The Nyāya and the Buddhist Logic on Perception: Revisiting the Controversy

by Debalina Ghosh

The perceptual process plays a foundational role in giving us cognition. Sense-perception is invariably taken as the principal among all the evidential means, for, all the means are, in some way or other, preceded by sensory perception, and must be, in the long run, authenticated by some sensory perceptual base. Of course, in accordance with the difference in world-views, different Indian philosophical systems do not consider that perceptible entities are of the same type. For example, the Nyāya system considers that we perceive physical objects

directly, and not through a veil of sense-impressions. This principle in combination with the thesis that the world is exactly as we know it in our normal perception and inference, yields a world-view that is physicalistic in the sense that the elements are physical items such as things and properties, parts and wholes. They choose the observable physical elements, consisting of the things and properties, and hence

their programme is to explain the phenomenal in terms of the physical. The Nyāya wishes to defend a common-sense version of realism, for it qualifies the ordinary physical things with visual, tactile, and other properties. The Buddhist, on the other hand, prefers the phenomenal object and argues that nothing beyond the phenomenal need be accepted, for we can explain everything in terms of

the phenomenal. The dharma doctrine of the Abhidharma can be seen as an attempt to carry on this programme of explanation. In this view, actual sensory perception is the non-conceptual and indubitable cognitive experience. The choice of one rather than the other type of elements as basic reflects a difference in the philosophical motivations of the Nyāya and the Buddhist.

Perception in Nyāya

Perception is the primary (jyestha/ parā/ upajīvya) causal way of knowing (pramāṇa) since other three pramāṇas, inference (anumāna), analogy (upamāna), and verbal testimony (śabda) are somehow dependent (upajīvaka) on it. One of the most extensive as well as controversial definition of perception in Indian epistemology is provided by Maharṣi Gautam in his Nyāyasŭtra 1/1/4: Indriya-artha-sannikarṣa-utpannaṃ jñānam avyapadeśyam avyabhicāri vyavasāyātmakaṃ pratyakṣam. Perception is a cognition generated from sense organ-object connection and is not caused by words, is unerring, and well-ascertained. The connection between sense organ and the object (indriya-artha-sannikarṣa) is regarded by older naiyāyikas as the chief instrumental cause of veridical perceptual cognition. Definition is in accordance with it. In case of perception the relation of the sense organ with the object characterised by the property which

figures as the qualifier in the perceptual cognition is said to be the virtue (guṇa) for producing veridicality of perceptual cognition. For example, when one perceives a white shell as white, our sense organ stands in appropriate relation with the object of perception, a shell in this case, and apprehends the property whiteness which characterizes the shell in question and thus gives rise to a veridical perception. All five external sense-organs, viz., cakşu (visual sense organ), śrotra (auditory), ghrāṇa (olfactory), rasana (gustatory) and tvak (tactile) and one internal sense-organ ‘manas’ are imperceptible. Located in different parts of the body imperceptible sense-organs produce perceptual cognitions. Objects (artha) of perceptual cognitions are grasped by senseorgans and for this objects must have capabilities or appropriateness (yogyatā) of being perceived. Sense-organs are made of substances and Gautama takes special qualities of respective substances by which sense-organs are made to be the object of perception or the object of sense-organs. In a wider sense, all the six positive categories accepted by

Nyāya-Vaiśeśika and even the negative one can be considered as the object of perception. As the mark of Nyāya realism, the object of perception must be an actually existing entity which is different as well as independent from the perceptual cognition itself. By the term ‘unerring’ or ‘non-deviating’ (avyabhicāri) truth or veridicality of perceptual cognition is ensured by distinguishing it from perceptual error. Perceptual errors (vyabhicāri or bhram pratyaḳṣa) are also originated from sense-object relations.

Perceptual awareness grasps reality by being non-promiscuous or 'non-deviating' ensuring physical soundness of perception, in the sense that it would not be destroyed or falsified by 'the way the world is’. If it is considered that error occurs only at the level of qualifiers (prakāra), and not at the level of qualificand (viśeṣya or dharmī) of a cognition then the term ‘avyabhicāri’ should be interpreted as ‘tadvatitatprakārakatva’ (the qualificand being qualified by the same qualifier in the same way as it is in the actual world). This version of what may be called the correspondence theory of truth was considerably revised and improved upon by the Naiyāyikas. Some sort of correspondence or 'faithfulness' of the awareness with the world outside is the

bedrock for distinguishing perception from misperception. Perceptual cognition is non-verbal (avyāpadeśya), though it may be verbalizable. Even though each entity has its own signifying word or name (vācaka), at the time of perception of an entity, the word is not the content of that perceptual awareness. And even if the perceptual awareness of an entity is expressible or verbalizable through words, words are not among the causal apparatus of the perceptual awareness. Perceptual cognition is well-ascertained (vyavasāyātmaka or niścya). It is not a doubtful cognition. Doubt is the apprehension of contradictory properties, viz. presence and absence with regard to the same substantive. It is an indecisive or uncertain cognition where the mind oscillates between two contradictory properties. But in a certain or wellascertained cognition (niśc̣ya) negation of a property does not act as a qualifier (tatabhāva-aprakāraka). For, navya-Naiyāyikas both the perceptual error and perceptual veridical cognition are certain. There is some disagreement among the Naiyāyika

commentators about the interpretations of the terms ‘non-verbal’ and ‘well-ascertained’. Vācaspati, Udayana, Viswanāth do not consider that these two terms form the part of the definition; rather they introduce two types of perception — indeterminate (nirvikalpaka) and determinate (savikalpaka) respectively. Vācaspati, for example, contends that by the term non-verbal, Gautama refutes the śābdika view that since cognitions are necessarily informed by language, words act as qualifiers in a perceptual cognition also and hence there is no unqualified and non-conceptual nirvikalpaka perception. But Gautama accepts

that type of perception. Again, the adjective well-ascertained need not be used to exclude doubt, as doubtful perception is also deviant from the actual object of perception, it is also a perception of what the object is actually not (viśaya vyabhicāri). So it is already excluded by the adjective non-deviating (avyabhicāri). Rather, the terms vyavasāyātmaka, vikalpātmaka, niścaya all stand for determinate perceptual judgment. He contends that by the term non-verbal, Gautama refutes the Grammarian view and includes non-conceptual perception and, by the term well-ascertained, he refutes the Buddhist view and declares conceptual or judgemental perceptions as veridical.

Perceptual cognition is immediate. A general definition of perception is provided by navya-naiyāyikas to account for both eternal and originated perceptions on the basis of immediacy character of all perceptions: perception is a cognition which has no other cognition as its chief instrumental cause (jňānākaraṇakam jňānam pratyakṣam). Except perception all other cognitions are brought about by the instrumentality of any antecedent cognition.

Historical Overview of the Conflict between the school of Buddhist Logic and the Nyāya The Nyāyasūtra of Akşapāda Gautam and the Bhāśya of Vātsyāyana show the influence of Buddhist critics like Nagārjuna and refute some of their charges. Dinnāga then sets Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

himself to criticizing Brahmanic doctrines as those of Akşapāda and Vātsyāyana. To answer the objections of Dinnāga, Uddyotakara writes his Nyāyavārttika. Brahmanic criticism on Dinnāga similarly induced Dharmakīrti to write the pramānavāttika-kārīkā, a matrical commentry upon the pramāņa-samuccaya effecting all possible improvements in their own defence. Dharmakīrti was again answered by Vācaspati, the great Brahmanic philosopher and commentator. Dharmakīrti was succeeded by a number of Buddhist logicians like Devendrabodhi, Vinītadeva, Jinendrabodhi, Śāntirakşita, Dharmottara, Arcaţa and Jetāri, many of whom

wrote commentaries and sub-commentaries on the treaties of Dinnāga and Dharmakīrti and occasionally criticized Brahmanic writers like Kumārila and Vācaspati. But they did not possess much originality of thinking like the two masters: Dinnāga and Dharmakīrti. Owing to these mutual conflicts and opposition Indian Logic had the opportunity of developing by a process of alternate criticism and construction. The continuity of Buddhist logic1 came up to about 1000 A.D. when the decline and fall of Buddhism in India sounded its death knell. During this time with the revival of Brahmanism Brahmanic logic being tinctured with Buddhistic influence came to be studied over again and thus was laid the foundation of the new school of Brahmanic logic Navya-Nyāya which flourished later on so luxuriantly in Mithila and Nadia.

Nature and Definition of Pratyakşa according to Dinnāga and Dharmakīrti According to Buddhist logic, instantaneous momentary particulars (svalakşņa) are the object of perception. Svalakşņa is the unique momentary nature of things and devoid of name, form etc. Momentary particulars are neither extended in space nor stretched in duration of time — they are instantaneous. Momentary particulars are the immediate and prima facie object of sense-perception without being the object of any perceptive

judgment. Svalakşņa is the object of indeterminate perception. It is the clear and distinct manifestation of the object of perception. Any judgmental definite knowledge must present its object as an object of volition that induces the knower to act accordingly. This constitutes the practical utility of a piece of true cognition. Only a constructed synthesis of successive momentary particulars (santāna) following the first moment finally becomes the object of perceptive judgement that induces the volition of the knower. In this way, there are two types, i.e, levels of objects as the content of a true cognition, perception or inference: (a) immediate, actual and prima facie (b) distant and judgemental. Even if svalakşņas are momentary, when they become members of a series we can have judgmental knowledge of them. Judgmental knowledge is not the true manifestation of the object of perception. Only after the indeterminate perception of Svalakşņa, we can have judgmental or conceptual cognition like ‘I am perceiving a pot’. This is called vikalpa jñāna.

Indeterminate perception is the actual perception since svalakşņa becomes the object of it and svalakşņa is the ultimate objectual reality. Vikalpa jñāna, arises out of a complex object with all of its name, form etc. But Buddhist logicians assert that conceptual judgmental cognition does not causes violation. That is why an object is capable of being reached by perception, when it is cognized as something directly perceived. Inference for indirect cognition differs in that it points out the mark (hetu) of the object, and by this indirectly makes sure of the existence of object of possible purposive (prāpan-yogya). Thus sense perception points out a definite (niyata) object which appear before us directly, and inference also points out a definite object by way of the mark it is connected with. These two methods of cognizing point out definite object. Hence there is the valid cognition (pramā). Knowledge is valid when it makes us reach the object, and it makes us reach it when it has pointed to an attainable object. Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

Now defining the perception Dinnāga asserts perception is devoid of all types of conception or imagination like that of name and form etc.2 Kalpanāpoḍha means that which is not of the nature of conceptual cognition. Pratykşa (Perception) etymologically means related to sense organs. Buddhists also accept this but ‘devoid of kalpanā’ is included to mark the Buddhist insight. Here in explaining the term kalpanā Dinnāga asserts that the addition of name and form etc. into the object is called kalpanā 3. In case kind words like cow form or genus is added; while in case of action-words like cook etc. cooking action is added. This type of addition of name, genus, action to the object is kalpanā. On the other hand the term ‘devoid of’ (apoḍha) intends to mean absolute absence. So kalpanāpoḍha means the knowledge where there is the absolute absence of any connection of name, genus, action etc. This type of bare perception is the actual perception.

Dharmakīrti adds another qualification to the definition of perception given by Dinnāga. It is ‘abhrānta’ (not illusion). Perception is a valid source of knowledge only under the condition that the knowledge produced by it does not represent an illusion of the senses. So the definition of perception by Dharmakīrti is ‘Tatra ptatyakşam kalpanāpoḍhambhrantam’ (Nyāyabindu). Dharmakīrti’s analysis of the term kalpanāpḍham is not the same as that of Dinnāga. Both qualifications in the definition, that is, ‘kalpanāpoḍham’ and ‘abhrāntam’ are taken in a negative way. According to the Buddhist the definition of what a thing really is can never be actually given. It is not possible to define the actual nature of a thing because if one knows the nature of a thing then the definition would be superfluous and if one does not know the thing then it is equally impossible to give the definition of that thing. In this way, positively defining the nature of a thing leads to a dilemma. That is why it is convenient to give the definition in a negative way. The cognition which is devoid of kalpanā and also devoid of illusion is called pratyakşa.

The term ‘kalpanāpoḍha’ means ‘devoid of kalpanā-nature’. Dharmakīrti defines kalpanā as ‘Abhilāpasaṃsargayogya pratibhāsā pratītiḥ kalpanā’ ( Nyāyabindu). Kalpanā is the cognition which is of the nature of a mental reflex capable of relating with a verbal designation. When a certain cognition is designated by a word, the form of the referent is then related to the form of the word. This is called abhilāpa saṃsarga, and the knowledge which is capable of abhilāpa saṃsarga indicates the general form of the object. A cognitive which can be verbalized or named is called kalpanā and verbal designation includes name, genus, quality, action and substance.4 This type of knowledge is always determinate but the knowledge through senses can never be determinate or capable of relating with the verbal designation. Knowledge through senses is produced by the conformity of the form of the cognition to that of the object arthasārupya. And this object is never of the nature of words. That is to say in the sensuous perception it is never the case that appearance of the object calls for the appearance of the verbal designation. Hence perception is kalpanāpodha.

Dharmakīrti in his analysis of the term ‘abhrāntam’ asserts that the cognition which is free from kalpanā as well as illusion produced by many external and internal factors, is called perception. Illusion is caused by many defects like the defects of eye, distance, desire etc. Perception should be free from illusion caused by these factors. A perceiver with the defect of eye has an indeterminate cognition of double moon. But this cognition is never free illusion. Hence even if it is an indeterminate cognition, it is not pratyakşa pramā. The term ‘abhrānta’ should be added to exclude these types of illusory indeterminate perception from the scope of the definition of the perception. Dharmottara further clarifies the meaning of the term ‘abhrānta’ by asserting that it is not opposite to the practical capability of an object. The general form of an object is the adjective of the object. Cognition by conforming to the form of the object finds its practical utility. A pot is capable of carrying water because of a special arrangement of atoms. In this way a specific form or a special arrangement of atoms becomes the practical utility of objects. Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

A perceptual cognition shows this capability of practicality. This is actually the nature of an illusion free perception or a veridical perception.

Determinate Perception: Arguments and Counter-arguments

The main Buddhist argument against determinate knowledge being regarded as a senseperception is that the determinate perception cannot be held to be true perception because the image (pratibhāsa) presented by it can be associated with a verbal expression. A cognition produced by the sense-object-contact which grasps a pure object can never be associated with a verbal expression, for neither there are words present in the object, nor are the words identical with the object. If it were so, a person ignorant of the meaning of words would also be capable of communicating through words like a person who knows their meaning. Neither could the verbal expression, which is in no way related to the real object, be explained as the attribute of cognition, because a verbal expression refers to the object and not to the cognition. And, therefore, a cognition produced by an object should present only the object and not

the word denoting it. Never does a perception produced by colour present the cognition of colour along with that of taste. Therefore determinate knowledge, which is merely a mental construction, is erroneously held to be a true perception. Determinate knowledge presents an object as associated with a name (abhilāpa), although in reality the object is independent of any association with a name. Determinate knowledge is caused by impressions of the past determinate knowledge (vikalpavāsanā), and it grasps an uncertain (aniyata) object, and not a fixed one; it differs with every individual, because it is not a real object, and is caused by imagination. The function of imagination remains concealed in determinate knowledge, but because that knowledge follows in the wake of direct sense-experience (anubhava), it puts the function of the latter, i.e., sensation (darśana) in the forefront and, therefore, determinate knowledge is supposed to be true sense perception (anubhava). Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

It may be said that words are associated with the objects which they denote, and that when the objects are cognized, the words denoting them are remembered; then there arises the knowledge of the objects as associated with the words. The Buddhist replies that even if it be so, only the objects which are associated with the words could call those words to memory, and what is associated with the words is the generalized universal which is not grasped by the sense. It is the unique particular (svalaksana), the ultimate real and the cause of cognition, which is grasped by the sense, and not the universal which, being devoid of all efficient function, is unreal. Words, thus, do not denote what is grasped by the sense, and what they denote is not grasped by the sense. Moreover, if what is grasped by the sense could be denoted by words, i.e., the real object could be called forth by a word, then the fire expressed by a word should be felt hot like the real fire grasped by the sense. In that case, it should be capable of removing cold even when expressed by a word.

The Buddhist considers that when two objects are apprehended by one single cognition, one of them cannot be the attribute of the other; and that the particular and the universal which are grasped, according to the Nyāya-Vaiśesika, by a single cognition, cannot therefore stand in the relation of the qualifier and the qualified. The Buddhist continues, if attributes like universals, etc., are held to be real attributes of an object, then in the case of tree called (siṃsapā), all the universals existence (sattā), substance-ness (dravyaatva), the universal of earth (pŗthivītva) the universal of the tree (vŗkşatva), and the universal siṃsapātva will be real attributes of the same object. When therefore a tree is comprehended from a distance as qualified by one of the attributes (say existence), there will arise the comprehension of the tree as qualified by all the above-given attributes because its nature of having subsistence of one of the universals is the same as that of having the subsistence of the other universals.

Moreover, the relation of the container and the contained (ādhāra and ādheya) implies some kind of service; for instance, a badara fruit which is likely to fall, when placed in a basket is prevented from falling, and therefore the basket does service to the badara as its container. Similarly, in the other instance, the substance (i.e., the tree which is the container) must render some service to the universals which are contained in it. Nor can it be said that it renders service not by its very nature but by capacities which are different in the case of each universal and therefore all the universals are not comprehended simultaneously. In that case even for containing one such capacity, another capacity for containing that capacity will have to be assumed, and the process will go an ad infinitum. It will, therefore, have to be admitted that the very nature of the substance produced therein by its causes is such that the substance renders service to its various attributes by containing them in itself. When, therefore, a substance (tree) is being cognized as capable of rendering service to the universal ‘existence’ (sattā), it has the same nature with reference to all other universals, e.g., substanceness, etc., and

therefore, by the determination of the universal ‘existence’ which relates to a real substance, all the attributes dravyatva, pŗthivītva, vŗkşatva and śiṃśapātva, etc., conditioned by the nature of that object, will be comprehended, and there will be no need of separate determinations of other universals. It has been declared by Dharmakīrti that when the object, which is identical with the power to render service of containing numerous attributes has been grasped in entirety (sarvātmanā), which one of the universals will remain uncomprehended? When one universal is comprehended, all the universals will be comprehended. When an object (the substance) rendering the service of containing attributes has been comprehended, there remains no other service to be rendered apart from the object. The object, therefore, being cognized, all the attributes (universals) will be cognized. Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

Now so far as Buddhists are concerned, whatever is cognized or whatever is determined by the ideas (vikalpas) recalled by the beginningless impressions of past experiences, both these (what is cognized and what is determined) being in the form of exclusion of others, are unreal, and do not in the least comprehend the object which is the ultimate real (svalakşana). Indirectly, of course, the determinate ideas, having concomitance with the real objects because determinate perception follows in the wake of sense-perception, do lead people to the real objects, enable them to reach those objects, and thus do not delude them. Ideas (vikalpas) of numerous universals not being real, there is no question of their being repeated, because, according to the Buddhist, they are not separate realities, but mere ideas which are different from one another.

The Buddhist points out further that when an object has been grasped by a sensation, it cannot produce determinate perception, because the operation of the sense is interrupted by subsequent remembrance of the world. It has been said by Dharmakīrti that if perception requires, even when the object has been contacted, the association of a world recalled by memory, the object (in that case) would become interrupted.5 Further, it cannot be maintained that the self-same sensation, which was produced by the sense, brings about the determinate perception when helped by remembrance. The Nyāya-Vaiśesika points out that the remembrance, being part of the act of perception, does not interrupt determinate perception because one’s own limb offers no interruption. The Buddhist replies that the position of the Nyāya-Vaiśesika is untenable because, as Dharmakīrti has stated, the object, which did not produce the determinate perception in the first instance, i.e., when it produced the indeterminate perception, will not produce it even afterwards, because the contact of the object with the sense is the same in both cases. Moreover, the past can never come within the range of the sense, and the sense cannot function even by thousand-fold efforts with reference to an object which is outside its range. Neither can memory, which grasps the past objects, grasp the present object

which has not been experienced before. If memory could grasp past objects as present, there would arise the impossible situation that the blind would also perceive colour, and therefore it has been said by Dharmakīrti that in that case there would be a visual perception even when the visual sense has been lost.

Concluding his argument the Buddhist declares that it has thus been proved that knowledge which involves, as an attribute, a name, the universal, qualities and movements cannot be a sense-perception. Similar is the case with the knowledge which involves, as an attribute, a substance, as in the instance a man with a stick. It has been said by Dharmakīrti that the qualifier, the object to be qualified, the relation, all these factors are cognized and arranged by the intellect; and then alone there is comprehension, and not otherwise.

The entire series of operations which can be accomplished only by a thinking agent (intellect) cannot be sustained by the sense-perception which, being produced by the power of the present object, is not a thinking agent. The Buddhist, therefore concludes that determinate knowledge cannot be held to be a sense-perception.

The Nyāya replies to these objections center around establishing common-sense realism in which the ultimate reality of svalakşaņa has been denied and the ordinary material objects with attributes have been accepted as realities. The Nyāya considers that an individual substance and its universals are two different real entities both of which produce determinate perception.

Then the Buddhist raises another objection that the fact is that the past state of an object is not in contact with the sense; how can it then be cognized by the sense? The Nyāya poses a counter-question: does it then mean that all that is in contact with the sense is perceived? If it were so, ether and atoms, etc., which are in contact with the sense, would also be perceptible. The fact, therefore, is that whatever comes within the Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

range of the knowledge produced by the sense is perceptible, and not all that which is in contact with the sense.

It may, however be asked: if there be no contact with an object, and how can the sense produce a knowledge with reference to that object, and how can that knowledge be held to be perception? And if that kind of knowledge be held to be perception, how will the definition of perception, viz., that which is produced by sense-object-contact, cover all the instances of perception? In that case, the above quoted instance (the perception of the past state of an object) itself will not be covered by that definition. The Nyāya meets the objection thus: although the past state of an object may not be grasped by the sense, yet that state is comprehended by the perceptive knowledge produced by the sense with the aid of the memory or impressions. It cannot be said that the knowledge which has been produced by the sense-object contact with the help of memory, is not produced by sense-object-contact; and, thus, the definition of perception will cover all cases. 

The Buddhist still persists that if the same object which was grasped by the sense could also be denoted by a word, the fire expressed by a word should also remove cold. But the objection is not tenable because although both, the determinate perception and the word, refer to a real object, the knowledge in the two cases is not of the same kind. On account of the difference of the causes, there is a difference in the knowledge, i.e., mediate and immediate. Removal of cold is caused by the contact with real fire, present only at the time of the immediate knowledge, and not by its mere verbal knowledge. As to the objection that when one universal adjunct, viz., existence has been comprehended, other adjuncts like substanceness, etc. should also be comprehended because the nature of an object of possessing one universal is the same as that of possessing another universal, the Nyāya gives the following answer: when an object as qualified by one universal is comprehended, it does not denote its comprehension as qualified by other universals, because it is the nature of the (universals, etc.) or the fact of

its being qualified by them is not its nature. It is not that whatever is related to the nature of an object becomes itself its nature. If that were so, it would not be held to be related to it, because a thing cannot be related to itself. Answering to the Buddhist objection as to how a sense perception which is not a thinking agent can arrange all the factors, the qualifier and the qualified, etc. the Nyāya rejoins: is even a mental knowledge capable of arranging various ideas? If it be said that it does so because the manas which is its cause can comprehend all kinds of things, it may be asked, there being no permanent thinking agent according to the Buddhist, if manas is in the form of the previous cognition, how will it comprehend all kinds of objects? So far as the Nyāya is concerned, manas, although capable of grasping all objects, is not a thinking agent because it is unconscious (acetana). The soul alone, as the substratum of all cognitions and impressions produced by them, recollects and arranges ideas.

Now the Nyāya proceeds to answer the question as to how the two simultaneous ideas, of which one cannot render any service to the other, can be arranged in the relation the qualifier and the qualified, and why that relation is not cognized at the very first moment of indeterminate perception. The Nyāya asserts that, although between two objects, the individual as well as the universal, comprehended by one single cognition, there can be no relation of one of them serving the other by making the other known, yet the indeterminate sensation in which the universal and the individual are cognized without mutual relationship and remembrance of the past are the causes of the determinate knowledge which establishes the relation of the qualifier and the qualified between them. Therefore, indeterminate perception and remembrance render a service to determinate knowledge by producing it. The two objects, the universal and the individual, although standing in relation of ‘form’ and ‘possessor of form, (rupa-rupi-bhāvena), have not been comprehended in that relation by the indeterminate perception which occurs at

the very first moment; they are comprehended by it in their own nature as unrelated. It is not necessary that all that exists should be completely cognized. Therefore, although indeterminate perception has cognized the universal and the individual only partially, i.e., without their mutual relationship, it does not become invalid for that reason. The determinate perception, of course, produced by the causes referred to above, i.e., the indeterminate perception of the unrelated universal and individual, and the remembrance of the past, presents the universal, etc., as ‘form’ (rupataya) and the substance (individual) as ‘possessor of form’ (rupitaya). And although determinate perception arises at a later stage, it is has been produced by sense-object-contact.

The Impact of Buddhist Logic on the Nyāya theory of Perception The impact of the Buddhist onslaught was such that the thinker of the caliber of Vācaspati Miśra, had to change the original definition of perception in order to defend the Nyāya position. He declares in Nyāyavārtika-tātparyatīkā: Yad eva indriyajasya jñānasya gocara tat pratyakşam, na tu indriya-saṃbaddham, that is, whatever comes within the range of the knowledge produced by the sense is the object of perception, and not merely that which comes in contact with the sense. It means that there may be an object of perception which is not in contact with the sense, and yet it may have been comprehended by a perception produced by the sense. Vācaspati Miśra, himself says in the same context that even if the relation of an object with its past state is not grasped by the sense, yet that relationship is comprehended in the perceptual knowledge produced by the sense with the aid of memory or impressions. It appears that the Buddhist objection is met here only by giving up the original Nyāya-Vaiśesika position, i.e., that alone which is in contact with the sense is perceived, and by adopting a new theory that whatever comes within the range of knowledge produced by the sense is the object of perception.

Here we notice that a new factor is in the process of being introduced in the Nyāya-Vaiśesika epistemology. When perceptual knowledge is produced by senseobject-contact, it smuggles in something which is not in contact with the sense. For example when we see from a distance that a plate of shāhi biriyani is being cooked, we also virtually smell its fragrance and sometimes even have a virtual taste of it. How does it do so? The answer is that it is comprehended with the aid of memory or impressions. There are two alternatives that either (i) memory of the past experience or (ii) simply the impressions (saṃskāra) of that experience without awakening memory (smŗiti) render aid to the sense, and the sense thus aided produces a perceptual knowledge which, in addition to cognizing the object in its present state, cognizes it as related to a past state as well. The main emphasis laid by the Nyāya-Vaiśesika school, is that the perceptual knowledge of the object as related to both the states, present and past, is one unitary knowledge. As held by the Buddhist, it is not composed of two parts, Viz., (i) sense-perception of the present state, and (ii) memory of the past. But if we hold that memory (and not saṃskāra) renders aid to the sense, it will mean that there is, in the first place, the memory of the relation of the object with a past state, and that secondly, the memory helps the sense, but itself disappears; it does not itself become part of the perceptive knowledge. For, otherwise the perceptual cognition would not be a unitary cognition. It is thereof held that the sense, aided by the memory, produces a unitary perceptive knowledge which comprehended an object as related to its present as well as to its past state. Even then the fact the perceptual knowledge is preceded by memory suggests a sort of alien element in the unitary perception. The later Nyāya-Vaiśesika writers, therefore, gave up this alternative; and adopted the second alternative that the sense was helped by the impression (saṃskāra) and not by memory (smŗti). The question of impressions helping the sense is usually taken up in connection with the perceptual knowledge, called recognition (pratyabhijñā), which is held, by the Nyāya-Vaiśesika writers, to be a special kind of perception, and differentiated from ordinary determinate perception.

It is tenaciously held by the later Nyāya-Vaiśesika that a recognition is not composed of two kinds of knowledge, viz., sense-perception grasping the present state of an object, and memory referring to its past state; but it is one unitary knowledge in which the past state is also apprehended by sense perception. According to the later writers, apprehension of the past state by sense-perception is due to the aid of impressions. It is held that the sense, unaided by impressions, grasps an object only in its present state, but when aided by impressions, it has the power to comprehend the past state also. The function of an impression (saṃskāra), as originally conceived, is that of bringing about remembrance. Praśastapāda, of course, mentions recognition (pratyabhijñā) as well as memory, both being produced by impression. He says that the quality of the soul called impression (bhāvanā, i.e., saṃskāra) is the cause of remembrance and recognition of the objects seen, heard, or experienced before. It appears that Praśastapāda conceives recognition as a special kind of remembrance which occurs along with some kinds of perception. He does not seem to have conceived that recognition is a unitary perceptive knowledge in which impressions are capable of presenting a past state in a senseperception. This theory seems to have developed only in the ninth or the tenth century.

Jayanta and Śridhara both hold that the past state of an object is revealed in a sense-perception when the sense is aided by impressions. They do not accept the other alternative that the sense is aided by remembrance. In fact, Śridhara differentiates between an object called forth by memory (smŗtyopanita), which, according to him, is of the nature of thought, and the object called forth by an impression (saṃskāropanita) which is of the nature of sense-perception. Śridhara says with reference to the cognition of a word that at the time of cognition of the last letter, the previous letters, although destroyed, are brought forth by memory, and they become the efficient cause of the comprehension of the meaning. If the efficient causality of a past object is possible, the inherent causality of some (past) objects may also be possible, the inherent causality of some (past) objects may also be possible6. But an object can be an inherent cause only if

it were actually existent. The existence of a past object at the present time, contradictory as it may appear, is brought about through the power of impression. An impression is held to be the cause of a sense-perception also, as it is that of remembrance.

The Buddhist objects that the sense always grasps a present object, that impression has always reference to a past object, and vice versa, the sense can never grasp a past object and that impression can never refer to a present object. Śridhara replies that although sense and impression, working separately, are incapable of grasping past and present objects respectively, yet, when working in concert, they produce an effect in the form of recognition. The later Nyāya-Vaiśeşika writers, however, hold that apprehension of a past state occurs only in the case of recognition which is only a special kind of determinate perception, but not in every determinate perception. Both the points that a determinate perception has reference to a past state and to other objects of the same class, appeared in the later period of Nyāya literature in the form of transcendental relations called jñāna-lakşaņa and sāmānya-lakşaņa

respectively. These two forms of transcendental relations were ultimately crystallized by later Nyāya-Vaiśeşika writers in response to a prolonged controversy over the following two questions posed by the Buddhists: (i) how can the past state of an object, or for the matter of fact, how can the fragrance of sandal wood, which is not in contact with the sense, be cognized by senseperception? and (ii) how can the concomitance between the inferential mark and the thing to be inferred be established by observation of their association in a large number of cases? But the introduction of extraordinary transcendental elements in explaining ordinary epistemological phenomena certainly robs the Nyāya theory of the rigor a common-sense realism should have. The Buddhists logicians, in my opinion, succeed in forcing the Naiyāyikas to lose some of the glories of their theory. Both the Nyāya and Buddhist agree that an ontological system must be epistemologically grounded in the sense that what is epistemologically prior should be Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

the starting-point at ontological enquiry, but the difference lies in what they consider to be epistemologically prior, how they define perceptual knowledge. The Buddhist argues that the phenomenal (the appearance at particular, colour, shape, taste, smell, touch, and so on) by its nature comprises the entire content of our immediate perceptual experience, and hence it is epistemologically prior, while physical objects or propertied things are far removed from ‘raw’ experience. The Nyāya claims, on the other hand, that physical things or propertied things are more directly accessible to our perceptual awareness.

We saw that according to Nyāya there are two stages in a perceptual cognition: there is the conceptual stage, in which we are aware of the object as being a certain way; but this is always preceded by a non-conceptual stage in which we are aware of the constituents of a perceptual judgment individually, without being aware of them as forming a relational complex. On the Nyāya account all the elements of this judgment are out there in the world. None is constructed by the mind. But the mind must first be aware of them by themselves before it can be aware of the relational complex they make up. Buddhist logicians do agree with Nyāya that perceptual cognition involves two stages, the non-conceptual and the conceptual. They also agree that perceptual judgments involve attributing some general nature to an object. They do not accept universals, though, so they have to deny that perceptual judgments reflect what exists outside the mind. Only non-conceptual perception can do that. My conceptual cognition of Dhenu as a cow involves attributing cowness to the real particular out there. But while the world contains real particulars, it does not contain universals like cowness. So if perception puts us in touch with the world, then the conceptual stage of perceptual cognition is not perception. We may call it a ‘perceptual judgment’, for it is always preceded by what is properly called ‘perception’. But it is a judgment, something involving concepts, and that makes it a kind of inference. Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

Seeing Dhenu as a cow feels like something we just do when we see Dhenu. This is what Nyāya is getting at when it says we see universal like cowness. But if there are no universals out there where Dhenu is, there must be some process of mental construction involved in our seeing her as a cow — a process that we are not ordinarily conscious of. Dinnāga started with the idea that this mental process involves associating what we actually see — the unique particular Dhenu — with a word, in this case ‘cow’. His real insight comes when he says that knowing what word to call something involves an inference. To know that what I see is called a cow, I need to know to infer that there’s a fire on the hill: that anything that’s like this hill in having smoke also has fire, while nothing that lacks fire has smoke. So if seeing Dhenu as a cow is a matter of associating what I see with the word ‘cow’, then seeing Dhenu as a cow must involve a kind of inference. We are not aware of performing this inference because we do it so quickly and automatically. It’s something we learned to do when we learned to talk, and ever since then we’ve been doing it constantly. So it’s not surprising that we are unaware of doing it, just as we are no longer aware of all we are doing when we stand upright and walk.

According to Buddhist logicians, two things happen in rapid succession when you are on the hill and see the fire there. First you have a non-conceptual cognition in which you are visually aware of that unique particular located on the hill. This is what is properly called ‘perception’. You then very quickly perform a kind of unconscious inference whereby you judge that what you see is the kind of thing that is called a ‘fire’. This result in the conceptual cognition that may be expressed as ‘This is a fire’. We can call this cognition a ‘perceptual judgment’. But it is important to remember that for Dinnāga and his school it is not perception, for it is about fire-in-general, which is not really out there. This does not make your cognition erroneous. Now that you are aware of the fire as a fire, you can use your knowledge about fire-in-general. One of the things you know about fire is that it can warm you when you are cold. Since you are cold and want to warm up, your perceptual cognition of the fire helps you satisfy your desire. We said Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

above that a means of knowledge must be able to lead successful practice. Even though your judgment is about a mental construction, not something in the world, it still leads to successful practice. So it counts as a means of knowledge. It just turns out that it’s an inference. Perception is also a means of knowledge. But because it is non-conceptual, Buddhist logicians claim that it does not give rise directly to successful practice. You can only know that the particular you see will warm your hands after you have conceptualized it as fire. And that involves inference. So why is perception a means of knowledge? Because it leads you to construct fire-in-general, and that construction leads to successful practice. Both perception and inference (including perceptual judgment) are means of knowledge because both bring about veridical cognitions. A veridical cognition is one that is not falsified by bringing about unsuccessful practice. Suppose I wrongly took the smoke on the hill as a sign of an impending volcanic eruption. I might try to save my life by running away. But this would be pointless, since no volcanic eruption threatens my life. This is a case of a faulty inference. Its faultiness lies in its leading to unsuccessful practice. Perceptions don’t give rise actions; they only lead to perceptual judgments. So a perception can not be directly falsified by an action. It can only be falsified through its leading to unsuccessful practice. This unique feature of the Buddhist logicians’ analysis of a perceptual cognition is innovative as well as practical.

Notes and References 1 The school of Buddhist logicians is founded by Dinnāga. Sometimes the school is called “Yogācāra-Sautrāntika” or simply “the school of Dinnāga”. There is no one name that classical Indian Buddhists used for all the thinkers of this tradition. In this essay we are mainly concerned with the views of Dharmakīrti and other later Buddhist logicians. Philosophy Pathways – Issue 227 – 16th November 2018 https://philosophypathways.com/newsletter/

                                                                                                                                                   2 Pratyakşam Kalpanāpoḍham nāmjatyādyasaṃyutam. (Pramāna Samuccya,). Dinnāga, 1966, Pramānasamuccaya, , Bhabnagar, Jaina Ātmānanda Sabhā 

3 Atha kalpana ca kīdŗśī cedāha. nāmayātyādiyojonā (Pramāna Samuccya) 4Abilāpinī pratītiḥ kalpanā ( Pramānaviniścaya). Dharmakīrti, 1966, Pramāņaviniścaya, edited by T. Vetter, I. Vienna. Dharmakīrti, 2007, Nyāyabindu (with Dharmottara’s tikā), Translated into Bengali by Sanjit Kumar Sadukhan, Sadesh

5 Akśadhīryadyapekśeta sa artham vyavahito bhavet (Pramāņaviniścaya)

6 Yathedam tathā samavāyikāraņatvam api keşāñcid bhavişyati. (Nyāyakandali). Praśastapāda, 1977, Praśastapādabhāsya (with Nyāyakandali of Sridhara), Durgadhara Jha (ed.), Sampurnānanda Sanskrit Mahāvidyālaya.