THE INDIAN CONCEPTS OF KNOWLEDGE AND SELF (Second instalment) KALIDAS BHATTACHARYYA The [Nyaya-Vai'sesika] doctrines that mental states are short-lived and that two or more such states cannot co-exit were examined in the last section. We arrived at the following conclusions: (i) There is no possible denial of mental states as emergent and cessant.(ii) Their cessation is due to no foreign cause, they are self-destroying; and continuation is not incompatible with self-destruction. (III) Co-existence of two or more mental states is not merely not impossible but often a fact. In the next section we propose to examine in detail the [Nyaya-Vai'sesika] concept of Object. SECTION III The concept of object examined A. [Nyaya-Vai'sesika concept of Object reiterated [Nyaya-Vai'sesika] has distinguished between object (visaya) and the real (padartha). A real becomes an object when it is known; and as the content of a possible (not actual) knowledge, it is a possible object. The real is that which as absolutely independent of my present knowledge has only been revealed by it. When it is so revealed (known) there occurs between it and the knowledge a relation which as belonging to the real is called its objectivity (visayata), but as belonging at the same time to the knowledge it is subjectivity(visayita) of that knowledge. Objectivity, unless it be only possible, is, in other words, and extrinsic relational property accruing to the real when it is known. This concept of objectivity was elucidated in further details in Section I. In that Section it was also shown that this objectivity is almost a tertiary property, in the sense that though it belongs to the real, and not, as objectivity, to the knownledge of the real, it yet, as a relational property, is constituted by that knowledge. For a proper understanding of this two questions which were not raised in Section I need here be examined. They are (i) whether the relation cannot be extrinsic in the sense that it iss not constituted by either term, and (ii) whether objectivity as aproperty (relational or not ) may not be due to knowledge as an efficient (nimitta) cause, not constituted by it. The reply to the first question would be thsi: Relation may often be extrinsic in the sense indicated, but never so in certain cases, particularly where it is between knowledge and the real that is known. Between the world of knowledge and that of reals there is nothing that is not included in either. Hence the relation between an instance of knowledge and the real known must belong to one of these worlds. As a matter of fact, it is found to belong to either alternatively: Knowledge is of the real and the real is known. In the former case the relation belongs to knowledge, in the latter it belongs to the real. The relation between knowledge and the real is not, in other words, a simple affair like that between any two reals. It may be asked if the dichotomy of the knowledge-world and the thing-world is metaphysically justified. Modern realists have questioned this, and we are told that [Nyaya-vai'sesika] also does not allow this. Is not knowledge known quite as much as other things? Knowledge indeed is known like other things. Yet the knowledge that is known is knowledge of a particular thing. No other thing is necessarily of another thing. So far knowledge is fundamentally different from other things. [Nyaya-Vai'sesika] onlyinsists that this type of thing is nevertheless revealed as an ordinary thing in another knowledge and, as so revealed, is an object. Knowledge, in other words, as necessarily of a thing, is necessarily subjective (visayin), and yet there is no metaphysical clash between this subjectivity and the objectivity (thinghood) of knowledge. Included in the sweeping world of things there are two entirely different types, viz., those which are necessarily subjective and those which are not so. The distinction between the two is more deep-seated than that between a tree and a blade of grass. Neither the tree nor that grass is necessarily subjective. The second question was whether knowledge to which objectivity is due is not its efficient cause. Our reply is: it is not an efficient cause, it is constitutive. The reasong is stated below: Objectivity, though a property of the real known, is also the relation between that real and the knowledge. It is a relation of the real to that knowledge. As of the real it belongs to the real, but as a relation to that knowledge it is constituted by the knowledge. The real is here the locus (anuyogin) of the relation, and knowledge its constitutive determinant (Pratigyogin). There is no mere relation, no relation that is without a constitutive determinant. A further peculiarity of the determinant of a relation is that it is never a class, unless the relation is specifically of a thing to a class. Object, from this point of view, may then be defined as that real which has for a property a relation constitutively determined by the knowledge ofthat real. The real here is not constitute by knowledge, because the relation in question is its extrinsic property. But objectivity and , therfore, object also are constituted by knowledge. 'Constituted by knowledge' may not mean that knowledge isan [upandana karana], but there is no denying the fact that objectivity is somehow constituted by knowledge. The very concept of object as the real that has been known involves reference to knowledge. No effect, on the other hand, involves in the very concept of itself reference to its efficient cause. This also proves that knowledge is not efficient cause (but constitutive of object). But though objectivity is constituted by knowledge this does not mean that the total knowledge-situation is to be interpreted idealistically. Objectivity belongs also to the real as its property, and this real is independent of the knowledge that reveals it. The reals as such are apprehended in non-judgmental perception (nirvikalpa-pratyaksa). This, again, is not the only reason why [Nyaya-Vai'sesika] sides with realism. There is another reason more fundamental. The fundamental postulate of knowledge is truly independent. Objectivity, though constituted by knowledge, appears independent. Hence it is truly independent. The only way to reconcile this independence with its being constituted by knowledge is to hold that the independent is the real as such and objectivity as consituted by knowledge belongs nevertheless to thsi real. The postulate is not dogmatic. It is capable of some sort of proor. If O appears independent of knowledge it is either really independent or not. But the negative alternative is untenable. If it were not really independent it was either the knowledge itself or constructed by it. But it cannot be either. To no corrective awareness is it ever felt that way. One cannot also insist that, whether felt or not, it is inferred that way. The apparent objectivity of O would go aganinst that inference. No cognition ever appears independent of itself, and no cognitive construction appears independent of the cognition that constructs it. it cannot be said, again, that the independence is an illusion. The independence as such cannot be an illusion. There is no illusory content which, or the like of which, was never presented as real. Object, then, is independent of the knowledge of it. This independence of object is the same thing as the fact that objectivity belongs to the real as a contingent property, which means that object being independent of knowledge does not clash with its being constituted by knowledge. Even if this were not the case, but object or objectivity were understood as itself independent of knowledge, even then there would be no grat difficulty. To be constituted by knowledge would than, it is true, contradict is forced upon us, and there is no way out, it has to be submitted to. Such cases, however, ought not to be multiplied for the mere luxury of specualtion. In spite, then, of being constituted by knowledge, object or objectivity is real. But there is yet another difficulty to remove. [Nyaya-Vai'sesika] has classified reals into severn original groups. But object or objectivity appears to belong to none of them. Forms of objectivity, ivz. [vi'sesyata], [Prakarata], etc., and therefore object also, are neither [dravya] nor [guna] nor [karma] nor [samanya], [samavaya], [vi'sesa] or [abhava]. If they do not belong to any of these they ought not to be called real. This is the difficulty. [Nyaya-Vai'sesika] has removed it in two ways. Most objects and there for also the forms of objectivity involved are the [svarupa] of reals; and some objects, particularly those which are flase, are only to bee analysed into real consituents where theform of objectivity is not substantive-adjective [samanadhi-karanya], but only [samsarga]. What is meant is that the total object of illusion is only a loose unity. To explaing. An Object as the content of knowledge is always a complex unity. The elements of this unity are reals (Padarthas) which as such are knowable in [nirvikalpa-Pratyaksa] only, and the relations that are added in [savikalpa] knowledge are, as seen, both knowledge-wise and reality-wise. As reality-wise they are taken as real, and unless contradicted they are also really real. The elements and the relation are thus equally real. If the relations cannot be places among the catalogued [Padarthas], this is because these relations, though real, are not additional realities. If a real A is really related to a real B, this does not necessarily mean that the relation is a third real entitly. The Buddhists too have admitted this when they hold that [santana] which is as real as teh [Ksanikas] is yet not other than these. Many Western thinkers also insist that relations, though really relating, are not other than relata. All the difficulty arises when the reality of relation is misunderstood as its being a third entity. If it is not third, if, in other words, the real relation is exhausted in the catalogued [Padarthas], there remains no difficulty in admitting its reality. [Nyaya-Vai'sesika] holds that the so-called additional relation we are aware of in [savikalpa-jnana] has this status only. They are exhausted in, another name of which is that they are the [svarupa] of, the [padarthas] they relate, not additional realities. Not that all entities which we call relation are of this type. In herence (samavaya) and contact (samyoga) are called relation and they are additional reals. Similaly when a fact or a series of facts which are normally treated as terms (as oppposed to relation) act as relation (e.g., between a father and a child) they, even as relation, are additonal entities. The additionality of inherence and contact follow from the fact that they are matters of [nirvikalpa-Pratyaksa], and that of the facts or the rseries is immediately evident. Where there is no such special reason or immediate evidence a relation need not be additional. A flower, its red colout and the inherence of the latter in the former are, according to [Nyaya-Vai'sesika], separate reals; yet in the perceptual judgment (savikalpa-Pratyaksa) of the form 'this flower is red' where the inherence of the red colour in the flowere stands as related to that flower and that colour, this second relation need not, because there is no special reason or immediate evidence, be a separate object. Not that it is therefore a subjective construction only. We have seen why, according to [Nyaya-Vai'sesika], it has to be taken as real. It follows that such relations are real and yet not other than thereals they relate. Such relation are the [suarupa] of the [Padarthas] related. The [above] is the account of the object of normal [savikalpa-pratyaksa]. the account of the false object (assuming that falsity has been detected) is different. In erroneous [savikalpa-pratyaksa] the total object is definitely known to be not real.Hence though, like the object of normal [savikalpa-Pratyaksa], is too is broken up into real elements and a relation, can the relation be taken as teh [svarupa] of the elements, seeing that the total object is not real? Ordinarily we should say 'No'.But [Nyaya-Vai'sesika] prefers to stick to the claim that [vikalpa] relations are the [svarupa] of the [Padarthas] related. They stick to it, only because it has followed from the fundamental postulate that whatever appears as independent is really independent. Object, everywhere, is to be analysed into the constituent reals and the [vikalpa] relations, whcih latter are everywhere exhausted in those reals. But how, then, could the total object be unreal here? [Nyaya-Vai'sesika] replies as follows: The total object here is no close unity. When the illusion is exposed the elements cannot be said to have been apprehended as realted in the way of substantive-adjective identity (samanadhikaranya). The unity here is loose, it is of the form 'P is inS', not of the form 'S is P'; and such unity is only nominal, no genuine unity. This, in effect, means that when the illusion is exposed we cannot say there was any genuine close unity of S (this) and P (snake). Not that 'P is in S' is never a close unity. Rather, normally it develops into that. 'P is in S' is easily translatable into 'P is as in S'='S is with P' which is a close unity. But such translation is sometimes impossible, particularly when it is known for certain that there is no real 'P as in S'. 'Horns are in the hare' cannot be translated into 'Horns are as in the hare'. While a denial of the former is intelligible it is impossible to deny the latter in the form 'Hornas as in hare are not'.Every judgment, whether affirmative or negative, presupposes that at least the subject-term stands for a reality, but 'horns as in the hare' stands from the begining as self-condemned. There is no suchdififculty, on the other hand, in the judgment 'Horns in the hare are not'. This judgment is only form 'Horns are not in teh hare', the corresponding affirmative judgment cannot but be in the form 'Horns are in the hare', not 'Horns are as in the hare'. Teh false object of an illusion corrected has also to be understood in this form. We cannot say 'This is snake' or 'The snake is as in the locus', we must say 'The snake is in the locus'. In the case of "hare's horn" or 'this snake' we are compelled to say this, only because stating the situation the other way about would stand self-condemened: we already know taht "hare's horn" or 'this snake' is not real. Denial of substantive-adjective identity does not, however, mean that there is no [vikalpa] relation here. Every [savikalpa-jnana] must involve [vikalpa] relations that are also asserted as real. But here the [vikalpa] relation is anything but identity. It is [samsarga], meaning any relation but identity. Teh 'in' in 'horns in the hare' or 'snake in the locus' is the [vikalpa] relation of [samsarga]. A distinction should be drawn between (a) [ghato nilah] (the pot is black), (b) [ghate nilah](black colour is in the pot) and (c) [ghato nilah (the pot is with black colour). In (a) the [vikalpa] relation is substantive-adjective identity (samanadhikaranya). In (b) it is [samsarga]. In (c) it is more complicated : there is a turn back to [samana-dhikaranya] through [samsarga]. Normally (b) and (c) coincide. But in cases like "hare's horn" or 'this snake' (b) fails to amount to (c). In the case (b) the content is peculiar. Though there is the [vikalpa] relation of [samsarga] the total objectis not a close unity. A pure case of (b) is not indeed a normal occurrence. We have to recognise it only where we are already assured that there is no real total object, as in the case of error. In [Nyaya-vai'sesika], object (visaya) is neither wholly reducible to knowledge and its phases, and is so far real, nor wholly equated to reality (padartha), though it is [svarupa] of that. Object as the real-that- is-known is as much real also. Wereit the real itself there would have been no occasion to distinguish between the real and the real-as known. But, again, even as not entirely the real, it si also exhausted in, i.e., the [svarupa] of the real. Also objectivity, though not wholly reducible to knowledge and its phases, is yet constituted by knowledge, being unintelligible apart from the fact that the corrsponding real is being known. Almost all Indian thinkers accept this view. Those who accept it differ only in further details. But most of the Western thinkers would reject it altogether. Western realists would never admit the intermediate object: they hold that knowledge is straight in relation with the real. Idealists and semi-idealists in the West would also, contrarily, deny object, reducing it to knowledge and its phases, and either reject the so- called real thing or admit it as never bodily knowable. A Berkeley would deny the real altogether, and a Kant or a Hegel would go the second way about. In defence of the intermediate object Indian thinkers would argue as follows: Awareness fo a real is either judgmental (savikalpa) or pre-judgmental (nirvikalpa). When it is [savikalpa] certain relation-forms of judgment-creep in . What is the status of these forms? Are they modes (or functions) of knowledge, or are they real, or both ? On the first alternative,realism, at least with regard to the content of [savikalpa] knowledge, is gone. On the third alternative there would indeed be a type of realism, but is would be more Indian than Western. The second alternative would only add difficulties. Are hypothetical and disjunctive forms and forms of inference real in the realistic sense ? They evidently involve subjective experiment; and so the contents of hypothetical and disjunctive forms and forms fo inference, embodythe experiment: the resulting propositions, and the conclusion are in the form 'if -then-', 'either-or-'and 'therefore-'. Attempts to get rid of such embodiment of the experiment have always looked forced. The reduction of the hypothetical propositionto the categorical may be a piece of skilful translation work, but no hypothetical proposition ever means a categorical fact only. It follows that the reduction of the disjunctive proposition to the categorical is equally a faliure, for such reduction is possible through another reduction, viz., of the disjunctive to the hypothetical. It is doubtful, again, if even the latter reduction is complete and natural. Even if a disjunctive proposition can be analysed into two or four (or whatever be the number) hypothetical propositions we must not forget that the disjunctive proposition is the unity of those hypotheticals, that unity being its specific characteristic. The attempt to get rid of the "therefore" in inference would also be equally abortive,that "therefore" being the very characteristic feature of inference. There is indeed something like "because- therefore" in the hypothetical proposition also; but it is only like that. In "because- therefore" the antecedent stands asserted. But it is not so asserted in "if-then.", unless "if-then" be only an apologetic softening of "because - there-fore". What, now,is true of these judgments and inference is true equally of categorical judgments, affirmative or negative judgment involves subjective experiment. There is such experiment so far at least as the possibility (yogyata) fo the negatum being related to its locus is concerned. The experiment is also embodied in the content, though not so obvertly as before. In hypothetical and disjunctive judgments, and also in inference, the embodiment was evident in the forms of "if-then", "either -or" and "because-thereofre"; but [Possibility] which is an embodiment of subjective experiment is not stated explicitly in a negative judgment. Yet if the negatum were not understood as a possible real relatable to the locus, there would be no negative judgmentat all. "S is not P' necessarily implies, though this implication does not come up to the surface, that a possible reality P relatable to S does not stand so related to it. Though negation, whether by way of identity or that of [samsarga], may be a reality the form of the negative judgment-which form is also inevitably asserted of the content- is not a reality in the realistic sense. As regards affirmative judgment, one type of it, viz., the universal, cannot have a form that can pass unchallenged as realistically real. Like negation, the universal judgment we do not merely assert a universal related to another universal. In the judgment "All men are mortal" we inevitably assert all individual men also (taken in denotation) as related to either mortality or mortal beings. How, now, are all individual men apprehended here ? We do not apprehend every man with his particular features, we apprehend him as only a case of the universal humanity. Individual men are, in other words, known through our knowledge of that universal. This need not be the [samanyalaksana-Pratyaksa] of the [Naiyayikas]. We may not [Perceive] all individual men, and this is possible if only we apprehend all men through our knowledge of the universal humanity. A subjective experiment is thus involved, and the experiment is embodied in the form "all". "All X's" cannot be a purely realistic fact. The Russellian idea of such "all". as an open class is unacceptable. In the judgment "All men are mortal" we do not mean that A,B, C,D, .........and so on are mortal. There is no sense of privation here. It does not mean that the men whom you and I have seen and those whom we have not seen are mortal. This would be unduly apologetic. What is positively meant is that all individual men are mortal. We mean, in other words, a closed class, as much closed and positive as any group of enumerable things, the only difference between the two being that while the number in the latter is finite that in the former is infinite (not negative infinite, but positive). Russell could at all interpret "all" in his way because he was predisposed to denying the connotative universal. His interpretation would have been legitimate were he able to account forthe total meaning of "all" without having recourse to the connotative universal. But in the interest of economy he sacrificed at least an important part of thetotal meaning. We mean by "all" a positively infinite number of individuals. Such an "all" is not an absurdity as a Russellian would have us believe. A closed class of a postively infinite number of individuals is intelligible if understood through (the presupposed knowledge of ) the corresponding connotative universal. Whether or not that connotative universal is itself also meant by "all" is not the point here. It is enough that at least the positively infinite number of individuals are meant. Likewise the simple categorical form "this S is P" cannot also taken as real in the realistic sense. In such judgments the predicate as almost always universal is to be understood, in the way of a universal subject, as somehow referring to all individuals, and therefore through a corresponding universal. Where the predicate is not a universal, or supposing that a universal need not be understood in denotation, there is still another reason-and that is more primary-why the form "this S is P" cannot be real in the realistic sense. The relation meant by the copula "is" embodies a subjective experiment, though only covertly. The relation meant is neither inherence nor contact nor any taht is a [Padartha] in the [Nyaya-Vai'sesika] sense. It is one that relates S, P and that [Padartha] relation into a unittary object, and is, therefore a [vikalpa]. Thsi [vikalpa] is not consciously felt as experiment. But it must be one such. We have already proved that every [vikalpa] is knowledgewise-ness is noother than the fact that a mode of knowledge is embodied in the content. Over and above S, P and the [Padartha] relation a second relation which as unifying the three has to be admitted cannot be real in the realistic sense. C. Some clarifications- There are two questions which should be answered at this stage. It may be asked if thsi relation also does not require another relation, and so on [ad infinitum]. It may also be asked if the original [Padartha] relation does at all require the seond relation. To the first question the reply is in the negative. The second relation was required only to relate into a unity three items of reality one of which happened to be a relation. Before that unification there were only three items. But now that they stand unified through the second relation which is a [vikalpa], there is no task left to relate thsi second relation again to the three items by further relations. The reply to teh second question is in the affirmative. The first original relation was not sufficient to have formed the unity that is meant by the judgment "This S is P'. Often it is no genuine relation, but only a quality or even a substantive-indeed, anything whatever-which is somehow taken as intermediate between S and P. As such it cannot by itself relate S and P and (itself also) into a unity. When a so-called relation is a reality of this kind anothere relation whcih is genuinely a relation is requisitioned to do that work. But whatever else there is in the world of reals, thsi genuine relation is not there till that S, P and the so-called relation stand unified in knowledge. "Unified in knowledge", we repeat, does not preclude the possibility that they stand unified in the world of reals also. Teh unsophisticated mind takes them as also forming a real unity, for such is the plain realistic import of the judgment. The fact that S,P and their so-called relation are unified in knowledge and that yet the unity formed is real may be understood in three ways of which one only is tenable. It may mean taht S,P and the so-called relation only appear to be really related. Secondly, it may mean that they had been standing as already really related before I had the [savikalpa] knowledge, but that this real unity comes to be revealed only with that [savikalpa] knowledge, alomst in the same way in which [Vai'sesika] understands [samanya]. Or, thirdly, it may mean that they were not standing as really unified, but become related and unified just when I know them in the [savikialpa] way. Of these, the first alternative is rejected on the plain ground that no appearance can be dismissed as mere appearance or false unless there is a reason here for the dismissal, viz., that the genuine [vikalpa] relation has come to be knownas a mode of knowledge. For we cannot overlook the other side, viz., that it is also asserted as real. To show merely that something is A does not prove that its appearance as B is unreal. For that another step is necessary. Either we must point to a clear defect (dosa) in that appearance or at least its being A is to be a matter of inference, itbeing presumed for the present that inference is a stronger [Pramana] than perception. But here the [vikalpa] relation to be a mode of knowledge is not a matter of inference. It is true the knowledgewise-ness of the [vikalpa] relationis not always evident ; but for one who has perceived that because-therefore, either-or, if-then, A as not B and all A are knowledge-wise it is not difficult to perceive that even the simple categorical form is also a mode of knowledge, particularly when it is realsied that S, P and their so-called relation cannot unify themselves. Knowledgewise-ness of the categorical form does not merely follow from the impossiblity of unification, it comes also to be immediately realised. There is, again, no specifiable defect in our awareness (which is quite primary) that the genuine relation is fact. Hence the dimissal of it as sheer appearance or false would be unjustified. Even if the knowledgewise-ness of the categorical [vikalpa] were merely a matter of inference, there is no reason why inference here should be preferred to the immediate knowlege that the [vikalpa] is real. Inference is preferred to immediate knowledge either when it not merely contradicts but definitely sublates (why, we may not say) the content of immediate knowledge, or when it is followed by the perception of a defect in that immediate knowledge, or when our point of view is that of [Pramanya], not of primary assertion which is present as much in inference as in perception. In the present case the inferred knowledgewise-ness of the [vikalpa] does nothing of the sort, and the point of view is [ex-hypothesi] not of [Pramanya] Inference is sometimes regarded as a stronger [Pramana] on the ground that it is supported by many cognitions that are involved in it. But the point of view of corroboration is that of [Pramanya], not of primary assertion. The [Pramanya] of a cognition may be extrinsic to that cognition as primary assertion, in whcih case it is doubtful if [Pramanya] has any metaphysical import. Or it may be intrinsic in which case the entire problem of [Pramanya] is a little more than explication. Either way the attitude of [Pramanya] is not very relevant to primary assertion. It would be useless to argue that when a cognition is confirmed from the point of view of [Pramanya] chances of its possible rejection are eliminated. Mere elimination of possible errorsdoes not make a cognition valied unless it were already so taken, though amidst a mass of confusions. A Particular cognition can also be dismissed as erroneous if it is succeeded by one which is its contradictory, the idea being that a cognition is the assertion of a genuine reality till it comes to be contradicted. [Uttarajnanapksapata] belongs, in this sense, to the very constitution of knowledge. But in the present case there is a strange phenomenon. The knowledge that the [vikalpa] relation is a mode of knowledge may be later than the assertion of that relation as real, yet when that later knowledge occurs the prior one is not sublated. Both continue with unabated primacy. The reality of this relation, then, cannot be false or sheer appearance. The second alternative mentioed in page 36 above, ivz., that S,P and their so-called relation had already formed a unity and is only reveraled in [savikalpajnana], has also to be rejected. The unity could not have been formed by the so-called relation, and a fresh relation whcih alone could form it could not have been there before the [savikalpajnana], because, as already shown, it is knowledge-wise. It has also been shown that the simple categorical form, quite as much as other forms of proposition, embody subjective experiment. Hence the third alternative alone is left. The [vikalpa] relation and the unity occur as real only when S,P and their so-called relation are known in the [savikalpa] way. This does not mean that the [savikalpajnana] as an efficient cause has producedsomething in the reals concerned. Waht is meant is that the propostional form, though knowledge-wise, comes to be asserted as involved in those reals. Though knowledge-wise, it comes to be asserted as real also; and as thsi is not self-contradictory, it can be taken as really real. But is not a real independent of the knowledge of it, and does this not imply that it existed before that knowledge occurred ? If something appears real only so long as it is known, is it not for that very reason called unreal ? The answer depends on what is meant by the word "reality". If it means 'that which exists and is independent fo the knowledge of it', the [vikalpa] relation and the unity are real, because even though they are constituted by knowledge they are yet asserted as existent and independent of knowledge, and we have seen how to be constituted by knowledge does not clash with this other character. It follows that to have remained prior to knowledge is not necessary for something to be real. Many reals may be so prior, but some need not be. Or, it may be said that the [vikalpa] and the unity had remained prior to knowledge, but as so prior they were not actually existent. They come to exist only when they are known. As subsistent, [vikalpa] relations remain in their self-contained aloofness, and relate, S,P and their so-called relation only when these latter come to be known, and through that knowledge. It is because they yet maintain their Platonic ideality that they refuse to be wholly identified with that knowledge and procalim themselves as prior to that knowledge; and it is because they now stand as relating, and therefore adjectival to, the actually real S,P and their so-called relation that they in that function come to be known as actually existent. This is more or less the Kantian view of [vikalpas]. whichever interpretation is accepted we have to admit grades of metaphysical status. In the first interpretation there would be two kinds of reality, one co-temporal with knowledge and the other transcending its duration; and as this distinction concerns the very existence, not the content, it is a distinction of metaphysical status. The distinction between subsistence (demand for existence) and actual existence is obviously a distinction of metaphysical status. The Indian position has been vindicated. It has been shown that over and above reality, though not necessarily separate from it, object has to be admitted. This has been established through an analysis of the metaphysical import of though-forms. Theonly conceivable way to get rid of this intermediate entity would be to deny that forms of thought have any metaphysical import. Logical Positivists have attempted this in their systematic campaign aganist thought. They consider forms of thought as either only means to analytical interpretation, the whole interpretation being only linguistic, or vicarious, misrepresenting a clever language-construction as pointing to a reality. But it is difficult to see why thought should be so unceremoniously guillotined. Mass hysteria is no logical justification. These Positivists ought to have seen that no judgment, not even the simple categorical, is either a mere analytical representation of a non-judgmental content-what to speak of non-perceptual judgments which are obviously not so ? - or, because of the extral element involved in it, vicarious, for we all believe that the total content of the judgment is real exactly in the form in which it appears in the judgment. We have already seen that in spite of being knowledge-wise the extra element is nevertheless felt as real and taht the two aspects do not clash. These Positivists have never explained why among the devils of judgmentsome, viz., a good number of perceptual judgments, are obedient slaves. We can understand Kant who has excluded a very limited number of judgments, and that on definite grounds. But these Positivists have stated with a bias. They have indeed shown extra-ordinary skill in translating non-perceptual judgments into the language of simple perception. But translation always falls short of the original: the original vitality is always missed and there is only vicarious compensation. Perceptual judgments do not merely analytically represent contents of simple perception. In simple perception there is a bare plurality of S,P and a so-called relation between them, all appearing either discrete or non-distinguished. But the judgment"This S is P" means that S and P, and sometimes their so-called relation also, are distinguished and yet related into a unity. We have said that in simple perception S, P and their so-called relation are [either ] discrete or non-distinguished. The former is the [Nyaya-Vai'sesika] view according to which simplw perception (nirvikalpa-Pratyaksa) is certified not by introspection but by inference, and the simple elements that are inferred as constituting a substantive-adjective complex perceived have to be inferred as discrete. But one is not compelled to accept the [Nyaya-Vai'sesika] view that simple perception has only to be inferred. One might hold that it is an introspectable stage. In this other view the simple constituents are not found as discrete, but in a sense non-distinguished. Let us explain, how. If there is any Psychological stage, called non-judgmental perception, it is of the form SP (this blue pot-nilaghatah) which differs from "this S is P" in that while in the latter S and P are both distinguished and related, the relation standing as a distinct entity, we do not find thsi phenomenon in the simple SP. Not that SP is therefore and amorphous homogeneity, as Bradley would have it. If there is at all a Psychological stage having the simple content SP, we are aware, at that very stage, of S and P also, the three contents-S, P and SP- alternating indeterminately, each, at the time it is apprehended, standing as absolute. When we are aware of S there is no question of either P or SP, and similarly with P, but as much an absolute entity astaht S or P. A whole, in simple perception, is never known as a whole of parts. For that apprehension the parts and the whole require to be related in a judgmental form of awareness. The very words "parts' and "whole" are relevant in a judgment context only. If A, B and C are three absolute entities, C is a whole, and A and B are parts, only when between C, on the one hand, and A and B, on the other, a certaing relation of dependence is asserted, when, eg., it is known that while A and B are dissociable from C, C is not so dissociable from A and B; and such knowledge cannot be simple perception. Similarly with regard to any other unity. A universal or a substance, e.g., is felt as dissociable from the relevant particulars or attributes, but not the latter from the former. In simple perception, then, S,P and SP are each absolute, and there is no question of a relation between them. But in the judgment "This S is P"S and P (and it may be, their so-called relation also) are related in a specifiable way in the unity SP. The indeterminate alternation of several absolutes is thus, in simple categorical judgment, replaced by determinate relation. In simple perception S,P and SP are each absilute. SP is not a unity, but as much an absolute entity as S or P. It may indeed be asked-Do not S and P stand involved in SP ? How otherwise could it be known as SP ? The reply is that in simple perception it is not known as SP, but merely as an absolute entity with a differential quality perceived. It is only retrospectively called SP, called that way from the point of view of the latter judgment "S is P." But, it may be asked again, is not that SP known, at least in this retrospective manner, as identical with the unity known in the judgment "S is P" ? We reply -Yes, there is only as much unity as between object and reality. We thus find that even simple perception is not so simple as Logical Positivists believe. It too involves an extra element, the as yet undefined differential quality. The logical form of the simple categorical judgment may be understood as linguisticdefinition of this quality. But it is not like definitions elsewhere. In other cases of definitions there is no line, except in the verbal presentation, between the definitum and the definition. Here there is such a line. Yet, however, the linguistic form is asserted as real without any sense of inconsistency. We have also seen that there is no contradiction in a though (language)-construction here being real. It may be asked-if in the world of reals there are inherence (samavaya), contact (samyoga), etc., are they not genuine relations relating S and P into a unity, even apart from my knowing them ? The answer should be prefaced by a more fundamental problem to be raised here and solved. If at a non-judgmental stage we can at all apprehend S, P and SP which are real, is there at that stage any object over and above those reals ? The problem, in other words, is if even in non-judgmental perception ther is the intermediate entity called object ? We reply- There is. If the reals here are S, P and SP, the object is these in indeterminate alternation. In [savikalpajnana] there are definite [vikalpa] relations binding relas into unities, but here in the place of thsoe definite [vikalpas] there is only indeterminate alternation, and therefore also an inddeterminate unity through that alternation. The unit that is effected by alternation is always indeterminate, as is evident in the case of disjunctive judgment. Here, however, in the present case, the alternation it self is indeterminate, and hence the unity effected is unlike one in disjunctive judgment. The unity here is not judgmental: the stage in question is below even simple categorical judgment. But there is still a unity, though at the vanishing point; and the vanishing unity is here the object. The object here is more coincident with reals than in [savikalpajnana]. The object and the real here are not [definitely] distinguishable. It may stall be asked if even at this stage the real SP is not apprehended as different from S and P, and, if so, whether the distinction can be anything but that thsi SP is a unity of S and P. The unity may be inderminate, but is it not a unity still? If so, has not the unity been effected by some elements in the region of reals, viz., inherence, contact, etc., is possible, why was it said before that these are only so-called relation, not relation that unify and, therefore, relate reals ? With this we come to the question asked at the beginning of thsi sub-section. The answer is that indeterminate unity is qualitatively different from one that is determinate. Indeterminate unity of S and P is little more than their alternation, as we find even in disjunction. When, again, the alternation itself is indeterminate, even SP which is the indeterminate unity of S and P alternates with that S and P. This latter means that though the difference between SP, on the one hand, and S and P, on the other, is now a little more defined the situation still remains indefinite. Indeterminate unity at the non-judgmental level, then, means either that S and P are only alternating with onw another or that SP comes to stand with just a differential quality, not further defined. Even where S and P merely alternate they stand eachwith a differential quality, and the quality is such that though it distinguishes S-with -that -emergent-quality from simple S, and P -with-that-emergent-quality from simple P, it is apprehended as somehow also the same in both. This vague sameness or identity of the differential quality, as appearing to transcend, on account of thsi identity, S and P comes to be represented as some sort of unity in the vague form of SP even here. The unities effected by inherence and contact, and the latter as relations, are to be understood in this light. When S and P in contact effect SP what is apprehended at the non-judgmental level is (i) that S and P have each a differential quality which is, only retrospectively from the point of view of a later [savikalpajnana], represented as S-in-cantact-with-P or P-inpcantact-with-S, and (ii) that somehow the contact is also felt as numerically one and the same, so that we also say tha S and P are in contact, the result being SP. The self-identical contact as standing between S and is never apprehended as an explicit definite real, what is expplicitly felt being only the indeterminate alternation of S-with-that-differentail-qualty and P-with-that-differential-quality. That this indeterminate alternation is at all felt, however vaguely, as the self-identical contact between S and P is no more than an incipient interpretation of the alternation in terms of [savikalpajnana](judgment). Judgment is so much a normal mode of knowing that even when we are aware that there is a non-judgmental mode, we, in spite of all caustion, involuntarily smuggle its form, though now in disguise, into the non-judgmental content. Contact is really a differential quality of each term, the contact of P with S being different from and alternating with the contact of S with P. Indian thinkers have always taken contact as qualities of S and P alternating. Contact includes a host of relations. Parts of a whole, e.g., are in contact with one another: the spatial relation of the parts with one another is, in other words, nothing but a form of contact. The spatial relation of up-down, right-left, etc., are in many cases forms of contact, with, of course, additional differential qualities at the level of non-judgmetal perception. The additional differentail quality is only retrospectively definable in terms of dik. Often, again, this differential quality alone is found, when, e.g., S and P are not in contact. As with spatial relation, so with corresponding temporal relations. Often, again, the contact is with the very principles of space and time. Into further niceties we need not enter. A host of other relations are represented by inherence. The relation, e.g., between a whole and a part, a universal and a particular, a quality and a substance, is inherence. But at the non-judgmental level it is not apprehended as a definite relation relating S and P. At that level it is only a differential quality of SP. SP no doubt alternates with S and P, but stands evident with that differential quality. The differential quality is only retrospectively specificable as teh fact that from the total situation SP either S or P is dissociable and the other not. at the non-judgmental level there is only a vague sense of thsi dissociability. A whole or a universal or a substance is only vaguely felt as dissociable from the total situation, and the parts, particulars or qualties are vaguely felt as dissociable from the total situation, and the parts, particulars or qualities are vaguely felt as undissociable. The total situation SP is felt with this differential quality. Some Western thinkers and the Buddhists have missed the differntial quality corresponding to what is called inherence and have accordingly denied the reality of the whole, the universal and the substance. Some of them have committed a further mistake of missing the reality-aspect of [vikalpa] relations, and thsi has led them to deny all reality to relation and unities. But, as is evident now, both these are exaggeration. The Buddhist position will be examined later. F. Object-reality distinction evident in correction of illusion The distinction between objec and the real will also be evident from an analysis of illusion as corrected. Before correction the content of illusion is felt as real object. But after correction it stands as an object minus the reality-aspect, so thatto the end it is still an object, though of a peculiar type, unconnected, or better, disconnected, with reality. This disconnection is not a normal feature of objects. But the illusory content is an abornal object, and because illusion is cancelled weare forced to admit such disconnection. Some believe that the corrected content as over and above reality is no object but subjective. [Vijnanavadi Buddhists] India and many thinkers in the West have held this view. The [Vijnanavadin's] view will be examined later. They have offered arguments, and these will be examined in due course. But the Western thinkers who have passed this as almost self-evident have only confused different issues. That appearance is distinct from reality is one issue, and whether what is distinct from reality is no sufficient reason that it is subjective. Further, these Western thinkers have misunderstood object as wholly identified with the real, and have naturally been driven to the conclusion that what is not real is, on that very account, not object, and is therefore subjective. But we have seen that object is neither unqualified real nor unqualifiedly subjective (knowledge-wise). There is another point to be considered in connection with the thesis that in correction of illusion we realise object as over and above reality. The object here is not necessarily the content of [savikalpa-pratyaksa]. It includes the content of [nirvikalpa-pratyaksa] as well, supposing there is such a stage evident to judgmental or non-judgmental-object, in spite of being knowledge-wise, is found coincident with the real, it is apprehended as loosened when a perception comes to be corrected. We have seen that in non-perceptual knowledge [vikalpas] and, therefore, objects are clearly felt as knowledge-wise (experimental), though not for that reason denied reality. But this knowledgewise-ness, we have also seen, is not so manifest in [savikalpa-perception], far less in [nirvikalpa]; object in these two cases is not clearly felt as distinct from the real. Our present thesis is that the distinction of the perceptual object-a determinate unity or an indeterminate whole-from the real stands exposed in correction. By implication it is admitted that even non-judgmental simple perception (nirvikalpa-Pratyaksa) can be erroneous. We are often told, particularly by Western thinkers, that is non-judgmental simple perception there is no question of falsity, all question of truth or falsity arising only when knowledge is judgmental (savikalpa). This is untenable. If the content of non-judmental perception be S,P and SP alternating, with a differential quality of either S and P or SP, there is no reason why this content should not be as much true or false as the content of judgment: all the difference between the two kinds of knowledge is that while in the latter there is explicit relation there is only a differential quality (or qualities) in the former. There is a kind of vague predication (unification), in the form of differential quality, in non-judgmental perception only.May not a simple content, not known as related with another be true or false ? When it is apprehended is it not asserted as real, and may not such assertion come in certain cases to be sublated later? The whole question as to whether the content of non-judgmental perception can or cannot be true or false depends on what is meant by the word "truth" or "falsity". If "truth" means that the content of knowledge exists, there is truth-claim in non-judgmental perception, for it too is asserted, i.e., taken as existent. Similarly if "falsity" means that the once-asserted existence of the content is now disbelieved - disbelief being not necessarily judgmental, but at lest in some cases the awareness ofa differential quality of the content- there is nothing against a non-judgmental congnition being false. An analysis of the very concept of judgmental rejection would corroborate this. Judgmental rejection=rejective judgment may be perceptive or non-perceptive-in Indian terminology, [savikalpa-pratyaksa] and [savikalpa-Paroksajnana]. Where it is perceptive there is in the content perceived a differential qulity corresponding to the [vikalpa] relation of contradiction, over and above that contradiction itself. It is only when the rejective judgment is non-perceptive (Paroksa) that there is no question of that diffeential quality, and rejection in such cases is either through a categorical or hypothetical inference or through a categorical or hypothetical inference or through testimony. Thus even perceptual rejective judgment is intelligible through a perceived differential quality, corresponding to the relation of contradiction, in the content rejected (though there is in the content the explicit relation of contradiction also). If so, the differential quality is, at lest in some cases, a sufficient ground for the rejection of the content. Why, then, may not the content of non-judgmental cognition be also rejected, when in it too a similar differntial quality comes to be perceived ? Truth, however, and therefore falsity also, may mean something else. Truth may mean that the existence of the content is explicitly asserted, as in the judgment 'SP exists', and such assertion is always the confirmation of a prior knowledge of the content. Truth, in this sense, is but the confirmedness of that prior cognition, so that the esixtence of the content has come to be specifically pointed to. Falsity would, from this point of view, be the [untenability[ of the prior cognition and, therefore, the explicit rejection of the content. Thsi is the problem of [Pramanya] in Indian philosophy, not always clearly distinguished in the West from the simple assertion of the existence or non-existence of a content/ If truth and falsity are understood from this reflective point of view it would be admissible indeed that only judgments can be true or false. But there should be a note of caution at the same time that this is not true of all judgments, so that judgmentality is one [sine qua non] of truth and falsity. Existential judgments and judgments of modality alone can be true or flase-judgments, namely, where existence or its near equivalent is stated as the explicit predicate. In other judgments there is no such explicit statement. In the judgment 'S is P,' for example, the copula 'is' represents more an explicit [vikalpa] relation than explicit existence of the content SP. That it appears to stand equally for both is an accident of English language. In Sanskrit we find that 'ghato nilah' is a sufficient expression, and the statement 'ghato nilo bhavati' is not required. There is logical ground also. Even in English language the existential import can be explicitly distinguished, as in the judgment 'SP exists'; and it is plain logic that if somethind can be distinguished it, where not distinguished, remains implicit and subordinate. The copula in 'S is P' thus only implicitly and subordinately conveys the existence of SP. Obvertly it represents a relation only between S and P. If it be insisted that after all the existential import is still present, though not explicitly, in the judgement 'S is P,' we reply that it is equally present in non-judgmental perception also. As much in the latter as in the former the total contentis known as existent. It has sometimes been urged that even judgments like 'S is P,' as distinguished from the non-judgmental awareness or SP, is against a doubt or challenge that S might not be P, so that as so against the doubt or the challenge it ismore reflective than the non-judgmental awareness of SP and, therefore, assertss so far the existence of the content explicitly. But this would be a wrong understanding of the actual situation. 'S is P' is certainly more reflective than SP, and perhaps non-judgmetal awareness is not reflective at all. But the reflectiveness of the former does not lie in its being against a doubt or challenge. Here there is neither an actual nor a possible doubt (or challenge). That there is no actual doubt can hardly can hardly be questioned. There is no possible doubt too; for a possible one is no more than that which I in judging that way only anticipate, and it is a fact that I did not anticipate one. Had I anticipated,the judgment would have been of the form 'S is P,' with an emphasis on the existential. The simple judgments would be hardly distingusihable from the existential. The simple judgment 'S is P' is reflective in the sense that it is against the background of a half-distinguished assumption of the abstract content 'S as P.' This 'S as P' as half -distinguished is no other than the unity-through-vikalpa-relation considered apart from its reality aspect. Sometimes, again, a thrid reason is offered why only judgment, and not non-judgmental awareness, can be true or false. It is said that as only judgment involves a sort of spontaneity, either because [vikalpa] relations are considered as acts or because a constructed general idea is papended to the subject, the question of the truth of the judgmentsal knowledge naturally crops up. But this, again, is both a too simple and a uselessly complicated account. To simple, because whether vikalpas be acts or not, and whether a general idea be a construction or not, there is also the undeniable fact that every judgment asserts the reality of the total content. To foreget this aspect and to insist on the vikalpa relations being subjective would be over-simplification. There is also unnecessary complication in that the vikalpas are taken as acts or, even by some, as forms of will, and general ideas are taken as mere constructions, whereas the peculiar character of judgment is intelligible even in the absence of any such theory. So there is no reason why judgments alone should be true or false, and non-judgmental knowledge outside this disinction. Both equally are true or false, if 'truth' means that the content is known as existent, and 'falsity' that it is rejected. Only when truth is understood as the confimedness of a cognition as against an actual or a possible challenge, and falsity as the corresponding rejection, can ordinary of truth and falsity. But as here we are not using the words 'truth' and 'falsity' in that sense we hold that all cognition can be true or false. With this we come back to the problem of the exact status of the illusory content, whether in judgment or in non-judgmental knowledge. G. Buddhist theory of [atmakhyati] examined Before an illusion is corrected the total content is taken as a real object. But after correction it is known as definitely not real and, therefore, to have been an object minus the reality-aspect. This is what is meant by rejection of the illusory content. It would be too much to claim, as some Buddhists have done, that even its objectivity is rejected. I fthey intend that both objectivity and reality are denied this would be unnecessary duplication. Rejection of any one of the two aspects is enough; so the other aspect has to be retained. It is enough for correction that the reality-aspect is rejected; hence objectivity ought to be retained. But why may it not be interpreted the other way about? May it not be said that the aspect of objectivity is rejected and the reality-aspect retained ? The Buddhists under consideration have, as a maater of fact, offered this interpretation. But this would only make the confusion worse confounded. If the content is real and yet not an object, it would be real as only a mdoe of knowledge. But does the corrective judgment assert this subjective reality ? Do we find that the illusory snake was not an object but an existent subjective idea ? Correction is either judgmental or non-judgmental. When judgmental, it is of the form 'this is not snake' coupled in a mysterious manner with another form, viz., 'this is rope'. The content 'this as not snake' is a unity, effected through a vikalpa relation, of a real this and either a real snake (when the vikalpa relation is negative) or the absence of snake. The content 'this as rope' is also a real unity of a real this and a real rope. In either case there is no escape from the this-element which is no subjective idea. The Buddhists in question have held that the content of correction is 'not this, but snake'. But even if thsi be allowed there is the other content 'this is rope' inseparably connected with it. In that other content this-element is asserted as existent, and it is also evident that this [this] is somehow non-different from the [this] in 'not this, but snake'. It is impossible that in the same correction the same [this] is both asserted as existent and rejected. That in the content 'this is rope' it is asserted as existent is beyond question. It follows that 'not this, but snake' is a mis-representation of the other content. That other content is either 'this, not snake' or 'this and snake, but no predicational identity of the two' or 'this and snake, but the two not consciously distinguished', etc., all of which are representable as 'this is not snake'. The Buddhist theory of [atmakhyati] cannot pass unchallenged. Even if we allow the form ' not this, but snake', it does not follow that the snake-aspect is subjective. That would presuppose that 'this' means to be now outside me. But 'this' does not mean that. Even an idea which no one can call outside is a this to me if it is now. The concept 'this' is highly intriguing and involves either now or here which are equally intriguing. To interpret it as ' to be now outside me' would only be too facile. The [Vijnanavadin] may argue that snake would still be subjective even if the content of correction were 'this, not snake'. 'Not snake' means that the snake is rejected, and the rejected snake as ousted, on the one hand, from the world of reals and asyet not zero, on the other hand, cannot but be subjective. But this too would be a hasty conclusion. In spite of being false, the snake appeared as object. A theory error which can retain thsi objectivity is to be prefferred to one which denies it too easily; and considering what has been sa8id so far about the distinction of object from reality, the presumption is against the idealsitic theory of the Buddhists. Correction may also be non-judgmental. But even there, as in all non-judgemental knowledge, the content is a presented rope with the peculiar flavour of denied presented snake, or an absent snake with the preculiar flavour of its being ousted by a now-presented rope, the once-presentedness of the snake being, of course, no more than a fringe round the flavour of being ousted. Which-ever way the content appears, there is no scope for the particular Buddhist theory. In every case the content is presented as an object. The rejected snake can in no way be taken as subjectively real. Indeed the phrase 'subjectively real' is often a camouflage. In what sense is a subjective idea real ? Is it real in the sense of being independent of its knowledge, or is it real in the sense of being just existent ? The Buddhists under consideration hold that in correction the outsideness only of the content is denied, and its reality is retained. But is the subjective reality of the snake its original pre-correction reality ? The pre-correction reality of the sanke included its having been independent of the knowledge of it, whatever else it might have included. But at lest that independence is now denied by these Buddhists. The subjectve snake is than real in some other sense. The reality of subjectivity is qualitatively different from that of a non-subjective content. While the reality of a non-subjective content is distinguishable from that content this is not the case, at lest according to the Buddhists in question, with the subjective. The subjective, at least according to them, is self-evident: to be subjective is [ipse-facto] to be real. In 'this flower exists' existence can be imagined as dissociable, as at least a universal belonging to thsi flower, or even as what may lapse. But in 'Iam' am-ness is the same thing as I-ness. I=I am. Contrariwise, the content fo the non-subjective is imaginable apart from existence (or non-existence), but not so the content of the subjective. If the subjective can at all be imagined apart from existence, there is no conceivable way of adding that existence ever to the content. The subjective is either ever a more content or ever with existence. Whichever way it is understood, it is evident that the reality of the subjective, if at all it is real, is qualitatively different from that of the non-subjective. To say, therefore, that the snake is subjectively different from that of the non-subjective. To say,therefore, that the snake is subjectively real is little more than saying that it is just subjecitve. The reality with which we contrast the false is the reality of the non-subjective. And yet these Buddhists persuade themselves that in correction the reality of the snake has been retained, as though it is the same reality which we had before correction. It is true that there is a natural tendency to take what is not real (in the realistic sense) as merely my imagination and, so far, subjective. But there is no assurance till now that the image, though subjective, does not stand outside. The false snake, detected as false, may have been a subjective image. But I saw it outside, and it is not yet certain whether this outsideness came to be cancelled. It might well be that its reality (existence) alone is cancolled, the snake being understood as a ghoslty outside entity, a floating adjective, as it were, of the rope that is real. An image to stand outside is not Prima facie absurd. In every normal perception where the content is presentative-representative the representative element, though imaginal, stands outside, tied to what is merely presented. If thsi be allowed, why may not an image, in illuion, stand outside, though unconnected or misconnected with what is presented ? That which in normal perception made the image an outside content is not the correctness of the perception, but only there being to that perception a presented content. Inilluion too there is a presented content is not, it is true, eveident in its full character. But there is no denying the fact that there is a presented content. The represented content, again, is not a real adjective of the presented element. Nevertheless it is an adjective, though false, false in the sense fo being really unconnected or misconnected. Alike in normal perception and illusion the image-element is outside. Imagination may be directed to a past thingor a given presentation, or to no thing whatever. When directed to a past thing, the insideness of the image is more evident than its outsideness. The thing no doubt is remembered,but as imagination has added nothing to the thing-as-it-was-perceived no special outsideness of the image is evident. What is evident ofn the other hand is that to a given presentation, however, the outsideness alone of the image is evident: the imagestands tied, it is said, to the presentation. The insideness of the image have amtured at all into a subjective image. The same thing occurs in illusion; only, here the image is not directed to anything-past or present-it is ever on the vanishing point and is kept steady,even as so vanishing, by words. In this case-we may call it idea, as distict from the two previous types of image-it stands evident as merely inside. The outsideness of the image is complete in the second case only. The complete outsideness in the second case and the much less outsideness in the first are equally due to the reference of the imagination to real things outside. The Buddhists under consideration have been deluded by the theoretical insideness of the image. They have not seen that except in the third case above there is also its outsideness, evident in its fulness as much in perception as presentative-representative as in illusion. The idealistic account of the false content is thus untenable. The false content has to be taken as non-subjective, i.e., an object, though it may not be a real object. All other Indian theories of error and the modern realistic theories of Alexanderand other realists agree in this point. In spite of this general agreement, however, they differ in some fundamentals, each having understood the concepts of object, reality and their realtion in a different way. These theories should be examined separately. H. Some modern realistic theories of error examined Some modern realists believe that an object as such is neither real nor unreal and that the reality of a normal object and the unreality of aone called illusory are equally unmetaphysical, being only contingent derivative characters. But this is over-simplication in various ways. Let us see, how. (1) An object that is rejected may be provisionally granted as subsisting on its own account and having unreality as a contingent derivative character. But the object of a normal cognition is never felt as subsiting aloof from reality. If is felt from the beginning to the end as absolutely coincident with the real-in other words, as unqualifiedly real. It is only where there is no assertion, where a content is merely entertained that one may say it subsists. But such content is in the face of it an abstraction, and actually felt that way. EVen doubt, question and suggestion are more or less assertive. In doubt and question there is still assertion, though it is either midway between or alternation of affirmation and denial, or the assertion here is vague and incomplete. It cannot be said that in doubt and question there is neither affirmation nor denial. Suggestion also is not without all assertion. Suggestion is the mere entertainment of a content-as-asserted. In all other types of cognition, except in error corrected, there is unambiguous affirmation or denial, though in the affirmation of one content there may remian involved (and subordinated) the denial of another content, and vice versa. In such cases the content is not felt the same content can be asserted, suggested, questioned, doubted, merely entertained or even rejected is enough to make one feel that it is at least dissociable, if not dissociate, from relaity. The content that is simply entertained is abstract and symbolic,but a content asserted is felt neither as that abstract one plus its assertedness nor as symbolic plus something else. No concrete can be broken up adequately into (several abstract features or)an abstract feature and a dark solod base. A cow is not analysable into cowhood and an indenifite solid base; that base is itself also a particular cow. Had not the base had a definite [svarupa] the universal cowhood could not be connected with it to the preference of any other particular, say, one to which doghood or horsehood belongs. The content asserted is, againg, real, and no reality is constituted by a bare symbolic possibility and something else. Possiblity may at the most be the essence of the reall and a corresponding possible there is nothing that is explicity common. If Y be a modification of X, it is X in another form, not X and another form, far less, therefore X and a dark ground. The relation of an asserted content C to a C that is merely entertained is true [mutadis mutandis] of its realtion to C's that are doubted, questioned and suggested. To all these attitudes there is necver the self-same content except in name, and even that name C is not the content of simple entertainment. Only the content of correction is absolutely the same as what was asserted. But of that later. The realists under consideration might still argue that as we ourselves have shown through all these pages that object is different from the real we ought not to take exception to their view. Should not object as distinct from the real be taken as neither existent nor non-existent ? We reply, we hold also that object is yet felt as coincident with the real, i.e., as itself the real. We have also shown that there is no reasong why one of these two apprehensions is to be preferred and the other rejected. To have preferred their distinction to the extent of rejecting their identity has been the over-simplification No.I of which these realsits are guilty. There are other acts of over-simplification also. (2) They have understood the illusory content too hastily. True, when error is corrected we come to doubt if the content was definitely either existent or non-existent. But this 'not definitely either existent or non-existent' does not amount to 'neither existent nor non-existent'. There is no evidence yet, nor even a reasonable suggestion, that it was definitely neither. The only case where there is definite absence of either is simple entertainment where the content is admittedly abstract; but the content of error even after correction, does not appear to be abstract. No one feels that the contents of error should be taken as having been merely supposed or simply entertained. It need not be denied that the content is not felt as definitely either existent or non-existent, but that does not mean that it is definitely neither. It is still asserted, though neither as existent nor as non-existent. The content corrected is still asserted in the sense that it is known as a sort of appearance of the real that is discovered in correction. After correction it is not felt as floating in the air. It is felt even then as somehow tagged to the real, not a self-subsistent content having nothing to do with the real. The question of unreality of that content at all arises only becausee ther e is such tagging: this appearance of the real is not a real appearance. (3) These modern realists are guily or yet another over-simplification. By treating object as such as neither real nor unreal and interpreting reality and unreality as equally pragmatic or linguistic or anything else they have missed a notable featureof the unreal object. In whatever is that which was once apprehended as real. If it were not understood as 'once apprehended as real, but now rejected,' even abstract contents (including even the neutral contents of these realists) would have to be called unreal. The central problem of erro ris how content can be both objective and unreal. If the denial of objectivity, as by the [Vijnanavadi Buddhist], has been too easy, so has been the attempt to treat reality and unreality as only extrinsic to the content. I. [Nyaya-Vai'sesika] theory of error The illusory content as both object and unreal could not be a problem at all if object in normal perception were not wholly coincident with the real. Object to be so coincident with the real is not merely what just happens when a perception is not erroneous. It follows, we have seen, from a fundametal postulate of knowledge, at least of perception. The problem, then, is this :-How can the same object be real and unreal at teh same time ? The problem can be formulated in another way. In course of examining the modern realstic theory of error we have shown that the content rejected is, even after correction, asserted, though neither definitely as existent nor definitely as real. How can the rejected content be yet an appearance of the real ? A can be taken as an appearance of B if between them there runs a bond of identity. But how can there be a bond of identityu between the false and the real when the false is definitely rejected as unreal ? [Nyaya-Vai'sesika] solves the problem characteristically in close touch with objective common sense. It holds that though prior to correction there was the awareness of a total object 'this snake' or 'this is snake,' correction of it entails that this awareness was wrong, another name of which is that the total content is unreal. Yet, however, the awarencess of it was [savikalpa-Pratyaksa], which implies that some reals (apprehended in nirvikalpa-Pratyaksa) were related into a unity by [vikalpas] which are knowledge-wise. The reals in the present case were this and snake, for nothing else could be related into the unity 'this is snake.' The this here was but the real rope perceived as mere this. Its [rope-svarupa] was not perceived, in other words, just a given substratum, no [svarupa] of it. The other real was snake, but not this snake or that snake. Not this snake, because there was no snake presented. Nor, again, that snake, i.e., a snake, of the past remembered in relative fullness as the snake there and then, for that snake could not be combined with a this substratum. What could be so combined is just snake (sarpamatra). Some past snake is no doubt remembered, for otherwise there could not be a question of snake at all; but it is not remebered as that snake. Only the [snake- svarupa] is remebered. As any past snake is real, so is also the [snake-svarupa (sarpamatra)] which is only a part of it. This [snake- svarupa] came to be combined with a this into the [savikalpa] unity 'this is snake' through a peculiar psychological mechanism, viz., that the very memory of the [snake-svarupa] acted as the contact between the sense and the real substratum. This psychological mechanism does not concern us for the present. Teh elements thsi and snake are real. The [vikalpa] relation that combined them into a unity is also real; this follows from the fundamental postualte of knowledge already mentioned. But unlike the elements and the vikalpa relation, the unity formed is not real. In correction this unity stands rejected. This last is the intriguing feature of illusion. Normally when the elements and the vikalpa relation are real teh unity effected stands also as real. The present case is an exception, only because the unity has been rejected in correction. Not that I was not aware of the unity before correction, nor that as an object then it was not apprehended as real. But correction contradicts just this prior awareness and therefore sublates this object. It follows that once it is sublated it cannot be taken to have been real even before. But if it cannot be said to have been real, how we say that it was yet an object ? Does not the reality of every object follow from the very fundamental postualte of knowledge ? The [Nyaya-Vai'sesika] reply is that it cannot be said to have been an object even. It was indeed felt as an object, but as a matter of fact it was not an object. Not that it was therefore wholly subjective. This idealsitic theory has been already refuted. Moreever, if the elements are real outside their unity cannot bemerely subjective. It cannot be said, again, that though the elements are real by themselves they yet as in the unity must partake of the nature of that unity. Here there is no question of the elements in the unity: in the unity there are no elements,there is only the unity, and nothing else, the elements being only inferred as having been apprehended in a prior [nirvikalpa] knowledge. The unity in question is neither merely subjective nor an object coincident with the real. Not that as neither subjective nor such object it is the neutral object of the modern reasists. Such neutral objects we have already dismissed. [Nyaya-Vai'sesika] is forced to conclude that after correction there is no talk of such unitary object. Though prior to correction some such unity appeared, correction is just its sublation. What is meant is this: After correction we cannot say 'This snake is (was) not'. Such negative judgment is impossible. EVery judgment, affirmative or negative, is possible if at least the subject is already known as real. 'A table is not in the room' presupposes that there is a table in the world (though not in the room). But before we are entitled to say 'This snake is not' we are already assured that this snake has been sublated. So there is no occasion to use 'this snake' as the subject of a judgment. It will be no use arguing that though the present this snake is sublated there were other this-snakes at other times. 'This refers primarily to one unique particular, one that is presented just here and now, and in comparison with it the use of the word 'this' as characterising other things which were so presented is abstract and symbolic, not a genuine living use. Whatever else may be called this, the primary and living use of the word is regarding a very unique particular entity. This snake is the very particular unique snake that was here taken as a real object and is now sublated in correction. 'This snake', so understood, cannot be the subject of a judgment, affirmative or negative. The negation of thsi snake, so understood, would be a case of [aprasaktapratisedha]. If this snake cannot be denied now, it cannot also be taken, from the point of view of correction, as what was affirmed before correction. From the point of view of correction, then, this snake was not an object. But do we not yet, even from the point of view of correction, say 'This snake was not' or 'This snake was apprehended as object', and do we not mean something by that ? [Nyaya-Vai'sesika] would claim that this is an unjustified use of language. We docertainly mean something, we mean that thsi snake is sublated. But sublation is not adequately representable in the form of a negative judgment. Sublation may include, imply or entail a negative judgment, but it is more than that. Even the negative judgment that is included, implied or entailed is not of the form ['this snake is not'], it is of the form 'no snake is here (in this), where the subject is not [aprasakta]. It follows that the unity effected out of this and snake through a vikalpa relation is not 'this is snake' or 'this snake', but 'sanke is in this' or 'sanke in this ', not even 'sanke as in this' (for in the statement 'snake as in this is not' the subject would be equally aprasakta). We have remarked earlier that though in normal cases 'P is in S' is translatable as 'P is as in S' this si not possible here. The unity effected here is loose, not a close one like 'this is snake'. It may even be said that this unity is little more than nominal. 'In S' in the judgment 'P is in S' does not characterise and is not, therefore, predicable, in any normal sense of predication, of P. The content 'snake in this' is not a unity except in name. What is apprehended here in [savikalpa-Pratyaksa] is the very reals snake, this and inness, and nothing else. The factual relation is here itself the vikalpa relation. Such is also the case with the content "hare's horn" which is rejected in the statement "hare'shorn is not". What is negated here is not truly "hare's horn", but 'horn in the hare'. Such interpretation in either case may appear circuitous. But it is inevitable, because otherwise there would be the impossible situation that a content-'this as snake' or "hare's horn"-is both rejected and yet a real object: If only a content is interpreted this way the difficulty would be removed : there would be an easy reconciliation of the rejection of a content with its being a real object. Because there was no genuine untiy of the form 'this snake' ='this is snake' [Nyaya-Vai'sesika] takes this-aspect as belonging to thsi rope, not to that apparent this snake. The snake that was real was not a this snake, it was merely snake; the rope alone was thsi rope, or, better, the rope was perceived (barely) as this. this, only because they believed that there was a total false object of the form 'this is snake'. Why they hold this and how far they are justified will be seen in connection withour discussion of the [Advaita] theory of error later. [Nyaya-Vai'sesika], for reasons we have seen, cannot subscribe to this view. According to [Nyaya-Vai'sesika], the this-aspect does not really belong to the apparent content 'this snake'. This does not, however, mean that the business of correction is only to drop the this-aspect and retain the mere sanke. It is only the [Vijnanavadi Buddhists] who argued that way and concluded that beacuse 'this' means 'to be now outside me' correction presents the illusory content as not so outside, and, therefore, as subjective. The [Vijnanavadin's] view has been dismissed already. [Nyaya-Vai'sesika] may add that correction does not drop this-aspect ; it only cancles [samanadhikaranya] of thsi and snake and presents the illusory content as 'snake in this'. J. Alexander's theory of error examined Alexander's theory, though largely in tune with the [Nyaya-Vai'sesika], differs from it in an important respect. Like the [Nyaya-Vai'sesika] thinkers, and practically on the same ground as theirs, he too holds that this and snake are each real, and asthe snake-as-here ('this snake') is rejected it must be a snake elsewhere. Error lies, according to him, in mis-connecting the elsewhere snake with a sensed this. But the main point, viz, about the exact status of the illusory content 'this snake', heleft untouched. He draws no distinction between object and the real, except admitting that this snake is a false appearance of the sensed rope and that the falsity of the appearance is due to the content being a joint appearance of the rope, on the onehand, and the percipient mind (or the physiological organism), on the other. If by this he means that the appearance is of the rope and yet consituted in whatever way by the mind, it would be what we have so long been terming object. But probably he does not mean this. He understands it as in line with his 'mere appearance ' where the constitutive factors are all physical. His 'mere appearance' is not object in out sense; and in false appearance it is, as appears from what he says, an accident thatone of the constitutive factors is the mind. Even as regards his 'real appearance', there is no contribution of the mind. By 'appearance' he only means a selected portion of reality. But in his doctrine of selection he errs in two ways. In the case of real appearance he has shown that the content of a perceptual knowledge is a portion of the reality-con-tinuum, knowledge being nothing but a selective response. But he does not show how 'mere appearance' is a selection. The factors constituting it are admittedly not selected from the reality-continuum, but neither so is the content called 'mere appearance'. The factors, again, are parts of the reality-continuum, though not selected ; but that content is nto even a part. It would be too much tocontend that the oval shape of a round coin is a part of the reality. It depends on the position of the percipient's body vis a vis the round coin. If it be contended that the round shape too depends on the position of the body, the conclusion shouldrather be that every appearance-real or mere (and a fortiori the unreal also)- depends on the subject and is, therefore, object in our sense. There is no ground to overlook this dependence in either case. As a matter of fact, even the real appearace depends on selection by the mind-depends, but for the content bring an appearance at all. This is not to be tabooed immediately as involving ego-centric predicament. We never deny that though the appearance so depends there is nevertheless an independent reality as the background, and we perceive not merely the appearance but also that reality. If Alexander wants to avoid thsi conclusion the only course left to him would be to hold that there is no appearance at all, but that knowledge as diaphanous directly reveals the real. But, then, there should be no talk of selection in the sense in which Alexander understands the term. If reality were a continuum selection whould change it into a definite discrete protion, and knowledge would not be diaphanous. If, however, reality were not a continuum, but a series of discretes, knowledge would indeed be diaphanous, and the word 'selection' might be used in the ordinary sense of the mind being directly in contact with one entire metaphysical structure which Alexander had built before he turned to epistemological problems. This is his second error. As for the concept of diaphanous knoweldge directly referring to definite discete reals, we have already seen its defects in Sec. I. Here we may add one more point. If knowledge were diaphanous, directly, in contact with definite discrete reals, how would perception, memory, inference, etc., be distinguished from one another ? We must say that either these cognitions are qualitatively distinct or their contents have perceivedness in one case, rememberedness in another, inferredness in a third, and so on, these being emergent differential characters of the contents themselves. But on the former alternative knowledge would no longer be diaphanous, and the second alternative would inevitably lead to a distinction between reality and object, that whcih has perceivedness, rememberedness, etc., being a real, and that reality as with the perceivedeness or rememberedness, etc., being objects. If it be contended that the qualitative differeence of types of cognition odes not militate against being diaphanous-each such type directly referring to the real-we would ask: Does this nto merely prove that there is a real (with such and such cahracters) ? From where, then, does the consciousness of that reality as object come ? It cannot be said that object is another name for there being a real. The real was there even before I knew it. Nor can it be said that object is only another name for that real being known, for while the 'real being known' is known as object even in the primary experience, commonly called introspection, the real is known as object even in the primary experience. Knowledge as diaphanous cannot explain this primary knowledge of a real as object. The much maligned representationism is in thsi point a better account than directrealism. The only defect-though that is serious- or repressentationism is that it has very sharply distinguished object and reality to the extreme point of their separation. They, we have so long been noting, are not separate. Except in erroneous perception, object cannot be dissociated from reality. Objectivity is a character accruing to the real and is itself, on that very account, believed as real. To put the matter more succinctly, object, except in false perception coincides with the real. K. [Prabhakara theory of error examined like the object of any normal perception, the false snake has to be taken as object,though it does not coincide with the real. But this non-coincidence, we have seen, is an anomalous phenomenon. [Nyaya-Vai'sesika] tried to remove the anomaly by recognising the consituents only of 'this snake' as real. This, snake and the [vikalpa] relation are alone, according to them, real; the total content is not real an, therefore, no unitary object even. The [Prabhakaras] have proceeded another way. They stick more closely to the basic doctrine that object (at least in perception) must coincide with reality. Object, everywhere, is nothing but a real as revealed by knowledge, objectivity being only the character of being so revealed. Because this character must belong to a real that is so revealed, there obviously cannot be an object in default of that real. Except in cases where a content is false or self-condtradictory, the [Naiyayika] has also held this view; he has excepted the false or the self-contradictory only because it has come to be rejected. He has rather been compelled to except it. But the [Prabhakaras] would argue that there is no such compulsion. There is another alternative: we may deny that the content has at all been rejected. The [Prabhakaras] would argue as follows: If once it is established that object is but a real as revealed by cognition it would be senseless to modify the position to teh absurd extent that there may be object even though it is not real. The false content is, of course, a challenge to thsi notion of object: it appears to be rejected in correction. But would it not be better, the [Prabhakaras] argue, to re-assess the correction-situation to see if that rejection is not only apparent, nothing serious, than abondoning the definition or objectalready established ? The [Prabhakaras] contend that in correction there is as a matter of fact no rejection. Rejection is always of a content which was known, i.e., taken as real object. But as in correction we come to know that the false content was not a real object, this means that it was not a known object. What reflection certifies is the true nature of a thing. Correction as reflection certifies that there was no cognitive object. Hence truly there was no cognitive object. [Nyaya-Vai'sesika] thinkers have also admitted this logic, though only partially. They too have contended that because in correction we come to know that there was no object in the form 'this snake' there really was no such object. But immediately after this, and uncritically enough, they have yet held that somehow they were aware of the object 'this snake' before correction. It is because of this their uncritical faith that they spoke of correction as the rejection (badha) of the content. The [Prabhakaras], on the other hand, hold that no content-not even the total content 'this snake'-is rejected. If at all anything is rejected it is only the knowness, the cognitive character, of the total content, the content remaining untouched. But even this cognitive character is not rejected. Rejection of it would imply that before correction the content 'this snake' was apprehended as a cognitive object. Correction certifies this much only that there was no cognitive object like 'this snake'. A cognitive object is ipso facto real (Paramarthika). Correction certifies only that it was not cognitive, but conative (vyavaharika). Hence even before correction we were aware of it as only a conative unity. This and snake were, however, cognitive and, therefore, real object; the question here is not about them, but about the content 'this snake' ='this snake'. But how is it, it may be asked, that when this and snake were known as real objects the total content 'this is snake' was nto a cognitive object ? The [Prabhakaras] reply that teh so-called total situation was, from the cognitive point of view, a sheer privation : we only did nto distinguish the two cognitions-one of this and the other of snake; or, better, the two cognitions remained undistinguished, and the so-called unitary content, cognitively speaking, is only their non-distinction. True, we acted according to this so-called total content, we fled when we sae 'this snake'. But the [Prabhakaras] argue that though such positive unitary content and positive awareness have to be admitted the unity and the awareness are not cognitive. 'This is snake' is, in other words, no object. It is either what is only referred to by conation or a mere verbal unity. Two things non-distinguished are often taken as one unity in the context of an act. It is the act which treats tham as though they are unified. Act or will is normally indeed a response to a cognitive unity. But even in every such normal act there are contents which are cognised as non-distinct and yet unified by that act. What is called object of will is primarily the object of the cognition that causes the will; but in every will there is inevitably reference also to the means and a purpose which do not stand cognised as related to that object or to one another. By 'purpose' here is meant the actualisation (bhavana) of the object. The object of will was cognised as only a future reality, but there was no cognition of it as to be actualised. Futurity of the object was no doubt cognised, but it means only future actuality, not the dynamic to be actualised which is a peculiar unification, through will only, of the object and its futurity. X, Y, Z which are means to that actualisation were also cognised, but not as means. Their means hood (upayata) is another peculiar unification, by will only, of X, Y, Z with that object of will. They might have been cognised as causes, but not as means. Means-hodd and purpose-hood are absolutely conativecategories. Action alone thus unifies contents which are cognised as non-distinct, e.i., unrelated to one another, relation necessarily presupposing that relata are known as distinct from one another. We have seen that the contents of [nirvikalpa-Pratyaksa] in [Nyaya-Vai'sesika] were also known as non-distinct from one another, 'non-distinct' meaning here, as also in the [Prabhakara] view under discussion, not that the contents are each known with its self-identity, but that they are not known as each being not another or each dissociated from another. We have also seen how in [savikalpa-Pratyaksa] these non-distinct contents get related to one another and turn into a unity (though these relations and that unity are not merely subjective). Such unity is cognitive. The [Prabhakaras] only contend that there is also anothere type of unity which, as described above, is conative. The conative unity is called by them [vyavaharika]. As in normal will, so also in illusion the unity 'this snake' is [vyavaharika] only. This and snake get unified in the context of act only. There is another possible account of the positive unity of the illusory content, and some [Prabhakaras] have admitted that. It is that the unity is only verbal. In a sense the [Naiyayikas] also regard the unity, not only here but even in normal [savikalpa-Pratyaksa], as verbal. Every [savikalpa-Pratyaksa] is, according to them, ['sabdanubiddha]. [Vikalpa] relations are necessarily semantic forms of language, forms of language as judgement. The language that is is dissociated from of language spoken, not heard, language that is spoken being, as spoken, undissociable from knowledge as judgment. The language that is dissociated from knowledge as judgment is the language which is heard, such language as heard being taken as a system of sounds or marks producing in the hearer another judgmental knowledge which, however, is not then spoken by the hearer implicitly or explicitly. The unity, thus, not merely in illusion but in every case of [savikalpa-Pratyaksa] is , according to the [Naiyayika], verbal. But the [Naiyaika] has not refrained from saying that as much in illusion as in every case of [savikalpa-pratyaksa] it is also read and, therefore, an object. These [Prabhakaras], however, here part company. They agree with the [Naiyaikas] that in normal [savikalpa-pratyaksa] the unity is an object and would even go farther and hold that there is such unity as object even in [nirvikalpa-Pratyaksa] where it remains in some latent form. But they would entirely disagree with them so far as the content of erroneous perception is concerned. The unitary content is in this case merely verbal, not real. Whichever way 'this snake' is interpreted-whether as non-distinction of thsi and snake or as a conative or a merely verbal unity of these there is no question of its rejection. What may be said to be rejected is only the positive cognitive character of 'this snake'. But, as already seen, even this is not rejected, we only deny it. Even before correction 'this snake' was not apprehended as a possitive cognitive object. The [Prabhakara] view is in perfect consonance with the doctrine that every cognitive object is real. But its weakness also is evident, and the weakness is fundamental. The [Prabhakara] contention that even the cognitive character of the object 'this snake' is not rejected, but only negated, does not appear to be a sound account of the business of correction. Whatever be the [Prabhakara] theory, we do feel that before correction we were aware fo 'this snake' as a cognitive object. It is too much to claim that we were aware of it as a conative or only a verbal object or as this and snake non-distinguished. The [Prabhakaras] were right in claiming that reflection offers a true account of the nature of the thing reflected on. But this does not mean that even before reflection we were aware of the thing in that correct way. Often the reflective account appears, without any hitch, as contradicting and often, again, as rejecting the unreflective account. There is no good reason why the [Prabhakaras] should discount the second contingency. Rejection (badha) is often an actual phenomenon, and it is no good fighting shy of it. But once we admit rejection it would