Buddhism as Philosophy

Bibliographic note

Mark Siderits, Buddhism as Philosophy: An Introduction (Ashgate / Hackett, 2007), 240 pp. The textbook-systematic statement of Siderits’s “Buddhist Reductionism” programme, written for upper-undergraduate philosophy courses (the material on Madhyamaka in Ch 10 was first delivered as the Matilal Lectures at King’s College London). Chapter 9, “Madhyamaka: The Doctrine of Emptiness” (pp. 180–207), is the direct interest for this wiki; Chapter 7 (“The Rise of Mahāyāna”) frames the Mahāyāna two-truths question that Madhyamaka inherits, and Chapter 10 (“The School of Diṅnāga”) supplies the pramāṇa-theoretical contrast.

Thesis

Of three possible interpretations of Nāgārjuna’s universal-emptiness claim — (a) metaphysical nihilism, (b) “reality is ineffable,” (c) semantic non-dualism — only the third is internally coherent and consistent with Nāgārjuna’s own statements. Semantic non-dualism reads the doctrine of emptiness as the rejection of the very idea of an ultimate truth in the Abhidharma sense (a statement corresponding to the ultimate, mind-independent nature of reality), retaining only conventional truth understood as semantic anti-realism (truth as warranted assertibility within useful conventions). Madhyamaka thereby resembles contemporary semantic anti-realism (Putnam, Dummett) more than any metaphysical position. The apparent self-contradiction of “the ultimate truth is that there is no ultimate truth” is dissolved by disambiguating two senses of ultimate truth: ultimate truth₁ = a fact that must be grasped to attain enlightenment; ultimate truth₂ = a statement corresponding to the ultimate nature of mind-independent reality. The doctrine of emptiness then reads: “the ultimate truth₁ is that there is no ultimate truth₂.”

Key claims

• The three interpretive options for emptiness (pp. 180–183). Once Abhidharma mereological reductionism is granted and Madhyamaka arguments succeed in showing that nothing has svabhāva, three options open: (a) metaphysical nihilism — “ultimately nothing whatever exists”; (b) “reality is ineffable” — what is ultimately real transcends our conceptual capacities (Yogācāra adopts a variant of this); (c) semantic non-dualism — the very idea of ultimate truth (in the Abhidharma sense) is incoherent. The first two are metaphysical interpretations; the third is semantic. Siderits argues throughout the chapter that (a) and (b) are blocked by Nāgārjuna’s own arguments and (c) is the only remaining live option.
• Nāgārjuna explicitly rejects metaphysical nihilism (pp. 188–189, on MMK V.6–8). The argument against the dhātus shows that no ultimately true statement can be made about space (it cannot be coherently described as either existent, non-existent, both, or neither). Nāgārjuna’s v. 6 — “When the existent is not real, with respect to what will there come to be non-existence?” — formally blocks the nihilist conclusion. Metaphysical nihilism is itself a metaphysical thesis and therefore falls under the same negation.
• Nāgārjuna explicitly rejects the “reality is ineffable” reading (pp. 190–191, on MMK XIII.7–8). The purgative simile (Candrakīrti’s quotation of the Ratnakūṭa Sūtra in commentary): emptiness is medicine that must itself be expelled. The “reality is ineffable” reading turns emptiness into a metaphysical thesis (“reality is forever beyond our conceptual grasp”) and so fails to be purged. “Those for whom emptiness is a [metaphysical] view, they have been called incurable” (XIII.8) is therefore directly aimed at the ineffabilist reading.
• The two senses of “ultimate truth” disambiguation (pp. 202–203). “ultimate truth₁: a fact that must be grasped in order to attain full enlightenment; ultimate truth₂: a statement that corresponds to the ultimate nature of mind-independent reality.” The doctrine of emptiness then says “the ultimate truth₁ is that there is no ultimate truth₂.” This is Siderits’s analytic-philosophical reconstruction of a move that Tsongkhapa achieves with the bden grub qualifier, but Siderits does not name the lineage.
• The conventional truth as paper-currency-off-the-gold-standard (pp. 202–204). Just as the value of paper currency does not require backing by gold (its value derives from its role in human exchange-practices), so the truth of conventional statements does not require grounding in things with intrinsic nature. This is the analogy by which Siderits answers the standard objection that conventional truth presupposes ultimate truth.
• Direct response to Burton and Hayes on the equivocation charge (pp. 192–195, fn. 5 explicitly names Burton Emptiness Appraised pp. 90–94 and Richard Hayes “Nāgārjuna’s Appeal” 1994). Hayes and Burton accuse Nāgārjuna of equivocation between compounded₁ (“composed of several distinct things”) and compounded₂ (“produced by causes and conditions”): the inference x is compounded → x lacks svabhāva works for the first sense but fallaciously assumes the equivalence holds for the second. Siderits concedes the point about MMK XV.1–2 in isolation but argues that MMK I (the causation chapter) supplies the missing argument: cause and effect, if both ultimately real, must exist at distinct times; only the mind can hold them together; therefore causation is a conceptual construction; therefore anything compoundedz lacks intrinsic nature. The argument runs via three-times reasoning + Bradley-style infinite regress on causal force. This is the wiki’s strongest-yet contemporary analytic defence of MMK against the equivocation charge that Burton presses.
• The MMK XV.3 argument against “everything-depends-on-everything” (pp. 200–202). A common gloss of emptiness — “everything is conditioned and so derives its nature from other things” — is itself ruled out by the MMK XV argument that there can be no extrinsic nature without intrinsic nature (“for it is said that extrinsic nature is the intrinsic nature of an other-existent”). So even the metaphysical “interconnectedness” reading of emptiness is blocked by Nāgārjuna’s own text. Pratītyasamutpāda must therefore be read at the conventional level only, not as an ultimate metaphysical thesis.
• Emptiness of emptiness as direct support for the semantic reading (pp. 203–204, on MMK XXII.11 and XXIV.18). “It is empty is not to be said, nor that something could be non-empty, … ‘empty’ is said only in the sense of ‘conceptual fiction.‘” That emptiness is itself empty — a prajñaptir upādāya — is the textual locus that “actually supports” the semantic non-dualist reading, after the negative case against the metaphysical readings has been made.
• Madhyamaka method = reductio ad absurdum from the assumption of svabhāva (p. 183). “They start from the hypothesis that there are ultimately real things, things with intrinsic natures, and they then show that this assumption has unacceptable consequences. They are all, in other words, reductio ad absurdum arguments.” Siderits’s reading is structurally Prāsaṅgika without using the term — he does not engage the Prāsaṅgika–Svātantrika distinction in the chapter, citing Dreyfus & McClintock 2003 only in Further Reading.
• The soteriology of semantic non-dualism: ultimate truth as useful fiction (pp. 204–207). The objection that semantic non-dualism is “too abstract” to be soteriologically efficacious is met two ways. (a) Realisation of emptiness corrects a subtle form of self-clinging that survives realisation of non-self: the “table-pounding” gesture by which one asserts an absolute objectivity is a form of self-assertion (the truth is on my side). (b) The relativism worry is met by treating ultimate truth as a useful fiction — a regulative ideal that prevents premature satisfaction with one’s own perspective without committing to a metaphysical absolute. Better and worse can exist without a “best.”
• MMK 18:6 invoked as Buddha’s own graded teaching (p. 206). “The Buddha might teach non-self to some audiences, while he will teach others in a way that leads them to believe there is a self. In both cases the Buddha is simply using expedient pedagogical methods.” This is the neyārtha / nītārtha device operative again — the same locus Siderits invokes in siderits-reality-altruism-2000 (BCA 8 vs BCA 9). He explicitly endorses the “progression involved here” in the commentaries.

Methodology

Hybrid: rational reconstruction (Hume, Bradley, Putnam, Quine, Dummett as analytic interlocutors) plus close reading of the Sanskrit verses against the Indo-Tibetan commentarial tradition. The Madhyamaka chapter cites the Akutobhayā, Buddhapālita’s Buddhapālitavṛtti, Bhāviveka’s Prajñāpradīpa, and Candrakīrti’s Prasannapadā at parity, treating them as a single commentarial voice for the purpose of explication and not arbitrating between them. He explicitly drops the textbook’s usual rule of “presenting only the original text” for MMK because the verses are too compressed without commentary, and provides his own running commentary (pp. 183–184). The pedagogical register is undergraduate-philosophy textbook — clarity is the dominant virtue — but the philosophical reading is original and substantive, particularly the three-way taxonomy and the two-senses-of-ultimate-truth disambiguation.

Methodologically Prāsaṅgika without naming the school: the standard reading of Madhyamaka argument as reductio ad absurdum from the hypothesis of svabhāva (p. 183), with no engagement of the Svātantrika position or the Buddhapālita–Bhāviveka–Candrakīrti exchange. The Prāsaṅgika–Svātantrika distinction is referred only to Further Reading (Dreyfus & McClintock 2003).

Notable quotes

• “The ultimate truth₁ is that there is no ultimate truth₂.” (p. 203, the canonical formulation of the semantic non-dualist reading)

Further note from the wiki author’s personal notes (2025)

tenpa-personal-notes-2025 develops a partial response to Siderits’s Ch 6 claim that the Two Truths distinction “was developed by commentators on the early Buddhist texts in order to solve an exegetical problem.” Tenpa agrees that the Buddha did not present the Two Truths as a system — the systematisation, and the neyārtha / nītārtha labels, are commentarial — but disagrees that the underlying doctrine was absent from the original teachings. The image: butter from milk. The commentators did not invent the butter; they churned the milk. The Buddha taught contextually for forty-five years to interlocutors of varying capacity and did not need to systematise; any later corpus-organiser working with that body of teaching would face an intellectual necessity of systematisation, and what they produced would resemble the two-truths methodology.

Convergent-evolution corroboration: both Śrāvaka and Mahāyāna lineages independently reach two-truths-style distinctions (Buddhaghosa’s paramattha / sammuti on the Theravāda side; Madhyamaka’s saṃvṛti / paramārtha on the Mahāyāna side). Field evidence from the Thai Forest Tradition supplies operational two-tier discourse from a non-Mahāyāna lineage — see the Ajahn Teera material in tenpa-personal-notes-2025 under “On Siderits and the systematisation of the Two Truths.”

The position leaves Siderits as a substantial ally on the graded-teaching device (which he operates without naming, both at MMK 18:6 here and at BCA 8 / BCA 9 in siderits-reality-altruism-2000) while parting ways on the origin of the Two Truths apparatus.

Connections

• Continues siderits-reality-altruism-2000 — same Buddhist Reductionism programme, now stated systematically rather than in defensive mode. The 2000 review’s reach for the graded-teaching device (BCA 8 / BCA 9, with cross-reference to MMK 18:8) is independently and more programmatically deployed at MMK 18:6 in this volume. The 2000 review remains primary for the BCA / Williams material; the 2007 textbook is primary for the Madhyamaka exposition.
• Direct rebuttal of burton-emptiness-appraised-1999 on the equivocation charge (pp. 192–195, fn. 5 names Emptiness Appraised pp. 90–94 explicitly). Pair the two sources for .
• Cited approvingly by westerhoff-candrakirti-2024 on “semantic insulation” between the two truths (MA 6:031) — though that citation is to the 2003 Personal Identity and Buddhist Philosophy monograph rather than to Buddhism as Philosophy. The two analytic readings converge on the constructed-conceal function of conventional truth.
• Methodologically opposed to kalupahana-mmk-1986 — Kalupahana wants Nāgārjuna pre-Mahāyāna and pre-systematic; Siderits reconstructs Madhyamaka as a sophisticated philosophical system continuous with contemporary semantic anti-realism (Putnam’s Reason, Truth and History is named in Further Reading at p. 207).
• Shares territory with oetke-remarks-interpretation-1991 on the rational-reconstruction-in-analytic-idiom project. Oetke’s “on the level of highest truth there is nothing of any kind” is structurally close to Siderits’s “ultimate truth₁ that there is no ultimate truth₂” but Oetke retains the formal-operator scaffolding while emptying the pedagogical content; Siderits retains the pedagogical content (graded teaching, two truths as disambiguated method) and is relaxed about formal scaffolding. Different positions on the gradient.
• Shares territory with sprung-lucid-exposition-1979 on the deflationary semantic reading. Sprung’s “the higher truth, in so far as it is a theory, falls within the lower truth” (Translator’s Introduction, p. 16) is the early Wittgensteinian predecessor to Siderits’s semantic non-dualism. Both arrive at the position from outside the Indo-Tibetan commentarial tradition.
• Affirms Buddhapālita’s reading style (reductio ad absurdum from the svabhāva-hypothesis) as definitive of Madhyamaka method (p. 183), without engaging the Buddhapālita–Bhāviveka–Candrakīrti dispute as a methodological question.
• Endorses Garfield 1995 as “the best English translation” of MMK (Further Reading, p. 207). Murti 1955 named as “the classic statement of the ‘reality is ineffable’ interpretation”; Ruegg 1977 (JIP 5) as a more recent formulation. Wood 1994 cited as a “charitable nihilist reading” alongside Burton 1999 as the “hostile” version.