A History of Indian Philosophy Volume 1

by Surendranath Dasgupta | 1922 | 212,082 words | ISBN-13: 9788120804081

This page describes the philosophy of perception (pratyaksha): a concept having historical value dating from ancient India. This is the sixteenth part in the series called the “the nyaya-vaisheshika philosophy”, originally composed by Surendranath Dasgupta in the early 20th century.

Part 16 - Perception (Pratyakṣa)

The naiyāyikas admitted only the five cognitive senses which they believed to be composed of one or other of the five elements. These senses could each come in contact with the special characteristic of that element of which they were composed. Thus the ear could perceive sound, because sound was the attribute of ākāśa, of which the auditory sense, the ear, was made up. The eye could send forth rays to receive the colour, etc., of things. Thus the cognitive senses can only manifest their specific objects by going over to them and thereby coming in contact with them. The conative senses (vāk,pāni,pāda,pāyu , and upastha)recognized in Sāṃkhya as separate senses are not recognized here as such for the functions of these so-called senses are discharged by the general motor functions of the body.

Perception is defined as that right knowledge generated by the contact of the senses with the object, devoid of doubt and error not associated with any other simultaneous sound cognition (such as the name of the object as heard from a person uttering it, just at the time when the object is seen) or name association, and determinate[1]. If when we see a cow, a man says here is a cow, the knowledge of the sound as associated with the percept cannot be counted as perception but as sound-knowledge (śabda-pramāfta). That right knowledge which is generated directly by the contact of the senses with the object is said to be the product of the perceptual process. Perception may be divided as indeterminate (nirvikalpa) and (savikalpa) determinate. Indeterminate perception is that in which the thing is taken at the very first moment of perception in which it appears without any association with name.

Determinate perception takes place after the indeterminate stage is just passed; it reveals things as being endowed with all characteristics and qualities and names just as we find in all our concrete experience. Indeterminate perception reveals the things with their characteristics and universals, but at this stage there being no association of name it is more or less indistinct. When once the names are connected with the percept it forms the determinate perception of a thing called savikalpa-pratyakṣa. If at the time of having the perception of a thing of which the name is not known to me anybody utters its name then the hearing of that should be regarded as a separate auditory name perception. Only that product is said to constitute nirvikalpa perception which results from the perceiving process of the contact of the senses with the object.

Of this nirvikalpa (indeterminate) perception it is held by the later naiyāyikas that we are not conscious of it directly, but yet it has to be admitted as a necessary first stage without which the determinate consciousness could not arise. The indeterminate perception is regarded as the first stage in the process of perception. At the second stage it joins the other conditions of perception in producing the determinate perception.

The contact of the sense with the object is regarded as being of six kinds:

  1. contact with the dravya (thing) called saṃyoga,
  2. contact with the guṇas (qualities) through the thing (saṃyukta-samavāya) in which they inhere in samavāya (inseparable) relation,
  3. contact with the guṇas (such as colour etc.) in the generic character as universals of those qualities,e.g. colourness (rūpatva), which inhere in the guṇas in the samavāya relation. This species of contact is called saṃyukta-samaveta-samavāya, for the eye is in contact with the thing, in the thing the colour is in samavāya relation, and in the specific colour there is the colour universal or the generic character of colour in samavāya relation.
  4. There is another kind of contact called samavāya by which sounds are said to be perceived by the ear. The auditory sense is ākāśa and the sound exists in ākāśa in the samavāya relation, and thus the auditory sense can perceive sound in a peculiar kind of contact called samaveta-samavāya.
  5. The generic character of sound as the universal of sound (śabdatva) is perceived by the kind of contact known as samaveta-samavāya.
  6. There is another kind of contact by which negation (abhāva) is perceived, namely saṃyukta viśeṣaṇa (as qualifying contact). This is so called because the eye perceives only the empty space which is qualified by the absence of an object and through it the negation.

Thus I see that there is no jug here on the ground. My eye in this case is in touch with the ground and the absence of the jug is only a kind of quality of the ground which is perceived along with the perception of the empty ground. It will thus be seen that Nyāya admits not only the substances and qualities but all kinds of relations as real and existing and as being directly apprehended by perception (so far as they are directly presented).

The most important thing about the Nyāya-Vaiśeṣika theory of perception is this that the whole process beginning from the contact of the sense with the object to the distinct and clear perception of the thing, sometimes involving the appreciation of its usefulness or harmfulness, is regarded as the process of perception and its result perception. The self, the mind, the senses and the objects are the main factors by the particular kinds of contact between which perceptual knowledge is produced. All knowledge is indeed arthaprakāśa , revelation of objects, and it is called perception when the sense factors are the instruments of its production and the knowledge produced is of the objects with which the senses are in contact. The contact of the senses with the objects is not in any sense metaphorical but actual. Not only in the case of touch and taste are the senses in contact with the objects, but in the cases of sight, hearing and smell as well. The senses according to Nyāya-Vaiśeṣika are material and wehave seen that the system does not admit of any other kind of transcendental (atīndriya) power (śakti) than that of actual vibratory movement which is within the purview of sense-cognition[2].

The production of knowledge is thus no transcendental occurrence, but is one which is similar to the effects produced by the conglomeration and movements of physical causes. When I perceive an orange, my visual or the tactual sense is in touch not only with its specific colour, or hardness, but also with the universals associated with them in a relation of inherence and also with the object itself of which the colour etc. are predicated. The result of this sense-contact at the first stage is called ālocana-jñāna (sense-cognition) and as a result of that there is roused the memory of its previous taste and a sense of pleasurable character (sukhasādhanatvasmṛti) and as a result of that I perceive the orange before me to have a certain pleasure-giving character[3]. It is urged that this appreciation of the orange as a pleasurable object should also be regarded as a direct result of perception through the action of the memory operating as a concomitant cause (sahakāri).

I perceive the orange with the eye and understand the pleasure it will give, by the mind, and thereupon understand by the mind that it is a pleasurable object. So though this perception results immediately by the operation of the mind, yet since it could only happen in association with sense-contact, it must be considered as a subsidiary effect of sense-contact and hence regarded as visual perception. Whatever may be the successive intermediary processes, if the knowledge is a result of sense-contact and if it appertains to the object with which the sense is in contact, we should regard it as a result of the perceptual process. Sense-contact with the object is thus the primary and indispensable condition of all perceptions and not only can the senses be in contact with the objects, their qualities, and the universals associated with them but also with negation.

A perception is erroneous when it presents an object in a character which it does not possess (atasmimstaditi) and right knowledge (pramā) is that which presents an object with a character which it really has (tadvati tatprakārakānubhava)[4]. In all cases of perceptual illusion the sense is in real contact with the right object, but it is only on account of the presence of certain other conditions that it is associated with wrong characteristics or misapprehended as a different object. Thus when the sun’s rays are perceived in a desert and misapprehended as a stream, at the first indeterminate stage the visual sense is in real contact with the rays and thus far there is no illusion so far as the contact with a real object is concerned, but at the second determinate stage it is owing to the similarity of certain of its characteristics with those of a stream that it is misapprehended as a stream[5].

Jayanta observes that on account of the presence of the defect of the organs or the rousing of the memory of similar objects, the object with which the sense is in contact hides its own characteristics and appears with the characteristics of other objects and this is what is meant by illusion[6]. In the case of mental delusions however there is no sense-contact with any object and the rousing of irrelevant memories is sufficient to produce illusory notions[7]. This doctrine of illusion is known as viparītakhyāti or anyathākhyāti. What existed in the mind appeared as the object before us (hṛdaye parisphurato’rthasya bahiravabhāsanam)[8]. Later Vaiśeṣika as interpreted by Praśastapāda and Śrīdhara is in full agreement with Nyāya in this doctrine of illusion (bhrama or as Vaiśeṣika calls it viparyaya) that the object of illusion is always the right thing with which the sense is in contact and that the illusion consists in the imposition of wrong characteristics[9].

I have pointed out above that Nyāya divided perception into two classes as nirvikalpa (indeterminate) and savikalpa (determinate) according as it is an earlier or a later stage. Vācaspati says, that at the first stage perception reveals an object as a particular; the perception of an orange at this avikalpika or nirvikalpika stage gives us indeed all its colour, form, and also the universal of orangeness associated with it, but it does not reveal it in a subject-predicate relation as when I say “this is an orange.” The avikalpika stage thus reveals the universal associated with the particular, but as there is no association of name at this stage, the universal and the particular are taken in one sweep and not as terms of relation as subject and predicate or substance and attribute (jātyādisvarūpāvagāhi na tu jātyādinām mitho viśeṣaṇaviśeṣyabhāvāvagāhīti yāvat)[10]. He thinks that such a stage, when the object is only seen but not associated with name or a subject-predicate relation, can be distinguished in perception not only in the case of infants or dumb persons that do not know the names of things, but also in the case of all ordinary persons, for the association of the names and relations could be distinguished as occurring at a succeeding stage[11].

Śrīdhara, in explaining the Vaiśesika view, seems to be largely in agreement with the above view of Vācaspati. Thus Śrīdhara says that in the nirvikalpa stage not only the universals were perceived but the differences as well. But as at this stage there is no memory of other things, there is no manifest differentiation and unification such as can only result by comparison. But the differences and the universals as they are in the thing are perceived, only they are not consciously ordered as “different from this” or “similar to this,” which can only take place at the savikalpa stage[12]. Vācaspati did not bring in the question of comparison with others, but had only spoken of the determinate notion of the thing in definite subject-predicate relation in association with names. The later Nyāya writers however, following Gaṅgeśa, hold an altogether different opinion on the subject. With them nirvikalpa knowledge means the knowledge of mere predication without any association with the subject or the thing to which the predicate refers. But such a knowledge is never testified by experience.

The nirvikalpa stage is thus a logical stage in the development of perceptual cognition and not a psychological stage. They would not like to dispense with it for they think that it is impossible to have the knowledge of a thing as qualified by a predicate or a quality, without previously knowing the quality or the predicate (viśiṣṭavaiśiṣṭyajñānam prati hi viśeṣaṇatāvacchedakaprakāraṃ jñānaṃ kāraṇaṃ)[13]. So, before any determinate knowledge such as “I see a cow,” “this is a cow” or “a cow” can arise it must be preceded by an indeterminate stage presenting only the indeterminate, unrelated, predicative quality as nirvikalpa, unconnected with universality or any other relations (jātyādiyojanārahitaṃ vaiśiṣṭyānavagāhi niṣprakārakam nirvikalpakaṃ)[14]. But this stage is never psychologically experienced (atīndriya) and it is only a logical necessity arising out of their synthetic conception of a proposition as being the relationing of a predicate with a subject.

Thus Viśvanātha says in his Siddhānta-muktāvalī,

“the cognition which does not involve relationing cannot be perceptual for the perception is of the form ‘I know the jug’; here the knowledge is related to the self, the knower, the jug again is related to knowledge and the definite content of jugness is related to the jug. It is this content which forms the predicative quality (viśeṣaṇatāvacchedaka) of the predicate ‘jug’ which is related to knowledge. We cannot therefore have the knowledge of the jug without having the knowledge of the predicative quality, the content[15].”

But in order that the knowledge of the jug could be rendered possible, there must be a stage at which the universal or the pure predication should be known and this is the nirvikalpa stage, the admission of which though not testified by experience is after all logically indispensably necessary. In the proposition “It is a cow,” the cow is an universal, and this must be intuited directly before it could be related to the particular with which it is associated.

But both the old and the new schools of Nyāya and Vaiśeṣika admitted the validity of the savikalpa perception which the Buddhists denied. Things are not of the nature of momentary particulars, but they are endowed with class-characters or universals and thus our knowledge of universals as revealed by the perception of objects is not erroneous and is directly produced by objects. The Buddhists hold that the error of savikalpa perception consists in the attribution of jāti (universal), guṇa (quality), kriyā (action), nāma (name), and dravya (substance) to things[16].

The universal and that of which the universal is predicated are not different but are the same identical entity. Thus the predication of an universal in the savikalpa perception involves the false creation of a difference where there was none. So also the quality is not different from the substance and to speak of a thing as qualified is thus an error similar to the former. The same remark applies to action, for motion is not something different from that which moves. But name is completely different from the thing and yet the name and the thing are identified, and again the percept “man with a stick” is regarded as if it was a single thing or substance, though “man” and “stick” are altogether different and there is no unity between them.

Now as regards the first three objections it is a question of the difference of the Nyāya ontological position with that of the Buddhists, for we know that Nyāya and Vaiśeṣika believe jāti, guṇa and kriyā to be different from substance and therefore the predicating of them of substance as different categories related to it at the determinate stage of perception cannot be regarded as erroneous. As to the fourth objection Vācaspati replies that the memory of the name of the thing roused by its sight cannot make the perception erroneous.

The fact that memory operates cannot in any way vitiate perception. The fact that name is not associated until the second stage through the joint action of memory is easily explained, for the operation of memory was necessary in order to bring about the association. But so long as it is borne in mind that the name is not identical with the thing but is only associated with it as being the same as was previously acquired, there cannot be any objection to the association of the name. But the Buddhists further object that there is no reascn why one should identify a thing seen at the present moment as being that which was seen before, for this identity is never the object of visual perception.

To this Vācaspati says that through the help of memory or past impressions (saṃskāra) this can be considered as being directly the object of perception, for whatever may be the concomitant causes when the main cause of sense-contact is present, this perception of identity should be regarded as an effect of it. But the Buddhists still emphasize the point that an object of past experience refers to a past time and place and is not experienced now and cannot therefore be identified with an object which is experienced at the present moment. It has to be admitted that Vācaspati’s answer is not very satisfactory for it leads ultimately to the testimony of direct perception which was challenged by the Buddhists[17]. It is easy to see that early Nyāya-Vaiśeṣika could not dismiss the savikalpa perception as invalid for it was the same as the nirvikalpa and differed from it only in this, that a name was associated with the thing of perception at this stage.

As it admits a gradual development of perception as the progressive effects of causal operations continued through the contacts of the mind with the self and the object under the influence of various intellectual (e.g. memory) and physical (e.g. light rays) concomitant causes, it does not, like Vedānta, require that right perception should only give knowledge which was not previously acquired. The variation as well as production of knowledge in the soul depends upon the variety of causal collocations.

Mind according to Nyāya is regarded as a separate sense and can come in contact with pleasure, pain, desire, antipathy and will. The later Nyāya writers speak of three other kinds of contact of a transcendental nature called sāmānyalakṣaṇa , jñānalakṣaṇa and yogaja (miraculous). The contact sāmānyalak-ṣaṇa is that by virtue of which by coming in contact with a particular we are transcendentally (alaukika) in contact with all the particulars (in a general way) of which the corresponding universal may be predicated. Thus when I see smoke and through it my sense is in contact with the universal associated with smoke my visual sense is in transcendental contact with all smoke in general. Jñānalakṣaṇa contact is that by virtue of which we can associate the perceptions of other senses when perceiving by any one sense. Thus when we are looking at a piece of sandal wood our visual sense is in touch with its colour only, but still we perceive it to be fragrant without any direct contact of the object with the organ of smell. The sort of transcendental contact (alaukika sannikarṣd) by virtue of which this is rendered possible is called jñānalakṣaṇa. But the knowledge acquired by these two contacts is not counted as perception[18].

Pleasures and pains (sukha and duhkha) are held by Nyāya to be different from knowledge (jñāna). For knowledge interprets, conceives or illumines things, but sukha etc. are never found to appear as behaving in that character. On the other hand we feel that we grasp them after having some knowledge. They cannot be self-revealing, for even knowledge is not so; if it were so, then that experience which generates sukha in one should have generated the same kind of feeling in others, or in other words it should have manifested its nature as sukha to all; and this does not happen, for the same thing which generates sukha in one might not do so in others. Moreover even admitting for argument’s sake that it is knowledge itself that appears as pleasure and pain, it is evident that there must be some differences between the pleasurable and painful experiences that make them so different, and this difference is due to the fact that knowledge in one case was associated with sukha and in another case with duhkha.

This shows that sukha and duhkha are not themselves knowledge. Such is the course of things that sukha and duhkha are generated by the collocation of certain conditions,and are manifested through or in association with other objects either in direct perception or in memory. They are thus the qualities which are generated in the self as a result of causal operation. It should however be remembered that merit and demerit act as concomitant causes in their production.

The yogins are believed to have the pratyakṣa of the most distant things beyond our senses; they can acquire this power by gradually increasing their powers of concentration and perceive the subtlest and most distant objects directly by their mind. Even we ourselves may at some time have the notions of future events which come to be true, e.g. sometimes I may have the intuition that “To-morrow my brother will come,” and this may happen to be true. This is called pratibhāna-jñāna, which is also to be regarded as a pratyakṣa directly by the mind. This is of course different from the other form of perception called mānasa-pratyakṣa, by which memories of past perceptions by other senses are associated with a percept visualized at the present moment; thus we see a rose and perceive that it is fragrant; the fragrance is not perceived by the eye, but the manas perceives it directly and associates the visual percept with it. According to Vedānta this acquired perception is only a case of inference.

The prātibha-pratyakṣa however is that which is with reference to the happening of a future event. When a cognition is produced, it is produced only as an objective cognition, e.g. This is a pot, but after this it is again related to the self by the mind as “I know this pot.” This is effected by the mind again coming in contact for reperception of the cognition which had already been generated in the soul. This second reperception is called anuvyavasāya, and all practical work can proceed as a result of this anuvyavasāya[19].

Footnotes and references:


Gaṅgeśa, a later naiyāyika of great reputation, describes perception as immediate awareness (pratyakṣasya sākṣātkāritvam lakṣaṇam).



Na khalvatīndriyā śaktirasmabhirupagamyate
yayā saha na kāryyasya sambandhajñānasambhavaḥ.

     Nyāyamañjarī, p. 69.


Sukhādi manasā buddhvā kapitthādi ca cakṣuṣā
tasya karaṇatā tatra manasaivāvagamyate...
...Sambandhagrahaṇakāle yattatkapitthādiviṣayamakṣajam
jñānam tadupādeyādijñānaphalamiti bhāṣyakṛtaścetasi sthitam

     Nyāyamañjarī, pp. 69-70; see also pp. 66-71.


See Udyotakara’s Nvāyavārttika , p. 37, and Gaṅgeśa’s Tattvacintāmarii, p. 401, Bibliotheca Indica.


Indriyeṇālocya marīcīn uccāvacamuccalato nirvikalpena grhītvā paścāttatropaghātadoṣāt viparyyeti, savikalpakosya pratyayo bhrānto jāyate tasmādvijñānasya vyabhicāro nārthasya,

     Vācaspati’s Tātparyaṭīkā p. 87.


Nyāyamañjarī, p. 88.


Ibid. pp. 89 and 184.


Ibid. p. 184.


Nyāyakandalī, pp. 177-181,

Suktisaṃyuktenendriyeṇa doṣasahakārmā rajatasaṃskārasacivena sādṛśyamanunmdhatā śuktikāviṣayo rajatādhyavasñyaḥ kṛtaḥ.”


Tātparyaṭīkā , p. 82, also ibid. p. 91,

prathamamālocito'rthah sāmānyaviśeṣavān.


Ibid. p. 84,

tasmādvyutpannasyñpi nāmadheyasmaraṇāya pūrvameṣitavyo vinaiva nāmadheyamartkapratyayaḥ.


Nyāyakandalī , p. 189 ff.,

ataḥ savikalpakamicchatā nirvikalpakamapyeṣitavyam, tacca na sāmānyamātram gṛhṇāti bhedasyāpi pratibhāsanāt nāpi svalakṣaṇamātram sāmānyākārasyāpi saṃvedanāt vyaktyantaradarśane pratisandhānācca, kintu sāmānyam viśeṣañcobhayamapi gṛhṇāti yadi paramidaṃ sāmānyamayam viśeṣaḥ ityevaṃ vivicya na pratyeti vastvantarānusandhānavirahāt, piṇḍāntarānuvṛttigrahaṇāddhi sāmānyaṃ vivicyate, vyāvṛttigrahaṇāuiviśeṣoyamiti vivekaḥ.


Tattvacintāmaṇi, p. 812.


Ibid. p. 809.


Siddkānlamuktāvalī on Bhāṣāpariccheda kārikā , 58.


Nyāyamañjarī , pp. 93-100,

Pañca caite kalpanā bhavanti jātikalpanā, guṇakalpanā, kriyākalpanā, nāmakalpanā dravyakalpanā ceti, tāśca kvacidabhede’pi bhedakalpanāt kvacicca bhede’pyabhedakalpanāt kalpanā ucyante.”

See Dharmakīrtti’s theory of Perception, pp. 151-4. See also pp. 409-410 of this book.


Tātparyaṭīkā, pp. 88-95.


Siddhāntamuktāvalī on Kārikā 63 and 64. We must remember that Gañgeśa discarded the definition of perception as given in the Nyāya sūtra which we have discussed above, and held that perception should be defined as that cognition which has the special class-character of direct apprehension. He thinks that the old definition of perception as the cognition generated by sense-contact involves a vicious circle (Tattvacintāmaṇi, pp. 538-546). Sense-contact is still regarded by him as the cause of perception, but it should not be included in the definition. He agrees to the six kinds of contact described first by Udyotakara as mentioned above.


This later Nyāya doctrine that the cognition of self in association with cognition is produced at a later moment must be contrasted with the tnputīpratyakṣa doctrine of Prabhākara, which holds that the object, knower and knowledge are all given simultaneously in knowledge. Vyavasāya (determinate cognition), according to Gaṅgeśa, gives us only the cognition of the object, but the cognition that I am aware of this object or cognition is a different functioning succeeding the former one and is called anu (after) vyavasāya (cognition),

idamahaṃ jānāmīti vyavasāye na bhāsate tadbodhakendriyasannikarṣābhāvāt kintvidaṃviṣayakajñānatvaviśiṣṭasya jñānasya vaiśiṣṭyamātmani bhāsate; na ca svaprakāśe vyavasāye tādṛśaṃ svasya vaiśiṣṭyaṃ bhāsitumarhati, pūrvaṃ viśeṣaṇasya lasyājñānāt, tasmādidamahaṃ jānāmīti na vyavasāyaḥ kintu annvyavasāyah.”

     Tattvacintāmaṇi , p. 795.