E P I S T E M O L O G I C A L N E G A T I V E D I A L E C T I C S OF I N D I A N L O G I C - ABH_~VA VERSUS ANUPALABDHI by DHIRENDRA SHARMA

Michigan State University

The problem of negation is primarily an epistemological one: H o w do we know the absence of a thing? On this question the indian LogiO is divided into two main groups: the Realist and the Idealist. The former consists of the systems in which the absence (abh6va) is conceived as a real non-entity (paddrtha) and as such is a real object of its corresponding negative cognition (abhdvadhi). In spite of certain differences the Ny~yaVaige.sika and the Bh~.tta-Mim~ .msg systems belong to this group. The second group comprises of the later Buddhist logicians led by Dharmakirti and the Pr~bhgkara-Mim~.ms~ thinkers who reject the objectivity ascribed to 'abh~va' by the Realists. According to the idealists negation is an inferential judgment and as such the cognition of absence of a thing is only a logical synthesis (vikalpa). Absence of a perceptible thing (dr@a) is inferred from its non-perception (anupalabdhi) and from the perception of something else, namely, the bare locus (bhf~talamdtra). The suggestion of the perception or the presence of the thing negated remains as an imposed ideal situation (d.rdyatvabuddhau samaropdt). N o w the question is: Can Negation be an independent means of knowledge (pramdn. a)?~ The view that it is an independent means seems to be very old. According to the Bhd.sya of Pra~astapgda the Vaige.sika Satra (IX.i.5.) rejects the view of the negative means: In the Nydya t The present investigation will be limited to the Buddhist and the Bh~ta-Mimar0s~ systems of Indian Philosophy. 2 The term 'pramh.na' is not well defined in Indian logical writings. It is used in the sense of either (a) the means of knowledge, or (b) the form of cognition, or (c) the means of proof. (Cf. Ganganath Jha, SadholalLectures, p. 28.) Here the term will be used in the first sense. It should also be noted that the term 'Negation' will be used to express Kumfirila Bha~a's theory of 'abhava-pramaoa', 'Non-opprehension' for the later BhaIIa's theory of 'yogyanupalabdhi', and 'Non-perception' for the Buddhist theory 'd.r~y~mupalabdhi'. However, in quotations of modern writers they may occur as synonyms. 8 Padarthadharma-sa~graha,ed. Vindhyeshvariprasad Dvivedin (with Ny~yakandali of ~ridhara) (Vizianagram Sanskrit Series) (Banares, 1895), p. 225.

292

DHIRENDRA SHARMA

Sf~tra (II.ii.2) there is refutation of negation as the means. Subsequently, with the exception of the PQrva-Mim~ .ms~ all schools of Indian logic rejected this theory. ~ II The theory of the negative means evidently belongs to the POrvaMim~ .ms~ system alone. This system accepts the six means of knowledge: (1) Perception (pratyak.sa), (2) Inference (anumdna), (3) Analogy (upamdna), (4) Verbal Testimony (gabda), (5) Presumption (arthdpatti), and (6) Negation (abhdva). 5 The Nygya-Vaige.sika admits only the first four while the Pr~bh~kara-Mimfi .ms~ rejects the last one, namely, the Negation. ~ The Buddhist logicians accept only the first two pram~n.as. 7 It is evident from the early texts of the POrva-Mim~m. s~ that it was this school of thought alone that propounded Negation as a praman, a. gabarasvhmin commenting on the Jaimini-Satra (I.i.5.) said that "Negation stands for the non-existence (or the non-operation) of the (other five) means of Knowledge; and it is what brings about the cognition that '... does not exist' in regard to things that are not in contact with the senses. ''s Sabarasvgmin appears to have endorsed the view of a Vrttik~ra, most probably Upavar.sa, who had postulated the six means in the system before ~abara. 9 Kumgrila Bhat.t.a in his exposition of the Sabara Bhg.sya affirms: "Validity (pramd~atd) of Negation (as the means of knowledge) is to apprehend the fact (sattd) of the (negative) entity (vastu) where the five (positive) means of knowledge in case of cognising the (negative) form of reality (vastu-rftpe) fail. ''~~ III Now the problem arises: How do we determine the validity of a negative 4 cf. A. B. Keith, Indian Logic & Atomism (Oxford, 1921), pp. 53-57. Nloka-vdrttika (henceafter ~l.V.) of Kumftrila Bhat~a, ed., with the commen. 'Nyftyaratnftkara' of Parthasfirathi Migra by Ram Shastri Tailanga (= Chowkhamba Sanskrit Series) (Banares, 1898), Codanft-sfitra, verse 111, p. 60. Cf, M. Hiriyanna, Essentials oflndian Philosophy (London, Allen & Unwin, 1949), pp. 99-100, 143. r Ny~ya-praveia of Dign~ga, Pt. I, ed. with notes by Anandshankar B. Dhruva (Baroda, Oriental Institute, 1930), p. 7. s ,~abara-Bhfi.sya (=Bibliotheca Indica) (Calcutta 1873), p. 10: abhavo'pi prfim~.ny~bh~vo n~stityasy~rthasy~sannik.rst,asyeti. a Cf. Damodarvishnu Garge, Citations in ~abara-Bh~sya, p. 11. lo ~l. IT., p. 473 (Abh~va ch. vers. 1-2): prama.napa~caka.m yatra vasturfipe na jayate. vastasattfivabodhdrtha~ tatrfibh~vapramd.natfi.
5

abhdva VERSUSanupalabdhi

293

judgment and what is the ground of its validity? According to the early Ny~ya and Vaige.sika logicians the validity of the negative judgment is established through Inference (anumdna). n The later Ny~ya-Vai~e.sika writers, most probably following Uddyotakara, invented an unique sense-object-contact called 'vi~e.sa.natg-sannikar.sa' and on its basis pleaded perceptibility for the negative judgment. TM Here we do not intend to go into the details of this theory of negative Perception. For the theory is of little significance, and the main contestants are the BMt.t.as and the Buddhists. The Bhh.t.tas maintain that the negative cognition is such that it cannot be a perceptual judgment. The cognition of an object is of two kinds: (a) positive and (b) negative. For example, the cognition of a cow is considered to be positive in relation to its own nature such as "This is cow." But the cognition of the cow in relation to a horse is regarded to be a negative one such as "This is not horse." This cognition 'not horse' is such that it is not derived from the sense-object-contact or from perception. For there is no positive entity called not horse which would come into contact with the senses and without the contact there can be no perception. Thus, Kumhrila Bhat..ta contends that the negative cognition "This is not . . . " cannot be brought about by Perception. The senses are capable of having contact only with the positive forms of reality. 18

IV Now the question arises: Can Inference (anumdna) be valid in the case of a negative cognition? Kum~rila maintains that Inference cannot. For an inferential judgment is possible only in the case where we can determine a logical mark or reason (li~ga = hetu). Since a logical mark is recognised only when the invariable concommitance of the universal relation (vydpti) between the mark or the middle term and the major or sddhya has been established - as is the case with smoke and fire - in the case of a negative object of cognition no logical mark of the universal relation is
11 Nyaya-S~tra, II. ii.2; Vaige.sika-Sfttra, 1X.i.5. 1~ Cf. Nyaya-Varttika of Uddyotakara, ed. Vindhyeshvariprasad Dvivedin (= Kashi Sanskrit Series) (Banares, 1915), on I.i.4., p. 31. Uddyotakara, I believe, is the first Naiyhyika who propounded a systematic formula of determining a six-fold senseobject-contract with negative object. He is followed by ~ridhara and others of the Ny~ya-Vaige~ikasystem. x~ ~I. V., Ch. Abhfiva, verse 17: pratyak.sOdyavatSras tu bhav~.mgo grhyate yadS. vyaparas-tad-anutpattir abhavar~Aejighrk.site.

294

DHIRENDRA SHARMA

possible. 14 i f it were possible we would infer all the instances o f negation by cognising simply one instance o f absence as we can apprehend all the instances o f smoke-fire relation by knowing one instance o f that fact. Furthermore, in the case of Universal-absence no valid means (Inference) is possible. Then, since the relation of the hetu and sftdhya is not established (vyabhicara), how can the particular instances of (absence) be known by that (Inference) Y5 Moreover, Kum~rila further argues, the three characters o f the logical m a r k la c a n n o t be established o f such an entity absence o f which has yet never been apprehended. Thus, for example, the cognition o f antecedent absence o f X, that is, the absence o f X prior to its coming into existence (prggabhava), cannot possibly be determined by inference. The inferential cognition is contended to be that cognition which is derived from the three-fold reason (trilak~al~a). But in the case of (the cognition of the absence of) the form of the Antecedent (an-utpattirQpasya), no cause is found anywhere (which can become the logical mark of the negative inference - as is the case in smoke-fire relation)3 ~ By perceiving the effect (smoke) we infer the presence of the cause (fire) at the given time and place. But with reference to the Antecedent-absence that is never p r o d u c e d by any cause, ~a inference c a n n o t be the m e a n s of knowledge. Further Kum~rila reasons in a naive realistic tone:
x4 Ibid., verse 29: 'na capyasyanumanatva.m li.ngabhavat pratiyate'. For a detailed exposition of Universal relation (vyapti) the reader is referred to Karl H. Potter's scholarly work Presuppositions of India's Philosophies (= Prentice-Hall Philosophy Series) (Englewood Cliffs, N.J., 1963), pp. 59-74. z~ Ibid., verse 39: na c@yabh6va-samanye pram~n, am upajayate, vyabhicarad viges.6s tu prat~yeran katha.m tay~. zs The three characters of the mark ([email protected] lihgasya) are (1) the existence only (never non-existence) in the Subject or thing denoted by the minor term; (2) existence in things which resemble the Subject only (never in things which do not resemble the Subject, i.e., in vipak.sas); (3) only non-existence (never existence) in things which do not resemble the Subject. Cp. H. N. Randle, Indian Logic in The Early Schools (Oxford, 1930), p. 181. z~ ,~I.V., op. cit., verse 44: trilak.sa.nena y~ buddhir janyate s~nume~yate, na c~nutpattir@asya [email protected] kvacit. a Kumftrila Bhat.~a postulates four-fold negative entity: antecedent or prior-absence (prag-abhava), posterior-absence (pradhvazns~bhava), mutual-absence (anyony~bhava) and absolute-absence (atyant6bhava). 1bid., verses 2-4. The antecedent absence is that which exists prior to the creation of the thing. Hence this absence has no beginning and no cause of its creation. However it has an end when it is destroyed by the creation of the thing. Cf. D. H. H. IngaUs, Materials for the Study of Navya-Nyaya Logic (=Harvard Oriental Series, Vol. 40) (1951), p. 54 (27). Also Nyaya Satra, II, ii, 12; Vai~e~ika S~tra, IX. i. 1. Nyayalflavat~, pp. 544-579.

abhdva VERSUS anupalabdhi

295

Just as Negation cannot be the means (pram~n. a) where the object of knowledge is a positive one, so also in the case of a negative object nothing positive can be the means of knowledge. 1~ The Mimg .msaka here seems to be unaware of the Method of Difference and the Principle of double negation where positive facts are cognised by applying negative methods. 2~

The strongest opposition to the Mim~ .msakas' view of Negation came from the Buddhist logicians led by Dharmakirti. To the argument of the Mima .msakas that the non-operation of all the positive means of knowledge, being real, ipso f a c t o becomes an independent means of negative knowledge, the Buddhist objects: H o w can it be said that the absence of knowledge is a self-established fact, and that the absence of the the object is determined by the absence of the means of knowledge? For as the presence of the means of knowledge, say, Perception requires no other means for its establishment - it is self-evident - so is the absence of the means also a self-evident fact. For instance, on the basis of the senseperception we know that the book is on the table. Similarly, when the b o o k is not present and therefore no sense-perception is produced in the intellect, the absence of the cognition of the b o o k is known by the same intellect by which we know when it is present before us. For if it were really an absence of all means of knowledge it would require some other means to establish its validity and that would lead to an infinite regress. ~1 Dharmakirti postulates three logical m a r k s (hetu), (1) Causation (kdrya), (2) Identity (svabhava) and (3) Non-perception (anupalabdhi), as the basis of all inferential judgments. The first two are to establish the knowledge of real (positive) things and the third one, Non-perception is the reason of all negation (prati.sedhahetu.h). ~2
19 ,~l. v., op. cit., verse 46: bhavatmake tatha (read: yathd) meye n~bhavasya prama.nata. tathdbhavaprameya'pi na bhavasya pram~.nata. 20 For the Method of Difference cp. Irving M. Copi, Introduction to Logic, 2nd ed.

(New York, Macmil/an Co., 1961), pp. 368ff. 31 Vide, I(ar.nakagomin's commentary on Prama.navarttika-svav.rtti of Dharmakirti [hereinafter PVS], ed. Rahula Sankrtygyana (Allahabad, 1943), p. 30: kevala.m yadi
svasantane ] ~ n a .msyad upalabhyetanupalambhad asad eva tad iti svata eva ]~anabhavaO siddha i.syate. 2~ PVS, ed. Raniero Gnoli (Rome, Instituto Italiano Per I1 Medio ed Estremo Oriente, 1960), p. 2: ta eta karyasvabhavanupalabdhi-lak.sa.nas trayo hetava.h. . . . tatra dvau vastusadhanav ekah prati.sedhahetu.h.

296

DHIRENDRASHARMA

Dharmakirti further explains the formula of Non-perception in the following: (Thesis): On some particular place there is no jar. (Reason): Because it is not perceived, although the conditions (lak.saea) of perception are fulfilled. If it were present it would have been perceived - as it cannot be otherwise. 23 Thus the Non-perception (anupalabdhi) becomes the reason of negation with reference to the object capable of being perceived (yuktopalambha). ~4 In this theory, non-perception of a non-perceptible (ad.rgya) is merely problematic (sam. gayahetu). ~ This definition of the negative reason (hetu) provides the fundamental principle governing all possible kinds of formulations of negative judgments. 2~

VI It has been stated above that Kum~rila Bhat.t.a, in his allegiance to the Pflrva Mim~m.s~ tradition, postulated Negation (abhdva) as an independent means (abhdvapramdn. a) of cognising negative facts. Also it has been demonstrated clearly that the Buddhist logicians led by Dharmakirti admitted negation but as an inferential judgment (d.rgydnupalabdhi) These two theories are diametrically opposed to each other, and their distinction is marked by the two distinct terms by which the logicians referred to their respective theories - abhdva and anupalabdhi. It is pertinent to point out here that the use of the two different terms by the two opposite systems is very significant, for they denote two different approaches to negation. In discussing different theories, especially in Indian Philosophy, one must strictly adhere to the terminology applied by the philosophers in the original texts. All through the history of indian Philosophy, the term Abhdva refers to an ontological situation, meaning N o n - E n s or a real negative category (paddrtha). Kum~rila Bha~t.a's position is unique inasmuch as he postulated Negation as the means (pramdn.a) of cognising its own corresponding
~3 1bid., pradegavige.se kvaein na ghat.a upalabdhilaks.a.napraptasyanupalabdhe.h, yadi syad upalabhyasattva eva sydn nanyath6. 24 1bid., p. 20: yuktopalambhasya tasya canupalambhana.m prati.sedhahetu.h. - yuktopalambha = d.r~ya = laks.a.napr6pta. ~ Nyaya-bindu of Dharmakirti [hereinafter NB], II, 48: viprak.r~tavi.sayanupalabdhi.h pratyaks.6numana-niv.rttilak.saea sa.mAayahetu.h. English translation by B. L. Steher-

batsky, Vol. II, p. 107. ~ For manifold negative formula in Buddhist logic the reader is referred to PVS, I, and NB, II, 32ff.

abhdva WRSUS anupalabdhi

297

negative object. To him abhdva is an object as well as the means of its own knowledge. On the epistemological question, How do we know a negative fact or the absence of a thing?, Kum~rila explained his theory of abhdva-pramdn, a in terms of 'non-existence of all other positive means of knowledge'. This absence of all positive means being 'absence real' ipsofacto becomes an independent means of negative judgment. ~7 The term anupalabdhi, on the other hand, denotes an inferential character of the problem. That is an-upalabdhi or non-perception = the denial of perception. This refers to the fact that the perceptibility of the negatum is the a priori cognition and the absence of the thing is known on the ground of its being non-perceived. This is the view upheld by the Buddhist logicians led by Dharmakirti. 2s On the question of the epistemological negative dialectics a confusion prevails in the history books of Indian Philosophy. A critical study of the history books would reveal the fact that Dasgupta, Radhakrishnan and Sinha, on the question of Negation, do not present a comparative study of the Buddhists and the Pftrva Mim~.ms~ systems - the two main rival schools of India. Their statements on Negation are ill-founded and misleading. They betray the whole historical development of the problem of negation in Indian logic. These scholars, I submit, have failed to recognise the very demonstrative influence of Buddhist logic on the thinkers of the Brahmanic tradition. Dasgupta writes: "In addition to the four positive pramhn, as, Kum~rila admits a fifth kind of pram~.na, viz., anupalabdhi for the perception of the non-existence of a thing. ''29 Radhakrishnan states: "Kum~rila, after V.rttik~ra, admits non-apprehension (anupalabdhi) as an independent source of knowledge. (See gabara on I, i. 5.) Dissimilarity is only want of similarity, and it is accounted for by the principle of non-apprehension. When we say "There is no jar in this place," we cognise the absence of the jar. Absence (abh~va) cannot be apprehended by perception, which stands in need of sense-contact with a present object, which is not possible in the case, (refers to: SlokaV~rttika, Abhhvapariccheda) nor can non-existence be apprehended by the other pram~.nas. Non-apprehension is a means of knowledge (mftnam) with reference to the object negated. We perceive the vacant space and think of the absence of the jar. We may say that the non27 See above note 10. 28 Cf. notes 21-25. ~9 Vide S. N. Dasgupta, History of Indian Philosophy, Vol. I (Cambridge University Press, 1922), p. 397.

298

DHIRENDRA SHARMA

existence of the jar is as much perceived as the vacant space, yet, since perception involves contact of an actual object with the sense, we cannot identify the act of non-apprehension with perception. We perceive the vacant space, remember the jar that is absent, and then we have the knowledge of the absence of the jar, which has no reference to the act of perception. Apprehension of non-existence is through anupalabdhi. (Refers to: ~stradipikfi, pp. 234ff). T M Here I may point out the fact that Sdstrad~pika is the work of P~rthas~rathi MiCra who lived in 900 A.D. about 300 years after Kum~rila. Radhakrishnan explains the nonperceptibility of abhdva, according to Pfirthasfirathi MiCra, and does not say why it cannot be an inferential cognition. In Kum~rila's theory repudiation of negative inference is the crux of his argument. Sinha observes: "Savara recognises non-apprehension (anupalabdhi) as an independent pram~n,a, and defines it as the absence of any means of valid knowledge...'31 In another section of his volume Sinha writes: "Kumarila holds that the non-existence of a jar on the ground is known by non-apprehension (anupalabdhi). T M According to these scholars the theory of anupalabdhi appears as propounded by the Pfirva Mimh .msaka philosophers, namely, V.rttikfira Upavar.sa, ~abara, and Kumfirila Bha.tt.a. In fact, none of these Mim~m.sakas show any awareness of the theory of anupalabdhi. My conclusions are as follows: 1) Dasgupta is wrong in stating that "for the perception o f the nonexistence of a thing, Kum~rila admits a fifth kind of praman.a, viz., anupalabdhi." Unfortunately, Dasgupta quotes no work in support of his statement. During my investigation i have not come across any work of the Mim~m.sakas or non-Mim~ .msakas where Negation is mentioned as thefifth pram~n,a. On the contrary, abhdva is frequently referred to as the sixth pramhn.a (.sas.t.ha.m kiledam, pramdn, am). ~ Moreover, Kumarila himself candidly opposed the perceptibility of the non-existence and said that abhdva is themeans in addition to the other five positive ones, which include perception, a4 Hence, Dasgupta's statement is completely erroneous. 2) The two terms, namely, abhdva and anupalabdhi refer to two entirely different theories on epistemological negative dialectics. In the early 30 81 32 83 34 S. Radhakrishnan, Indian Philosophy, Vol. II (London 1948, reprint), p. 394-395. J. Sinha, History of Indian Philosophy, Vol. I (Calcutta, 1956), p. 789. Ibid., p. 309. B.rhati-rjuvimal~, p. 120. ,~I.V., op. cit., verses 1-2.

abhfiva VERSUSanupalabdhi

299

writings of the Pfirva Mim~m. sh system we find no mention of the theory 'anupalabdhi'. Neither V.rttik~ra Upavar.sa, nor Sabara or Kumfirila ever referred to anupalabdhi? 5 Sinha has quoted Sabara (on JaiminiSfttra I.i.5) but missed the point that he mentioned only abhdva, not anupalabdhi. 3) P~rthas~rathi MiCra (circa 900 A.D.) seems to be the first Mim~ .msaka who, after about 300 years of Kum~rila, introduced 'anupalabdhi' to the system. Defending Kumgrila's tradition against the Buddhist reasoning, Pfirthasfirathi modified the Mimg.msfi theory of Negation (abhdvapramd~a), and admitted that, in fact, the negative cog~nition is an inferential judgment. Thus, he contended that the Mimg .msaka theory of Negation may also be termed as d.r~yadargana or yogyanupalambha. He still hesitated to use the Buddhist term d.rgyanupalabdhi. Instead he split the Buddhist term and coined new expressions, 'd.rgya-adar~ana' and 'yogya-anupalambha'. P~rthasgrathi contended that these terms are synonymous and the theory expressed by them is the same as the theory referred to by the term 'pramgt.n~bhgva' in the Bhd.sya of Mimfim. s~. s6 4) Admission of anupalabdhi by the later Mim~ .msakas was indeed a great blow to the original stand taken by the early Mim~ .msakas. The view of an independent negative means lost its vigour, and the theory of abhavapram~.na, evidently under Buddhist influence, was transformed into an inferential theory called yogyanupalabdhi. For example, G~g~bha.t.ta, a later follower of the Bhgt..ta school of the Pfirva Mim~ .msg, explaining the logical process of the modified negative theory, observes that the non-perception of the counter-entity or the negatum (pratiyogin) is the reason (kara.na = lihga) of negation and the process is similar to that of inference. ~ Here we must bear in mind that this view of negation is identical with the Buddhist logician Dharmakirti's view that the nonperception of the negatum is the negative reason (hetu) of inference, ss Further, Ghb~bha.tt.a explicitly admits that as a matter of fact there is no difference between the non-apprehension (anupalabdhi) of the Mimfi.msaka and the principle of inference, z~ 85 Vide G. Jha, Parva Mimfi.msfi in its Sources, pp. 163-165. 3~ ~astradipika, ed. Laxman Shastri Dvavid (= Chowkhamba Sanskrit Series, 188) (Banares, 1916), p. 234: d.rgyadargana-yogy~nupalambh6di parydyo bh6.sye pramfi.nabh~va-gabdenokta.h. 87 Bhat..ta-cinta-ma.ni(= Chowkhamba Sanskrit Series, 25 & 27) (Banares, 1933), p. 47: atra pratiyogipratyak.sabhfiva.h kara.nam anurnitir ivantaravy6para.h. 38 Cf. PVS (G. ed.), p. 20, verse I. 29(R. ed., p. 85, I. 31): yuktopalambhasya tasya c~nupalambhana.m prati.sedhahetu.h. 89 B. Cin., p. 47: ...anum~nac-ca bhedar~ n6kalayamab.

300

DHIRENDRA S H A R M A

Hence, I may conclude that this modified theory of the Bhgt.t.a Mim~m. sakas called Non-apprehension (yogydnupalabdhi) reduces Negation (abhdvapramd~a), the earlier theory of the Mimfi.msakas, to a mere negative form of Inference which is equivalent to the Buddhist theory of Non-perception; that the term anupalabdhirefers to the inferential character of the negative cognition; and that there is no evidence to support the view that the theory of anupalabdhi was propounded by the early Pflrva Mim~m.s~ philosophers. 4~

~o I gratefully acknowledge my indebtedness to Dr. D. Friedman of the School of Oriental and African Studies, University of London, for his constructive suggestions and discussions concerning many problems of Negation in Indian Logic.