Vaiśeṣika

Vaiśeṣika (བྱེ་བྲག་པ་)

Vaiśeṣika supplied the ontological vocabulary that Buddhist epistemology then spent three treatises taking apart. Substance, quality, activity, universal, particularity and inherence are its six categories, and Dignāga, Dharmakīrti and Sakya Paṇḍita argue against every one of them in words the school coined. Two fronts carry the weight in our material. The first is the whole (avayavin, ཡན་ལག་ཅན་): the claim that a composite is a further substance over and above the parts it is made of, which dharmakirti answers with a hand, a paint-brush and a pair of scales. The second is the real universal held to be a substance distinct from its instances, against the Sāṃkhya version that makes it one substance with them; pramanavartika I.150–156 works through three named Vaiśeṣika teachers in succession, and sakya-pandita disposes of the position in a single line. dignaga gives the school a refutation section of its own in four of the six chapters of the pramanasamuccaya, and answers a Vaiśeṣika objection in a fifth.

Overview

The school is named after one of its own categories: viśeṣa, particularity, rendered བྱེ་བྲག་ in Tibetan, so that བྱེ་བྲག་པ་ is roughly the particularists. Tibetan sources more often use the nickname. Its founder Kaṇāda carried the epithet Ulūka, “owl”—tradition says he studied all day and ate once a night, like an owl—and the school is accordingly the Owl School (འུག་ཕྲུག་པ་) or, in Sakya Paṇḍita’s colophon, the school of the Owl’s Son (འུག་པའི་བུ་); the Sanskrit adjectival form Aulūkya appears in pramanavartika II.19. “Kaṇāda” itself means eater of grains or of atoms (Wikipedia).

Kaṇāda is dated anywhere between the sixth and the second century BCE, with the Vaiśeṣikasūtra systematised between 200 BCE and the start of the common era; both figures are a spread rather than a date (Wikipedia). Praśastapāda’s Padārthadharmasaṃgraha (c. 4th century) is less a commentary than an independent restatement, and it is Praśastapāda whom Gedun Drub groups with the Nyāya when reporting their theory of direct perception (Gedun Drub’s commentary on PV III). Candramati’s Daśapadārthaśāstra survives only in Chinese translation (648). Śrīdhara’s Nyāyakandalī is dated 991, and Udayana, in the tenth or eleventh century, wrote on the texts of both schools and completed the fusion with Nyāya; from then on the joint system is Nyāya-Vaiśeṣika, and Vaiśeṣika has no separate life (Wikipedia; web search on the merger). The Vaiśeṣikasūtra itself does not mention a god. Later commentators concluded that unchanging atoms and the working of karma could not by themselves yield an ordered world, and took over the Nyāya’s creator (New World Encyclopedia, via web search)—which is why the theistic opponent of pramanavartika II is labelled an Īśvara-proponent even where he borrows Vaiśeṣika machinery.

Two features of the school’s relation to Nyāya matter for reading our texts. First, the six categories are Vaiśeṣika in origin and were adopted by the Nyāya at an early stage, along with other realist schools such as the Mīmāṃsā; inherence in particular is “a kind of cosmic glue that can never be observed on its own” (dreyfus-recognizing-reality, p. 58). Second, and less often noticed, the theory of perception that our texts attack under the Nyāya’s name was, on Dreyfus’s account, first propounded by the Vaiśeṣikas and only afterwards became the standard Indian account (p. 345). The division of labour between this page and nyaya is therefore a convenience of our sources, not a clean historical line. Where they do differ, the difference is epistemological: “Where Naiyāyikas accept four means of valid knowledge, Vaiśeṣikas, like Buddhists, only accept direct perception and inference” (gorampa-pramanavartika, translator’s introduction)—a point on which Wikipedia agrees.

The school was already unfamiliar to Tibetan scholars by the nineteenth century. Dreyfus quotes Ngak-wang-bel-den’s complaint that although scholars of his day were slightly familiar with the Sāṃkhya system, they were not familiar with the Vaiśeṣika system, which Jam-yang-shay-ba’s Great Exposition of Tenets does not explain extensively (p. 52). Our own material bears this out: the Vaiśeṣika is named constantly in the root texts and commentaries and almost never described.

Core doctrines

The six categories

Substance (རྫས་, dravya), quality (ཡོན་ཏན་, guṇa), activity (ལས་, karman), universal (སྤྱི་, sāmānya), particularity (བྱེ་བྲག་, viśeṣa) and inherence (ཡང་དག་འཕྲོད་པ་, samavāya)—six real kinds of thing, to which later Vaiśeṣikas added a seventh, absence (Wikipedia; the list is also given at gorampa-pramanavartika, note 214). There are nine substances: earth, water, fire, air, ether, time, space, self and mind. Gorampa reports the Tibetan understanding of how the substances divide: self, mind, the directions, time and space pervade everywhere; the four elements do not, and the measure of a non-pervading substance is one of its qualities, which they label solidity (gorampa-pramanavartika, on PV II.19). Sound, for the Vaiśeṣika, is a quality of space (sodargye-pramanayuktanidhi, on chapter 4).

The categories are not a private classification but a claim about what perception delivers—that a sense cognition hands over substance, quality and action together with the universal and the particular under which they fall. Dignāga states the position in the act of refusing it:

སྤྱི་དང་ཁྱད་པར་ལ་བལྟོས་པ་དང་། རྫས་དང་ཡོན་ཏན་དང་ལས་ལ་བལྟོས་པ་ནི་མངོན་སུམ་མོ་ཞེས་སྦྱར་བར་མི་བྱ་སྟེ།

[However,] one cannot assert that [cognition] which depends on universals (sāmānya) and particulars (viśeṣa), or which depends on substance (dravya), quality (guṇa), and action (karma), is direct perception, for— (pramanasamuccaya I)

In Fazun’s exposition the same line runs: “Considering the universal and the particular, considering substance, quality, and action, is perception”—this should not be said. Everything in the refutation follows from taking the claim seriously.

One word needs care. བྱེ་བྲག་ is the glossary’s preferred rendering of viśeṣa, “particular”, in its ordinary sense of an instance falling under a universal. As a Vaiśeṣika category it is narrower: the ultimate individuator that distinguishes one eternal substance from another. Here “particularity” is used for the category and “particular” for the instance.

Atoms, permanent and partless

The four elements reduce to indivisible atoms (རྡུལ་ཕྲན་, paramāṇu) which, unlike the Buddhist ones, are permanent. Atoms combine into dyads and triads before yielding perceptible bodies, and since they are inert in themselves the motion is imparted by the unseen force (adṛṣṭa, མ་མཐོང་བ་) of merit and demerit residing in individual selves (Wikipedia; New World Encyclopedia, via web search). That adṛṣṭa surfaces once in our material, borrowed by the Śaiva ritualist of pramanavartika II.263 to argue that since ritual destroys the unseen, there is no further going to rebirth (the chapter-2 attribution map, translations/pramanavartika/translation-notes.md).

The permanence of atoms is what a Buddhist proof at pramanavartika II.19 targets. Gorampa glosses the Aulūkya as the Vaiśeṣika, “who assert that subtle particles are permanent”, and takes the exchange as a lesson about debate rather than about atoms. The proof runs: a subtle particle is impermanent because it has solidity and so forth. The Vaiśeṣika will not apply the word solidity to an atom, since for them it names a quality of the non-pervading substances at ordinary scale; but they grant its meaning, tangibility. Since a reason works by its meaning and not by its label, the proof stands (gorampa-pramanavartika, on PV II.19).

The whole over its parts

A composite is not the parts. It is a further substance, inhering in them, produced by their conjunction, and—this is the load-bearing part—possessing properties of its own. Gedun Drub states the thesis with unusual precision: the combined thing generated by atoms mutually combining has substance-character together with three qualifying properties, namely bearing weight, the downward-falling action that weight draws, and being the cause of conjunction; and it exists as a separate substance over its perceived parts (Gedun Drub’s commentary on PV IV). A whole without those three, he adds, would be pointless to posit: it is exactly in order to explain obstruction and the rest that the school wants it.

Sodargye gives the same doctrine in its familiar form: the whole is a real thing pervading all its parts, so that the body exists on the head, on the eyes, on the hands and on the feet (sodargye-pramanayuktanidhi, on chapter 3).

Perception as a conjunction of four factors

The definition Dignāga cites, in our translation of the Tibetan:

བྱེ་བྲག་པ་རྣམས་ནི་བདག་དང་དབང་པོ་དང་ཡིད་དང་དོན་དུ་ཕྲད་པ་ལས་གང་གྲུབ་པ་དེ་གཞན་ཡིན་ནོ།

The Vaiśeṣikas [assert]: [Pramāṇa is] the contact of self (ātman), sense faculty (indriya), mind (manas), and object (artha); the cognition of the object established from that [contact] is other [than the pramāṇa—i.e. it is the phala]. (pramanasamuccaya I)

Fazun’s exposition renders the same definition “Formed from [the combination of] self, faculty, mind, and object, [perception] is something distinct”, and records a second Vaiśeṣika formulation: “Born only of [such] conjunction, it is perception with respect to substance.” Which conjunction is the valid cognition (ཚད་མ་) divides the school: some make it the conjunction of faculty and object, taking that as the distinctive cause; others make it the conjunction of self and mind, on the ground that this is the primary one (pramanasamuccaya I, refutation of the Vaiśeṣika position on perception).

The associated account of valid cognition and its result is reported by Sakya Paṇḍita in a survey of four rival schemes: the Vaiśeṣikas “speak of the distinction and the distinction-bearer” (ཁྱད་པར་ / ཁྱད་པར་ཅན་) (pramanayuktanidhi chapter 9). Sodargye glosses this as cognising the entity, the pillar, as the valid cognition, and cognising one of its distinguishing qualities, its being red or white, as the result (sodargye-pramanayuktanidhi, on chapter 9).

On mind and on inner states the school is a thoroughgoing realist about the self. Mind is atom-sized, permanent and one, so that when it attaches to the nose-faculty it is not available to the others and only one consciousness arises at a time; and pleasure and pain are qualities belonging to the self, cognised not by themselves but by a consciousness that is one substance with the self (Gedun Drub’s commentary on PV III). That last thesis is a direct denial of self-awareness (རང་རིག་).

Two means of valid cognition, and five relations that ground inference

Unlike the Nyāya the school accepts only direct perception and inference. What it adds instead is an account of why a reason indicates what it indicates. Dignāga reports the position: “‘This is the effect of this’; ‘cause’; ‘connection’; ‘combined in one locus’; and ‘contrariety’—these [five relations] all arise from ‘cause’.” Connection subdivides into conjunction (saṃyoga—the relation our Pramāṇavārttika translation renders “combination”, ལྡན་པ་), as of fire and smoke, and inherence (ཡང་དག་འཕྲོད་པ་, samavāya), as of cow and horn (pramanasamuccaya II, refutation of the Vaiśeṣika).

Words express the type-universal

Since a name cannot be attached to each of infinitely many individuals, the Vaiśeṣika holds that a word expresses the type-universal (jātisāmānya, རིགས་སྤྱི་), and reaches the individual through it. Sakya Paṇḍita gives the position two moves. The first is that since words express the type-universal and not the individual, no fault of unnameable particulars arises; the second, when pressed, is that “We apply [the name] to the specifically characterised phenomenon that is connected with the universal, and by that the specifically characterised phenomenon is understood” (pramanayuktanidhi chapter 5). Behind it lies a semantic realism the Buddhists take to be the school’s real commitment: words express by the power of the entity rather than by the speaker’s wish to express (pramanavartika I.65 with gorampa-pramanavartika ad loc.).

The refutations

Dignāga

Chapters 1 to 4 of the pramanasamuccaya each carry a section titled against the Vaiśeṣikas, chapter 3 carrying two; chapter 5 answers a Vaiśeṣika objection without a section of its own; and chapter 6 adds nothing of its own, sending the detailed refutation of the various sham rebuttals back to the examination of the principles of the Vaiśeṣika and of Akṣapāda’s school (pramanasamuccaya VI).

Perception (PS I). If a sense cognition delivers substance, quality, activity, universal and particularity together, then the qualifier and the qualified will belong to different faculties: “this sweet scent” would be perception, though scent is the nose’s object and sweetness the tongue’s. Suppose the reply is that one substance is grasped by many faculties. Then colour and the rest should be grasped by many faculties too, which is never observed. Suppose instead that the faculties are different but see only the substance, by an undifferentiated cognition—then it is not the faculty that sees, and positing several faculties becomes pointless, since each would reach the others’ objects. Dignāga presses the same fork on the notion of determinateness: if what fixes each faculty to its object is the presence of colour-ness, tangible-ness and so on, then substance, number and activity, having none of these, are not grasped by any faculty at all, and the categories drop out of perception entirely. The cognition that takes substance as non-different from its qualities is therefore a recollection, and mental: “Such a distinction is made only by mental consciousness, and only so is it reasonable.” His verdict is that the Vaiśeṣika perception-doctrine “is extremely difficult to establish” (pramanasamuccaya I).

The relations behind inference (PS II). This is the most consequential of the four, because Dignāga’s own account of the reason is what replaces what he tears down. If conjunction is what makes smoke indicate fire, then conjunction holds of both relata equally and fire should indicate smoke as readily—yet a red-hot iron ball is fire without smoke. The Vaiśeṣika answer, that the single counter-instance shows the relation runs one way only, is exactly what conjunction cannot deliver: “the conjunction itself has no differentiation: whatever holds of one relatum holds likewise of the second”. Worse, if conjunction grounds the inference then every part of smoke down to its substance-hood is a reason, and every particular of fire down to its brightness is inferred, since nothing about the thing falls outside the conjunction. Inherence fares no better—cow and horn are not thereby reason and inferendum—nor does combination in one locus, nor contrariety. What does the work in every case is “without-which-not” (avinābhāva, མེད་ན་མི་འབྱུང་): the reason and the thing possessing the reason are determined not by conjunction but only by without-which-not (pramanasamuccaya II). The relation that grounds inference is thus removed from the ontology altogether, which is the move the whole tradition rests on.

The reason (PS III). The Vaiśeṣika definition is compact: “The property in conjunction with that is the reason”, where “that” is the probandum. Dignāga asks what the definition ranges over. If the aggregate, then the universal impermanence-property becomes the reason for “sound is impermanent”, since a universal has no other property, and “it is a horse because it has horns” comes out equally well. If the bare property, its produced-ness is nowhere to be found; if the property-possessor, it is already established for both parties and proves nothing. And the definition licenses “sound is eternal, because audible”, since audibility is certainly a property conjoined with sound; adding “in conjunction with that, similar-property, [and] non-errant” to exclude it concedes the original formulation was empty. A definition that must be repaired by importing the three modes is not a rival account of the reason (pramanasamuccaya III).

The fallacious reason (PS III). Their three kinds are the unestablished, the undisplayed and the doubtful-non-existent. Dignāga answers that a merely non-existent reason makes nothing known, so it is not a species of fallacy but the absence of a reason; that “horned” as a reason for “a horse” does not produce doubt but establishes the opposite, and so is contradictory rather than doubtful; and that the list omits the uncommon reason, where the mark is absent from both the similar and the dissimilar class. The exposition ends by conceding difficulty: “These meanings still require study” is Fazun’s own note in the middle of the passage [unverified] (pramanasamuccaya III).

The example (PS IV). “That which is fully established for both [parties] is the example.” Then, since thesis and reason are both established with respect to space as its qualities, everything becomes an example with respect to space; and an example that does not display the concomitance of reason and probandum is a pseudo-example, so the definition has picked out the wrong feature. Fazun records that this chapter’s exposition “is overly brief and diverges considerably from Xinhui’s translation; it requires further study”, so the reconstruction here rests on a text the translator himself would not settle [unverified] (pramanasamuccaya IV).

Universals (PS V). The Vaiśeṣika argues that universals have differentiations by their own nature, since a thing depending on its universal is what displays it—head, back, belly and hand display the cow. Dignāga’s reply is that displayer and displayed have been defined by each other: ask what substance is, and it is what displays its universal; ask what the universal is, and it is what its substance displays. Neither can show its own essence, and if each is many and one by turns, mutual display multiplies them without end (pramanasamuccaya V).

Dharmakīrti

The whole, and a pair of scales. The refutation of the whole runs through pramanavartika II.84–102 and returns in chapter 4, and it is the place where this school’s specific commitments, rather than realism in general, are being attacked.

Chapter 2 reaches it from an unexpected direction. The Cārvāka has claimed that mind arises from the body, so Dharmakīrti asks what the body is: one coarse substance, or many particles? If one, then moving a hand moves everything, since contradictory activities—moving and not moving—cannot belong to one partless thing.

།ལག་སོགས་གཡོ་ན་ཐམས་ཅད་དག །གཡོ་བར་འགྱུར་ཕྱིར་འགལ་བ་ཅན། །ལས་ནི་གཅིག་ལ་མི་རུང་ཕྱིར། །གཞན་དུ་ཐ་དད་གྲུབ་པར་འགྱུར།

[Were the body one whole,] then if a hand or some other part moved, all would move; and since contradictory acts [moving and not moving] cannot belong to one [thing], it is instead established that [the parts] are distinct. (pramanavartika II)

Cover the head and the feet are covered; paint the face and the whole body is painted; or, if not, then one part is dyed while the rest is not, and the single substance has come apart. Gorampa adds the escape the Vaiśeṣika must not take: if you say the part moves while the possessor of parts does not, you have said that the possessor is a substance distinct from the parts, which is the thesis under examination (gorampa-pramanavartika, on PV II.84–86).

The Vaiśeṣika counter is the strongest card the school holds: if bodies and pots are nothing but many atoms, then they are as imperceptible and as unobstructive as the atoms they came from. Dharmakīrti denies the identity rather than the inference. Atoms that have come into contact with others are not the atoms that had not:

།དུ་མ་ཉིད་ན་སྔ་མ་བཞིན། །ཁྱད་མེད་ཕྱིར་དང་ཕྲ་རབ་ཕྱིར། །རྟོགས་མིན་ཞེ་ན་ཁྱད་མེད་པར། །མ་གྲུབ་ཁྱད་པར་ཅན་རྣམས་ནི། །དབང་ཡུལ་དེ་ཕྱིར་རྡུལ་ཕྲན་མིན།

[Objection—Vaiśeṣika:] “If [it is instead] many [atoms], then, as before, being mutually indistinct and subtle, it would not be apprehended.” [Reply:] It is not that [the atoms] are indistinct: bearing [their] distinctive features [in aggregation], they are objects of the senses—hence [what is perceived] is not [a coarse whole over and above] the atoms. (pramanavartika II.87)

Quicksilver in gold and mottled stone then appear as they do because contacted atoms present that aspect, not because a further substance has arrived. Offered combination (saṃyoga, ལྡན་པ་) as the thing perceived, he answers that its substrate is not seen, so nothing is left to perceive it by, and that one combination belonging at once to taste and to form contradicts the school’s own rule that a quality bears no quality. Offered designation instead, he answers that the school has then granted the Buddhist case for one class of features and owes a reason for withholding it elsewhere—and that “long” said of a row of a hundred birds inheres in no member of the row (pramanavartika II.88–90 with Gorampa).

Chapter 4, as Gedun Drub expounds it, brings the dispute down to a measurement. If the composite bears a weight of its own and the downward action that weight draws, put it on a balance. Take two equal masses of pellets, one merged in stages into a single lump, the other left unmerged: the merged pan should sink, and it does not. The defence—that the extra substance’s weight is real but too small to register, as cotton-fluff added a wisp at a time never moves the scale—is answered on its own terms. Cotton-fluff is perceptible and its weight is registered on the scale, whereas the further substance has no perceived character at all and no detectable effect; and if the whole’s weight were merely small, a progressively combined pile with many part-possessors should still outweigh an at-once-combined pile with one. Nor does any other quality of the whole show up: no increase of colour is seen either, though the eye that sees the parts should see it. The conclusion is stated flatly: the Vaiśeṣika thesis “a separate substance-part-possessor apart from parts exists” is not a true thesis, and falls to perception, inference and common usage as applicable (Gedun Drub’s commentary on PV IV).

One limitation is recorded in the same passage, and it is worth keeping. Gedun Drub says the refutation targets the qualifier of the thesis—the whole as distinguished by the three properties—and “does not challenge the separate-part-possessor as such”. The argument shows there is no weight-bearing, conjunction-causing extra substance; a bare part-possessor with no properties of its own is left untouched, which is precisely what the school itself said would be pointless to posit.

The universal as a distinct substance (PV I.150–156). Dharmakīrti’s opening dilemma applies to the Sāṃkhya and the Vaiśeṣika at once: if the class is one thing, whether the same as its instances or other than them, no cognition of it can arise in an instance that did not exist before: “For what is seen in the one [earlier instance] cannot [then] be seen in another [new one]” (pramanavartika I.150–151). The half aimed at the Vaiśeṣika is the second: even if other, it would not depend on the new instances, and so is not in them.

What follows is a sequence of three named positions, and Gorampa is our only witness to the names. Kaṇāda holds the universal to be one, unmoving, partless, and fit to appear as pervading all particulars. His pupil Pethara, feeling the force of “because it is not a support”, allows that there is no universal until an instance exists, and says the universal of the earlier particulars pervades the later ones. Dharmakīrti’s answer is that what abides does not stir from where it abides, and that a universal arriving in a new place is not what the school means:

།མི་འགྲོ་དེ་ན་ཡོད་པའང་མིན། །ཕྱིས་ཡོད་ཆ་དང་ལྡན་པའང་མིན། །རྟེན་སྔ་མ་ནི་ཡོངས་མི་འདོར། །ཀྱེ་མ་ཕོངས་པའི་རྒྱུ་ཉིད་ཡིན།

It does not travel [there], nor is it present in those [later instances]; it has no part in the later [ones]; yet it does not relinquish its earlier support—alas, what a cause of ruin [for your thesis]! (pramanavartika I.152)

Beluka takes the other branch: the class exists everywhere already, so a new instance finds it waiting. Then, being partless and illustrated by the instances in one place, it would be seen in every place, including where no instance is (pramanavartika I.155). The last retreat—it is everywhere but seen only where instances display it—reverses the school’s own order of knowing: if the illustrator must be cognised for the illustrated to be cognised, you have made the universal depend on the particulars, having begun by making the particulars known through the universal (pramanavartika I.156 with gorampa-pramanavartika ad loc.). The three names, Kaṇāda apart, appear in our material only in Gorampa; the translation notes list Pethara and Beluka (also given as Pailuka) as glossary candidates, and no source we hold identifies them further [unverified].

Dharmakīrti’s positive alternative is put without ceremony:

།ཐ་དད་མེད་དོན་མེད་པར་ཡང་། །འཚེད་པ་པོ་སོགས་བརྗོད་པ་ཡིན། །ཐ་དད་ཕྱིར་ན་ལས་རྒྱུ་མིན། །འདི་ཡི་རིགས་མིན་ལས་འབྲེལ་ཕྱིར།

[Reply:] Even with no undifferentiated [universal], “cooks” and the like are [still] expressed [of many]. [And the act of cooking is not the cause:] since [the acts] are distinct, the work cannot be the reason; nor [can] this class “work”, since its connection is [only] to the work—(pramanavartika I.157)

What groups the cooks is not a shared entity but the nearness or non-nearness of a class-exclusion—the apoha account.

Quality and number as separate substances (PV II.91–96). The Vaiśeṣika makes number, conjunction and activity substances distinct from what bears them, and argues from language: we speak of the pot and of the pot’s number separately, so the two are separate. Dharmakīrti answers that separate isolates are enough to explain separate words, since word and cognition “label by way of a conceptual cognition that follows separate entities”; and that the school has already conceded the principle, since it treats the number of a quality as merely designated. If designation works there, the reason for refusing it everywhere else is owed (pramanavartika II.93–94). Restricting real distinctness to things with distinct features fails, because the names that let one comprehend numbers without error just are the names of those numbers. And if even the number of the quality white is an intrinsically separate entity, then quality is no longer different from substance and the categories collapse into one another (pramanavartika II.96).

Inherence has nothing to hold up (PV II.68–69, II.230). A pot supports water by keeping it from flowing away; what would a motionless universal, quality or activity need a support for? The same reasoning removes inherence itself, which was posited as the support that keeps an already existing thing from disintegrating (pramanavartika II.68–69). The point returns in the section on the antidote to the view of self: the objector says that seeing attachment as suffering will remove the cognition of what belongs to self, and Dharmakīrti replies that the connection of inherence from which that cognition arises “will remain as it was” whether it is seen as suffering or not. Gorampa’s note is unusually direct: “This is a refutation of the Vaiśeṣika school” (pramanavartika II.230 with gorampa-pramanavartika ad loc.).

The genitive case (PV I.65–67). If property and property-possessor are not really separate, says the Vaiśeṣika, then the genitive case and the plural have nothing to mark. Dharmakīrti concedes the objection has force—against those who share the premise: for the Vaiśeṣika and the Grammarian, who hold that words are expressed by the power of the entity and not primarily by the wish to express, “it would be correct for them to argue by way of the sixth case, word declensions—that is, plural forms—and so on”. It is not correct against a position on which the speaker’s wish to express fixes what a word does (pramanavartika I.65 with gorampa-pramanavartika ad loc.).

Nonapprehension (PV I.20, I.22). The Vaiśeṣika reason that neutral tactile sensation is not present in the visible elements, because it is not observed there, is used by Dharmakīrti as a counter-example: if mere non-observation in the dissimilar class established a negative pervasion, that reasoning would be faultless, which it is not. A self in the body, consciousness in earth, and the Sāṃkhya’s unmanifest effect are lined up in the same verse for the same purpose (pramanavartika I.20–22 with gorampa-pramanavartika ad loc.). The school here is a source of examples rather than a target.

Against the self as knower (PV III). Two Vaiśeṣika theses fall in the perception chapter. That pleasure and pain are cognised by a consciousness one substance with the self is answered by showing that they arise from the three conditions of a cognition and not from the self’s quality and action—present when those conditions are, absent when they are not—and so are cognitions themselves, which is self-awareness. That mind is a permanent atom-sized unit, entering one faculty at a time, is answered by the ordinary fact that several consciousnesses arise at once, and by the general principle that a permanent thing performs no action at all (Gedun Drub’s commentary on PV III).

Sakya Paṇḍita

The root text of the pramanayuktanidhi names the Vaiśeṣika in eight of its eleven chapters (1, 3, 4, 5, 6, 7, 9 and 11) and argues against a position of theirs in six, briefly in each. Its longest engagement with the school’s central thesis is not with the school at all, but with Tibetans whom Sakya Paṇḍita takes to have reinvented it.

The universal (chapter 3). The refutation is one line long:

།ཁ་ཅིག་ཐ་དད་ཡིན་ཞེས་ཟེར།

[Opponents—the Vaiśeṣikas:] Some say [that universal and particular] are distinct substances.

།ཐ་དད་སྣང་རུང་མེད་པས་ཁེགས།

[Reply:] This is refuted: a distinct [universal] that ought to be perceptible is nowhere observed. (pramanayuktanidhi chapter 3)

If a real universal stood on its particulars as a further substance, it would be the kind of thing that shows up, and it does not. The same reason disposes of the Sāṃkhya’s one-substance universal in the next line, so both classical realisms fall to nonapprehension of what should appear (སྣང་རུང་མ་དམིགས་).

The three consequences that follow in the chapter are aimed elsewhere, and this is worth being exact about, because the Vaiśeṣika appears in the first of them as the measure of absurdity rather than as its target: “If universal and particular are one substance, on what grounds could [you] refute the [Vaiśeṣika] whole-with-parts (avayavin)?” Sodargye draws out why the argument cannot be meant for the Sāṃkhya—a Sāṃkhya would simply reply that he never refuted the whole and is happy to keep it—and concludes that the target is the early Tibetan epistemologists, who accept a real universal pervading its particulars and then join Buddhists in refuting the Vaiśeṣika whole. The other two consequences, that the senses would become conceptual and that all contradictories would be one substance, run the same way (sodargye-pramanayuktanidhi, on chapter 3; see realism-vs-antirealism).

The chapter’s concluding verse ranks the positions: “The Owl School (འུག་ཕྲུག་པ་, the Vaiśeṣikas) hold the universal to be a distinct substance; the Sāṃkhyas hold it to be one substance; the Snow Landers (early Tibetans) follow them; to hold it as exclusion is the tradition of the wise” (pramanayuktanidhi chapter 3).

Words and the universal (chapters 4 and 5). The Vaiśeṣika objects that if there is no universal the word for it has no cause. The reply refuses the demand rather than meeting it: words are deployed by the force of the wish-to-express (བརྗོད་འདོད་), so they do not depend on their objects—otherwise “barren woman’s son” could not be said (pramanayuktanidhi chapter 4 with sodargye-pramanayuktanidhi). A second objection is that without a shared universal the properties of a blue utpala could not converge in one bearer, and the notion of a common locus (གཞི་མཐུན་) collapses; the reply grants the convergence and relocates it, since exclusion divides and gathers in conception’s perspective, and a common locus is therefore not the domain of sense cognition at all (pramanayuktanidhi chapter 4). In chapter 5 the type-universal as the direct expressed is dismissed in two steps: there is no universal in the object, and even if there were, applying a name to it would have “neither need nor power”—so apply the name to the specifically characterised phenomenon directly. The school’s own reason for introducing the universal, that particulars are limitless, then rebounds: “for this [position of yours] too the same [fault] holds” (pramanayuktanidhi chapter 5).

Smaller engagements. In chapter 6 the Sāṃkhya, holding the self sentient, and the Vaiśeṣika, holding it insentient, both posit a connection of agency between the permanent self and its activity; Sakya Paṇḍita answers that no such connection needs subsuming under his two, because a permanent self is not established even by conventional valid cognition (pramanayuktanidhi chapter 6 with sodargye-pramanayuktanidhi). In chapter 7, on how one thing expels another, the Vaiśeṣika answer is that on contact the weaker is transformed; if it is transformed it is not permanent, which is what the school asserts it to be (pramanayuktanidhi chapter 7). In chapter 9 the distinction-and-distinction-bearer account of valid cognition and result is refused along with the other three rivals, all of which construe the pair as producer and produced; Sodargye adds the internal objection that if distinction and distinguishing property are wholly separate, the school’s own claim that one knows the entity through its quality is lost (pramanayuktanidhi chapter 9 with sodargye-pramanayuktanidhi). The colophon to chapter 11 names the Owl’s Son among the six opponents the treatise was written to answer (pramanayuktanidhi chapter 11).

By chapter 10 the school has become a fixture of the debate manual rather than a live opponent. Explaining the fourth case of a reason that pervades-or-not in both the similar and the dissimilar class, Sodargye’s commentary sets the argument “sound is permanent, because it is intangible” before Vaiśeṣika opponents specifically, because they alone hold the atom to be permanent and tangible, which is what makes “intangible” fail to reach part of the similar class (sodargye-pramanayuktanidhi, on chapter 10).

The commentators

The Vaiśeṣika whole reappears in Gorampa as a charge against Tibetans. Some who count themselves scholars of the Seven Treatises, he says, verbally refuse a possessor of parts distinct from its parts while holding that a coarse form built from atoms is a separate intrinsically existing entity, obstructive, perceptible, and possessing a weight of its own that is not the weight of the atoms. Quoting Jamyang Gekpo’s commentary: “Therefore they accept form as the possessor of parts and of this extra weight, as asserted by the Vaiśeṣika” (gorampa-pramanavartika, on PV II.87). What the Sautrāntika holds instead, on Gorampa’s reading, is that the coarse aspect is the entity of the sense consciousness, not a coarse form in the object; and he records without settling it the Vijñaptika’s question to that position, “How do these particles create such a coarse form appearance?”

sakya-chok-den makes the same kind of charge in the vocabulary of isolates: what several distinguishers assume on the basis of a single substance are individuations, and the universal is the mere elimination of what is not that substance; to say instead that the different substances are the individuations is to take up the position of non-Buddhists such as the Vaiśeṣikas (dreyfus-recognizing-reality, p. 170). Dreyfus makes the terminological point behind both charges: the basis of which a property is predicated is called substance (རྫས་, dravya) by the Vaiśeṣikas and a distinguisher (ལྡོག་པ་, vyāvṛtti) by the Buddhists, while “subject” (ཆོས་ཅན་, dharmin) is accepted by both sides as ontologically neutral (p. 158).

What the debate turns on

Every Vaiśeṣika thesis in this material posits a further real entity to explain a feature of cognition or language: a real universal because one word covers many things, a real whole because we see one pot and not a swarm, real inherence to hold the two together, a real conjunction to make smoke tell us about fire. The refutations are more uniform than the theses. In each case the posited entity is looked for where it would have to show up and is not found—the universal in a new instance, the whole on the balance, the substrate of the conjunction, the distinct universal beside its particulars—and in each case it turns out not to do the work it was hired for, since a permanent universal cannot be produced by the instances that were supposed to display it, and a relation with no direction cannot tell smoke from fire.

The Buddhist replacement is deflationary on both fronts at once: aggregated atoms bearing features are what the senses take, and exclusion is what a general term picks out. Neither is free of cost, and our material records both bills. Gorampa concedes that the coarse appearance still needs an account, and leaves the Vijñaptika’s question standing. And Dharmakīrti’s most concrete argument against the whole, on Gedun Drub’s own reading, refutes the whole as qualified by weight and obstruction while leaving a bare part-possessor formally unrefuted—a gap the school could not exploit, since its whole was posited to explain weight and obstruction in the first place, but a gap in the argument as stated.

One attribution should not be tidied. pramanavartika I.194–196, against the view that a pot’s destruction waits on an external later cause such as a hammer, is a classical Nyāya-Vaiśeṣika position, and the Buddhist reply is central to momentariness; Gorampa names no school there, and the chapter-1 attribution map records it as unspecified. It is left unattributed here for the same reason.

Related pages

• problem-of-universals—the realist side, of which this school holds the distinct-substance version
• apoha—the exclusion account that replaces the type-universal
• nyaya—the sister school, which took over these six categories and merged with this one in the 11th century
• samkhya—the other real-universal position, one substance with its particulars
• realism-vs-antirealism—the intra-Tibetan continuation, where the whole and the real universal reappear as charges
• perception, self-awareness—the Buddhist accounts that replace the four-factor definition and the self as knower of its own qualities
• inference, can-inference-be-valid—Dignāga’s replacement of the five relations by “without-which-not”
• momentariness—against permanent atoms and externally caused destruction
• dignaga, dharmakirti, sakya-pandita, gorampa, sakya-chok-den—the refuters, and the commentators who turned the charge on Tibetans