Pramana Wiki

Charvaka

32 min read

Charvaka (རྒྱང་ཕན་)

The Charvakas held that cognition is a product of the body. No other school takes up as much of pramanavartika chapter 2: the argument with them runs unbroken through II.35–119 and returns at II.148–176 and II.187–189, some 117 of the chapter’s 285 verses, of which only II.165–167 belong to another school. (dharmakirti gives the Mīmāṃsā a slightly longer unbroken stretch, PV I.224–339, but that is about scripture, not mind.) The length has a reason. Chapter 2 establishes that the Buddha is valid cognition (ཚད་མ་) by tracing his compassion back through many lives of training; if the body is what produces mind, there are no earlier lives and the chapter has nothing to argue. dignaga does not engage the school anywhere in our pramanasamuccaya material. sakya-pandita names the Charvakas in the colophon of the Pramāṇayuktanidhi and meets them on two questions: whether a beginningless continuum of mind can be established, and whether there is more than one means of valid cognition. Our translations spell the name Cārvāka, following the Sanskrit, and quotations keep that spelling.

Overview

The school has three names and these texts use two of them. Cārvāka and Lokāyata (“directed towards the world”) are the usual Sanskrit pair; Bārhaspatya attaches it to Bṛhaspati, its traditional founder, whose Bṛhaspatisūtra is lost and whose founding role some scholars dispute (Wikipedia). The Tibetan is རྒྱང་ཕན་, the glossary’s headword and the term Gorampa’s translator uses. A second Tibetan name, ཚུ་རོལ་མཛེས་པ་, appears in the Pramāṇayuktanidhi colophon; Khenpo Sodargye renders it “beauty of this world” and glosses it as valuing only the enjoyments of the present life (Pramāṇayuktanidhi chapter 11, with sodargye-pramanayuktanidhi on the same chapter). Dreyfus gives the Sanskrit and Tibetan together in his survey of Dharmakīrti’s opponents, describing “the Materialists (cārvāka, rgyang ‘phen), who reject any form of knowledge other than perception” (dreyfus-recognizing-reality, p. 51).

Almost nothing of the school’s own writing survives. Discussions of it appear from the sixth century BCE onwards, in Buddhist and Jain literature that treats it as an established opponent; the earliest named exponent is Ajita Kesakambali; five later commentators are known by name, among them Purandara, Aviddhakarṇa and Udbhaṭa. The two texts usually cited for its doctrines are Jayarāśi Bhaṭṭa’s Tattvopaplavasiṃha (eighth century, and its Charvaka affiliation is disputed) and the opening chapter of the fourteenth-century Sarvadarśanasaṃgraha, which is a doxography by an opponent. The system is reported to have disappeared after the twelfth century without continuity, and it is not a living tradition; what is known of it comes largely from criticism (Wikipedia). That matters for reading the pages below: the Charvaka these texts refute is a Charvaka reconstructed by Buddhists, and there is no surviving Charvaka text against which to check the reconstruction.

Gorampa’s translator gives the doctrine as Dharmakīrti’s tradition received it. The Charvakas were materialists who “espoused no design or purpose to the world”; they asserted no immortal creator, no abiding soul and no law of karma, and held happiness and suffering to be random. Consciousness is a by-product of the body and not separate from it, “as a mural is essentially part of the wall”; so there is no past or future existence, “just as a mural ceases when the wall is destroyed”. And “for the Cārvāka, sense perception alone is valid. For them, seeing is believing” (translator’s introduction to gorampa-pramanavartika). Dreyfus states the epistemological half of this in the same terms: “In accordance with their materialism, the Cārvāka school accepts only one type of valid means of cognition, perception” (dreyfus-recognizing-reality, p. 293).

One historical qualification bears directly on the arguments in this page, and the sources here do not carry it. The Charvaka commentators known to modern scholarship were not uniform in rejecting inference outright. Purandara admitted inference for worldly matters where perceptual experience is available, and refused it only for the transcendental (a future life, the law of karma, anything outside possible perception), and Aviddhakarṇa and Udbhaṭa are reported to have agreed with him (web search on Purandara and the Cārvāka commentators; the statement goes back to Dasgupta’s A History of Indian Philosophy, vol. 3). These texts argue against an opponent who denies inference as such. Whether that opponent ever existed outside the doxographies is a question this material cannot settle, and the refutations in the next section should be read with it in view.

Core doctrines

Cognition is a property of the body

Gorampa opens the whole exchange by giving three ways the dependence might be understood, and the three differ: “cognition is in the nature of the body in the same way alcohol possesses the power to create drunkenness. Or, like a lamp and its light, cognition is a creation of the body, or like the sweetness of sugar, it is a quality of the body” (gorampa-pramanavartika, on PV II.34). Intoxicating power, lamplight, sweetness: an inseparable capacity, a produced effect, an inhering quality. Dharmakīrti’s replies take all three, and one of them (the wine and its power) the Charvaka later tries to use as a defence, which is why it is worth keeping the list separate at the outset. The consequence in every case is the same: “when the body is destroyed, the continuum of cognition ceases”, so compassion cannot be accomplished by training across lives.

The doctrine has a second, stranger form in the commentaries. Both Gorampa and Gedun Drub report the Charvaka as holding the producing body to be permanent. “According to the Lokāyata materialists, mental cognition is produced from a permanent body” (gorampa-pramanavartika, on PV II.43), and Gedun Drub sets the Charvaka’s “permanent great-element body generates many consciousnesses” beside the Sāṃkhya puruṣa and the Vaiśeṣika self as three variants of one error, all refuted by the principle that permanent things perform no action (Gedun Drub’s commentary on PV III). A materialism of four elements holding the body permanent is an odd position, and neither commentator explains how the school arrived at it; the reports are consistent with each other, and the page records them as they stand.

A related claim shows up in pramanavartika I.22, where Dharmakīrti is arguing that mere failure to apprehend something does not establish that it is absent. The list of positions that prove the point includes “consciousness in earth [as the Cārvāka hold]” (pramanavartika I.22; Gorampa names “the Lokāyata school” at that line). The same doctrine returns in chapter 2, where Gorampa glosses the Charvaka as asserting “that beings are born within earth and so on” (gorampa-pramanavartika, on PV II.171). Sodargye records a further split within the school: one faction holds that the first moment of mind in the womb arises from the body and every later moment from that first mind, the other that each day’s moments of cognition arise in dependence on the body (sodargye-pramanayuktanidhi, on chapter 9).

Perception is the only means of valid cognition

Gedun Drub states the argument for the single valid cognition compactly, as an inference from ontology to epistemology: “The Cārvākas say: the universal is not an object of pramāṇa, since it is unreal. Hence the objects of pramāṇa are determined to one—the particular; thus the pramāṇas are determined to one—perception” (Gedun Drub’s commentary on PV III). It is a good argument to put to a Buddhist, because a Buddhist agrees with the premise. Dharmakīrti holds that only the specifically characterised phenomenon (རང་མཚན་) performs a function and is therefore real, and that universals are conceptual constructs. If the objects of valid cognition are counted by what is real, the count comes to one.

In practice the school’s denial of inference (རྗེས་དཔག་) is not silent but argumentative. Gedun Drub gives two of its standard theses in the form of reasons: “[the Buddha is] not omniscient” because of speaking, and “there is no other world” (Gedun Drub’s commentary on PV III). Sodargye gives the second in the same shape: past and future lives and the working of karma do not exist, and the reason offered is “Because we have not seen them” (sodargye-pramanayuktanidhi, on chapter 8). Both are negative reasons of the form not observed, therefore absent, which is the structure the pramanavartika devotes its opening chapter to regulating.

Desire is a product of the body

The materialist account of mind is applied to the afflictions in pramanavartika II.148 onwards. Desire and anger are properties of the humours, wind, bile and phlegm; or of the four elements generally; or of semen, increasing when the humours are in balance; or of form, since it is a beautiful form that occasions desire. Objection and reply alternate verse by verse across II.148–II.176, so the stretch has no clean division between the school’s case and Dharmakīrti’s. Gorampa frames the whole stretch as the argument of the nihilist-materialist and names the Cārvāka at II.171. Two things about the attribution should be said plainly. The humoral and semen sub-objections at II.152–II.156 are introduced anonymously in Gorampa, with no medical or Sāṃkhya label attached, and the translation here keeps them inside the materialist thread on the strength of the surrounding context rather than on a naming (Pramāṇavārttika translation notes, ). And the substance-and-quality strand at II.159–II.163 is Vaiśeṣika-styled but again unnamed.

Body and mind are one entity

Gedun Drub reports a Charvaka argument from pramanavartika chapter 4 built on a single-entity thesis: to disprove future lives, the Lokāyata sets up a person whose body and mind are one entity. “If body and mind are one, when the body perishes the mind too perishes, so no future life—leading to no liberation. This is annihilationist” (Gedun Drub’s commentary on PV IV). The two inferences the school builds on that thesis are barely intelligible as Gedun Drub renders them; his verdict on them is quoted below.

Causeless arising: an attribution the sources dispute

Chapter 1 of the Pramāṇayuktanidhi lists the rival accounts of what things come from:

།ངོ་བོ་ཉིད་དང་དབང་ཕྱུག་དང༌། །གཙོ་བོ་རྡུལ་དང་རྣམ་རིག་དང༌། །རྟེན་འབྲེལ་ཡིན་ཞེས་གྲུབ་པའི་མཐའ། །ཐ་དད་རང་གཞན་སྡེ་པ་འདོད།

That [the cause of things] is intrinsic nature, Īśvara, primal substance, atoms, cognisance or dependent arising: these are the different tenets held by schools of our own and others. (Pramāṇayuktanidhi chapter 1)

Who holds that things come from “intrinsic nature” (ངོ་བོ་ཉིད་)? The auto-commentary names no school: it speaks only of certain non-Buddhists who propound intrinsic nature as the cause, for whom everything arises from its own nature (auto-commentary on Pramāṇayuktanidhi chapter 1). Sodargye names the Charvakas, both here (“The Cārvākas hold that everything arises naturally”) and again in chapter 6, where the consequence of denying the “without-which-not” relation between fire and smoke is that smoke could “arise naturally without any cause, as the Cārvākas hold” (sodargye-pramanayuktanidhi, on chapters 1 and 6). Gorampa does not make that assignment. At pramanavartika II.181, where causelessness is refuted, he glosses the speaker only as “those who propound the entity itself as a cause” (proponents of intrinsic nature, svabhāvavāda) and does not name the Charvaka (Pramāṇavārttika translation notes, ). The disagreement is between the two commentaries and nothing here resolves it [unverified]; this page follows Gorampa where the Pramāṇavārttika is at issue and reports Sodargye’s attribution as his.

The refutations

Dignāga

Nothing. The school is not named, and no position recognisable as its own is refuted, anywhere in the translations here of the pramanasamuccaya or in either the raw or the wiki commentaries on it. Dignāga’s titled refutation sections go to the Sāṃkhya, the Nyāya, the Vaiśeṣika and the Mīmāṃsā; the Charvaka is absent from all six chapters. Whether that is a fact about the Pramāṇasamuccaya or an artefact of what we hold of it is not settled here.

Dharmakīrti

The consequence arguments (II.35–II.39)

The reply opens by denying the dependence outright and then extracting consequences from it. If breath, the sense faculties and cognition arise from the body alone, without depending on anything of their own kind, then they arise from matter as such, and matter is everywhere:

།གང་དུ་དྲོད་གཤེར་ལ་སོགས་པ། །སྐྱེ་བོ་སྐྱེ་བར་མི་འགྱུར་བའི། །ས་སོགས་ཆ་དེ་འགའ་ཡང་མེད། །དེ་ཕྱིར་ཐམས་ཅད་ས་བོན་བདག

There is no portion of earth and so on in which living beings would not be born from heat, moisture and the like; therefore all [the elements] would have the nature of seeds [of mind]. (pramanavartika II.37)

Gorampa records the school’s attempt to block this: some elements are seeds of mind and others are not. Dharmakīrti’s answer is that no such difference has been given: if one portion of matter can become a sense faculty without a prior faculty of its own kind, every portion can, “because there would be no difference in the reasoning” (gorampa-pramanavartika, on PV II.37–II.38).

Then the direction of dependence is reversed. Damage an eye and the mental cognition is not damaged; grieve, and the faculties alter. So the faculties depend on mind rather than the reverse (II.39). This is the point at which the ordinary evidence the Charvaka relies on (drunkenness, illness, fever, blows to the head) is taken over rather than denied. Dharmakīrti grants that the body affects the mind and then locates the effect one step back. The body’s flourishing and decline produce pleasure and pain in the body consciousness, and those in turn intensify desire; fever and alcohol impair the mind not directly but by altering the inner physical sensations, from which the sense cognition brings about the change (pramanavartika II.75–II.76 with gorampa-pramanavartika ad loc.). The body is a cooperating condition, and a cooperating condition is not a substantial cause.

Why the body cannot be mind’s substantial cause (II.60–II.62)

This is the load-bearing argument of the chapter, and it turns on a definition rather than on evidence. A substantial cause (ཉེར་ལེན་) is what turns into its effect; it does not stand alongside it unchanged.

།ཉེར་ལེན་འགྱུར་བ་མེད་པར་ནི། །ཉེར་ལེན་ཅན་དག་འགྱུར་བར་ནི། །བྱེད་པར་མི་ནུས་དཔེར་འཇིམ་པ། །འགྱུར་བ་མེད་པར་བུམ་སོགས་བཞིན།

།དངོས་གང་འགྱུར་བ་མེད་པར་ནི། །དངོས་པོ་གང་ཞིག་རྣམ་འགྱུར་པ། །དེ་དེའི་ཉེར་ལེན་མི་རིགས་ཏེ། །བ་གླང་བ་མེན་ལ་སོགས་བཞིན།

།སེམས་དང་ལུས་ཀྱང་དེ་བཞིན་ནོ།

Without the substantial cause changing, what has it as its substantial cause cannot be made to change, just as pots and so on [cannot change] without the clay changing. When one entity is transformed without another entity changing, the latter cannot be the substantial cause of the former, as with an ox and a gayal and so on, and the same holds for mind and body. (pramanavartika II.60–II.62)

Clay becomes the pot; the ox does not become the gayal, though an ox may father one. Gorampa applies it in a line: “Mind and body, like an ox and a gayal, are not material cause and that possessed of material cause because one does not necessarily follow the other” (gorampa-pramanavartika, on PV II.61–II.62). The Charvaka’s reply is to appeal to coexistence (body and mind do go together), and the answer concedes the coexistence and re-describes it: a result arises from its own cause together with a cooperating condition and then abides alongside that condition, as fire and the molten state of copper abide together (II.62 ff.). Copper melts in fire and not from fire being copper.

The same requirement generates the succession argument. A cause with no earlier and later moments cannot yield an effect with earlier and later moments:

།རིམ་ཅན་མིན་ལས་རིམ་ཅན་མིན། །ཁྱད་པར་ཅན་མིན་ལྟོས་པའང་མེད། །ལུས་ལས་རིམ་གྱིས་བློ་འགྱུར་ན། །དེའི་ཡང་རིམ་པ་གསལ་བར་བྱེད།

From what is not successive [, a permanent body, as the Charvakas hold,] nothing successive [arises]; and what is not changed [by conditions] has no dependence either. If the mind arose in succession from the body, that would make [the body’s] succession plain as well. (pramanavartika II.43)

Gorampa runs this against the permanent-body version and gets the concession he wants. If the body is permanent it has no succession to lend. If the school protects itself by saying the body produces mind successively in dependence on successive cognition, then the body has earlier and later moments after all, and once it does, “the previous body and mind being a cause of subsequent body and mind would be seen at all times, even at the time of death” (gorampa-pramanavartika, on PV II.43–II.44). The opponent’s repair establishes the conclusion he was resisting.

The corpse (II.54–II.58)

If the body produces mind, a corpse should think. The Charvaka answers with an unfitness clause (a wick that is wet will not light), and Dharmakīrti pushes it:

།གལ་ཏེ་སྡོང་བུ་ལ་སོགས་བཞིན། །ཉེས་པས་མི་མཐུན་ལུས་རྒྱུ་མིན། །ཤི་བས་ཉེས་པ་ཉམས་གྱུར་ཚེ། །སླར་ཡང་འཚོ་བ་དག་ཏུ་འགྱུར།

།གལ་ཏེ་མེ་ནི་ལྡོག་ན་ཡང་། །ཤིང་གི་འགྱུར་བ་མི་ལྡོག་བཞིན། །དེ་ནི་ལྡོག་པ་མེད་ཅེ་ན།

If you say, “As with a [damp] wick and so on, a fault [such as illness] makes [conditions] unconducive, so the body is not the cause”, then when the fault had subsided through death, [the body] would live again. You may say, “Just as, even when the fire is gone, the change in the wood is not reversed, so that [death] is not reversed”. (pramanavartika II.54–II.55)

The burnt-wood analogy is the school’s best move here, and the answer to it is empirical rather than formal: some changes are irreversible and some are not, and the reason we know the difference is that medicine works. Fire acts on wood irreversibly and on gold reversibly, and melted gold hardens again (II.56–II.57). A poisoning is a reversible change of the second kind, so if the body were the cause of mind, then once the poison was neutralised or the bitten part cut away, the dead body, “now rid of the causes of [further] change”, should live again (pramanavartika II, following II.57). Gorampa notes the obvious rejoinder from the medical literature, that some illnesses are described as incurable, and gives Dharmakīrti’s answer: those cases turn on the difficulty of finding a remedy or on the patient’s lifespan being spent, not on the nature of the change, since “were it [a matter of] the flaw alone, nothing would be incurable” (II.58).

Not seeing (II.82)

The school’s central evidence is that nobody sees anyone go or come. The verse that answers it is the one everything else in this exchange depends on:

།གལ་ཏེ་འགྲོ་འོང་མ་མཐོང་ན། །དབང་པོ་མི་གསལ་ཕྱིར་མ་མཐོང་། །དཔེར་ན་མིག་མི་གསལ་བ་ཡིས། །དུ་བ་ཆུང་ངུ་མི་མཐོང་བཞིན།

You [Charvakas] may say, “Going [to the next life] and coming [from the last] are not seen”; [however,] they are not seen because [your] faculties are not clear, just as one whose eyesight is not clear does not see a faint [wisp of] smoke. (pramanavartika II.82)

Gorampa sharpens it into a question about whose eyes are being appealed to. If the claim is that nobody at all sees it, it is unproved; if the claim is that you do not see it, then what follows is a fact about the Lokāyata’s faculties and not about rebirth (gorampa-pramanavartika, on PV II.82). The reply is the applied form of the principle stated at pramanavartika I.22: non-apprehension establishes non-existence only where the thing, if present, would have been apprehended.

A second thread runs the same way. The Charvakas argue that an ordinary being’s final mind does not link to a successor, “like the mind of an arhat at death”, and the example is Buddhist property. Dharmakīrti’s reply is not a defence of the doctrine but a question about standing: on what valid cognition does a school that rejects Buddhist tenets rely when it borrows one? “Why should [we] follow a tenet whose import is not established by valid cognition?” (pramanavartika II.46). Gorampa is explicit that this is a self-consistency argument and not a statement of the Buddhist position: “in order to prove an internal contradiction in their own beliefs, we are just asking what reasons the Lokāyata have”; it “is not saying that our own position is that it does connect” (gorampa-pramanavartika, on PV II.46).

Breath (II.103–II.115)

The longest technical stretch answers a materialist proposal that mind is caused by breath. It is a good test case for the whole dispute, because breath, unlike the body at large, is undeniably successive. The replies work through the options one at a time. Breaths are many and simultaneous, so they do not determine a single mind; the many minds would arise at once. If a single mind is caused by several breaths together, then when one is missing no mind should arise at all. If the breaths cause the mind “however they are”, then the cognition should vary as they vary, “and what is not differentiated by the differences in that [cause] is not its result”. If cognitions arise from the body all at once at birth and thereafter from their own kind, then what cut off the body’s capacity in the meantime? (pramanavartika II.103–II.113.) The last objection in the sequence is the one with real force: without a body to support it, mind alone would be left, which the Charvaka takes to be absurd. The answer is what stops the Buddhist account from leaving a mind with nothing to inhabit: the five sense-sources of this life are themselves among the causes that produce the next body (pramanavartika II.114–II.115).

Desire (II.148–II.176)

The refutation of the humoral account has one shape, repeated. If desire is a property of the elements and the elements are alike in all bodies, everyone should desire alike; if the elements vary, desire should track the variation, waxing and waning with bodily vigour, which it does not. When the cause increases the effect cannot decline, “as with heat and so on” (II.152), but a flourishing body is not a more desirous one. The image Gorampa attaches to II.171 is exact: within the cognition of “cow”, arising from the similar nature of cow, there are no gradations of “cow”, “large cow”, “very large cow”; likewise, if beings were born within earth, there would be no gradations among beings.

The individual proposals fall to counter-examples. Semen: desire is seen in those out of balance and absent in those in balance, and persists where the semen is spent (II.154). Beautiful form: desire arises where no form has been ascertained as desirable, and in those who do not hold beauty a worthy quality; and if holding it a worthy quality is itself an element, everyone with the element should so hold it (II.155–II.156). The Charvaka objects that desire and anger cannot coexist on a single elemental basis; the reply is that dissimilar natures are exactly what one would expect not to appear together (II.157). The wine analogy from the school’s opening statement is then answered directly: the intoxicating power is not a separate entity from the wine, and if the base entity were destroyed while its support remained intact, the supported power would not be what was destroyed, whereas mind and the elements are apprehended as of distinct appearance, so they are distinct (pramanavartika II, at II.159–II.163). Dharmakīrti’s own account puts desire’s cause in its own homogeneous predecessor and in imprints (བག་ཆགས་): “Whoever’s desire and the like operate in dependence on [their] similar kind and on the particularities of [their] imprints (vāsanā)—for them this fault does not follow” (II.158).

The self-refutation (II.187–II.188)

The chapter’s last exchange with the school is the one that turns. The Buddhist argument for craving as the cause of birth runs from a negative observation: birth is not seen in those free of desire. The Charvaka takes the same form and turns it round: desire is not seen in those without a body, so desire arises from the body too.

།ཆགས་བྲལ་སྐྱེ་བ་མ་མཐོང་ཕྱིར། །ཞེས་ནི་སློབ་དཔོན་རྣམས་ཀྱིས་གསུངས། །ལུས་མེད་ཆགས་པ་མ་མཐོང་ཕྱིར། །ལུས་ལས་ཀྱང་ནི་འདོད་ཆགས་འབྱུང་།

།རྒྱུ་མཚན་ཁས་བླངས་ཕྱིར་ན་འདོད། །ཉེར་ལེན་ཉིད་ནི་བསལ་བ་ཡིན། །རིགས་འདི་རྗེས་སུ་འདོད་ན་ནི། །རང་ཉིད་རང་འདོད་ལ་གནོད་འགྱུར།

“[Desire is the cause,] since birth is not seen in those free of desire”, the masters have said. [The Charvakas object:] “Since desire is not seen in those without a body, desire comes from the body as well”. We agree, since [the body as] a reason [, a cooperating cause,] is accepted; it was [only as] the substantial cause that it was refuted. If you follow this reasoning, you harm your own position [that there are no past and future lives]. (pramanavartika II.187–II.188)

Gorampa spells out the damage. The concession the Charvaka has just made is that a previous body and a previous desire are causes of a subsequent desire and a subsequent body, and a school whose thesis is that there are no past and future lives cannot help itself to previous bodies (gorampa-pramanavartika, on PV II.188). The Buddhist concedes the small point in the same breath: a prior body is granted as a cooperating cause of later desire, which was never what the earlier argument denied. What was denied was that the body is the substantial cause.

Against the single valid cognition (PV III)

Chapter 3 answers the epistemological thesis, and the reply Gedun Drub reconstructs has four steps. First, the concession: the Charvaka need not prove to a Buddhist that only the particular performs a function: “I too accept this”. Second, the reversal. The Charvaka’s own thesis, that the universal is not an object of valid cognition, has been determined; the reason nonexistent is therefore inconclusive as a ground for denying that universals are objects of comprehension, because that denial was itself the work of a valid cognition, and not of perception. “The second pramāṇa, inference, is already established, since the unreal universal is determined by a pramāṇa” (Gedun Drub’s commentary on PV III).

Gedun Drub then follows the escape routes. If the Charvaka says the unreal is directly experienced by sense-perception, the answer is that sense cognition arises from the force of its apprehended object and so cannot take an unreal one; if he says it does not depend on the object’s function, then sense cognition should arise through walls. If he says he determines universals to be non-existent merely by excluding the perception that would otherwise cognise them, then a non-implicative negation is being asked to do determining work, which it cannot; and if he takes that exclusion as a reason, then the reason has subject-property, exclusion and following-along, which is to say it has the three modes (ཚུལ་གསུམ་) and is an inference. There is also an argument from the school’s own practice: “Even Cārvāka teachers determine the existence of someone’s mind-nature, because [they observe] speech etc. occurring”. Cognition of another mind is inference, and they engage in it.

The fourth step states what denying inference costs. If the cognition of a universal is not valid cognition, then presently unseen places do not exist, “like other worlds”. The Charvaka distinguishes: unseen places can be seen at other times, other worlds never. The answer takes an example he cannot refuse: the eye faculty itself is never an object of sense cognition, so on his own principle there is no eye faculty, and eye cognition has no cause. What establishes the eye faculty is inference (Gedun Drub’s commentary on PV III).

Dreyfus reads the same passage as a general result about the ambiguity of existence in Dharmakīrti, and states the structure cleanly: the entailment between “nonexistent” and “not being an object of comprehension” must itself be known by a valid cognition, and its terms would then be objects of comprehension, so “the adversaries are forced to acknowledge that there are nonexistents, namely, universals such as not being a real existent, which require a special type of valid cognition” (dreyfus-recognizing-reality, p. 74). He also notes what it costs Dharmakīrti: denying that conceptual constructs exist in any sense “would involve Dharmakirti in the same self-defeating paradox as the Materialists”, which is why constructs have to be granted existence in the secondary, epistemic sense (p. 74).

A limit is admitted in the same chapter, and it is the limit that matters for the rebirth argument. Refuting-reasons work through non-perception, and non-perception determines nothing about what is extremely hidden. Gedun Drub applies this against the Charvaka’s own negative theses: “you Cārvākas cannot determine the existence or non-existence of omniscience by perception, since omniscience is extremely non-manifest to you” (Gedun Drub’s commentary on PV III). The principle is symmetrical, and it is stated as a general bar rather than an ad hominem: “non-perception cannot make such a determination, as said before” (same passage). The Buddhist therefore does not claim to perceive past lives either; the whole weight is carried by an inference from the nature of mind.

The body–mind inference (PV IV)

Chapter 4 disposes of the two inferences the school builds on its single-entity thesis. Gedun Drub gives one of them as “the pot is, on the body-mind-as-one-entity person’s body or pot, one or the other of resistance-things—since it is not a non-blue-lotus, like an existent pot or wall”, and the reply refuses the subject rather than the reason: “the pot has no meaning beyond pot, and the body-as-one-with-mind-and-pot is unestablished”. The judgement is blunt: the statement is “empty speech driven by ignorance” (Gedun Drub’s commentary on PV IV). The inferences lack positive concomitance on the example and their reasons are contradictory.

Sakya Paṇḍita

The Pramāṇayuktanidhi does not repeat the Pramāṇavārttika’s campaign. It intervenes on the two fronts named above, and the first opens against fellow Buddhists rather than against the school.

Self-awareness cannot prove past and future lives (Pramāṇayuktanidhi chapter 9). Some masters had argued that both limits of the mental continuum can be established by self-aware perception (རང་རིག་), on the analogy of fire and a lamp: the present cognition, known reflexively, excludes the claim that no cognition preceded it, as a lamp now shining excludes the claim that lamps do not shine.

།འདི་ལ་ཁ་ཅིག་རྟེན་བརྟེན་པ། །རིག་པའི་ཐོག་མཐའ་རྣམ་པ་གཉིས། །མེ་དང་སྒྲོན་མ་དཔེར་བྱས་ནས། །མངོན་སུམ་ཉིད་ཀྱིས་འགྲུབ་ཅེས་ཟེར།

།ཐུག་མེད་རང་རིག་གིས་མི་འགྲུབ། །མེ་དང་སྒྲོན་མ་དཔེར་མི་རུང༌།

[Objection:] Here some say: in dependence on the support (རྟེན་བརྟེན་པ), the two limits—the beginning and end of cognition (རིག་པའི་ཐོག་མཐའ)—are established by perception (self-awareness) itself, using fire and a lamp as examples. [Reply:] The endless [continuum] is not established by self-awareness; the fire and lamp examples are untenable. (Pramāṇayuktanidhi chapter 9)

Sodargye’s exposition gives the reason the examples fail, and it is the sharpest thing in the Pramāṇayuktanidhi’s treatment of this school: both similes hand the Charvaka his own position. Fire producing smoke is a thing of one kind producing a thing of another kind, which is exactly the Charvaka claim that luminous, clear-and-aware cognition is produced from a non-luminous body of four elements. The lamp is worse: once the oil is spent the light stops, which is the Charvaka account of death. And self-awareness reaches only the immediately preceding and following moments, not a beginningless prior limit; since the Charvakas already grant this life’s moments of mind, an argument from them proves nothing the opponent denies (sodargye-pramanayuktanidhi, on chapter 9). What is needed instead is an inference complete in the three modes, and Sakya Paṇḍita gives two, one for each limit:

།སེམས་ལས་གཞན་ལ་ལྟོས་མེད་ཀྱི། །རྟགས་ཀྱིས་སྔོན་མཐའ་ཐུག་མེད་འགྲུབ།

།རྒྱུ་ཚོགས་ཚང་ཞིང་གགས་མེད་པའི། །རྟགས་ཀྱིས་ཕྱི་མཐའ་ཐུག་མེད་འགྲུབ།

By the reason “[mind arises] depending on nothing other than [a preceding] mind,” the beginningless prior limit (སྔོན་མཐའ) is established. By the reason “the assembly of causes is complete and there is no obstruction,” the endless future limit (ཕྱི་མཐའ) is established. (Pramāṇayuktanidhi chapter 9)

Body and parental seed are admitted as concurrent conditions; what is denied is that anything but mind is mind’s substantial cause. Sodargye is careful about the scope: the claim concerns the essence of cognition and not its distinctions, so that whether a person is bright or dull may well depend on external conditions (sodargye-pramanayuktanidhi, on chapter 9).

Reasons that prove a prior cause (Pramāṇayuktanidhi chapters 6 and 10). Chapter 10’s second kind of effect-reason infers a cause that no longer exists (because smoke is curling in the sky there must have been fire, though the fire has burnt out), and Sodargye says it is posited precisely to refute the Charvaka denial that the present consciousness had a prior consciousness (sodargye-pramanayuktanidhi, on chapter 10). The same chapter states the condition on non-observation reasons that the school violates: the principle applies only where seeing is possible, so “past and future lives do not exist, because I have not seen them directly” is no better than “before me there cannot be a demon, because I do not see it directly”, when even tomorrow’s appearances are imperceptible to an ordinary being (same chapter). And chapter 6 supplies the positive requirement the Charvaka cannot meet: to establish that x causes y one must find x and y both absent, then x present, then y arising from it. Nobody has observed a body present and mind then arising from it; what is observed is that mind changes when the body acts, which is compatible with the body’s being a cooperating condition (sodargye-pramanayuktanidhi, on chapter 6).

Two objects, two valid cognitions (Pramāṇayuktanidhi chapter 8). The count is fixed from the object side in a single line:

།གཞལ་བྱ་གཉིས་ཕྱིར་ཚད་མ་གཉིས། །དེ་ལས་གྲངས་གཞན་བཀག་ཕྱིར་རོ།

Because the objects of comprehension (གཞལ་བྱ) are two, valid cognitions too are two; for any other number has been refuted. (Pramāṇayuktanidhi chapter 8)

The verse is aimed at both directions of error at once, at schemes with more than two and at the Charvaka’s one, and Sodargye supplies the argument for the lower bound: rejecting inference leaves every hidden matter undecidable, and in practice the school does not manage without it, since “Not seen, therefore non-existent” is, in Sodargye’s words, “in fact a pseudo-inference”, and so an indirect admission of the very instrument being denied (sodargye-pramanayuktanidhi, on chapter 8). He adds that Mipham Rinpoche refutes the school at length in his commentary on the Madhyamakālaṃkāra, which is outside the material held here. Sodargye also records Dharmottara’s use of the same point: non-deception in essence, the first of his three features of valid cognition, “eliminates the view of the Cārvāka and others, who hold that only the grasped thing is valid cognition and reject all inference and all hidden things” (same chapter).

The colophon. The Charvakas are named among the authors whose “bad logic” the treatise was written to crush:

།ཐུབ་པ་སེར་སྐྱ་རྐང་མིག་པ་དང་འུག་པའི་བུ། །མཁའ་གོས་ཅན་དང་ཚུ་རོལ་མཛེས་པའི་གཞུང་འཛིན་པ། །ཐོས་སྒྲོག་པ་དང་གངས་རིའི་ཁྲོད་གནས་སྨྲ་རྣམས་ཀྱི། །རྟོག་གེ་ངན་འཇོམས་བསྟན་བཅོས་ཆེན་པོ་འདི་བྱས་སོ།

To crush the bad logic of the sage Kapila (སེར་སྐྱ, Sāṃkhya), of Akṣapāda (རྐང་མིག་པ, Nyāya), of the Owl’s Son (འུག་པའི་བུ, Ulūka/Kaṇāda, Vaiśeṣika), of the holders of the Sky-Clad (མཁའ་གོས་ཅན, Jain) and Cārvāka (ཚུ་རོལ་མཛེས་པ, Lokāyata) systems, of the Vedic reciters (ཐོས་སྒྲོག་པ, Mīmāṃsā) [translation uncertain], and of the proponents dwelling amid the snow mountains (the Snowland masters)—he composed this great treatise. (Pramāṇayuktanidhi chapter 11)

The commentators

Two of the school’s uses in the later tradition are worth separating from its refutation, because in both the Charvaka has stopped being an opponent and become a measure.

The first is Sodargye’s, and it is straightforwardly homiletic: the lectures behind sodargye-pramanayuktanidhi address a modern audience for whom materialism is the default, and the Charvaka position is repeatedly identified with it (“today’s materialists and Cārvākas always hold the human body to be the mind’s substantial cause”). The identification is the lecturer’s, not the Pramāṇayuktanidhi’s, and nothing in the root text or the auto-commentary licenses it.

The second is polemical and is aimed at Sakya Paṇḍita. Dreyfus records that Sakya Paṇḍita’s remarks suggesting that generally characterised phenomena are not objects of knowledge drew the charge that his view “is not able to escape the self-defeating paradox afflicting the Materialist view” (Bo-dong’s formulation), and that Gyeltsap made the same point more mildly: those who hold that generally characterised phenomena cannot be objects of comprehension “are no better than Materialists, who reduce knowledge to perception”. Gyeltsap then runs the Charvaka dilemma against him: is the thesis that generally characterised phenomena are not objects of comprehension itself validated by a valid cognition? If it is not, “it is false and its contradictory is true; hence, constructs are objects of valid cognition”; and if it is, “it has to be a real thing” (dreyfus-recognizing-reality, pp. 78–79). Dreyfus is candid that this is polemic rather than fair description, and that Tibetan debate of this period tends “to dramatically overstate” its cases (p. 78). The argument Dharmakīrti built against the Charvaka became, four centuries later in Tibet, a weapon against a Buddhist reading of Dharmakīrti, which is the shape of the dispute treated at realism-vs-antirealism.

What the debate turns on

Every argument on this page is an argument about what failing to see something establishes. The Charvaka’s whole case rests on non-apprehension: nobody sees anyone go or come, nobody sees a mind without a body, and what is not observed is not there. Dharmakīrti does not answer by producing an observation. He answers by regulating the inference: non-apprehension establishes absence only where the thing, if present, would have appeared, and a past life is not that kind of thing; the wisp of smoke is invisible to poor eyesight and still smoke. But that reply is a two-edged instrument, and PV III concedes the edge that cuts back: about what is extremely hidden, non-perception determines neither existence nor non-existence. The Buddhist therefore cannot claim to perceive past lives either, and the entire positive case has to be carried by an inference from what mind’s substantial cause must be. That is why the definition of a substantial cause at II.60–II.62 is the argument’s real foundation and the corpses, wicks and breaths are its illustrations: the conclusion is only as strong as the claim that a cause which does not itself change into its effect is no substantial cause of it.

The second front reverses the polarity. Against the Charvaka’s denial of inference, the Buddhist argument is not that inference works but that the denial cannot be stated: the thesis that universals are not objects of valid cognition is itself known, and known by something that is not perception. This is a strong argument and it has an exposed premise, which Dreyfus names: it forces Dharmakīrti to grant conceptual constructs a secondary, epistemic mode of existence, and every Tibetan quarrel about generally characterised phenomena descends from that concession.

Two things this page cannot settle. The material surveyed here gives Dignāga no engagement with the school at all, which is either a fact about the Pramāṇasamuccaya or a gap in the sources gathered. And the opponent refuted here, a Charvaka who rejects inference outright and so refutes himself the moment he argues, may not be the historical school: Purandara and others are reported to have accepted inference for worldly matters and to have refused it only for the transcendental, which is precisely the class the rebirth proof falls into. Against that opponent the self-refutation argument does no work at all, and what remains is the argument from the substantial cause, standing alone [unverified].

Graph View

Backlinks