Perception in Dignāga’s School of Philosophy
Dharmakīrti in his Nyāyabindu, a representative work on Buddhist Logic and Epistemology, has divided all true knowledge into two broad classes, viz., (1) perception and (2) inference. All human activities depend for their success in the last analysis, on true and authentic knowledge and Dharmottara, the author; of an authoritative commentary on the Nyāyabindu, defines, this true knowledge in his commentary as knowledge which is capable of verification, or in bis own words, which does not disagree with the objective reality represented in it.. Correspondence of knowledge with reality is regarded as the test and warrant of its validity and this correspondence is attested when knowledge leads to the actual attainment of the object creating.a volitional urge for the object presented.
So the purpose of knowledge is served when it reveals an objective, reality in its true character; and the actual attainment of the object, which takes place by reason; of a chain of psychical facts, beginning with desire and volitional urge, and ending in actual physical endeavour, is only a bye-product. This intermediate link between knowledge and attainment has only a psychological importance and though they have an important bearing on the problem of truth, the logical value of these intermediate psychical states is only mediate and derivative. Dharmottara explicitly asserts that the function of an accredited instrument of knowledge (pramāṇa) is completed when the object is apprehended. The volitional urge and the attainment follow as necessary consequences. It follows, therefore, that an instrument of knowledge fulfils itself by making known an object which is not cognised before. A [Page 274] cognition, which reveals an object which has been known before, is redundant and so is not an independent pramāṇa.1
Dignāga omitted to put pratyakṣa under the rubric of “Valid knowledge” (sa myagjñāna) as Dharmakīrti has done and Uddyotakara has made capital out of this apparent omission.2 There is, however, no room for honest doubt that Dignāga proposed to give a definition of pratyakṣa as a species of valid knowledge and could not mean anything else, Śāntarakṣita also did not care to supply the word jñāna (cognition) in his definition of pratyakṣa and Kamalaśīla observes that the word jñāna has not been read in the definition as the negation of kalpanā (ideal or conceptual constructions) perforce indicates that it must be knowledge, which is alone liable to be associated with conceptual elements.3
This appears to be a trifling matter and is stated here only with a view to drawing the attention of the readers to the trivial and frivolous character of some of the criticisms of the Brāhmaṇical writers. Most of these criticisms are misleading as evidence of Buddhist doctrines and unless they are corroborated by the original writings of Buddhist authors themselves, the only course of action for an honest student of Buddhist [Page 275] philosophy will be to hold his judgment in suspension. There has been a good deal of conscious or unconscious misrepresentation and suppression of facts and suggestio falsi and this should be regarded as sufficient warning against placing implicit reliance on the evidentiary value of such testimony.1
Dignāga’s definition of Pratyakṣa
Pratyakṣa has been defined by Dignāga as “pratyakṣam kalpanāpoḍhām,” which in simple English can be rendered as “perception is (a cognition) which is free from conceptual constructions.” This single adjective has been deemed sufficient to exclude inference, which is invariably associated with ideal constructions (kalpanā). It is also competent to exclude errors and illusions (bhrama) from the category of perception, as errors and illusions are never in harmony with facts though they may be free from ideal elements. Perception, however, being, a species of authentic knowledge presupposes as a necessary condition this harmony of fact with knowledge and as illusions do not admit of verification, which is the only test of this harmony, there is no possibility of confusing them with valid knowledge, much less with perception which is only a subdivision of the same. So we see that the definition of pratyakṣa, as propounded by Dignāga, is self-contained and self-sufficient.
1It must be observed here that there are a good many Brāhminical writers who possessed first-hand knowledge of Buddhist philosophical works and who have tried to criticise the Buddhist position on fair grounds. Barring a few inaccuracies here and there, the account of Buddhist doctrines, as given by Kumārila, Vācaspati Miśra and Jayanta Bhaṭṭa in their works, appears to be a faithful representation of the Buddhist position and so will continue to attract the attention of students of Buddhist philosophy, particularly so when the original works of Buddhist writers have been lost for the most part.
Dharmakīrti’s definition of Pratyakṣa
[Page 276] Dharmakīrti, however, has added another element, namely, abhrānta (non-erroneous) to Dignāga’s definition with a view to excluding errors from the category of perception. This additional qualification, however, is redundant as we have seen that Dignāga’s definition is competent to exclude such contingencies. This addition, however, has been a source of confusion and has led to polemic among the commentators. We have it on the authority of Śāntarakṣita that there were some thinkers who regarded illusions as purely mental facts, having nothing to do with sense-perception; and so these thinkers objected to the inclusion of the adjective ‘non-erroneous’ (abhrānta) in the definition of pratyakṣa, as uncalled for. But Śāntarakṣita has stoutly opposed this view on the ground that as illusions occur on the operation of particular sense-organs and cease when this operation ceases, they should be regarded as sensuous aberrations and not pure mental errors. They arise only when there is a defect in sense-organs concerned, and if organic defect is not held to be responsible, these errors would disappear in spite of this defect, if the person is logically persuaded of his error.
But however much a man might be satisfied by reasoning, his illusory perception does not disappear so long as the organic defect is not removed. A jaundiced person, though persuaded of the error, does not cease to see things yellow until the jaundice is cured. But mental illusions, such as belief in the existence of supernatural beings or of universals (bhāvasāmānya) as objective categories, however obstinate and confirmed by habits, are seen to disappear when the deluded person is properly schooled in philosophic thinking. But the mirage or the double moon will not cease to be presented unless the physical defect is removed. Moreover, the vivid presentation; of false objects in illusions cannot be accounted for unless they, are regarded as sensuous presentations. Śāntarakṣita, therefore, concludes that illusions being perceptual knowledge and [Page 277] being free from ideal constructions could come within the category of perception, unless the saving clause is added to Dignāga’s definition.4
Vinītadeva, an older commentator on the Nyāyabindu, however, gave a different interpretation of the expression ‘abhrānta.’ He interpreted ‘abhrānta’ as meaning ‘not lacking correspondence with reality’ (avisaṃvādaka). But this alone would be wide enough to include inference as the latter too does not lack this correspondence. So the other clause “free from ideal constructions” is added for the exclusion of inference, which is invariably attended with ideal elements.
“Abhrānta should not be construed,” says Vinītadeva,
“as meaning a cognition which is contrary to and so erroneous in respect of the object. This interpretation of the word ‘ abhrānta’ would make the definition absolutely futile as. all knowledge, let alone perception, is erroneous with regard to its object according to the Yogācāras (Buddhist subjective idealists) and accordingly this definition has been so worded as to meet their position also.”
This interpretation of Vinītadeva has been strongly animadverted upon by Dharmottara. Dharmottara observes that this interpretation of the word ‘ abhrānta’ as “not lacking correspondence with reality” is itself futile, as from the context which treats of ‘true and authentic knowledge’ and of perception as a sub-species of the same, we have it that perception must not be incongruent with fact, because authentic knowledge connotes this very congruence and not anything else.
So Vinītadeva’s interpretation would make the definition tautologous, as [Page 278] the definition in relation to the context would read as follows;
“The cognition which is not incongruent and is free from ideation (kalpanā) is not incongruent.”
But this reiteration of ‘not incongruent’ does not answer any purpose. So the word ‘abhrānta’ should be taken to mean that which is not contrary to the real object presented in it. But what about the position of the idealist? The definition so interpreted will not meet their purpose. The author of the sub-commentary assures us that there is absolutely no difficulty as the definition has been propounded from the Sautrāntikas position and not from the idealistic standpoint, though the former is not the orthodox position of the master (ācārya).5
Is the adjective ‘abhrānta’ absolutely necessary even from the Sautrāntika standpoint ?
If we look deeper into the meaning of the definition, we shall see that the adjective ‘abhrānta’ is not necessary. Perception being a species of valid knowledge must be free from discrepancy with fact and this is adequate to exclude ‘errors,’ as errors are invariably discrepant with reality. The adjective ‘abhrānta’ is, therefore, useless whether it is taken in the sense of ‘non-discrepant’ (avisaṃvadaka) as Vinītadeva suggests, or [Page 279] in the sense of ‘non-erroneous’ as proposed by Dharmottara. The idealistic position has been severely left alone and the Sautrāntika standpoint can be fully met even without this qualification. The question pertinently arises—what led Dharmakīrti to propose this amendment? We have the answer from Dharmottara’s commentary and its confirmation from the Tattvasaṅgraha. Dharmottara observes that the twofold qualification is introduced in the definition to combat a prevailing misconception and not for the exclusion of inference, as for this the adjective “free from ideal constructions” is sufficient.
If the second epithet was not added, such experiences as of moving trees and the like could be regarded as true perception, as these are free from ideation and capable of satisfying the pragmatic test. But these experiences are absolutely false and so cannot be included in the category of valid perception.6 Śāntarakṣita and Kamalaśila too observe that there were certain thinkers among the Buddhists themselves who held even these abnormal experiences to be valid knowledge inasmuch as they satisfied the pragmatic test. But both Śāntarakṣita and Dharmottara rightly point out that what constitutes validity is not pragmatic fitness alone, but that plus harmony of presentation with reality. So such presentations as that of the light of jewel for the jewel itself, or of yellow conch-shell for a really white conch-shell, or of moving trees for trees which are really fixed and stationary are not valid perceptions though there is actual verification. Mere verification and pragmatic satisfaction cannot however be accepted as the test of [Page 280] validity; but verification of presentation with reality is the criterion.
What is presented is the light of the jewel or the white conch and what is actually attained is not the yellow conch or the light of the jewel, but something different. In the mirage, too, what is presented is the refracted light of the sun and the determinate experience is of water. In the case of the jewel’s light which is mistaken for the jewel itself, the presented datum is the light, though the experience is of the jewel. Here of course there is correspondence of experience with reality. But the test of truth is not correspondence of experience with reality either, but of presentation (pratibhāsa) with experience (adhyavasāya) and of presentation with reality.
And this correspondence is lacking in the case of the jewel’s light. The pragmatic utility and partial congruence of such experience, which have given rise to this misapprehension of its validity, are due to previous experience of the white conch, the memory-imposition of which makes this false experience possible. There were some thinkers, who held that discrepancy in respect of colour was immaterial, as the idea of contrary colour was an imposition of the imagination due to memory-association and as there was congruence in respect of the shape and configuration, these experiences should be allowed as valid. But this view is open to grave objection, as no shape or configuration is detachable from its colour and so these should be regarded as identical.7 Disagreement, therefore, in respect of colour is tantamount to disagreement of the entire presentation with reality.8
It has become perfectly clear that Dignāga’s definition of perception is complete and sufficient by itself. The addition of the adjective ‘abhrānta’ has no logical necessity or justification, as the sine qua non of valid experience is agreement with reality in all respects and as experiences of yellow conch-shell and the [Page 281] like do lack this all-round correspondence, they are excluded eo ipso from the category of valid perception. But the misapprehension prevailed in certain quarters and Dharmakīrti felt it imperative to clear this misconception. It is fully evident from the testimony of Dharmottara and of śāntarakṣita that the introduction of this objective ‘unerring’ (abhrānta) was not made by way of improvement, but was dictated by a practical necessity to rebut a prevailing misconception among a section of Buddhist philosophers, which, perhaps on account of its volume and strength, called for this amendment.
Kalpanā—What is its meaning?
[Page 282] Dharmakīrti defines kalpanā as a cognition, the content (pratibhāsa) of which is competent to be associated with verbal expressions.9 This association takes place when the content and the verbal expression are cognised in one sweep, so the two are felt to be one inseparable whole.10 The word ‘competent’ (yogya) is advisedly put in to include even the conceptual cognitions of children, who have not yet learnt the use of language, but whose knowledge has reached the state of judgment and so would have been actually associated with articulate words. Even the knowledge of the baby born on the very day is not free from ideation, as the baby, too, recognises the mother’s breast and ceases crying when its mouth is applied to it. This recognition presupposes an act of relationing a present sense-datum with a past experience and this recognition of identity has all the competence for verbal association, which is invariably the mode of relational thought in adult psychology.
The actual employment of words is, at best, symptomatic of conceptual thought and does not constitute its essential character. The criterion of conceptual thought is found in the indefinite, blurred presentation of the content (aniyatapratibhāsatvāt) and this indefiniteness is due to the absence of sense-datum, which alone is the cause of a definite invariable presentation. But as the objective datum in question is not present before the eyes and the conceptual [Page 283] thought arises independently of this objective reality, the presentation of the content lacks tbe distinct richness and vividness of direct perceptual cognition. Conceptual knowledge (vikalpa) has a past and a future reference and identifies the past and the present datum of experience and so is authentic being based upon and determined by a living fact. Conceptual thought or experience mixed with conceptual thought is independent of a live fact and so is unauthenticated and unreliable as evidence of objective reality.11 The unreliability of conceptual and relational thought will be made fully clear in a later section and for the present we propose to examine Dignāga’s definition of kalpanā. and to see whether it differs from Dharmakīrti s. definition or not.
Dignāga in his Nyāyamukha, a work on Buddhist logic,12 has on the other hand defined kalpanā as the association of [Page 284] class-character (jāti), quality (guṇa), of action (kriyā), of substance (dravya) and of name (saṃjñā). Critics have found in this definition of Dignāga an inexcusable flaw, inasmuch as class-character and the rest are all imaginary constructions and not objective existences and so cannot be associated with a real object, since association is possible only between two real substances like milk and water.
And even these realists, as for instance Kumārila, who believe in the reality of class-character and the rest, have got to admit that kalpanā is possible only through verbal association. It is therefore logically economic to hold kalpanā to be a verbal association actual or potential. Dignāga, therefore, lays himself open to the charge of looseness of expression or confusion of thought or perhaps both by resorting to this tortuous formulation. Śāntarakṣita has taken elaborate pains to save the master from this unenviable position by resorting to familiar scholastic devices, which the elasticity of Sanskrit idiom easily lends itself to. It will serve no useful, purpose to elucidate these textual manipulations and it will suffice to say that Śāntaraksita with all the aids of scholasticism in his armoury had to admit at last that verbal association alone is sufficient to characterise kalpanā and the association of class-character and the like has been mentioned only out of regard for others’ views which have found wide currency.13
Kalpanā—why should it be unreliable?
The next question arises as to why should verbal association be tabooed from the category of authentic knowledge? Verbal expressions are necessary for relational thought and unless relations are wholesale condemned as false appearance, use of words cannot be placed under an indiscriminate ban, as suggested [Page 285] by the Sautrāntikas. The Sautrāntika replies that relational thought, which, of necessity, is carried on by the use of words, cannot be a true measure of reality, since an entity is unique and unrelated (svalakṣaṇa), being entirely cut off from the rest of the world of similar as well as dissimilar things. What, however, is perceived in direct experience is this unique, self-characterised real, which has nothing in common with others. All reals are momentary point-instances, absolutely independent of each other and they only emerge into being under the inexorable law of pratītyasamutpāda (causality) and exercise a causal efficiency, which is peculiarly individualistic. Relations, therefore, are only ideal constructions (vikalpas) and have nothing corresponding to them in the objective world.
These constructions are purely subjective and independent of both sense-data and sense-organs. It cannot be urged that as this relational thought arises in the train of sense-object contact, it should be valid as much as non-conceptual. and nonrelational (nirvikalpa) cognition. Because, this sequence is purely accidental and is not contingent on sense-object contact, as relational thought is seen to arise even in the absence of such contact. And even in the event of sense-object contact there can be no relational thought, unless and until words expressive of the objects perceived are actually or implicitly associated with the latter. If sense-object contact bad competency for the generation of relational thought, it could not fail to do so even in the first instance. There is no reason why an intermediary, viz., the remembrance of word-relations, should be in request for the purpose. And even if the sense-object contact is seen to persist, the determinate, relational knowledge cannot be set down to its credit, as the act of remembrance, which is a non-sensuous and purely psychical fact, would detach the resultant experience from the objective reality.14
[Page 286] It has been urged that the recalling of the conventional relation of word and object is only an auxiliary factor, which merely reinforces the sense-faculty and, therefore, does not obstruct its operation. The relational knowledge is, therefore, purely a case of sense-perception as the sense object contact does not cease to function. If the sense-faculty was inoperative, we could have relational knowledge even if we did shut up our eyes just after the primal contact.
But as this is not the case, there is no logic in regarding conceptual experience as untrue.15 There are varieties of conceptual knowledge and all are not invalid. Conceptual knowledge, which arises as the result of sense-object contact should be regarded as the true perceptual experience and this is endorsed by popular behaviour.16 But this contention is illogical. The idea of assistant and assistable is only intelligible if there is any actual supplementation from reciprocal services; but this supplementation, even if conceded, gives rise to logical comlications, which are insurmountable.17
Moreover, a relational knowledge, as for instance, of the staff-bearing-man (dāṇḍī’ti vijñānam) is a complex, made up of varying factors and this cannot be the result of primal sense-object contact, but on the contrary presupposes a considerable number of psychical operations. Thus, in the case cited above the complex knowledge arises only after the adjectival factor (the staff), the substantive element (the person), the relation (e.g., the conjunction, saṃyoga) and the conventional mode of usage have been perceived severally and jointly and not otherwise.18 But it is too much to expect all [Page 287] this of the first indeterminate experience engendered by sense-object contact.
As regards the distinction of one class of conceptual knowledge from another class, viz., of imaginary constructions, which are independent of objective reality, from relational knowledge, which is supposed to be contingent on an objective sense-datum and is substantiated by verification, it must be observed that the distinction is not based on reality at all. All conceptual knowledge, which moves through the machinery of word-associations, is devoid of objective basis without exception. The logical value of such knowledge is, therefore, really nil.
The objective reference and relative vividness of the conceptual thought arising in the trail of sense-experience is due to the preceding non-conceptual cognition, which is generated by an objective reality. The verification and pragmatic satisfaction offered.by such knowledge is therefore mediate and derivative and cannot be claimed as a matter of right. The contention of Kumārila and Jayantabhaṭṭa—that verbal association is the condition of perception of class-character as much as the sense-faculty, light, attention and so forth are the conditions of perception of colour, etc., and so remembrance of verbal convention cannot be regarded as a barrier between sense-function and the object is an untenable sophistry. Well, the object is a single entity and, being amenable to perception, is cognised in its entirety by the first sense-perception. There cannot possibly be any part or aspect that may be left uncognised by the original experience.19 The assertion of the Naiyāyika that class-character [Page 288] and the like are also cognised by the first indeterminate perception is only a dogmatic statement, unsupported by experience and logic alike.
The proposition that the content of determinate knowledge is determined by that of indeterminate perception is an unwarranted supposition, which takes for granted that all our knowledge is derived from sense-data and the mind is only a passive register with no contributions of its own.20 But this supposition is contradicted by logic, as class-character and the like, which are thought to be cognised in perception, are fictions of the imagination. The relation of class-character with the individual object cannot be either one of identity or of difference. If the two are different and distinct, there is no reason why they should be found together and that for all time. Nor can they be identical, as they are possessed of contradictory characters. The class character or the genus is, therefore, a subjective idea and has no existence outside the subject’s consciousness.21
All conceptual knowledge refers to false, ideal constructions, having nothing whatever to do with reality. These ideal constructions are fivefold, to wit (1) genus, (2) quality, (3) action, (4) name and (5) substance. These are regarded as ideal constructions, as they proceed on the assumption of difference [Page 289] where there is identity and of identity, where there is difference. Thus,
(1) the genus or class-character (Jāti) is not anything distinct from the individual, but it is fancied to be distinct.
(2 and 3) The same is the case with quality and action, which are really non-distinct from the substratum, but they are imagined to be distinct and hence are called ‘false constructions.’
(4) Name and the individual, on the other hand, are actually distinct and different, one being a word and the other being a substantive object. But the two are regarded as identical, as is evidenced by such expressions as ‘he is Caitra,’ ‘Caitra’ being a mere name. The identification is so complete that a man invariably responds when his name is called out.
(5) The last variety is illustrated by such verbal usage as ‘He is a staff-bearer’ (daṇḍy ayam). Here the staff and the man are distinct as poles apart, but there is identification of the two.
Jayantabhaṭṭa, however, takes exception to these forms of understanding being regarded as false constructions on the ground that the relation of identity or of difference actually obtains between the objective facts concerned. But the distinction of class-character and the like has been proved to be false by the dialectic of relations. Now, as regards his contention that name and the individual are never confounded as identical, as the usage is not of the form ‘he is the name Caitra’ but ‘he is Caitra,’ it must be said that this is only a cavil and blinks the fact that ‘Caitra’ is a name none the less. His next objection is that there is no perversion in the usage ‘He is a staff bearer,’ as staff-bearer means the man and not the staff, which is cognised to be distinct as it actually is.
But this, too, has no force against the Buddhist, as the Buddhist does not admit any relation outside of the terms. There is no point in the argument that the relation of the staff is predicated and not the staff itself, as the relation and the relatum are not different. Moreover, language does not always conform to the experience of man and to make a contention on the basis of linguistic usage is not psychologically correct. Thus, for instance, it is an allowable [Page 290] expression to say ‘The boy is fire’ (agnir māṇavakaḥ). But our perceptual experience does not conform to the import of language. We do not perceive the boy to be identical with fire, though metaphor gives out such a meaning.22 There can be no gainsaying that all conceptual knowledge, which proceeds through the machinery of verbal expressions, gives false appearance and not truth.
But it may be asked that if these are only constructions of the imagination and perversion of relations and are all false experiences, then why should not there be any occasion of their being invalidated by a true experience just like the false experience of silver in the mother-of-pearl? The answer is that conceptual constructions, though false, are not on a level with errors and illusions. An illusion arises when one entity is perceived to be another, as in the case of the sun’s rays being perceived as water. But class-character and the like are not distinct entities from the individual and there is no chance of these being cognised as distinct entities.
Tie false conception of identity or difference centres round the individual entity itself and does not refer to a distinct entity. Hence, no experience sublative of the relational knowledge can possibly arise, as such experience can arise if there is confusion of one thing with another. These conceptual constructions are placed in a different category from the categories of truth and error. They cannot be authentic, as class-characters etc, are not objective realities; nor can they be levelled with errors, as there is no sublative experience possible.23
[Page 291] To sum up: it has been proved beyond the shadow of a doubt, we believe, that determinate perception, which invariably arises in the form of judgment, being essentially relational, only gives us false appearance. Conceptual constructions, e.g., class-character and the like are, at best, “working errors” and their pragmatic value is only a meretricious show. Reality is revealed only in the primal simple experience and the truth of such experience is attested by verification of the presentation with reality and pragmatic satisfaction is only symptomatic of such truth.
Is all knowledge determinate and conceptual?
Bhartṛhari, the grammarian, poet and philosopher in one and the author of the Vākyapadīya, a work on the philosophy of grammar and a product of wonderful learning and extraordinary genius, has propounded the theory that the whole universe has been evolved out of ‘Word,’ which is the eternal, imperishable Brahman.24 It is for this reason that all knowledge is interpenetrated with words, and a cognition, which is free from word-association, is an impossibility.25 The contention of the Buddhists that the simple, non-conceptual cognition, free from verbal association, is the only true knowledge, therefore, has no legs to stand upon.
Knowledge and word are co-extensive and one without the other is an idle abstraction, which is logically and psychologically absurd. This theory of the grammarians has been vehemently opposed by the philosophers of other schools; but with the metaphysical side of this doctrine we are not concerned in the present context. We shall only review it as a theory of knowledge and see how far its claims can be psychologically maintained.
Harivṛṣabha, the commentator, observes that unless words are actually preseut in the perceptual cognition, an object cannot be distinctly known and so there would be no memory, as memory comprehends only the thing that was perceived before.26 Bhartṛhari is emphatic that ‘word’ is the life and light of consciousness and consciousness, minus word, is comparable to light without its illumination and as ‘word’ refers to something beyond its own self and is, thus, by its very constitution relational,27 all knowledge is therefore a fortiori relational.
Jayantabhaṭṭa remarks that this doctrine embodies height of unreason; how can there be a cognition of word in ocular perception? There may be cases of perception, where the conventional relation of word and the object has not been cognised before; and even if previously cognised, the relation might be forgotten, or the memory-impression might remain in the subconscious level for want of stimulus. How can there be an impression of word-association in such cases?28 Harivṛṣabha, however, contends that even the primal cognition is not free from word-association, though the verbal expression may be of a very general kind. Thus, though the particular verbal expression may not be known, the perceived object will at least be referred to in its most generic verbal character, e.g., it is a substance or so. But the full individuality cannot be revealed unless the specific [Page 293] word-element presents itself.29
This dispute about the very matter of experience reminds us of the pregnant remarks of Jayantabhaṭṭa in another context.
“It is strange that these divergent views should be entertained with regard to the very object of perception. A dispute regarding an unperceived object is set at rest by perceptual evidence; but what can decide a controversy in the matter of perception itself? In a dispute about the matter of perception, a man can seek to convince his opponent only by swearing by his own experience.”30
But Śāntarakṣita does not think that the position is so desperate and seeks to clinch the issue on logical grounds. He argues that the vivid perception of an object presented through a sense-organ, when the mind is occupied in the determinate perception of another object, is clearly a case of simple, non-conceptual experience, free from verbal association. Here obviously a simple cognition synchronises with a determinate, conceptual experience and the former is indisputably an instance of non-relational knowledge. It cannot be urged that there is one determinate experience, as in that case there would be two distinct verbal references, or. the previous verbal association would be surrendered in deference to the newcomer.
But as this is neither one nor the other, since two verbal references are impossible to abide in one cognition, we have to conclude infidelity to experience that the determinate knowledge is synchronous with a simple indeterminate cognition. The contention of the Naiyāyika—that two distinct cognitions cannot synchronise in one perceiving mind and that the idea of synchronism is due to quickness of succession and so is essentially illusory—is opposed to experience. It is a matter of experience how various cognitions do appear simultaneously in the mind of a person witnessing a dancing performance. He sees the movement of the eye-brow of the dancing girl, hears the music, tastes the flavour of spices, smells the fragrance of flowers, feels the cool breeze of the fan waved overhead and contemplates the presents he will make and all these at one and the same time. And does not a man perceive the cooling sensation, fragrance and savour of a delicacy simultaneously when he eats it?
Moreover, quickness of succession cannot be a cause of this illusion of simultaneous perception. If rapidity of career could be an obstacle to perception of real succession, we could feel no succession to the movement of thoughts and feelings, which only last for a moment Likewise there would be no distinction of such words as ‘ rasaḥ’ (taste) and ‘ saraḥ’ (lake), as sounds are momentary and the order of syllables is one of unbroken succession. The example of the whirling fire-brand producing the illusion of a circle of fire is not apposite either. The illusion is not due to the quick succession of the flames, which as though perceived in succession are mistaken to be grasped in one sweep, as the Naiyāyika would make us believe.
The fact ©f the matter is that it is not a case of many cognitions being lumped together by memory; it is one cognition by one sense-organ. If it had been a confusion of memory and perception, the presentation of the circle would have been faint and blurred, as memory only cognises past objects and the representation of past objects by memory would lack the rich colourful vividness of sense-perception. The fire-circle in the fire-brand is not a mental illusion, as the Naiyāyika would make out; on the contrary it is a case of perceptual illusion devoid of order and sequence. So synchronism of manifold cognitions being established in perceptual experience, the simple cognition of an object in conjunction with a determinate experience [Page 295] cannot be disputed. Moreover, the appeal to experience is not the only resource of the Buddhist, but there is strong logical proof in favour of indeterminate experience being possible.
A determinate experience is always a relational knowledge, in which the individual is related to an ‘universal.’ But as ‘universals’ (jātis) are pure fictions of the imagination, they cannot be supposed to enter as constituents of the presentative data. They are absolute nonentities and a nonentity cannot be envisaged by perception, which takes stock of a really existent fact only. Again, relational knowledge is possible if there is a previous knowledge of at least one of the relata and this previous knowledge must be non-relational, otherwise regressus ad infinitum will become unavoidable. Besides, the whole contention of the grammarian is pivoted on a misapprehension.
All reals are unique, momentary individuals, having nothing to do with any other real, preceding or following it. Such reals are from their nature repugnant to word-association, as the conventional relation of word and object is only a fictitious relation which cannot subsist between facts. But the object previously cognised cannot last a moment longer, much less till the time when the word-relation will be comprehended. So words only relate to ideal fictions and not real entities. And the primal sense-perception, which takes stock of reality as it is, cannot, therefore, be amenable to word-association unless it is degraded to the rank of an unreliable vagary. But this is absurd on the face of it, as it sounds the death-knell of all relational activities. Unfailing correspondence and pragmatic satisfaction are, as we have observed before, the test of true knowledge and when these two tests are applicable only to the first non-conceptual experience, the first experience is alone regarded as reliable evidence of reality.
Classification of perception
Dharmakīrti has divided perception into four categories, viz.,
and lastly supersensuous perception of Yogins.
[Page 296] The category of ‘sense-perception’ (indriya-vijñāna) stands for the entire class of perception of objective realities, which are presented to consciousness through the medium of sense-organs. The sense-organs being live in number, sense-perception can be classified under five beads, to wit,
ocular perception (cakṣurvijñānam);
auditory perception (śrotra-vijñānam);
olfactory perception (ghrāṇendriyajavijñānam);
tactual perception (kāyendriyaja-spraṣṭavyavijñānam);
and lastly gustatory perception (rāsanavijñānam).
The classification is based on the variety of the media or the channels of perception and does not in any sense invest the sense-organs with agentive powers. Their function only consists in creating a connecting link between the subjective consciousness and the objective reality lying outside. This function is exhausted when the object is presented to consciousness and does not continue thereafter. So the first presentation is alone authoritative with regard to the objective reality.
But Kumārila would contend that even subsequent cognitions are equally valid, as they only tend to set the first cognition on a ground of certitude; and this certitude being excluded from the category of errors should be regarded as valid knowledge like inference. Śāntarakṣita, however, observes that as the certifying knowledge, which arises in the trail of primal cognition, does not exclude any misapprehension, it cannot be put on the level of inference. Mere exclusion from the category of errors cannot be the ground of validity. Inference, too, is valid not because it is distinguished from error, but because it removes error and misapprehension, which were actually present. But in this case of determinate perception, the determinate knowledge does not remove any misapprehension, because no such misapprehension is felt to exist.31
Is auditory perception free from verbal association?
Perception has been defined to be a cognition, which immediately takes stock of reality in all its uniqueness and so is free [Page 297] from verbal associations. But auditory perception, which cognises word, cannot be free from word-association and if word-association is condemnation of perceptual knowledge, auditory perception must be eo ipso invalid. Moreover, ‘words’ stand in a different category from all other objects of perception. A word is not a self-contained and self-enclosed entity. It has a reference beyond its own self. It not only reveals the object it stands for, but also reveals itself. In this respect, word, light and consciousness stand in a category altogether distinct from the rest of knowables. These three have a double aspect and a double reference. They express themselves and express others.32
Bhartṛhari too has emphasised this dual character of ‘word’ and he has sought to bring home this peculiar trait possessed by it by the analogy of consciousness and light. A word reveals not only the meaning, the fact meant by it, but also its own identity, quite in the same way as knowledge reveals the object cognised by it and also its own self. In reality these two aspects or powers do but represent one identical reality and are not factually different. But still they are cognised as distinct functions or powers by reason of an inherent differentiating propensity just as light appears to have two functional powers or energies, to wit, its power of revealing itself, its self-luminosity and its power of revealing an other.
Thus, word is self-regarding and other-regarding like light and cognition, though these two functional traits are in reality one identical energy or the fact itself, the difference being only an appearance.33 By [Page 298] virtue of this double functional energy a word is distinguished from pure material objects like jar, etc., which possess the power of only being revealed (grāhyatva) and also from sense-organs, which are seen to possess and exercise only the energy of revealing (grāhakatva).34 A word, therefore, being possessed of a double facet, in other words being both an expression (vācaka) as well as the content of expression (vacya), all auditory perception must by its very nature be associated with verbal expression and so cannot be valid knowledge.
Śāntarakṣita observes that the difficulty exists only in the imagination of the opponent. Word, as an objective reality, is as unique and self-contained as other entities are The double aspect of a word, which consists in its being both an expression and the expressed content, does not belong to the unique, selfcharacterised, momentary word, which alone is real. The expressive power does not belong to a real word quâ real word. This relation obtains between two purely ideal fictions and has nothing to do with the real word.35 Dharmottara also observes that even if the self-identical, unique word-individual is assumed to be possessed of expressive power, there would still be no difficulty, as this twofold character Of a word is cognised only when it becomes the subject of conventional relation (sic, of word and object).36
Harivrṣabha too seems to endorse [Page 299] the view of Dharmottara by saying that a word reveals its own self when it expresses, by reason of its objective reference, its necessary factual incidence in the objective reality. The obvious implication of Harivṛṣabha is that the self-regarding character of a word becomes manifest only when its meaning-reference, its factual incidence, is in evidence and not otherwise.37 But this relation of word and meaning is not understood in the primal auditory perception and only becomes manifest in the determinate knowledge that follows in its wake. Auditory perception, therefore, has no reason to become a bugbear.
avisaṃvādakaṃ jñānaṃ samyagjñānam...: ata eva cā’ rthādhigatir eva pramāṇaphalam adhigate cā’ rthe pravarttitaḥ puruṣaḥ prāpitaś cā’ rthaḥ. tathā ca saty’ārtbādhigamāt samāptaḥ pramāṇavyāpāraḥ, ata evā’nadhigataviṣayaṃ. yenai’va hi jñānena prathamam adhigato’rthas tenai’va pravartitaḥ puruṣaḥ prāpitaś cā’rthaḥ. tatrai vā’rthe kim anyena jñānenā’dhikaṃ kāryam. tato’ dhigatav ṣayam apramāṇam.
N. B. T, p. 3.
The meaning of pramāṇa and pramā will be made clear later on.
“atha svarūpato na vyapadeśyam ity eṣa kalpanāpoḍhaśabdārthaḥ? sarve’rthāṣ tarhi pratyakṣāḥ prāpnuvanti.
N. V., p. 42.
Cf. na hi yathā samyagjñānam adhikṛtya pratyakṣādilakṣaṇaṃ kṛtaṃ Kīrtinā na tathā Dignāgena, yenā ’dhikārāj jñāne vyavatistheta kalpanāpodha iti bhāvaḥ.
Tāt. Tī., p. 154.
kalpanāpratiṣedhāc ca jñānasva sāmarthyalabdhatvāt, avatsā dhenur ānīyatām iti yathā vatsapratiṣedhena godhenoḥ, ity ato jñānaṃ noktam.
T. S. P.. p. 367.
etac ca lakṣaṇadvayaṃ vipratipattinirākaraṇārtham, na tv anumānanivṛttyartham. yataḥ ka’panāpoḍhagrah (-?-) ṇenai’vā’numāṇam nivarttitam. tatrā’saty abhrāntagrahaṇe(?) gacchadvṛkṣadarśanādi(?) pratyakṣaṃ kalpanāpoḍhatvāt syāt. tato hi pravṛttena vṛksamātiam(?) avāpyata iti saṃvādakatvāt samyagjñānam...... tannivṛttyartham abhrāntagrahaṇam, tad dhi bhrāntatvān na pratyakṣam.
N.B.T., p. 9.
etac ca lakṣaṇadvayam ityādinā...... Vinītadevavyākhyā...... dūṣitā. tena tv evam vyākhyātam. “abhrāntam iti yad visaṃvādi na bhavati, evaṃ saty anumānasyā’py etal lakṣaṇaṃ prāpnotī’ti kalpanāpoḍhagrahaṇaṃ tannivṛttyarthamyady evam vyakhyāyate, ālambane yan na bhrāntaṃ tad abhrāntam ity ucyamāne sarvaṃ pratyakṣaṃ jṅānam ālambane bhrāntam iti na kasyacit pratyakṣatvaṃ syāt. tathā cā’ha, ‘sarvam ālambane bhrāntaṃ muktvā tathāgatajñānam’ iti Yogācāramate, tad apy atrā’cāryeṇa saṃgṛhītam” iti. tad ayuktam........................................ nanū’ktam Yogācāramatam asaṃgṛhītaṃ syād iti. ucyate. bāhyanayena Sautrāntikamatānusāreṇā’cāryeṇa lakṣaṇaṃ kṛtam ity adoṣaḥ.
N.B.T.T., pp. 18-19.
pītaśaṅkhādibuddhīnāṃ vibhraṃe’pi pramāṇatām |
arthakriyāvisaṃvādād apare sampracakṣate ||
tan nā’dhyavasitākārapratīrūpo na vidyate |
tatrā’py arthakriyāvātpir anyathā’tiprasajyate ||
T. S., śls 1324-25.
kecit tu svayūtthyā eva’bhrāntagrahaṇaṃ ne’cchanti. bhrāntasyā’pi pītaśaṅkhajñānasya pratyakṣatvāt............ pramāṇaṃ cā’visaṃvāditvāt. ata evā’cārya Dignāgena lakṣaṇe na kṛtam abhrāntagrahaṇam.
T. S. P., thereunder.
This distinction of colour and form and the premium put upon the latter remind us of Locke’s familiar distinction of Primary and Secondary qualities.
T. S., śls. 1325, 1327.
N. B. Ch., 1.
abhilāpena saṃsarga ekasmin jñāne’bbidheyākārasyā’bhidhānākāreṇa saha grāhyākāratayā śīlanam.
N. B. T., p. 10.
kathaṃ punar etad vikalpo’rāthān no’tpadyata iti? arthasannidhinirapekṣatvāt. bālo’pi hi jāvad dṛśyamānaṃ stanaṃ sa evā’yam iti pūrvadṛṣṭatvena na pratyavamṛṣaiti tāvan no’paratarudito mukham āropayati stane. pūrvadṛṣṭaparadṛṣṭaṃ cā’rtham ekīkurvad vijñānam asannihitaviṣayaṃ, pūrvadṛṣṭasyā’sannihitatvāt, asannihitaṃ cā’rthanirapekṣam, anapekṣaṃ ca pratibhāsaniyamahetor abhāvād aniyatapratibhāsam, tādṛśaṃ cā’bhilāpasaṃsargayogyam. indriyavijñānaṃ tu sannihitamātragrāhitvāt arthasāpekṣam. arthasya ca pratibhāsaniyamahetutvān niyatapratibhāsam.
N. B. T.,p. 11.
The Nyāyamukha is a work on logic composed by Dignāga. It has been referred to by Śīntarakṣita and a passage has been quoted from it by Kamalaśīla expressly and other passages seem also to have been quoted though the name of the author or of the book is not mentioned. The Nyāyamukha is lost in Sanskrit but is preserved in a Chinese translation. It is really a matter, of gratification that Prof. G. Tucci, Ph.D, of the University of Rome, has translated the Chinese version into English. The whole world of Buddhist scholars will be grateful to the learned Professor for having made this important work of Dignāga accessible in one of the most widely known modern languages of Europe.
Vide T. S: and the Pañjikā, śls, 1224, 1228-1237.
Vide T. S., śls. 1219-38.
parāparaprasiddhe’yam kalpanā dvividhā matā |
Ibid, śl. 1221.
satyaṃ lokānuvṛttye’dam uktaṃ nyāyavide’dṛśam |
iyān eva hi śabde’smin vyavahārapathaṃ gataḥ ||
Ibid, śl. 1228.
arthopayoge’pi punaḥ smārtaṃ śabdānuyojanam |
akṣadhīr yady apekṣeta so’rtho vyavahito bhavet ||
N. M., 92.
yaditvā(?)’locya saṃmīlaya netre kaścid vikalpayet |
na syāt pratyakṣatā tasya(?) sambandhānanusārataḥ |
S. V., p., 174, śl. 128.
evaṃ samāne’pi vikalpamārge | yatrā’kṣasambandhaphalānusāraḥ |
pratyakṣatā tasya, tathā ca loke vinā’py ado lakṣaṇataḥ prasiddham(?) ||
S. V., p., 207, śl. 254.
See ante, Ch. I.
viśeṣaṇaṃ viśeṣyaṃ ca sambandhaṃ laukikīṃ sthitim |
gṛhītvā sakalaṃ cai’tat tathā pratyeti nānyathā ||
N. M. p., 93.
The laukikī sthiti (the conventional mode of usage) is also a determining factor, by virtue of which ‘the staff’ is cognised as an adjectival adjunct to tbe ‘man,’ though the relation is the same and the order could be reversed but for this.
ekasyā’rthasvabhāvasya pratyakṣaṣya sataḥ svayam |
ko’nyo na dṛṣṭo bhāgaḥ syād yaḥ pramāṇaiḥ parīkṣyate ||
yat tu keṣāñcid vikalpānām idantāgrāhitvaspasṭatvādi rūpam tad arthāvinābhāv nirvikalpakadarśanapṛṣṭhabhāvitvāvāptatacchāyasaṃsārgajanitam. na tu teṣām arthasparśaḥ kaścid asti, arthātmano niyatātmano nirvikalpenai’va mudritatvāt.
N. M., P. 93.
nirvikalpānusāreṇa savikalpakasambhavāt |
grāhyaṃ tadānuguṇyena nirvakalpasaya manmahe ||
N.M., p. 98.
yad eva savikalpena tad evā’nena gṛhyate |
iha śabdānusandhānamātram abhyadhikaṃ param |
viṣaye na tu bhedo’sti savikalpāvikalpayoḥ ||
Ibid, p. 99.
tattvānyatvobhayātmānaḥ santi jātyādayo na ca |
yad vikalpakavijñānaṃ pratyakṣatvaṃ prayāsyati ||
T. S., śl. 1304.
(also,) vyaktayo’nā nuyanty anyad anuyāyi na bhāṣate |
jñānād avyatiriktañ ca katham arthāntaraṃ vrajet ||
Quoted in the Pañj., op. cit.
Vācaspati Miśra in his Bhāmatī proves that there is no distinction between quality and substance and this is attested by experience which takes them to be identical. It cannot be said that quality here stands for the substance qualified and so there is cognition of identity. Because, our experience does not obey the dictate of linguistic usage.
Cf. na ca śuklapadasya guṇaviśiṣṭaguṇiparatvāt evaṃ prathe’ti sāmpratam, na hi śabdavṛttyanusāri pratyakṣam. na hy agnir māṇayaka ity upacaritāgnibhāvo māṇavakaḥ pratyakṣeṇa dahanātmanā prathate.
Bhā. under Br. Sū., II. 2. 17.
Vide N. M., p. 94.
anādinidhanaṃ brahma śabdatattvaṃ yad akṣaram |
vivartate’rthabhāvena prakriyā jagato yataḥ ||
Vā. Pad., Ch. I. I.
na s o’sti pratyayo loke yaḥ śabdānugamād ṛte |
anuviddham iva jñānaṃ sarvaṃ śabdena bhāsate |
ibid, 1. 124.
śabdaśaktyanupātinā jñāneno’pagṛhayamāno vastvātmā vyaktarūpapratyavabhāso jāyata ity abhidhīyate, tādṛśa eva ca smṛtiviṣayo bhavati, anubhavasamānākāratvāt tasyāḥ.
Com. on 1, 124. Vā. Pad.
vāgrupatā ced utkrāmed avabodhasya śāśvatī |
na prakāśaḥ prakāśeta sā hi pratyavamarśinī ||
op. cit., 1, 125.
kathaṃ ca cākṣuṣe jñāne vāktattvam avabbāsate |
agṛhīte tu sambandhe gṛhīte vā’pi vismṛte |
aprabuddhe’pi saṃskāre vācakāvagatiḥ kutaḥ ||
N. M., p. 99.
yo’pi prathamanipātī bāhyeṣv artheṣu prakāśo viśeṣanimittāparigrahe’pi vastumātram idaṃ tad iti pratyavabhāsayati vāgrupatāyañ ca satyāmutpanno’pi prakāśo viśeṣavāgrūpam asvīkurvan prakāśakriyāsādhanatāyāṃ na vyavatiṣṭhate.
op. cit., Com. under I, 1, 126.
pratyakṣaviṣaye’py etāś citraṃ vipratipattayaḥ |
parokṣārthe hi vimatiḥ pratyakṣeṇo’paśāmyati |
pratyakṣe hi samutpannā vimatiḥ kena śāmyati ||
idaṃ bhāti na bhātī’ti saṃvidvipratipattiṣu |
parapratyāyane puṃsāṃ śaraṇaṃ śapathoktayaḥ ||
N. M., p. 9 8.
Vide T.S., śls. 1299-1302.
śabdasvalakṣaṇaṃ kiñcid vācyaṃ kiñcid vācakam.
N. B., p. 11.
also, cf. nanu jñānaśabdadīpās trayaḥ prakāśāḥ svaparaprakāśā ity āhuḥ.
N. M., p. 542.
ātmarūpaṃ yathā jñāne jñeyarupaṃ ca dṛśyate |
artharūpaṃ tathā śabde svarūpaṃ ca prakāśate ||
Vā. P., 1.50.
also, grāhyatvaṃ grāhakatvaṃ ca dve śaktī tejaso yathā |
tathai’va sarvaśabdānām ete pṛthag iva sthite ||
yady api ghaṭādayo grāhyā eva cakṣurādīni grāhakāṇy eva, tathāpi tejo yathā upalabdhau viṣayibhāvam āpannam eva viṣayopalabdhau karaṇatvam pratipadyate, tathā śabdo’pi. te cā’sya pratipādyapratipādakatvaśaktī nityam ātmabhūte pṛthag iva pratyavabhāsete.
Com. ad 1.55, ibid.
nāmā’pi vācakaṃ nai’va yac chabdasya svalakṣaṇam |
svalakṣaṇasya vācyatvavācakatve hi dūṣite ||
adhyāropītam evā’to vācyavācakam iśyate |
anāropitam arthaṃ ca pratyakṣaṃ pratipadyate ||
T.S., p. 542.
śrotrajñānaṃ tarhi śabdasvalakṣaṇagrāhi, śabdasvalakṣaṇaṃ kiñcid vācyaṃ kiñcid vācakam ity abhilāpasaṃsargayogyapratibhāsaṃ syāt, tathā ca savikalpakaṃ syāt. nai’ṣa doṣaḥ. saty api svalakṣaṇasya vācyayācakabhāve saṅketakāladṛṣṭatvena gṛhyamānaṃ svalakṣaṇaṃ vācyaṃ vācakaṃ ca gṛhītaṃ syāt. na ca saṅketakālabhāvidarśanaviṣayatvaṃ vastunaḥ sampraty asti tataḥ pūrvakāladṛṣṭatvaṃ apaśyac chrotravijñānaṃ na vācyavācakabhāvagrāhi.
N.B.T., p. 11.
yathā jñāne jñeyam......... gṛhyate, jñānasvarūpaṃ ca svaprakāśatvāt, tadvac chabdo’py abhidheyatantras tadrūpopagrāhī svarūpam api pratyavabhāsayati.
Prakāśa ad 1.50, Vā. P.