two [types of] valid cognition in consideration of the [two] modes [in which cognitions] realize [their objects].” — Sa-paṇ, used by Ngak-cho to soften Sa-paṇ’s antirealism (p. 79).
• “Sense consciousness is like the fool who sees. Conception is like a blind skilful speaker. Self-cognition is like [a person] with complete senses, who introduces one to the other.” — Sa-paṇ on the bridge function of self-awareness (Treasure, quoted p. 397).
• “No perception ever ascertains [anything, for] what is apprehended [by perception] is not ascertained. … The appearance [of the object] to it [the perception].” — Dharmakīrti, PV, cited by Sa-paṇ against Cha-ba’s “elimination of superimposition” (p. 393).
• “They confuse the functions of perception and conception. That’s all!” — Sa-paṇ’s flippant summary of Cha-ba’s mistake (p. 394).
• “With reference to this [Auto-]Commentary, some of my mighty and learned teachers stated that [this text] is difficult to posit [as Sa-paṇ’s work] … ‘this text is not [Sa-paṇ’s] Auto-Commentary’. Nevertheless, I have [explained Sa-paṇ’s text] in accordance with [his own] commentary.” — Gyel-tsap on the authenticity dispute over the Rang-‘grel (p. 26).
• “Those who know [should] speak having unerringly realized the meaning of those points [of Dharmakīrti’s system]. Nowadays in this Land of Snow I am the only [such informed speaker].” — Yak-don’s claim against Cha-ba’s followers (p. 25).
• “It is difficult to resist the impression that the Buddhist, especially as explained by Dharmakīrti, sets his standards almost impossibly high. It should be apparent that no one … can in practice inferentially or empirically test for himself all rationally analyzable propositions on which he must make a decision.” — Tom Tillemans, quoted approvingly by Dreyfus on the over-restrictiveness of Dharmakīrtian validity (p. 451).
• “Philosophy is the education of grownups.” — Stanley Cavell, quoted as Dreyfus’s closing frame (p. 450).

Translation-relevant material — FLAG

Dreyfus does not translate the pramanayuktanidhi in full, but the book contains a high density of paraphrased and short directly-translated passages from the Tshad ma rigs gter and its Rang-‘grel, especially in chapters 3, 8, 24, and 25. These should be revisited during dedicated translation sessions on PYN:

• PYN on the concept of existence — pp. 76–80 (Ch. 3). Includes Sa-paṇ’s lists of synonyms for svalakṣaṇa and sāmānyalakṣaṇa (p. 77).
• PYN on disintegration and the inference from existence — pp. 64–65 (Ch. 2).
• PYN on Sa-paṇ’s critique of Cha-ba’s sevenfold typology — pp. 389–94 (Ch. 24). Direct paraphrases of the relevant verses with Dreyfus’s English renderings.
• PYN on perception not determining its object — pp. 392–94 (the “fool / blind / sighted” simile of self-awareness).
• PYN on explicit vs. implicit realisation — pp. 395 (rejects the distinction).
• Sa-paṇ glosses on PV stanzas about disintegration, p. 64.

Several short verses from pramanavartika are translated in passing throughout, particularly PV II on the proof of Buddha as pramāṇa (Ch. 27), and PV III on perception and self-awareness (chs 19–25).

Relevance to the three main texts

pramanayuktanidhi (Sakya Paṇḍita’s Tshad ma rigs gter)

This is the central text of Dreyfus’s analysis — more so than the title might suggest. The book reconstructs the philosophical position of the Treasure through a sustained comparison with gorampa’s commentary, the Auto-Commentary (Rang-‘grel), and Sākya Chok-den’s Defeater (Sde bdun gyi ‘dod tshul rnam par ‘byed pa). Distinctive contributions to our understanding of PYN include: (1) historical reconstruction of why the Treasure was a “problematic text” for the first 150 years after Sa-paṇ’s death and only became the basis of Sa-gya logico-epistemological study after Yak-don’s c. 1400 revival (pp. 23–25); (2) detailed exposition of Sa-paṇ’s rejection of Cha-ba’s seven-fold typology of mental states as the central polemical thrust of the Treasure (Ch. 24); (3) the discrepancies between root verses and Rang-‘grel and the implications for which text is authoritative (pp. 26–27); (4) Sa-paṇ’s self-awareness-as-bridge solution, treated as the Treasure’s most distinctive positive contribution (p. 397).

pramanavartika (Dharmakīrti’s Pramāṇavārttika)

Dreyfus reconstructs Dharmakīrti’s system primarily from the Pramāṇavārttika but draws also on the Pramāṇaviniścaya, Nyāyabindu, and Vādanyāya. He notes Steinkellner’s important observation that the PV’s argument for impermanence from “disintegration” (vināśitvānumāna) is replaced in later works by the argument from “existence” (sattvānumāna), and that traditional Tibetan commentators tend to overlook this evolution (pp. 65–66). He argues against treating Dharmakīrti as a strict pragmatist by combining the PV’s two seemingly conflicting definitions of pramāṇa via Dharmottara’s reading of the Pramāṇaviniścaya (pp. 289–93). On PV II (Establishment of Buddha as Valid), Dreyfus situates Dharmakīrti’s soteriological intent without overstating it (pp. 437–42), against both Conze’s secularising reading and the Ge-luk maximalist reading.

pramanasamuccaya (Dignāga’s Pramāṇasamuccaya)

Dreyfus engages Dignāga only briefly and through Dharmakīrti’s lens (p. 1 of Intro I). His most substantive contribution to our understanding of pramanasamuccaya is the argument — developed in chs 11, 13, and 15 — that the Sa-gya tradition’s reading of apoha is in fact closer to Dignāga than to Dharmakīrti, because both Dignāga and Sa-paṇ treat language as primarily an internal articulation of the conceptual realm (signifier/signified), whereas Dharmakīrti adds a causal-experiential bridge to reality in response to Hindu critique (pp. 274–75). Dreyfus also presents Candrakīrti’s critique of Dignāga’s epistemology — that its strict criteria of validity do violence to common usage — as the philosophical backdrop against which Tibetan Madhyamikas read both Dignāga and Dharmakīrti (pp. 451–56).

Critical assessment

Dreyfus’s greatest strength is his ability to make Tibetan scholastic distinctions philosophically intelligible by translating them into the vocabulary of analytical philosophy (Quine on ontological commitment, Sellars on the Myth of the Given, Wittgenstein on philosophical riddles). This is also his characteristic limitation: the Western philosophical analogies are sometimes pressed harder than the Tibetan material warrants (the Quine/Dharmakīrti comparison is more illuminating than the Meinong/Dharmakīrti one on the existence of universals, p. 75). His Ge-luk training shows in his familiarity with Gyel-tsap and Kay-drup, but he is unusually fair to Sa-gya readings — Recognizing Reality is the work that put Gorampa and Sākya Chok-den on the map for English-language scholarship. His historical reconstruction of the 14th–15th century Sa-gya / Ge-luk polarisation as a political as well as doctrinal process (pp. 35–37) remains broadly accepted but is sometimes too sharply drawn for the early phase. The book is a study, not a translation, and readers wanting Tibetan or Sanskrit texts in full must go elsewhere; many key passages are in the footnotes.