• co
18
-in
S
;CD loo
THE HERITAGE OF INDIA
, .
P , ., manors
The Right Reverend V. S. Azariah,
Bishop of Dornakal. j N Farquhar> M A>> D>Litt. (Oxon.)
Already published.
The Heart of Buddhism. K. J. SAUNDERS, M.A.
Asoka. REV. J. M. MACPHAIL, M.A., M.D.
Indian Painting. PRINCIPAL PERCY BROWN, Calcutta.
Kanarese Literature. REV. E. P. RICE, B.A.
The Sarhkhya System. A. BERRIEDALE KEITH, D.C.L.,
D.Litt.
Psalms of Maratha Saints. NICOL MACNICOL, M.A., D.Litt. A History of Hindi Literature. REV. F. E. KEAY, M.A. Hymns of the Tamil Saivite Saints. KINGSBURY and
PHILLIPS.
Subjects proposed and •volumes tinder preparation.
SANSKRIT AND PALI LITERATURE.
Hymns from the Vedas. Prof. A. A. MACDONELL, Oxford. Anthology of Mahayana Literature. Prof. L. DE LA VALLE
POUSSIN, Ghent. Selections from the Upanisads. F. J. WESTERN, M.A.,
Delhi.
Scenes from the Ramaya_na._ Selections from the Mahabharata.
THE PHILOSOPHIES.
An Introduction to Hindu Philosophy. J. N. FARQUHAR,
and JOHN McKENZiE, M.A., Bombay. The Philosophy of the Upanisads. Sankara's Vedanta. A. K. SHARMA, M.A,, Patiala Ramanuja's Vedanta. The Buddhist System.
FINE ART^AND MUSIC.
Indian Architecture. R. L. EWING, B.A., Madras, Indian Sculpture.
The Minor Arts. Principal PERCY BROWN, Calcutta. Indian Coins. C. J. BROWN, M.A. (Oxon.).
BIOGRAPHIES OF EMINENT INDIANS.
Gautama Buddha. K. J. SAUNDERS, M.A., Rangoon.
Ramanuja.
Akbar. F. V. SLACK, M.A., Calcutta.
Tulsi Das.
Rabindranath Tagore. E. J. THOMPSON, M.A., Bankura.
VERNACULAR LITERATURE.
The Kurral. H. A. POPLEY, B.A., Erode.
Hymns of the Alvars. J. S. M. HOOPER, M.A., Madras.
Tulsi Das in Miniature. G. J. DANN, M.A., Patna.
Hymns of Bengali Singers. E. J. THOMPSON, M.A., Bankura.
Kanarese Hymns. Miss BUTLER, Bangalore.
HISTORIES OF VERNACULAR LITERATURE.
Bengali. C. S. PATERSON, M.A., Calcutta.
Gujaratl. R. H. Bo YD, M.A., Ahmadabad.
Marathi. NICOL MACNICOL, M.A., D.Litt., Poona.
Tamil. FRANCIS KINGSBURY, B.A., Bangalore.
Telugu. P. CHENCHIAH, M.A., Madras, and RAJA
BHUJANGA RAO, Ellore.
Malayalam. T. K. JOSEPH, B.A., L.T., Trivandrum. Sinhalese. Urdu. B. GHOSH AL, M.A., Bhopal.
NOTABLE INDIAN PEOPLES.
The Rajputs.
The Syrian Christians. K. C. MAMMEN MAPILLAI, Alleppey.
The Sikhs.
VARIOUS.
Modern Folk Tales. W. NORMAN BROWN, M.A., Ph.D.,
Philadelphia.
Indian Village Government.
Poems by Indian Women. MRS. N. MACNICOL, Poona. Classical Sanskrit Literature.
Indian Temple Legends. K. T. PAUL, B.A., Calcutta. Indian Astronomy and Chronology. DEWAN BAHADUR L. D.
SWAMIKANNU PILLAI, Madras.
EDITORIAL PREFACE
" Finally, brethren, whatsoever things are true, whatsoever things are honourable, whatso ever things are just, whatsoever things are pure, whatsoever things are lovely, whatsoever things are of good report ; if there be any virtue, and if there be any praise, think on these things."
No section of the population of India can afford to neglect her ancient heritage. In her literature, philosophy, art, and regulated life there is much that is worthless, much also that is distinctly unhealthy; yet the treasures of knowledge, wisdom, and beauty which they contain are too precious to be lost. Every citizen of India needs to use them, if he is to be a cultured modern Indian. This is as true of the Christian, the Muslim, the Zoroastrian as of the Hindu. But, while the heritage of India has been largely- explored by scholars, and the results of their toil are laid out for us in their books, they cannot be said to be really available for the ordinary man. The volumes are in most cases expensive, and are often technical and difficult. Hence this series of cheap books has been planned by a group of Christian men, in order that every educated Indian, whether rich or poor, may be able to find his way into the treasures of India's past. Many Europeans, both in India and elsewhere, will doubtless be glad to use the series.
The utmost care is being taken by the General Editors in selecting writers, and in passing manuscripts for the press. To every book two tests are rigidly applied : every thing must be scholarly, and everything must be sympathetic. The purpose is to bring the best out of the ancient treasuries, so that it may be known, enjoyed, and used.
THE M/.^.
KARMA-MiMAMSA
BY
A. BERRIEDALE KEITH, D.C.L., D.LITT.
(of the Inner Temple Barrister-at-Law, Regius Professor of Sanskrit and Comparative Philology at the University of Edinburgh)
AUTHOR OF "THE SAMKHYA SYSTEM," ETC.
. 3
ASSOCIATION PRESS 5, RUSSELL STREET, CALCUTTA
LONDON: OXFORD UNIVERSITY PRESS
NEW YORK, TORONTO, MELBOURNE,
BOMBAY, CALCUTTA AND MADRAS
1921
CONTENTS
PAGE
PREFACE
I. THE DEVELOPMENT AND LITERARY HISTORY OF
THE KARMA-MIMAMSA . . . . . . . . 6
The Origin of the System. — The Mirnamsa Sutra. —
The Vrttikara. — Prabhakara and Kumarila. — Later
Writers. — Other Literary Sources.
II. THE PROBLEM OF KNOWLEDGE .. .. .. 17
The Validity of Apprehension. — Mode of Appre hension of Cognition. — The Forms of Cognition. — Perception. — Inference. — Comparison. — Presumption. — Negation. — Scripture and Verbal Testimony.
III. THE WORLD OF REALITY .. .. ..44
The Refutation of Buddhist Nihilism and Idealism. — The Categories of Prabhakara and Kumarila. — Substance. — Quality. — Action or Motion. — Generality. — Inherence. — Similarity. — Cause. — Non-existence.
IV. GOD, THE SOUL, AND MATTER .. .. ..61
The Refutation of the Doctrine of Creation.— The Doctrine of the Soul. — The Destiny of Man. — The Purpose of Sacrifice.
V. THE RULES OF RITUAL INTERPRETATION . . 79
The Authority of Scripture. — The Brahmanas as Vidhi, Arthavada, and Namadheya. — The Mantras. — Smrtis, Vedangas, Grammar, and Custom. — The Classes of Actions enjoined. — Originating Injunc tions. — Injunctions of Application. — Injunctions of Performance. -Injunctions of Qualification.— Original, Restrictive, and Limiting Injunctions. — Prohibitions and Exceptions. — Transfer and Modification of Details. — Repetition and Option.
VI. THE MIMAMSA AND HINDU LAW.. .. .. 97
The Mimamsa and the Law Schools. — The Inter pretation of Injunctions. — Religious and Secular Factors in Ownership. — Inheritance and Partition. — Adoption and Partnership. — Criminal Law and the Law of Evidence.
INDEX . ; 108
PRINTED AT
THE WESLEYAN MISSION PRESS, MYSORE CITY.
I
THE DEVELOPMENT AND LITERARY HISTORY OF THE KARMA-MIMAMSA
NOT rarely in the Brahmanas, especially in later texts like the Kausttaki, the term Mimarhsa occurs as the designation of a discussion on some point of ritual practise. The sacrifice left innumerable opportunities for divergence of usage in detail, and the texts decide in favour of one or the other alternative, on the strength of the reasons- familiar to the Brahmanas, in special the symbolical significance attaching to the action recommended. There is a vital difference between this form of Mimamsa and that of the classical Karma-Mimarhsa school, in the fact that in the former the appeal to authority, and the necessity of reconciling apparent discrepancies of authority, are entirely lacking. But the tendency to surrender judgment in favour of tradition may be traced in the care with which in the Satapatha and the Kausitaki Brahmanas the name of the teacher is adduced in support of the doctrines expounded ; in the older style the reasonings stand by themselves, commended by their intrinsic value.
The process by which the Brahmanas came to be regard ed as texts of incontrovertible accuracy, and speculation on the sacrifice ceased to be independent, cannot now be traced in detail. The account of the sacrifices given in these texts, supplemented by the collections of Mantras in the Samhitas of the various schools, would, obviously, never have sufficed to enable priests to carry out the sacrifices, and there must have been a full and precise oral tradition regarding the mode in which the sacrifices, which formed the subjects of the mystical speculations of the Brahmanas, were to be performed. This tradition, however, in the course of time
2 THE KARMA-MIMAMSA
seems to have become obscured, just as the tradition of the interpretation of the Mantras fell into confusion, and in its place in some degree supervened an attempt, on the ground of reasoning, to deduce from the Sarhhitas and the Brah manas, taken together, rules for the regulation of the performance of the offerings. The difficulties of such a course were considerable; there are real divergences between the Samhitas and the Brahmanas, which we may justly attribute to change of ritual, but which in the opinion of the priests admitted of other explanations. Thus, in some cases, the order of the Mantras is patently different from the order of actions contemplated in the Brahmanas, a divergence which the new MImamsa decided in favour of the order of the Mantras on the ground that, as they were recited in the sacrifice they were more directly connected with the sacrifice than the Brahmanas, which were not immediately employed in the offering. With more plausi bility, the new doctrine held that if a Brahmana mentioned an action out of its natural order, such as the cooking of the rice grains before the husking, it was nevertheless to be assumed that the normal sequence was to be followed. More legitimately still, the new science devoted itself to such problems as the determination of the person by whom the several actions enjoined, without specification of the actor, fell to be performed; the connection as principal and subordinate of the many details of the offering ; and the precise mode of performance of the Vikrtis, or derivative forms of the main sacrifices, the particulars of which are seldom adequately indicated in the sacred texts.
The antiquity of the new science is vouched for by the Dharma Sutras. Apastamba in two passages1 disposes of contested points by the authority of those who know the Nyaya, a term which is the early designation of the Karma- Mimarhsa and persists through its history in its generic sense of " reasoning," while the Nyaya philosophy proper borrows it, and applies it more specifically to denote the syllogism. What is still more convincing is that Apastamba
1 Biihlcr, Sacred Books of the East, II, xxviii, xxix ; XXV, xlvii, Hi.
DEVELOPMENT AND LITERARY HISTORY 3
uses arguments which are to be found in the Mimamsa Sutra ; thus he maintains that no text can be inferred from a custom for which a secular motive is apparent, and that a revealed text has superior validity to a custom whence a text might be inferred. The corresponding rules in the Mimdmsa Sutra (I, 3, 3-4) do not textually agree, and we may fairly conclude that at this date, probably not later than the middle of the third century B.C., the Sutra did not exist in its present form, but it is plain that the science itself was in full vogue, and a Mimamsnka appears to have been deemed a necessary member of a Parisad. The influence of this discipline can plainly be discerned in the existing Sutra texts; the works of Asvalayana, Sankhayana, Apastamba, Hiranyakesin, Latyayana, and Drahyayana have been composed under its influence, and the same con sideration applies even to texts like those of the Baudhayana and Manava schools, which show greater affinities to the Brahmana style. We need not, of course, assume that the old sacrificial tradition was entirely lost, but we may be certain that it has been largely transformed in the process of remodelling.
Simultaneously with the remodelling of the Sutras, there must have proceeded the definition of the rules of interpre tation until they were finally codified in the Ml-mamsa Sutra,1 which passes under the name of Jaimini, but the details of this process must remain unknown to us. What is certain is that the Mlmamsd Sutra presupposes a long history of discussion, and that its aphorisms, which often assume, without expressing, general rules of interpretation, deal largely with difficulties affecting individual Vedic texts, which had long been the subject of dispute. This characteristic is shown clearly in the mode of discussion followed in the text; the essential subdivision is the Adhikarana, which, according to the school, is to be deemed to fall into five parts; these Madhava reckons as the subject of investigation (visaya), the doubt (samsaya) , the first or prima facie view (purvapaksa) , the
1 Edited, Bibliothecalndica, 1873-1889 ; trans, of Adhyayas I-III, by Gafiganatha Jha, Sacred Books of the Hindus, vol, X,
4 THE KARMA-MlMAMSA
answer or demonstrated conclusion (siddhdnta) , and the relevance of the topic (samgati), but the last head is else where reckoned as the third, and a more natural division1 omits it, and regards as the fourth and fifth members the answer (uttarapaksa), and the conclusion (nirnaya). Thus in the first Sutra of the text there are two Adhikaranas; is the study of the Veda necessary for the three upper castes, and is Dharma a proper subject of study ? The latter alone needs full discussion, the reply to the former being self- evident. The subject then is formed by the two Vedic pre cepts, " One should study the Veda," and " One should per form the final bath after studying the Veda." The doubt is whether one should, after learning the Vedic text, perform the bath and end one's studentship, or remain longer with the teacher to study Dharma. The priina facie view is that the bath should follow immediately on the learning of the text, but the reply is that real study of the Veda is not satisfied by mere reading of the text, and the conclusion, therefore, is that the final bath is to be postponed for a time in order that the student may complete his learning of the text by a study of Dharma. Of all this, however, the Sutra itself has nothing, consisting merely of the words, " Now, therefore, an enquiry into Dharma," and, though in some cases there is more full development of an Adhikarana, it is to the commentators that we must look for enlightenment on the exact issues in dispute. It is not, of course, to be supposed that at any time the Sutra was handed down without oral explanation, but, as usual, the authentic version was early obscured.
Of these Adhikaranas there are in the Sutra in Madhava's reckoning about 915, divided into twelve books with sixty Padas, the third, sixth, and tenth having eight each in lieu of the normal four, and 2,652 aphorisms (2,742 in another reckoning). Jaimini is the chief authority cited, but men tion is made also of other names, such as those of Badari, Atreya, and Badarayana, who occur also in the Vedanta Siitra, and of Labukayana, Aitasayana, etc. Who Jaimini was we cannot say. A Jaimini is credited with the author-
1 Cowell in Colcbrookc, Essays I, 326,
DEVELOPMENT AND LITERARY HISTORY 5
ship of a Srauta and a Grhya Sutra, and the name occurs in lists of doubtful authenticity in the Asvalayana and Sdiikhdyana Grhya Sutras; a Jaiminiya Samhitd and a Jaiminiya Brdhmana of the Sdma Veda are extant. As an authority on philosophy Jaimini appears in the Vedanta Sutra and often in later works,1 but it is significant that, while it is possible that the Mahdbhdrata recognises the existence of the Mimamsa it does not refer to Jaimini as a philosopher, but merely as an ancient sage.2 Neither Buddhist or Jain literature throws light on his personality or date, and the period of the Sutra can be determined, therefore, merely on grounds of comparison of its contents with those of other works.
It is probable that the Mimdrhsd Sutra is the earliest of the six Darsanas preserved to us. The Yoga Sutra is not an early work; it seems to recognise the Vijnanavada school of Buddhism, which, in all probability, belongs to the fourth century A.D., and the popular identification of Patanjali with the author of the Mahdbhdsya is clearly unten able.3 The lateness of the Sdrhkhya Sutra is admitted, and the theory that its contents include early matter has been controverted. The Vaisesika Sutra has no point of contact with the Mimamsa such as would render any con clusion possible, but the Nydya Sutra (II, 1,61) is familiar with the Mimamsa terminology, and it is improbable that, had the Nydya existed before the Mimdrhsd Siltra took form, it would have been ignored by the latter as it is. The relation to the Vedanta Sutra is less clear; the mention of Jaimini and Badarayana in both texts affords some ground for the view that the two works were simultaneously redacted, but this conclusion is by no means assured. We have no valid reason for assuming that the Sutras were actually redacted by Jaimini and Badarayana themselves,
1 His death, caused by an elephant, is recorded in Pancatantra II, 34, but not in the Tantrakhyayika. The name is strange, but is ignored in Panini and the Mahdbhdsya, which, however, knows of Mimarhsakas, probably adherents of this school (Indische Studien, XIII, 455, 466).
* Hopkins, Great Epic of India, p. 97.
8 Sdmkhya System, pp. 56, 57.
6 THE KARMA-MlMAMSA
and, unless this is established, the argument for contempor aneity is invalid. It is true that it is impossible to deduce from the style of the Mimamsa as Purva-MImamsa a relation of temporal priority for Jaimini's work; the Mimamsa is prior to the Vedanta because it deals with the sacred rites, the knowledge of which, in the view of one school of Vedanta, is an indispensable preliminary to the knowledge of the absolute, though Sarhkara declines to accept this view and insists instead on the diverse character of the ends of the two disciplines, which renders it impossible to treat the former as the normal or necessary prelude to the latter.1 Nonetheless it remains true that we must assume that the Mimamsa as a science developed before the Vedanta.2 The former was plainly necessitated by the development of the sacrificial ritual with which it is immediately connected, and it serves an important practical end; the latter is proof of the growth of a philosophical spirit, which sought to comprehend as a whole the extremely varied speculations which are scattered in the Aranyakas and Upanisads. While, of course, it is not impossible that the redaction of the two Sutras was contemporaneous, despite the earlier development of the Mimamsa, the probability surely lies in favour of the view that the Mimamsa Sutra was redacted first and served as a model for the other schools.
Even if this view is accepted, it remains difficult to assign any definite date to the Sutra. It contains no certain reference to Buddhist tenets of any kind, for the term buddha, in I, 2, 33, has not this signification, and we need not with Kumarila read a reference to Buddhism into I, 3, 5, and^. The Vedanta Sutra is of uncertain date ; if we believe Sarhkara, it criticises (II, 2, 28-32) the Vijnana- vada school of Buddhism, but this doctrine is probably wrong3 and we need see only a reference to the Sunyavada of Nagarjuna. The date of this school is uncertain ; if we accept the opinion that it was not enunciated before Nagarjuna in such a manner as to invite criticism in the
1 Deussen, Vedanta, ch. I.
2 Thibaut, S.B.E., XXXIV, ix ff .
3 Jacobi, J.A.O.S., XXXI, Iff; Keith, J.R.A.S., 1914, pp. 1091 ff ; see below, pp. 46, 47.
Veddnta Sutra, that work cannot be earlier than the third century A.D., for Aryadeva, Nagarjuna's contemporary, refers to the zodiacal signs and the week-days, which were not known in India until that epoch. But it is possible that the Sunyavada, which can be recognised in Asvaghosa, was of older fame than Nagarjuna, though on the whole it is more likely than not that it was the dialectical ability of that teacher which made the doctrine the object of Vedantic confutation. It is, then, a plausible conclusion that the Mimamsd Sutra does not date after 200 A.D., but that it is probably not much earlier, since otherwise it would have been natural to find in the Mahdbhdrata some reference to it and to its author.
As we have seen, the Sutra must from the first have been accompanied by a comment, which in course of time was lost or became defective. The first commentator of whom we have certain knowledge is a Vrttikara, from whose work a long extract is made in the Bhdsya of Sabarasvamin on Mimdrhsd Sutra, I, 1, 5, in which the author attacks and refutes Buddhist views. If we believe Kumarila, the dis cussion is directed in part against the Vijnanavada school, in part against the Sunyavada, but in this case we have every reason to distrust his assertion, for, plainly by error, he ascribes the major portion of the discussion to Sabara svamin, and not to the Vrttikara. It is, therefore, not improbable that he is also in error in finding any reference to the Vijnanavada, for the passage seems to deal with one topic only, and that the Sunyavada. It follows, accordingly, that the date of the Vrttikara was probably not later than the fourth century A.D., since, had he lived later, he would hardly have omitted an explicit discussion of the tenets of the idealistic school of Buddhism.
The name of the Vrttikara is uncertain. The conjecture1 that he was Bhavadasa, mentioned in one place by Kumarila, may be dismissed as wholly without support. The current opinion makes him to be Upavarsa, who, we know from Sarh- kara (Veddnta Sutra, III, 3, 53) wrote on both the texts. To this the objection has been brought that in the passage cited
1 Gaftganatha Jh&, trans, of Slokavdrttika, p. 116.
8 THE KARMA-MlMAMSA
from the Vrttikara by Sabarasvamin there is a reference to Upavarsa with the epithet Bhagavat, implying that he was in the eyes of the Vrttikara an author of venerable authority. It is probable, however, that the citation from the Vrttikara is only a resume,1 not a verbatim quotation, and that Sabara- svamin is responsible for the reference1 to Upavarsa, the Vrttikara's proper name, and for this view support may be derived from the mode in which the Vrttikara and Upavarsa are referred to by Kumarila elsewhere (II, 3, 16). If this view is rejected, it is possible that he is Bodhayana, who certainly wrote on the Vedanta Sutra, but this theory is a bare and unnecessary conjecture, seeing that Bodhayana nowhere else appears as a Mimamsa authority. Of other, presumably early, commentators we hear of Bhartrmitra2 and Hari,3 but there is no reason to identify either of these with the Vrttikara.
The extract from the Vrttikara proves that an im portant addition has been made to the teaching of the Mimamsa in the shape of the introduction of discussions of the validity of knowledge and its diverse forms. The Sutra itself is content with the denial of the validity of perception for the purpose of the knowledge of Dharma, and the exaltation of Vedic injunctions as the source of the necessary knowledge; under the influence, perhaps, of the Nyaya the earlier doctrine is now elaborated into a critical examination of the nature of evidence, its validity, and the forms of proof. It is not illegitimate to assume that the Vrttikara e indulged also in metaphysical discus sions; at any rate Sabarasvamin enters into a long discussion of the nature of soul, despite his predilection for brevity in treatment of the Sutra. The Mimamsa therefore by this time enters into the whole field of philosophy, while maintaining its primary duty of expounding the rules by which the ritual can be reconstructed from the Brahmanas and the Samhitas.
1 In II, 3, 16, he clearly describes the Vrttikara as bhagavdn dcaryah ; cf. Ill, 1, 6. These passages Jacob! has overlooked.
1 Parthasarathi on Slokavarttika, p. 4 (v. 10) ; he is cited on the organ of sound, Nydyamanjarl, p. 213.
• Sastradipika, X, 2, 59, 60.
DEVELOPMENT AND LITERARY HISTORY 9
Of Sabarasvamin, whose commentary forms the starting point for later discussion, we know practically nothing It is an idle fiction which makes him the father of the mythical founder of the Vikrama era (57 B.C.}, of Bhartrhari, and Varahamihira, his true name being Adityadeva, while he adopted the soubriquet Sabara, when he disguised himself as a forester to avoid Jain persecution. The form of his name and his relation to the Vrttikara suggest that 400 A.D. is the earliest date to which he can be assigned; the later limit is vaguely indicated by his priority to Prabhakara, Kumarila, and Sarhkara.
With Prabhakara and Kumarila there comes a sharp divergence in the unity of the teaching of the school, whose followers henceforth are divided between the adherents of Prabhakara, or the Guru par excellence, and the supporters of Kumarila Bhatta. It does not, however, appear that Prabhakara initiated the views which he became noted for expounding ; he cites in the Brhatl, his ' great ' exposition of Sabarasvamin's Bhdsya, the opinions of a Varttikakara who presumably must be regarded as the true founder of the school. The Brhatl itself is unfortunately only extant in imperfect form ;x it was commented on by Salikanatha in his Rjuvimala, formerly erroneously regarded as simply a commentary on the Sabarabhdsya, while in his Prakara- napancikd2 the same author deals with the more important epistemological and metaphysical views of his teacher. The Brhatl seems to have passed comparatively early into oblivion, though a passage from it is cited in the Mitdksard and its author's views were well known in MImamsa circles.
The relation of Prabhakara and Kumarila is represented by tradition as those of pupil and teacher ; the tradition is fairly old, as it occurs in the Sarvasiddhantasamgraha (I, 18, 19 ; VII, 15) attributed, doubtless wrongly, to Sarhkara himself. But the discovery of the text of the
1 See Ganganatha Jha, The Prabhakara School of Purva MImamsa (1911).
2 Ed. Benares, 1903-04 ; the text is defective, and the Prameyapdra- yana is lacking. He uses Uddyotakara (e.g. p. 44) and Dharmakirti, and therefore is not before 650-700 A.D., but is probably before Kuma rila. Prabhakara thus dates about 600-650 A.D.
10 THE KARMA-MfMAMSA
Brhatl has enabled us to correct this error. Prabhakara follows Sabarasvamin closely ; he does not refute the opinions of Kumarila ; in one passage (IV, 1, 2), when he does criticise an opinion of the latter, the form of words used by him in adducing it differs entirely from those in which the view of Kumarila is expressed, showing clearly that he is dealing with some older author, whom Kumarila has followed. On the other hand, Kumarila frequently diverges from the views of the Bhasya; he criticises (I, 2, 31 ; 3, 2 ; 4, 1) views which are expressed by Prabha kara, and asserts independent views. There is a clear difference of style between the two authors ; Prabhakara is comparatively simple, vivid and direct like Sabara svamin; he seldom uses long compounds; he avoids the com bination of various reasonings in a single clause ; in lieu of the formal terminology of objection and reply (nanu . . . ced, na or sydd etat . . . tad ayuktam) he adopts the form of question and answer, which, however, has the disadvan tage of leaving at limes the meaning in doubt. In all these aspects Kumarila shows a richer, more varied, and elaborated style, which is reminiscent of the Sarirakabhasya of Sarhkara. Kumarila's great exposition of the Sabarabhasya falls into three parts, the first, the Slokavdrttika,1 in verse, deals with Pada I of Adh}aya I of the Bhasya, and is of the greatest value as an explanation of the metaphysics and epistemology of his system. The second, the Tantravart- tika,2 covers the remaining three Padas of Adhyaya I and the whole of Adhyayas II and III. The third part, the Tuptlka, consists merely of scattered notes on the last nine Adhyayas. Each part has been commented on ; the first by Parthasarathi Misra in his Nydyaratndkara, and by Sucarita Misra in his Kdsikd; the second by Somesvara, son of Mahadeva, in his Nydyasudhd or Kanaka; the third by Venkatesvara Diksita in his Vdrttikdbharana. Kumarila's date is determinate within definite limits;3 he used the
1 Ed. Benares, 1898-99; trans. Gangiinatha Jha, Bibliothcca Indica, 1900-8.
2 Ed Benares, 1890; trans. Gahganatha Jha, Bibliotheca Indica, 1903-20.
3 Pathak, J.B.R.A.S. XVIII, 213 ff.
DEVELOPMENT AND LITERARY HISTORY 11
Vdkyapadlya of Bhartrhari; neither Hiuen-Thsang nor I-tsing mentions him; he was before Samkara; he attacked the Jain theory of an omniscient being as propounded in the Apt aim mams a of Samantabhadra, but is not answered by Akalanka in his Astasati, which comments on the Apta- mlmdrnsd. On the other hand, he is freely attacked by Vidyananda and Prabhacandra, who both lived before 838 A.D. Vidyananda assures us, doubtless correctly, that he criticised the Buddhist Dharmakirti, and Prabhakara, on the latter point agreeing with the result above arrived from internal evidence. The upper limit of date is, therefore, not earlier than 700 A.D. The lower limit depends on his precise chronological relation to Samkara and the latter's exact date. Later tradition, the Samkaravijayas of Madhava and the pseudo-Anandagiri, would make him an older contem porary, but the interval may have been considerably longer.
Only slightly later than Kumarila was Mandana Misra, author of the Vidhiviveka, a treatise on the significance of injunctions, and the Mimdmsanukramani, a summary of Sabarasvamin's Bhdsya. The tradition of the Samkara vijayas makes him out^ to be identical with Suresvara, a pupil of Samkara, but Anandagiri's account insists that he was also a pupil of Kumarila. The identification with Suresvara, which might be suspected because of the lateness and inferior character of the authorities, is to some extent confirmed by Vidyananda's description1 of Mandana Misra as Vedantavadin, which could hardly apply to him unless he were the author of the works ascribed to Suresvara. His direct connection with Kumarila, however, need not be insisted upon. His lower limit of date is fixed by the fact that the famous Vacaspati Misra devoted the Nyayakanikd to the exposition of his Vidhiviveka,2 and Vacaspati pro bably lived about 850 A.D. He wrote also the Tattvabindu3 on Kumarila's views.
Of the later writers the most f important is perhaps Parthasarathi Misra, who wrote the Sdstradipikd* to explain
1 Ibid. p. 228.
1 Ed. Pandit, XXV-XXVIII, 1903-6.
3 Ed. Benares, 1892.
4 Ed. Benares, 1891. He is earlier than Madhava.
12 THE KARMA-MlMAMSA
the Sutra; on it commented, in 1543 A.D., Ramakrsna Bhatta, son of Madhava, in the_ Yuktisnehaprapurani, Somanatha, son of Sura Bhatta, an Andhra Brahman of the Nittal family, in^ his Mayukhamalikd, Vaidyanatha (1710 A.D.), Bhatta Sarhkara, Bhatta Dinakara, Kamalakara, and others. His Tantraratna comments on points in the last nine Adhyayas of the Sutra and the Bhdsya, while his Nydyaratnamdla1 is an independent treatise on which Rama- nuja, apparently the great Vedantist, has written a comment, the Ndyakaratna.
Much later in date is Khandadeva, who died at Benares in 1665 A.D. His works, the BhattadipikS, and on a larger scale the Mtmdmsdkaustubha* deal fully with the^ Sutra; the former was commented on in 1708 by his pupil Sambhu Bhatta. Value attaches also to the Mlmdmsdsutradidhiti or Nydydvalldldhiti of Raghavananda Sarasvati, pupil of Advaya, pupil of Visvesvara, and to the Mlmdmsanaya- viveka of Bhavanatha Misra, which deals also with Sabara- svamin. Yet other commentaries are recorded, including works by Mahadeva Vedantin, Kamalakara and Vaidya natha, son of Ramacandra, the Subodhinl3 of Ramesvara Suri, the Bhdttacintdmani of Visvesvara or Gaga Bhatta, etc.
Apart from the Sutra there was developed a considerable literature which aims, as did Mandana Misra, at dealing systematically with the doctrines of the school. First in importance, perhaps, is the Jaimimyanydyamaldvistara* of the famous Madhava written in the fourteenth century, which, however, is merely a summary in verse, with a prose comment, of the Mlmdmsd Sutra. At the end of the sixteenth century Appayya Diksita wrote his Vidhirasdyana,5 a disquisition on the nature of injunction, adding himself a commentary, the Sukhopayojim. This text was refuted by Gopala Bhatta in his Vidhirasdyanabhusana, and by Samkara Bhatta in his Vidhirasdyanadusana. The same author, who was of the same period as Appayya, wrote
1 Ed. Benares, 1900.
1 Ed. Conjeveram, 1902 (I, 2 only). The Dipikd is ed. in the
Bibliotheca Indica.
* Ed. Pandit, XVII-XXI. 4 Ed. London, 1878.
" Ed. Benares 1901.
DEVELOPMENT AND LITERARY HISTORY 13
a commentary on the Sdstradlpikd, and the Mlmdmsdsdra- samgraha,1 in which he enumerates 1,000 Adhikaranas, allotting to each a quarter verse. An extended version of this work forms his Mimamsabalaprakasa? on which there is a commentary by Kesava, son of Visvanatha. Appayya himself wrote also the Upakramapardkrama, a treatise on the comparative importance of the commencement and end of a continuous Vedic passage.
The most popular introduction to the Mimamsa is pro bably the Mimdmsdnydyaprakdsa3 of Apadeva, son of Anantadeva, and pupil of Govinda. His date is determined by the fact that his son, Anantadeva, wrote his Smrti- kaustubha under a prince who lived in the middle of the seventeenth century. Anantadeva commented on his father's work in the Bhdttdlamkdra, and his brother, Jivadeva, discussed in the Bhdttabhdskara the divergent views prevalent in the schools. Even better known, perhaps, is the Arthasamgraha* of Laugaksi Bhaskara, which seems to be based in part on the work of Apadeva, and, if so, must belong to the seventeenth century. This date would suit adequately the probable period of his popular Nyaya-Vaisesika treatise, the Tarkakaumiidl. Another short text is the Mimdmsdparibhdsd5 of Krsna Diksita, and the Mlmariisdratna of Raghunatha, who uses the Kdsikd, contains some information of value on the views of the opposing schools. Narayanatirtha Muni's Bhdttabhdfd- prakdsa0 is an exposition of the terminology of the Mimamsa, while Ramakrsna Udlcya Bhattacarya's Adhikaranakau- mudi7 expounds a selection of interesting Adhikaranas. Khandadeva's Bhdttarahasya8 deals with the mode of determining which is the leading word in a text under discussionL More interesting is the fact that the famous Vallabha Acarya is credited with a Purvamimdmsdkdrikd, an epitome in 42 verses of Jaimini's views, written with reference to the doctrine of faith which Vallabha expounded,
Ed. Benares, 1904. * Ed. Benares, 1902.
Ed. Calcutta, 1901 ; Benares, 1905.
Ed. and trans. Thibaut, Benares, 1882.
Ed. Benares, 1904. • Ed. Benares, 1900.
Ed. Calcutta, 1885. 8 Ed. Conjeveram, 1900,
14 THE KARMA-MlMAMSA
and a Jaiminisiitrabhdsya which deals with the first chapter of the second book of the Sutra. The well-known scholar, Venkatanatha Vedantacarya, in his Mimdmsdpddukd,1 in verse, discusses the Adhikaranas in the first chapter of the first book of the Sutra, and in his Sesvaramlmamsa seeks to combine the Mimamsa with the Vedanta. Another writer from southern India, Venkatadhvarin, deals with the threefold classification of injunctions in his Vidhitrayaparitrdna, while in his Mimdmsdmakaranda he discusses the authoritative character of Arthavadas. Narayana of Kerala, the well-known author of the Ndrdyamya, who flourished at the end of the six teenth century, gives in the first part of the Mdnameyodaya2 an account of Rumania's views on the nature of proof; he purposed completing his task by adding an account of the same author's views on the world of reality, but this part of his work was never carried out, and was supplied at a later date by another Narayana, who was patronised by Mana- deva, king of Sailabdhi; the work is interesting as showing how far the school of Kumarila went in appropriating the views of the Nyaya-Vaisesika philosophy.
Of the other systems it is the Nyaya, and later the com bined school of Nyaya-Vaisesika, which throws the most light on the Mimamsa. The Nydya Sutra deals critically with the Mimamsa doctrine of the eternity of the word, and Kumarila and Prabhakara3 alike appear to have developed their philosophical tenets under the influence of the contro versy on logic which took place between the Nyaya school and the Buddhists, especially Dignaga and Dharmakirti on the other hand; Kumarila attacked both of these writers and was clearly aware of the Nydyavdrttika of Uddyotakara, in which the orthodox Nyaya view was set out in refutation of Dignaga's onslaughts. On the other hand, the Mimariisa views are freely disputed in Vacaspati Misra's comment on Uddyotakara and in Jayanta Bhatta's Nydyamanjari, Varada- raja's Tdrkikaraksd, and Udayana's works, much of the Kusumdnjali being expressly devoted to dealing with Mimamsa criticisms of the doctrine of the creation of the
1 Ed Conjeveram, 1900. » Ed. Trivandrum, 1912.
s Prakaranapancikd, pp. 47, 64, discusses Dharoiaklrti's views of perception and inference.
DEVELOPMENT AND LITERARY HISTORY 15
world. The Tattvacintdmani of Gangesa repeatedly attacks the Mlmarhsa views of the nature and validity of proof, and the controversy is continued in the voluminous literature based on that important text, and in the short text-books of the combined school of Nyaya-Vaisesika. In his commentary on Prasastapada's Bhasya Sridhara, from the point of view of the Vaisesika deals freely with Mimamsa views; moreover, the Jain Haribhadra (ninth century) includes in his Saddarsan- asamuccaya, commented on by Gunaratna, an account of the Mimamsa, and there are chapters upon it in theSarvasiddh- dntasamgraha, falsely ascribed to Samkara, and in Madha- va's Sarvadarsanasamgraha. The former work deals separ ately with the doctrines of Prabhakara and Kumarila; it betrays its late character by its attempt to show that Prabha kara was the pupil of Kumarila, and by converting the doc trine of Kumarila into a form of the Vedanta. The work of Madhava gives a long specimen of the conflicting views of the two schools as to the interpretation of the opening of the Sutra, and contains an interesting exposition of the arguments for and against the eternity of the Veda, and the self-evidence of cognition.
Jayanta Bhatta's work1 is of special interest, as it is the product of a member of a family skilled in the Mimamsa, and its author freely attacks Prabhakara and his followers, and repeatedly cites the Slokavdrttika. The author's grand father was confirmed in his faith in the efficacy of sacrifice by obtaining as the result of one offering the village of Gauramulaka, doubtless from a king of Kashmir, for Jayanta's great-grandfather, Saktisvamin, was a minister of Muktapida, better known as Lalitaditya. Incidentally Jayanta affords a welcome confirmation of the Hate of Vacaspati Misra, whom he quotes (pp. 120, 312), for, as Lalitaditya's reign ended about 753 A.D., it is impossible to place Jayanta later than the second half of the ninth century, and hence the disputed era of the year 898 given by Vacaspati himself as the date of his Nydyasuci
1 Ed. Benares, 1895. His quotation from Vacaspati on Sutra II, 1, 32, is found at p. 312. His son, Abhinanda, wrote the Kadambarika- thdsdra and lived c. 900 A.D.; Thomas, Kavindravacanasamuccaya • p. 30.
16 THE KARMA-MlMAMSA
must be taken as falling in the Samvat reckoning as 841 A.D. This date, it may be added, tells^strongly against any effort to bring down the date of Samkara,1 on whose Sanrakabhasya Vacaspati wrote the Bhdmati, and the same conclusion is favoured by the view that Mandana Misra, on whose work Vacaspati also commented, was a pupil of Samkara.
Varadaraja also claims, with obvious truth, to have been an expert in Mimamsa;2 he was evidently familiar with Salikanatha's work, and his commentator fortunately preserves for us a fragment of the Frame yap dray ana chapter of the Prakaranapancikd, no MS. of which has yet been discovered, which gives an authentic list of Prabhakara's categories.
1 S. V. Venkateswara (J.R.A.S., 1916, pp. 151-62) ignores this evidence in giving Sarhkara's date as 805-897 A.D. He cannot have died later than 825 A.D. or so.
a Ed. Benares, 1903, p. 364.
II
THE PROBLEM OF KNOWLEDGE
LIKE the Nyaya, the doctrine of Prabhakara and Kuma- rila accepts a distinction between valid and invalid know ledge, but the basis of the distinction is different in the two cases, and the ground on which validity is asserted is also diverse. To the Mlmamsa apprehension (anubhuti) is intrinsically valid, while remembrance (smrti) is intrin sically invalid, since it rests on a previous impression; the Nyaya equally disregards remembrance, but it does not accept the intrinsic validity of self-evidence (svatahprd- mdnya) of apprehension; apprehensions may be valid (pramd) or invalid (apramd), the proof being given ab extra. Remembrance both schools distinguish from recogni tion (pratyabhijnd), which is not regarded as depending solely on a previous mental impression, and therefore is exempt from the fatal defect of remembrance.
The intrinsic validity of apprehension is a cardinal doctrine of the Mlmamsa and Kumarila1 defends it at length against the obvious objections to which it is exposed. It is impossible, it is argued, for apprehension to possess the opposing characteristics of validity and non-validity as part of its nature; nor can the validity or non-validity of appre hension depend on the ascertainment of the perfection or defect of the cause of the apprehension, since this would imply that, prior to such ascertainment, apprehension was devoid of character of its own. To assume that some cases of apprehension are intrinsically valid, and other intrinsi cally invalid, is open to the objection that the criterion
1 Slokavarttika, I, 1, 2, vv. 21 ff ; Sastradipika, pp. 15, 31; Mdnameyodaya, pp. 1-4, 74-78; cf . Nydyamanjan, pp. 160-89; Tdrkika- raksd pp. 19 ff.
18 THE KARMA-MlMAMSA
between the two sets of instances can only be supplied by reference to an external factor, which is fatal to the belief in intrinsic validity or the reverse. Therefore, the Nyaya argues, it is best to accept the doctrine that apprehension or cognition is intrinsically unauthoritative, and its authorita- tiveness in any special case is derived from the perfection of the cause of the apprehension. This doctrine, it is pointed out, explains the case of dream consciousness; it is invalid because there is no perfection in its cause, while the waking consciousness may be valid if it is due to a perfect source; if, however, the source is vitiated, when the sense organs are defective, the apprehension is doubtful or erroneous, while in the case of non-apprehension there is no defect in the cause, but absence of cause.
The reply of Kumarila is that, if apprehension were not in itself valid, it could not be made so by any external power. Apprehension needs, indeed, an originating cause, but it does not depend on any external cause for its power of ascertaining the true nature of things. The conclusive argument is that, if the validity of a cognition is deemed dependent on the perfection of its source, then there must be another cognition to guarantee the correctness of the source, and so ad infinitum, and such a process is illegitimate, at least in the eyes of Kumarila, who does not appreciate the possibility of regarding truth as a complete system, in which all parts are dependent on one another, and there is no simple primary truth. All cases of apprehension, therefore, are prima facie valid, and, if cognitions are erroneous or doubtful, that is, due to defects in their causes, while non- apprehension is due to the absence of any cause, as on the Nyaya theory, with which Kumarila agrees in this regard.
The recognition of the non-validity of an apprehension establishes itself most simply when a subsequent cognition sublates an earlier cognition, for instance, when the erroneous judgment, "This is silver," is supplanted by the correct judgment, " This is mother-of-pearl." More indirectly the former judgment can be sublated by another judg ment, based on the recognition of the defect of the cause; thus the proposition, " The shell is yellow," may be sub lated by the further judgment, " The eye is jaundiced."
19
Normally, however, a judgment is valid, and is accepted as valid without question; only if, for any reason, such as distance, doubt-is possible, are further cognitions sought; if then a sublating cognition is found, and on further investiga tion it is not sublated either directly or indirectly, then the falsity of the first cognition appears; if, on the other hand, the sublating cognition is itself sublated, the validity of the first cognition is fully established. Thus, in lieu of the regressus in infinitum of the Nyaya theory, no more than three or four cognitions are necessary to establish the validity of any cognition, or, to put it more precisely, to negate the objections which may be adduced to impair its normal validity.
Prabhakara1 similarly maintains the validity of all cogni tions as such, and illustrates, in an interesting manner, the diverse modes in which apparent non-validity arises. When mother-of-pearl is mistaken for silver, the error is due to the fact that the percipient observes in the object presented to him the qualities common to the shell and the silver, and omits to notice those which differentiates the two; memory thus brings back to him the cognition of silver, and this cognition is itself real, leading no less than the actual perception of silver to the normal action of seeking to take up the object. Memory here plays the percipient false, for it does not present the silver as connected with something formerly perceived, thus differentiating it from the object actually before the eyes, and this failure is due to a certain weakness of the mind. Similarly, memory is to blame when we mistake one direction for another; the real direction is not seen, and the wrong is remembered. In the dream-state the cognitions which arise are erroneous, in as much as the things seen seem to be directly apprehended, whereas they are only remembered. The factor of apprehension on a previous occasion is lost sight of, thus obliterating the essential distinction between what is apprehended and what is remembered. The presentation of impressions in sleep is
1 Prakaranapancika, pp. 32-38 ; Bhandarkar Commemoration Volume, pp. 167-70.
20 THE KARMA-MlMAMSA
due to the operation of the unseen principle, that is the destiny begotten of man's previous acts, which thus secures to man pleasure or pain in due course.
In other cases the explanation rests, not on the interven tion of memory, but on fusion of impressions. Thus the white shell appears as yellow as a result of jaundice, the cognition being a blend of the shell perceived without colour, and the yellowness of the bile in the eye, perceived without its substratum. So the bilious man feels sugar bitter, because his taste is a blend of the sugar and bile. The vision of two moons is due to a lack of co-ordination of the rays of light which issue from the eyes and bring back the images.
In the case of merely doubtful cognitions the explanation of their character is that some object is seen as possessed of a quality which produces two discrepant remembrances; thus, seen at a distance a tall object may be either a pillar, or an ascetic buried in meditation and motionless.
As the Mimamsa differs from the Nyaya in its view of the validity of cognitions, so it differs in its attitude to the mode in which a cognition itself is apprehended. In the Nyaya view this is an act of mental perception (manasapraty- aksa), and the Vijnanavada school of Buddhism holds the opinion that one cognition is known by another, though, going further than the Nyaya, it draws the conclusion that, if the first cognition is to be apprehended by the second, it must have form, and form therefore does not belong to any external reality, as the Nyaya holds. The Mimamsa as early as the Vrttikara1 maintains that in apprehension it is the object that is perceived, not the cognition (arlhavisaya hi pratyaksabuddhih, na buddhivisayd}. As expounded by Prabhakara, consciousness (samvit), which is self-illumined, is cognised, but not as an object of cognition, but as cogni tion (samvittayaiva hi samvit samvedyd, na samvedyataya) . To say that the cognition is unknown is absurd, since the cognition of things is possible only if the cognition is known. The mode in which cognition is known is inference; in inference we grasp the existence of a thing only, not its
1 Mimamsa Sutra, p. 9, 1. 16 ; cf. Prakaranapancika, pp. 56-63 ; SaddarSanasamuccaya, pp. 289, 290.
THE PROBLEM OF KNOWLEDGE 21
concrete form; we learn the presence of fire on the mountain from its smoke, but we do not see the actual form of the fire. Cognition, therefore, we infer from the fact that we know things; it, therefore, may be classed as an object of proof (prameya), since it is arrived at by the use of infer ence, which is a means of proof (pramana), but it is not an object of direct apprehension. In Rumania's doctrine1 also this view appears, though the doctrine of self-illumination is rejected; the perception of any object does not result in a further cognition of the perception, but in the direct apprehension of the object, and every act of perception involves a relation (sambandha) between the self and the object; this relation implies action on the part of the self as agent, and this action constitutes the cognition, which is inferred from the relationship between the self and the object.
From this point of view it is possible to understand the definition of the valid apprehension given by Parthasarathi Misra as that which, being free from discrepancies, appre hends things not previously apprehended.2 This definition does not really derogate from the principle of the self- evidence of cognitions; the qualification of freedom from discrepancies merely lays stress on the fact that it is the absence of a sublating cognition which assures us in case of question of the validity of a cognition, while the condi tion that the thing in question should not have been previously apprehended is not a new factor, but merely a formal expression of the essential nature of apprehension. The exact process of cognition as explained in the Siddhantamukt avail* consists in the production in the object of the quality of being cognised (jnatata) , and, however often we cognise the same object, nevertheless in each instance the quality in question is generated anew.
The precise character of the doctrine was, it is clear, largely determined by the desire to avoid the difficulty of
1 Cf. Sastradipikft, p. 37; Manatneyodaya, p. 103 ; TSrkikaraksp, pp. 39 ff.
3 $5stradipika, p. 28.
'- P. 118 ; cf. Kusumafijali, IV 1.
22 THE KARMA-MIMAMSA
the infinite regress, which seemed to be involved in the theory that a cognition could only be known through the instrumentality of another cognition, and perhaps still more by the aim of avoiding the conclusion, which was derived from this doctrine by the Idealist school of Buddhism, that there existed no self, but merely a series of cognitions, held together by no substantial unity. To the Mimamsa such a doctrine was naturally anathema, since the essence of the sacrificial ritual lay in the fact that there was a self who could profit by the performance of sacrifices, not merely in this world but after death. It might have been hard to convince men that sacrifices were worth performing, if the only reward held out had been success in this life, for facts would too often have controverted the claim that sacrifices were availing; when the reward was predicted for the next world, the issue was removed from empirical verifica tion. But the denial of the possibility of introspection thus necessitated was obviously a real difficulty, and rendered the Mimamsa view less plausible than that of the Nyaya, which accepted cognition (vyavasaya} and as supervening upon it consciousness of cognition (amivyavasaya). The dis advantage of the Nyaya view was that it tended to ignore the fact, which was strongly emphasised in the Mimamsa, of the necessary implication of the subject in all cognition. The distinction between the cognition and the subject, which possesses it, is illustrated clearly in the case of sleep; in it, the school holds, there is no cognition normally, and apparently no cogniser or object of cognition, yet the existence of both, despite sleep, is proved by the fact of remembrance of dreams. The knowing subject, therefore, is not, like the cognition, self-illumined, though as to its exact character Prabhakara and Kumarila are far from agreed.
Of forms of apprehension or cognition Prabhakara recognises five: perception, inference, analogy, scripture or verbal testimony, and presumption; while Kumarila accepts also non-perception or negation, in accordance with the view of the Vrttikara, who thus supplements the bare mention of perception in the Sutra (I, 1, 4), where it is defined as the contact of the sense organs with the object, which must be actually present. The analysis of perception given by
THE PROBLEM OF KNOWLEDGE 23
Prabhakara shows on every hand clear trace of derivation from the views of the Nyaya and Vaisesika, which again are ultimately based on popular psychology, such as appears fitfully in the Upanisads and in Buddhist texts. The essential feature is contact between the object and the organ of sense, which is essentially something real; but the unity of consciousness makes it clear that there must be a further contact between the organ and the self, whether directly or mediately. The fact that, despite the presence of objects in contact with the senses, there may be no cognition of them, proves that the contact cannot be direct, but must be mediated by an instrumentality called mind. It is this which prevents all facts being always and at once present to the self, and it is this which perceives pleasure and pain and brings them home to the self. It is through the mind also that the self experiences desire, aversion, and volition. But mind has no qualities, such as colour, smell or taste, and therefore for the cognition of colour it needs the aid of an organ which possesses that quality, namely, the eye, which to possess colour as its distinctive quality must be possessed of light; similarly there must be the nose, composed of earth, for the cognition of smell ; the tongue, composed of water, for the cognition of savours; the skin, the crgan of air, for the cognition of touch; and the ear, consisting of the ether, for the cognition of sound; the organs themselves being imper ceptible.
This doctrine, of course, rests on metaphysical grounds and assumes in its treatment of the organs the doctrine that like must be known by like. The deduction of the exis tence and atomic size of mind by Prabhakara rests on the basis of a doctrine of causation1 which is different from, but allied to, that of the Nyaya, and which is applied to explain the partial and evanescent characteristics of our experience. Causes are either material or immaterial, the latter head covering all the circumstances which, in conjunc tion with a material cause, result in an effect. The immaterial or non-inherent cause may subsist either in the
1 Prakaranapancikd, pp. 52-54 ; cf. Slokavdrttika, I, 1, 4, vv. 157 ff.
24 THE KARMA-MlMAMSA
material or inherent cause, or in the material cause of that cause; thus, when by contact with the fire smell is generated in a substance, the immaterial cause is the contact with the fire, and the contact subsists in the substance itself, while, in the case of the colour of a mat, the colours of the yarns which cause the colour of the mat subsist in the yarns, which are the material cause of the mat. In the case of perception the soul is the material cause, and, as the soul is uncaused, the immaterial cause must subsist in it; in a substance, like the soul, only a quality can subsist, and therefore the immaterial cause of perception must be a quality of the soul, and this can only be some contact with an independent substance, just as the colour of the earth atom is produced by contact with fire. This independent sub stance cannot be all-pervading like space or time, contact with which is from their nature as all-pervading out of the question; it must therefore be atomic, and the only substance which fulfils the necessary condition is mind, residing in the body ensouled by the self, and possessing the power of swift motion, by which it can form a rapid series of contacts, giving the appearance of simultaneity in our mental life. The deduction is ingenious, but unconvincing; it is significant of the consciousness of the gap between the self and the body, which it seeks to bridge by the mediation of the atomic and therefore corporeal, but yet eternal substance, mind.
Of greater philosophical significance is the attitude of the school to the vexed question of the nature of perception as determinate or indeterminate (savikalpaka or nirvikal- paka). The Nydya Sutra (I, 1, 4) poses the problem in its famous definition of perception as knowledge produced by the contact of the sense organ and the object, consist ing of a determination which does not require definition by name (avyapadesya} and is not discrepant (avyabhicari). The precise of this declaration is far from certain, as the ambiguities of the commentors, Vatsyayana, Uddyotakara, and Vacaspati Misra, clearly show, but Dignaga and Dharmakirti developed a perfectly definite theory - in which a clear distinction was drawn between the element of sense in perception and the function of imagination. In the narrowest sense perception is without imagination
THE PROBLEM OF KNOWLEDGE 25
and is unerring(kalpandpodham abhrantam), but this merely gives us a momentary contact with something real, but utterly inexpressible, the momentary unit of experience. All our knowledge is based on this contact, but its content is supplied by the imagination (vikalpa), acting by rules which it itself imposes, a conception which has obvious analogues with the Kantian doctrine of perception.1 Dignaga's view did not prevail, but the problem had been brought by his efforts into clear light, and the later Nyaya seeks in various ways to explain the mutual relations of the indeterminate and determinate forms of perception. Kumarila2 happily ex presses the primitive form of perception as bare observation (dlocana) pertaining to the object pure and simple, and resembling the cognitions that a newborn child has of its environment. Prabhakara's doctrine3 is that the indeterminate perception apprehends both the class character and the specific individuality of the object, but, inasmuch as neither class character, nor individuality can be fully realised save by comparison, the first apprehension, since it is made without such comparison, is indeterminate in character. Deter minate perception arises when the self determines the perception by recalling both these things which it resembles and those from which it differs, thus recognising both its class character and its specific individuality. The part thus played by memory in the determinate perception suggests that it must be deemed as invalid, since the theory of error adopted by Prabhakara finds the source of mistake con stantly in the intervention of memory. But Prabhakara does not accept this objection as applicable to his own view of perception, and it may be argued that the action of memory in this case does not apply to the perception, but to the things which agree with, or differ from, the object perceived. There does not appear to be any very real difference between the view of Prabhakara and that of
1 T. von Stcherbatskoi, Museon, 1904, pp. 129 ff ; Keith, Logic and Atomism, pt. II, ch. ii ; SaddarSanasamuccaya, pp. 33-41.
2 Slokavartti-ka, I, 1, 4, vv. Ill ff ; Manameyodaya, pp. 8-10. The determinate form adds specification as substance, class, quality, motion, or name ; cf. Nyayamaiijarl, pp. 93-96.
3 Prakaranapancika, pp. 54-56.
26 THE KARMA-MlMAMSA
Kumarila, though their verbal expression differs ; Kumarila holds that in cognition in the form of indeterminate perception neither the genus nor the differentia is presented to conscious ness, and that all that is present is the individual in which both these characteristics subsist. Like Prabhakara, he holds that determinate perception is no less valid than indeterminate perception, since it merely makes explicit what is implicit in the indeterminate form.
The views of the school are best understood when brought into contact with the metaphysical doctrine to which they correspond. The essence of that doctrine accepts generality as a real existence which is perceptible as much as individual things, and in the simplest form of perception, therefore, the two aspects of reality are equally present.
The objects of perception include, besides generalities, substances, qualities, and, in the view of Kumarila, but not of Prabhakara, motion. The Nyaya holds that there are six forms of contact in perception; substance is perceived by conjunction; qualities by their inherence in what is in conjunction, and so also the generality of substance ; generality of quality by inherence in that which inheres in that which is in conjunction ; sound as a quality of ether, a portion of which forms the organ of hearing, is perceived by inherence, and its generality by inherence in that which inheres, while negation and inherence itself are perceived by a peculiar and artificial variety of contact, styled the relation of qualification and qualified. Prabhakara, though he accepts the doctrine of inherence, denies genus to quality, motion, and sound, and so contents himself with recognising the first, second, and fourth forms of contact as valid, and with pointing out that to perceive qualities, there is requisite the contact of the substance and the organs, of the organs and the qualities, of the organs and mind, and of mind and the self. Sub stance and qualities, he holds, may be perceived apart. In Rumania's school, however, which denies inherence, the contacts are reduced to simple conjunction, and identity with what is in conjunction (samyuktatddatmya), the second covering perception of generality of substance, quality and
THE PROBLEM OF KNOWLEDGE 27
motion, while the generalities of these two can be perceived by a relationship of identity with that which is identical with that which is in conjunction.
A further technicality, also found in the Nyaya, is the discussion of the exact nature of the means of proof and its result. If the term Pramana is understood as " means of proof," then perception denotes one or other of the contacts between object and organ, organ and mind, mind and soul, each of which is essential to the result (phala) , in this case the mental percept. If, however, Pramana denotes the cognition itself, then perception signifies the mental percept, and its result is the attitude of acceptance, rejection, or indifference of the subject to the object presented to him in the cognition.
Inference in the view of the Vrttikara1 is the apprehen sion of a thing not before the subject, by reason of the perception of some other thing, between which and the first object we know an invariable connection to exist. The relation, according to Prabhakara, must be both general and constant; examples are the relation between the class and the individuals; substance and quality; the qualities of the same substance; or cause and effect. Smoke stands in an invariable relation to fire, but not vice versa, for on the Indian view glowing iron emits no smoke. Even in dividual events may thus be related in Rumania's view; thus the sight of the constellation Krttika suggests the proximity of Rohini. How, then, is this relation to be recognised? The Nyaya view, when it realised the question as a result of the introduction by Dignaga and, following him, Prasastapada of the conception of a universal relation ship (vyapti) in lieu of mere reasoning by analogy, found refuge in the development of a transcendental perception (alaukika pratyaksa2), by which in perceiving, for example, fire and smoke, the percipient recognised not merely the
1 Mimdmsd Sutra, p. 10; Prakaranapancikd, pp. 64-87; Sloka- vdrttika, pp. 345-405 ; Mdnameyodaya, pp. 11-46; Nydyamanjarl, pp. 109-41 ; Logic and Atomism, pt. II, ch. iii.
2 The Mimarhsa rejects wholly the perception of Yogins, which is the precursor of this idea in the early Nyaya ; cf Nydyamanjarl, pp. 93 ff.
28 THE KARMA-MlMAMSA
connection of the individual fire and smoke perceived by him, but that of fire and smoke in their general aspect. Prabhakara, however, does not recognise this view, the elaboration of which is characteristic of a later epoch. He denies that sense perception can give the knowledge of a universal connection, since it deals only with particular times and places; he also rejects the view that the connection can rest on inference or presump tion, since obviously thus there would be a regressus in infinitum; nor will he accept the view that it is due to mental activity only, as suggested by the doctrine of Dignaga, since if the mind had this power, why is man not omnis cient ? His own view is that fire and smoke are perceived by sense as in relation to each other, as qualified by certain conditions of place and time. By repeated experience the impression is gained that, while the presence of smoke is always accompanied by the presence of fire, the reverse relation does not hold, but is qualified always, unlike the former, by special conditions of place and time. Hence emerges the recognition of the permanent relation of smoke and fire, so that the sight of smoke immediately produces the conception of fire. He admits that we do not by inference arrive at any knowledge which we had not before, but he does not admit that this is any defect to the inferen tial process, which does not involve novelty of result. The school of Kumarila, however, in accordance with its defini tion of apprehension as involving knowledge of something not previously apprehended, points out, with perfect truth, that the actual inference gives us much more than the mere knowledge of the connection of smoke and fire, which is already known; it enables us to infer the presence, at a particular time and place beyond our vision, of the existence of fire as result of the perception of smoke. Cidananda1 recognises also the part played by the reductio ad absurdum in arriving at the knowledge of the universal connection.
The relationship, however, which affords the basis of inference, need not refer merely to things which fall within the limits of perception (drstasvalaksana); matters which
1 Manamcyodaya, p. 15,
THE PROBLEM OF KNOWLEDGE 29
are supersensuous (adrstasvalaksana) may equally be in ferred; thus Prabhakara deduces from the general principle of the relation of cause and effect the existence of the capacity, e.g. of fire to burn. In the Vrttikara the distinc tion appears as pratyaksato drstasambandha and samanyato -drstasambandha, terminology reminiscent at once of the Nyaya Sutra and of Prasastapada; the latter is illustrated by the inference to the sun's movement from the observation of a man's change of place as following on movement.
Following Dignaga and Prasastapada, but in disagree ment with the orthodox commentators on the Nyaya Sutra, the Mlmamsa distinguishes between the inference for oneself, which is the true logical process, and that for another, which is in reality enunciation for another person of the process of reasoning, which leads to his drawing the conclusion already arrived at by the first person. In inference for one's self the process is that something is perceived, and recognised as invariably connected with something else, which thus is recalled to the mind; in inference for another a formal order of statement is usually adopted. First the proposition to be established is enunciated, e.g. " The mountain is fiery," the enunciation serving to bring before the mind any contrary judgment which might sublate it. Then the ground for the conclusion thus set out is given in the form of a general rule, supported by a corroborative instance, e.g. " Where there is smoke, there is fire, as in a kitchen." Finally, the necessary link between the conclusion and the general principle is supplied by the statement that the middle term exists in the subject, e.g. " The mountain is smoking." The order of the propositions is not regarded as of importance by Prabhakara or the other members of the school, who agree in rejecting the more complicated scheme of the Nyaya in which, with a certain redundancy due to its origin in dialectic, the argument is expounded in the five propositions, e.g. "The mountain is fiery; Because it is smoking; Where there is smoke there is fire, as in a kitchen; And this (mountain) is so (possessed of smoke with which fire is invariably concomitant) ; Therefore is it thus (fiery)." The omission of the last two members is no material injury to the scheme, while Buddhist logicians
30 THE KARMA-MIMAMSA
reduce the scheme to two members only. The retention of the example is due to the origin of inference as a process of reasoning by simple analogy; even when the necessity of a universal connection was asserted by Dignaga and adopted by Prasastapada and his followers, the example was religiously attained, and it is not until the latest days of the Nyaya that we find Laugaksi Bhaskara declaring that the example is a mere superfluity. But Prabhakara and the school of Kumarila are agreed in insisting on the use of positive instances only, rejecting the process of argument from such a general proposition as, " Where there is no fire, there there is no smoke, as in a lake," though Kumarila himself recognises its utility, though not its necessity, or, as in the Buddhist view, sole validity.
In the case of inference also there arises the problem, already seen in regard to perception, of the exact force of the term Anumana and the corresponding result. If Anumana is used as equivalent to " Inferential Cognition," which is more precisely designated Anumiti, then the fruit or result is the attitude of acceptance, rejection or indiffer ence assumed by the knowing subject to the inferred result. If, however, Anumana is referred to the means by which the cognition is attained, there is a divergence of view as to the exact process to which the name should be applied. The most immediate cause of inference is the perception of the middle term or minor proposition, e.g. "The mountain is smoking," but a more scientific Nyaya view accepts as the true Anumana the whole mental process, including the consciousness of the relation between the middle and the major terms, through which the major term comes to be pre dicated of the minor term, e.g. fire of the mountain. The result in either case is the inferential cognition itself.
The doctrine of fallacies is deduced both by Prabhakara and by Kumarila from the definition given by the Vrttikara of the nature of inference. Thus Prabhakara holds that the condition, that the relation between the two terms whence the inference is deduced, must be previously known, precludes all those cases styled in logic cases of the too restricted middle (asddhdrana) , where the middle term, which it is proposed to use as a basis of proof, is connected with the subject
THE PROBLEM OF KNOWLEDGE 31
of the inference alone, thus permitting no further conclusion. Earth, for example, has odour, but nothing further can be derived from this unique relationship. Again the relation must be universally valid, a rule which excludes the too general middle (sddhdrana). It is impossible to prove that sound is eternal because it can be known, since many things can be known and yet are not eternal. The necessity of some relation existing excludes the variety of middle term known as annulled (bddhita); to prove sound eternal because it is a product is impossible, since the character of being a product is flatly inconsistent with eternity. Finally, the necessity, that the middle term should be perceived as the basis of the attribution of the major to the minor, excludes the variety of middle term known as unreal (asiddha); thus the per ception by the Buddha of righteousness and unrighteousness on the ground of his omniscience is an illegitimate argu ment, since the omniscience of the Buddha has never been perceived. No other form of fallacy of the middle is accepted by Prabhakara ; he rejects the Nyaya view of the fallacy of the counter-balanced middle (satpratipaksa); which balances against the argument, e.g. of the impercep- tibility of air because of its lack of colour, the argument of its perceptibility because of its tangibility. Prabhakara's argument is that it is not possible for contradictory predi cates, such as lack of colour and tangibility, are thus assumed to be, to exist in respect of one subject; hence one of the two alleged inferences is wholly invalid, and there is no true counterbalancing. He holds that really contradic tory inferences are possible only of some subject whose nature is unknown, in which case, however, in the absence of the essential known relation, no true inference is attain able.
The views of Kumarila do not differ materially from those of Prabhakara ; he classifies the too restricted and the too general fallacies under the head of doubtful (anaikdntika) , and adds as a third class the case of con flicting inferences, which he accepts, contrary to the views of Prabhakara. Of the unreal (asiddha) and the contra dictory types of fallacy he gives various sub-divisions. In
32 THE KARMA-MlMAMSA
this and in his elaborate examination of the generality (samanya), which lies at the basis of inference, he shows plainly his close relation to the Nyaya and his polemic against the Buddhist views. In accord with the older view accepted in Buddhist logic, Prabhakara recognises not merely fallacies of the ground (hetu), but also of the minor (paksa), the example (drstdnta), and even of the proposi tion (pratijna), which in the Nyaya view are all reduced to special cases of fallacies of the ground.
Analogy or comparison is accepted by both schools of MImamsa with the Vrttikara,1 but their view of the exact nature of this form of proof differs from that of the Nyaya generally, which accepts analogy as a distinct form of proof. In the Nyaya view the process results in the cognition that an object, hitherto unknown, when brought within the range of perception, is recognised, by reason of its similarity to something already known, to be the object designated by a name communicated by some person of experience. Thus a man who has never seen a buffalo in his life is informed by a forester that the buffalo is like the cow ; on entering the waste he sees an animal similar in appearance to the cow, and formulates the judgment, " This thing is a buffalo." The precise force of the judg ment is disputed in the school, but the best opinion is that it applies not merely to the single animal seen, but that the precipient acquires a correct apprehension of the specific nature of the whole class buffalo. Thus, as Udayana2 says, the effect of this means of proof is to give a clear under standing of the meaning of a word, though he rejects the view, held by Bhasarvajfia3 and his followers in the Nyaya school, that analogy can be reduced to a particular instance of verbal testimony (sabda), as well as that of the Vaisesika, which reduces analogy to inference. The Mimamsa view of the analogical cognition is that it consists
1 P. 10; Prakaranapancikd, pp. 110-12 ; Slokavarttika , pp. 433- 50; Mdnameyodaya, pp. 47-51 ; Saddarfanasamuccaya, pp. 292, 293.
2 KusumdnjaH, III, 8-12 ; Nydyamanjari, pp. 141-49 ; Tdrki- karaksd, pp. 84-93.
* Nydyasdra, pp. 30, 31. The Jain view (SaddarSanasamuccaya, pp. 205, 206) reduces it to recognition, a form of Parok^a.
THE PROBLEM OF KNOWLEDGE 33
in the recognition in an object not presented to the senses of similarity to an object which is actually perceived. Thus, in the instance taken above, the judgment arrived at is, " The cow which I saw in the city is similar to the animal1 I now see in the forest." Both schools agree in this view, though Prabhakara regards similarity as constituting a distinct metaphysical category, a position denied by the school of Kumarila, who treat it as a quality arising from the fact that more than one object possesses the same set of qualities. The separate character of analogy as a means of proof is deduced by distinguishing it from the other means which it resembles. Thus it is not perception, since the cow is not perceived at the time the judgment is formulated; it is not verbal testimony, for it involves only the perception of simi larity; it is not inference, as the mental process is quite different; nor is it mere memory, for the similarity is not remembered. The Nyaya view is declared erroneous; the assertion that the buffalo is like the cow cannot be assumed as a basis for the conclusion, since, as a mere human utterance, it may be untrustworthy; the cognition of the buffalo and its similarity to the cow is pure perception; the conclusion that the animal seen is what is denoted by the word " buffalo " is merely inference, so that, if the Nyaya view is adopted, there is no real independent form of proof called analogy (upamdna), or true analogical judgment (upatniti). While the polemic against the Nyaya is not unsuccessful, the discussion makes it clear that there is no real separate sphere for analogy as a means of proof.
Unlike the Nyaya, both schools of Mimamsa accept, with the Vrttikara, presumption (arthdpatti) as a separate means of proof.2 But Prabhakara's analysis of this form of demonstration differs radically from that of Kumarila. Pre sumption in his view arises when it is necessary to assume some fact in order to avoid inconsistency in respect of some thing which is actually perceived. Thus, if we know that a
1 For " the buffalo " ; the Mimamsa view does not recognise the previous information as to the likeness of the cow and buffalo.
2 Mimdmsd Sutra, p. 10; Prakaranapancika, pp. 11318; Slokavdrttika, pp. 450-72 ; Manameyodaya, pp. 51-57; cf. Nyayaman- jari, pp. 36-48; SaddarSanasamuccaya, pp, 293-95.
3
34 THE KARMA-MlMAMSA
man who is alive is not in his house, we must assume that he has gone out, in order to make our thinking consistent with our perception. To give rise to presumption there must, Prabhakara holds, be doubt, which the presumption removes, and this element serves to distinguish presumption from inference, since inference can only begin when a certain fact, e.g. the existence of smoke, is known with perfect certainty. On the other hand. Rumania's view is that pre sumption is impossible, if the original fact were in doubt; it is only because the absence of the man from his house is for certain known that it can come into operation; the origin of presumption lies rather in the apparent inconsistency of two equally certain facts, in this case, the man's absence and his being alive, which leads to the enunciation of a presumption to reconcile the apparent discrepancy, and it is this reconciliation of apparent discrepancies which marks out presumption from inference. The Nyaya on the contrary finds place for presumption under the purely- negative (kevalavyatirekin) form of inference, in which it is impossible to adduce a positive instance of the general rule, but the Mimamsa could not accept this view since it declined to regard the use of the negative form in inference as satisfactory.
Unlike the Nyaya the Vrttikara1 accepts non-existence (abhava), or, as it is also termed, non-apprehension (anupalabdhi) , as a separate means of proof. The argu ment in favour of this view adopted by Kumarila is that the absence of any thing, e.g. of a jar on a particular spot of ground, cannot be the object of direct perception, which admittedly, according to the definition of the Mimamsa Sutra, requires a present contact with the organs of sense, nor can it be arrived at by inference, analogy, presumption or verbal testimony. It can only arise into an object of knowledge through the fact that none of the normal methods of cognition can come into operation, and this peculiarity distinguishes it from any of these means. Prabhakara,
1 P. 10; Prakaranapancika, pp. 118-25 ; Slokavarttika, pp. 473- 92 ; Manameyodaya, pp. 58-62, 114-18 ; cf. Nyayamaiijari, pp. 49-54 ; §addarSanasamuccaya, pp. 295-98; it is refuted from the Jain stand point, ibid. pp. 206-7,
THE PROBLEM OF KNOWLEDGE 35
with the Nyaya and Vaisesika, declines to accept non- apprehension as a distinct means of proof. When we say, " The jar is not on the ground," all that we mean is that, if the jar were on the ground, we would perceive it there, but that as a matter of fact we see the ground alone. The seeing of the ground is mere perception, and the further statement is merely a qualification of what is perceived in terms of something which, formerly seen along with it, is not now present. In this there is no separate mental process leading to proof. The Nyaya also escapes the difficulty by adopting a peculiar doctrine of its own, under which non-existence, regarded as a positive entity, is perceived by a peculiar mode of contact known as the relation of qualifier and qualified.
Whether, however, four, with Prabhakara, or, with the Vrttikara and Kumarila, five means of proof other than verbal testimony or scripture are reckoned, all these means of proof are subject to the defect that they do not avail to determine the nature of Dharma, man's duty and righteous ness. This is established by the Sutra (I, 1, 4) for the case of perception; that means of proof deals only with existing things which can be brought into contact with the organs of sense, but duty is a thing which is not already existing, but needs man's action to bring it to fruition, and duty is not tangible so as to be able to come into contact with the organs. Inference, analogy, presumption, and non-apprehension, all have relation to perception, and for that reason are vitiated by the defects of the latter, as we gather from the Vrttikara, who thus supplements Jaimini. On the other hand, Jaimini declares that the relation of the word to its meaning is natural and eternal, and Vedic injunc tions are, therefore, the source of knowledge of duty, which is something not open to ordinary means of apprehension. Such injunctions are authoritative, according to Badarayana as cited in the Mlmdmsd Sutra, because of their independ ence. In the definition of the Vrttikara1 scriptural cognition (sastra) is the cognition of some thing, which is not percept-
1 P. 10 ; Prakaranapancika, pp. 87-110, 131-40, 161-70; S.oka- varttika, pp. 405-33, 498 ff, 728 ff ; Mtnameyodaya, pp. 40-47; cf. Nyayamanjari, pp, 150 ff, 205 ff.
36 THE KARMA-MlMAMSA
ible, through the instrumentality of intelligible sounds, that is words, whose meaning is known. The further analysis of Prabhakara shows that each word is composed of letters which are severally apprehended, impressions of the earlier letters blending with that produced by the cognition of the last letter to bring about the idea of the whole word, which alone has the power to bring about the comprehension of a single definite meaning. The letters, then, are the means of verbal cognition, since it is they which by combination compose the word and bring about the comprehension of its meaning. With Rumania Prabhakara agrees in disregard ing the grammatical school's doctrine of Sphota, an entity which is invented to meet the difficulty felt by the grammar ians as to the possibility of any combination of impressions from individual letters producing the unity, which enables us to comprehend the meaning of a word, and in this view the Vedanta, Nyaya, Vaisesika and Samkhya are at one with the Mlmamsa, leaving the Yoga only to support the doctrine of the grammarians.1
The meaning of words is declared by Jaimini to be natural (autpattika), and Prabhakara insists on the fact that words cannot be supposed to owe their meanings to conven tion, whether human or divine. The view of the school in this regard can hardly be regarded as anything else than an attempt to bring the doctrine of verbal testimony into har mony with their traditional beliefs in the nature in the Veda, which doubtless long preceded their speculations on the nature of the relation of word and meaning. The Nyaya view, that meanings were given to words by a convention due to the action of God, offended the Mimarhsa belief that the Veda had no creator, and that no God, as understood by the Nyaya, existed. The alternative of human convention con tradicted flatly the Mimarhsa belief that the essential function of the Veda was to lay down injunctions for the performance of actions, whence arose an invisible potency (apurva) leading to a desirable end, and that this potency was a thing of
1 Cf. Max Miiller, Six Systems, pp. 527-44 ; a full refutation of the doctrine of Bhartrhari that Sabda is the source of the world and is the lower form of the absolute, Brahman, is given in Nyayamanjan, pp. 531-36; cf. SarvadarSanasamgraha, ch, xrii,
THE PROBLEM OF KNOWLEDGE 37
which no person, save through the Veda, could have any knowledge. The Nyaya argument in favour of convention, derived from the case of proper names, is met by the admission that in the case of such names convention is active, but that common names stand on a different footing. In the former case, we know that the persons or things so called have a beginning in time, and that some person must have applied the names to them; in the case of common names we have no warrant for finding a begin ning in time for either the things or the words. There has been no beginning of the world or of men, and they must have from the first talked of the things of the world, just as in actual life it is from observing the conver sation of his elders, or by their instruction, that a youth learns the meanings of words. What is still more conclusive evidence is that, unless we recognised, as we do, that words possess of their own nature meanings, we could never form the conception of conventional meanings, which is a later development.
The eternity of the word is established formally and at length by Jaimini in a systematic refutation (I, l,6-23)of the objections directed against the doctrine by the Nyaya school in particular. The Nyaya1 holds that the eternity of the word is precluded by the fact that it is perceptible only after effort; that it is evanescent; that in common parlance men talk of producing a sound, just as they speak of produc ing any ordinary article; that the same word is pronounced by many people and in many places; that words have changes in form, such as dadhy atra for dadhi; and that, when uttered by many people, the volume of sound is increased. The reply of Jaimini insists that the apparent production of sound, regarded by the Nyaya as a creation, is only a manifestation of a pre-existing entity, a fact in harmony with the disappearance of words on the cessation of the manifestation, while products proper remain in being. The analogy of the sun refutes the argument from simultaneity of perception by many persons; the change to
1 The Sutra (II, 2, 23-59) deals with the topic, but in such a way as to ghow in all likelihood posteriority to the Mim&msa Sutra.
38 THE KARMA-MIMAMSA
dadhy atra is not a modification of the letter i, but the substitution of a quite different form; increase of magnitude refers to the tone, not to the word itself. Positive argu ments for the eternity of the word are not lacking. If it were not so, it would fail in its purpose, the conveying of a meaning to another. Again, we do find in point of fact that men recognise words as being the same when uttered on diverse occasions by diverse people. Language supports the MImarhsa case; when a word is repeated, we talk of ten repetitions of the word, not of ten words. Moreover, no cause for the destruction of words is adduced, and in non- eternal things causes of destruction are always to be found. Finally, there is Vedic authority for the doctrine and no valid counter authority.
The word then exists ever, but only from time to time by effort on the part of some being is it made manifest to us. But effort is not enough; the deaf do not hear, and the effort must be supported by a suitable organ which aids in the cognition of the word. Through the effort on the part of the speaker, the air from his lungs rises upwards and comes into contact with the vocal chords, by which it is modified in character. Passing, then, out from the mouth, it reaches the ears of those near enough to be affected, produces in their ears a change favourable to audition, and passes out, bringing to a close the audition. The ear cavity contains a layer of air, upon which the air current issuing from the speaker's mouth impinges, producing the condition on which audition supervenes. Thus the MImamsa rejects the primi tive conception under which, as light from the eye travels to its object and brings back vision, so the sound travels in some form to the source of the sound, as held by the Jains, and the Samkhya view that the sense of hearing, as all- pervading, reaches the place of the sound. It also rejects the Buddhist view that actual contact is unnecessary for hearing, and the Nyaya-Vaisesika doctrine of propagation of sound on the analogy of waves, or the filaments of the Kadamba flower, in the ether until it reaches the ether enclosed in the ear cavity, which, on that view, constitutes the organ of hearing. To this opinion Kumarila objects that, the ether being one and indivisible, if one ear is
THE PROBLEM OF KNOWLEDGE 39
affected, all ears should equally be affected, and every sound be heard by every one; or, again, if one man is deaf, everyone should no less be deaf. The Mimamsa evades this objection by the doctrine that the ear cavity contains air, and that it differs in size and shape from man to man. A further objection to the wave theory is also based on the fact that sounds travelling with the wind are heard at further distances than sounds travelling against the wind, which is inexplicable if the propagation of waves takes place in the ether, which, of course, is unaffected by wind. The essential character of the word is, in the view of Jaimini, not mere denotation, but injunction, a view which clearly stands in close relation to the doctrine that the meaning of words is largely learned by the young from the observation of intercourse among the old ; one addresses the other, and the other acts as a result; one says, gam dnaya, the other brings the cow. Hence, as against the Vedanta, it is denied that the essence of Vedic texts lies in the making manifest of the sole existent Brahman, and asserted that, even when this seems to be the case, the real import of the text is an injunction to meditate on the Brahman. From this view Prabhakara proceeds to develop a conclusion, which is in harmony with the view of Sabarasvamin, that words themselves have no meaning, and obtain it only in sentences, properly injunctive clauses; gam by itself is nothing, but attains meaning when conjoined with anaya, the whole then signifying generically the genus cow as con nected with bringing. This view in the school obtains the name of the theory of signification in syntactical combination (anvitabhidhdna) , in opposition to the view of Kumarila, who admits that words possess a meaning independently of combination in injunctive sentences, and whose theory accepts, therefore, the combination of significant terms (abhihitdnvaya) . The two schools, however, are at one in holding that the signification of words is a class signifi cation (I, 3, 30-35), as the theory of the eternity of words demands. The modern Nyaya, on the other hand, insists that the import of words is always the concrete individual, while the older Nyaya (II, 2, 61-71) adopts the doctrine that the word expresses the class (jdti), individual (vyakti),
40 THE KARMA-MlMAMSA
and characteristic mark of the class (dkrti), all at once. The Buddhist view, again, insists on negative determination only, on the ground that it is impossible to determine positively the specific nature of anything, and that all that can be attained is a series of negations, a view which Kumarila refutes at great length,1 insisting that the doctrine would mean the impossibility of distinguishing between any two things, since, for instance, both the cow and the horse are negatives of the elephant, and could never deal with individuals, each negation being necessarily general. The defender of Apoha thus accepts the existence of negative classes, which must be all identical, since he cannot rely, ex hypothesi, on any positive means of discrimination between them.
The case for the claim that words denote individuals, as put in the Sutra (1,3,33) in the view of Prabhakara, is that, if it were not so, all injunctions, Vedic or profane, would become meaningless; number and gender would be out of place with regard to nouns; there could be no words to express qualities, and agreement between noun and adjective would be impossible. Prabhakara replies by insisting that, as indicated by the Bhasya, if words had individual meanings, such a sentence as, " One should pile the fire altar in the form of a kite," would be meaningless, as it cannot be supposed that such an injunction was intended to refer to an individual kite, while its plain meaning is a reference to the class " kite." Without this element of generality all injunctions are absurd, and the necessary individual reference in certain cases is obtained through the generality, with which it is inseparably connected. Kumarila adds that this view is supported by the fact that the word " cow," as experience shows, does not suggest to us an individual cow, but the class; if individuals were denoted by words, a generic idea like " cow " would be impossible, and even if possible would merely consist of the impression of all the peculiarities of all cows known to the thinker. Again, the word cannot denote all the individuals, since this would mean that the
1 Slokavarttika, pp. 566-614; Nyayamanjarl, pp. 303-8; contrast Ratnakirti's Apohasiddhi (Six Buddhist Nyaya Tracts, pp. 1-19).
THE PROBLEM OF KNOWLEDGE 41
word possessed as many potencies of denotation as there were individuals, and, as all the individuals could never be known, the word would never fully be comprehended. Nor could the word denote an aggregate of individuals since all individuals cannot be known, and therefore an aggregate could not be known, while, even if it could be known, as its constituent parts would be ever perishing or replaced, the signification of the word would be a constant flux. Finally, if the word meant a single individual only, there could be no eternal connection between word and meaning, and, in the absence of any means of discovering which individual was meant, action would be impossible. The difficulty regarding sacrificial operations in regard to a class instead of an individual, is disposed of by pointing out that such actions must be performed by substances, and that the injunctions specify, not the individual substance to be used in any special case, but the class of sub stance, individual portions of which are applied by each sacrificer.
As regards the authority attaching to words there is a sharp distinction between the older school, followed by Pra bhakara and Kumarila. Prabhakara holds that the only authoritative testimony to things beyond the reach of the senses and other means of proof is the scripture (sdstra). Other words deal only with matters cognised by perception, inference, etc., and have no inherent cogency; if they give us true information, it is merely because we believe the speaker to be trustworthy. Thus, like the Vaisesika school,1 Prabhakara holds all cognition of this kind to be based on inference, the argument being, " This man says something ; he must know what he is talking about ; what he says, therefore, must be true." From another point of view, human words are of no higher value than re membrance, which is admittedly no source of valid universal judgments. Thus the sole possibility of the validity of verbal testimony lies in the Veda, which has no author, and therefore is not vitiated by the doubts as to trustworthiness and ability of correct expression, which precludes us from
* Cf. Nyayamanjari, pp. 152 ff.
42 THE KARMA-MIMAMSA
according implicit belief to the assertions made to us by any merely human authority.
There is an obvious difficulty in this reasoning, when it is remembered that Prabhakara, like the Vrttikara, insists on the self-evidence of cognitions, from which it would seem to follow that the assertions of any man are prima facie valid, until sublated by better evidence. Kumarila, who is always anxious to accommodate the views of the school to popular beliefs, is at the same time more in harmony with the tenets of the school in adopting a doctrine, which does not involve the general denial of the validity of human testi mony. He adopts, therefore, the plan of distinguishing testimony as human and super-human (apauruseya) , while admitting both as valid, though for different reasons. In the case of the Veda there is no author, and there fore the possibility of defects is absolutely precluded. In the case of human testimony its validity may be impaired by defects in the speaker, but the presence of excellencies in him precludes the presence of defects, so that if we are assured of the latter we can be assured that the defects do not exist. But it must not be understood that the excellencies positively contribute to the validity of his utterances, which they possess of themselves; the excellencies are of service merely in assuring us of the absence of those defects, which might cause his testimony to be suspect.
The Veda, however, has special claims on our regard, and the Mimamsa Sutra (I, 1, 24-28) meets detailed criticisms of its claim to eternity. Thus it is argued against its validity that parts of it bear names of men, or refer to human beings, to which Jaimini replies that passages bear names of persons who studied them in detail, and that apparent human names in the Veda are really mere cases of homonymns; thus, as Sabarasvamin points out, Pravahana is not the name of a man, but an epithet, " The excellent carrier." Similarly, apparently absurd statements, such as "The cows performed a sacrificial session," are to be under stood merely as emphasising the value of some ritual action by way of hyperbole, not as showing that the authors of the Veda were foolish mortals. The eternity of words, and the fact that it alone serves to reveal the unseen potency, which
THE PROBLEM OF KNOWLEDGE 43
results from obeying its injunctions, are conclusive proofs of the eternity of the Veda, and the alternative view of a creator is needless and unsatisfactory both in regard to the Veda and to the world.
Other forms of proof, both Prabhakara and Kumarila expressly reject.1 Sambhava, which is variously interpreted as probability, e.g. that ten is included in fifty, or much more probably as inclusion pure and simple, is regarded as merely a form of inference. Rumour, which like Sambhava is claimed as a means of proof by the Pauranikas, is patently useless for purposes of proof; its source being uncertain, it is quite impossible to afford its contents any measure of credit. Gesture (cesta) , which is given as a means of proof by the Tantra school, the Mimarhsa ignores.
The relation between the various means of proof is developed by Kumarila; the use of any means of proof such as inference is debarred if there is a more direct mode of cognition, e.g. sense perception, or if the contrary of what is sought to be established is established in advance by the use of some simpler means of proof.
1 Prakaranapancika, pp. 125, 126 ; Mdnameyodaya, pp. 64, 65 ; Slokavarttika,' p. 492 (vv. 57. 58); Tdrkikarahsd, pp. 116, 117; SaddarSanasamuccaya, p. 207.
Ill
THE WORLD OF REALITY
THERE is nothing to show that the question of the reality of the world had ever occurred to the framers of the Mlmdmsd Sutra, but in Sabarasvamin's Bhasya we find the problem definitely faced in answer to the onslaught made by the Nihilist school of Buddhism on the whole conception of the reality of existence as we know it. The doctrine of Nagarjuna,1 doubtless an effective restate ment of tendencies earlier manifested in the Buddhist schools, denies at once the reality of the external world, and of the ideal world which seems to present us with the know ledge of external reality. That much of its dialectic is sophistic is true, but its novelty of view and the energy with which Nagarjuna, an eastern parallel of Zeno, urged his paradoxes, evoked from the orthodox schools elaborate replies, both the Nydya and the Veddnta Sutras seeking to refute heresies so dangerous to their own tenets. The reply of the Mimamsa, in keeping with what appears to be the early character of that Sutra as compared with the Veddnta or Nydya Sutras is given only in the Vrttikara as cited in the Bhdsya.2 An opponent objects, in his version, to the validity of our waking perceptions, on the ground that in a dream we have cognitions which all admit to be without foundation, and, if this is true of one set of cognitions, it may be assumed to be equally true of another. The reply of the Vrttikara is, in effect, that the argument assumes what is to be
1 Mulamadhyanmkakarika, ed. Bibliotheca Buddhica, 1903-1913 ; Max Walleser, Die Mittlere Lehredes Nagarjuna, Heidelberg, 1911 and 1912. Cf. SarvadarSanasamgraha, ch. II; Sarvasiddhantasameraha, ch. IV.
1 Pp. 8-10.
THE WORLD OF REALITY 45
proved, namely, that all cognitions as such are invalid. On the contrary, we can form the idea of the invalidity of dream cognition simply from our having waking cognitions which afford us a basis for discrediting the dream cognitions, and we can explain the defects of dream cognitions by the assumption that the mind in dream is weak and does not act effectively, a view which we can support by the fact that in deep sleep the mind is wholly absent, suggesting that in the dream state it is in a condition intermediate between its effective waking presence and its disappearance. The opponent, however, continues the argument by urging that the object of the cogni tion is really a void, thus discrediting the validity of the cognition. There is, he says, no difference between the object of perception and the idea; the idea is directly perceived, and there is nothing in reality corresponding to an external object. The Vrttikara replies that this view rests on the erroneous assumption that an idea must have a form; it really is without form, which, on the other hand, the external object possesses. What we perceive is not our idea, but something localised as outside ourselves; no idea can perceive another idea, for each has a momentary existence only, whence one cannot be present to another. The opponent contends that the second idea has a certain con tinuity with the first ; as it originates, it becomes known to the first and reveals to it the object, just as a lamp illumines and thus makes known things. Or, put in another way, it is the idea which first originates, and then the object becomes known, having no anterior real existence. The Vrttikara refutes this by insisting that, though the idea originates first, it is not known first; as we have seen, the idea is known by inference from the fact of our cognition of an object, and the actual knowledge and the knowledge of the idea cannot possibly be simultaneous. Though we know an object, we sometimes say we do not know it, that is, that we are not conscious of having an idea about it. Further, ideas are essentially connected with names, while perception is essentially immediate knowledge, in which naming is not necessarily involved. Moreover, if the idea and the object had the same form, as is assumed in the opponent's argument, this would sublate the idea, not the object, which is directly
46 THE KARMA-MlMAMSA
perceived, but in truth the idea is formless and known by inference, while the object is endowed with form and is an object of sense perception. Or, again, the reality of an external world is shown by the fact that we have the idea of a mat only when threads form its material cause; if otherwise, then a man might form the idea of a jar despite the use of threads in the composition of an object; put more broadly, our ideas are not the free result of our mental activity; they are imposed upon us as regards their content by external reality. The argument as a whole thus falls into two parts, the first dealing with the contention that ideas have no foundation (nirdlambana) , and the second with the view that external reality is void (sunya). Both these con tentions are the tenets of the Nihilism of Buddhism, and there is no real ground for doubt that the arguments of the Vrttikara are directed against this contention. Kumarila,1 however, or some predecessor, has interpreted the passage otherwise, treating the first part of the argu ment as directed against the Vijnanavada of Vasubandhu and Asanga,2 which admitted the reality of ideas, while denying that of the outer ^ world, and the second part he treats as a refutation of the Sunyavada of the Madhyamika school of Nagarjuna. Precisely the same fate3 has over taken the corresponding discussions of the Sunyavada in the Nydya and Vedanta Sutras; Vatsyayana still interpreted the former (IV, 2, 25-33) in its true sense, but Vacaspati Misra reads into part of it an attack on the Vijnanavada; in the case of the Vedanta Sarhkara turns the whole pass age (II, 2, 28-32) into an attack on that school, while Ramanuja treats it as refuting both Buddhist doctrines. The causes for these vagaries of interpretation are obvious ; the Sunyavada in its refutation of external reality used the arguments which the Vijnanavada later employed,
1 Slokavarttika, pp. 217-67, 268-345. Prakaranapancikd, pp. 141 ff, 171 (a fragment only); cf. Nyayakanika, pp. 253 ff, Manameyo- dqya, pp. 119-22; NySyamanjarl, pp. 536 ff (Vijnanavada), 548 ff (Sunyavada) .
2 Mahayanasutralamkara, ed. and trans. S. Levi, Paris, 1907-11 ; Sarvadarfanasamgraha, ch. I ; Sarvasiddh&ntasanigraha, ch. IV (ii); SaddarSanasamuccaya, pp. 40, 41, 47.
' • Jacobi, J.A.O.S., XXXI, Iff,
THE WORLD OF REALITY 47
but it supplemented the conclusions it arrived at regarding external reality by t demolishing the value of our ideas. Any reply to the Sunyavada must therefore include an answer which would apply to the Vijnanavada, and later authors like Kumarila naturally thought that the discussion must deal with the more recent and more convincing school of Vijnanavada. But the Vrttikara shows no knowledge of the peculiar terminology of the Vijnanavada, such as its distinction between the Alaya- vijnana, the quasi-permanent consciousness which constitutes the individual until he attains Nirvana, and the particular presentations which are thence derived (pravrtti-vijndna). Moreover, the argument from the dream condition is not peculiar to the Vijnanavada; on the contrary it is a special favourite of the Madhyamikas, occurring in the Mddhya- mika Sutra (VII, 34) and in other texts cited in the Vrtti on that text.
The view of Prabhakara is in accord with the Vrttikara and iheBhasya, but Rumania's interpretation of the passage has the advantage of eliciting from him a most interesting exposition of, and attack upon, the Buddhist Vijnanavada and Sunyavada theories. The discussion shows the close affinity of the two doctrines, and the form of the argument is often complicated by the resort to elaborate syllogistic reasoning, but the whole makes a very creditable effort to refute either the extreme scepticism of the Madhyamika or the extreme idealism of the Yogacaras. The reality of an external world is vehemently insisted upon as the only foundation of the common facts of life, including such distinctions as those of virtue and vice, teacher and pupil. If there were nothing but ideas, all our views would be false, since they essentially rest on the belief in external reality. Moreover, there is a complete counter argument; cognitions, we hold, have real substrata in the external world ; this notion of ours is correct, because it is without contradiction, like the notion of the falsity of dream cog nition. If you reply by denying the validity of the pro bative example which we adduce, then the doctrine that dream cognitions are false would disappear, and you would lose the chief argument adduced against the reality which
48 THE KARMA-MlMAMSA
underlies cognitions as a whole. Moreover, in dream cognitions, which you adduce as examples where there is no underlying reality, we find on examination that there is always a real substratum, however much distorted and disguised. If, again, you argue that the unreality of our waking cognitions is revealed by the fact that the Yogin sees reality far otherwise, we retort by denying the validity of his perception, and citing against him the visions of our Yogins. Nor can we accept the arguments of the Buddhist logicians, such as Dignaga, who assert that the activity of the mind can supply the full complement of notions, which appear to us to reflect reality;1 without an external world all these mental conceptions would be meaningless, for we deal not with conceptions, but with the facts of life.
Against the conception that cognition alone exists to the exclusion of cogniser and cognised, Kumarila contends with special energy. The case for this conception is set out by him with much care as the prelude to his reply to the Sunyavada. It rests on the difficulty of understanding how cognition and cognised can be related. There cannot really be two entities, one formless and one possessing form, for in memory, when no object is present, we still have cognition of form, showing that the cognition has form, and rendering the hypothesis of an external reality mere superfluity. How, again, can there be contact between the incorporeal cognition and the external object ? An object can be perceived only if it has form, but again the form does not exist until it is perceived, which involves contradiction. Again, even if contact were possible, how could two things, in themselves without form, acquire it in this way ? Moreover, the idea we have of a double moon is admittedly erroneous, and therefore cannot rest on reality. So also we use a variety of words of varied gender for the stars, and a masculine word (dardh) for a wife, which would be impossible if reality controlled our ideas. The same thing, e.g. a lovely woman, raises very different feelings in the mind of the ascetic, the
1 Cf. RatnakaraSanti's treatment of inference as internal only, Antarvyaptisamarthana (Six Buddhist Nyaya Tracts, pp. 103-14).
49
lover, and the dog. The same object appears to us in one aspect short, in the other long, and so forth. All this refutes the possibility of an external reality. The true theory is one suggested by the doctrine of impressions, which the Mimamsa itself uses to explain memory and dream cogni tion. There is one thing only, the cognition, but, as the result of impressions left by previous cognitions, there appears the distinction of cogniser, cognised, and cognition, in place of the unity. Each idea is momentary, but it can and does impress its successor; there is no substantial reality like the soul, but a never-ceasing series of momentary ideas, impressed each by the former, gives man the sem blances which we regard in ordinary life as the outer world and the soul.
The reply of Rumania to this ingenious suggestion emphasises the impossibility of belief in the momentary character of ideas and the continuance which the theory requires. If each idea is really momentary, and perishes utterly, as the Buddhists assert, how can it affect the subsequent idea, contemporaneity of ideas being negated by the Buddhist theory? How, again, can impressions create new sensations, as opposed to mere reviving memories? How can the essential distinction of cogniser and cognised be sublated? How can each cognition in an interminable series contain in itself the whole of the past, when manifestly it does not make any attempt to do so? In what sense can our cognition, say of one animal, followed by that of another animal, be said to involve the conditioning of the second by the quite disparate first? We have many cognitions which are not the result of impressions at all. The only possible explanation of the unity and continuity of our mental life, lies in the recognition of a substantial unity in the self.
Kumarila insists that no idea can comprehend itself, and also that no idea can be comprehended by another. He rejects, therefore, not merely the Buddhist view but the Nyaya-Vaisesika conception by which our ideas are the object of mental perception, and the allied Sautrantika1
1 $addarSanasamuccaya, p. 47 ; SarvasiddhSntasarhgraha, IV, ii, 1-7.
50 THE KARMA-MlMAMSA
speculation, which holds that the form of the object is impressed on the cognition. The objection to the Nyaya- Vaisesika view appears to be that the idea is understood by the school to be perceived simultaneously with the object, and, as the perception of the idea requires that it should be provided with visible form, that is, colour and extension, there would be no possibility of demonstrating the existence of the external object, since, the form being cognised with the idea, an external reference would be needless. The objection, it must be noted, is not cogent against the developed form of the Nyaya doctrine, in which it is held that on the actual cognition (vyavasaya) there supervenes the mental perception of the cognition (anuvyavasdya) ; the cognition thus brings reality immediately before the mind, while in a secondary act the cognition itself is made the object of introspection, as in the accepted theory of modern psychology. The Mnnamsa, by ignoring this possible view, renders it necessary to hold that a cognition can never be the object of introspection; it is an entity which is inferred from the fact of cognition; its existence is known, but not as an object of sense-perception of any kind. Mental perception, which the school admits, is thus restricted to those forms of mental activity which are not cognitive.
There remains, however, yet another contention of the Sunyavada which Kumarila seeks to refute. It is based on the view that atoms are invisible, that aggregates of atoms are invisible, that all objects, being composed of such aggregates, are invisible and incomprehensible, and therefore void. The weight of this argument lies in the fact that the Mlmamsa gives a more or less hearty acceptance to the doctrine of atoms, though Kumarila is careful not to bind himself definitively to it. The conglomeration of atoms, it is urged, is impossible, since atoms have no extension, or at any rate no parts, and no contact between them is, therefore, conceivable. More generally, it is also contended that no whole of parts can really exist. If it did, it must either reside in its entirety in each of the component parts, which is positively absurd, or it must reside collectively in all the parts; in this event, even if it can be assumed that it is something over and above the parts, it would be perceived only when all the parts had
THE WORLD OF REALITY 51
been perceived, which would be normally impossible, abso lutely so in the case of a whole of imperceptible parts like atoms. This dialectic, which the Nyaya Sutra (IV, 2, 7-14) also seeks to face, is met with the argument that, as there is an interminable dispute between the opposing schools, the Buddhists who deny the difference of the whole from its parts, and the Nyaya who assert the distinction, the safe course lies in the via media of admitting that a whole is in one sense different from, and in another sense not different from, its constituent parts. A whole, therefore, is not of a simple and absolute character, and resembles an object with variegated hues, but it is not the less real for that. Invalidity applies to doubtful ideas, not to ideas of an object which in itself is not absolute in character. The stock argument of the Buddhists, that if any composite thing is investigated no whole remains after deduction of the com posing parts, e.g. the threads of a mat, is met by the rejoin der, in harmony with the Nyaya, that the same result is achieved on the Nyaya view, which regards the whole as different from the parts ; the whole, in their view, only exists when -there is an agglomeration of parts; if, mentally, you take away the parts, naturally the whole, despite its difference from the parts, disappears also. The further hypothesis, that what is really seen is merely atoms without real unity but visible in numbers, though singly invisible, is naturally rejected as devoid of cogency. Finally, the argument is used that the attempt to ask if a whole resides in the parts, as an entirety in each or collectively in all, is mistaken. The whole is impartite, and the idea of its relation to its individual constituents in whole or in part is a question which arises only in respect of the individual elements, and is meaningless as applied to the whole.1
The value of Rumania's refutation of the Buddhist schools is not inconsiderable; he brings out fully the grave difficulties which meet any effort to account for the facts of life without accepting some permanent entity, and the objec tions to the effort to evade this problem by creating the fig-
1 Slokavarttika, pp. 632-34 (vv. 75-83) ; cf. Nyayamanjari, p. 550 ; Avayaivinirdkarana (Six Buddhist Nyaya Tracts, pp. 78-93),
52 THE KARMA-MlMAlVISA
ment of an unending series of ideas, each of which must be supposed to take upon itself in some form the impressions of the whole previous history of the series. He insists, also rightly, on the impossibility of accepting any purely subjective idealism, but he does not seem to have appreciated the possi bility of discarding this attitude, but accepting an objective idealism. A suggestion to this effect was implicit in the doc trine of knowledge adduced by Dignaga,1 which insisted that inference and other mental acts dealt with ideal contents, but Kumarila was able to reply to this doctrine that the whole scheme was meaningless, as it assumed that there was nothing truly real beyond the unreal play of ideas in the mind. No true objective idealism was, therefore, before his mind, and he is content to assert absolutely the reality of an external world, which is not the product of intellect, but which is known by us, the relation of knowledge to reality being of a peculiar and unique type, involving an activity on the part of the cogniser which does not, however, create the object.
In their positive doctrines as to the nature of the universe there are considerable differences between Prabha- kara and Kumarila. The former admitted, it is clear,2 no fewer than eight categories, while the latter accepted five only. They agreed in regard to substance, quality, action or motion, and generality, but, while Prabhakara accepted the category of inherence from the Nyaya-Vaisesika, and added the three of potency or capacity (sakti), similarity (sddrsya), and number (samkhyd) , Kumarila rejected the three additions of Prabhakara, and also, in this case in agreement with his predecessor, the particularity (visesa) of the Nyaya-VaiSesika. Finally, inherence was also rejected by him. On the other hand, the texts ascribe definitely to him the acceptance of the category of non- existence (abhdva), with a fourfold division of prior negation, subsequent negation or destruction, absolute nega-
1 Slokavarttika, p. 258 (v. 167) . The invalidity of all but indeter minate perception is asserted in Saddarfanasamuccaya, p. 41.
a Prameyaparayana in Mallinatha Tdrkikaraksd, p. 164 ; Mana- meyodaya, pp. 65, 114 ff ; Prakaranapancika, pp. 110, 111 (over looked in Prabhakara School, p. 89).
THE WORLD OF REALITY 53
tion, and mutual negation, sub-divisions which, of course, are simply transferred bodily from the Nyaya-Vaisesika doctrine. Non-existence stands in definite opposition to the other four categories accepted by Kumarila; though regarded as real, it is nevertheless admitted to be essentially relative to the four categories of being (bhava). Prabhakara, however, rejects non-existence, as might have been expected from his rejec tion of non-existence or non-apprehension as a means of proof. The only reality, in his view, in the absence of a pot from a spot of ground is the spot of ground. The particularity of the Nyaya-Vaisesika, which serves to differentiate such things as the ultimate atoms and selves, has no foundation as a separate category, as the differentia tion can be based on the ordinary qualities which these things possess.
Substance1 is that in which qualities reside, and Prabha kara reckons the number as nine: earth, water, air, fire, ether, the self or soul (fitment), mind, time, and space. Kumarila is credited with admitting also the substantiality of dark ness and sound, while others accept gold as a twelfth. Of these earth, water, air and fire all possess colour and tangibility, and accordingly are the objects of the senses of sight and touch, but only when in non-atomic form, for some degree of magnitude is recognised by Prabhakara, as by the later Nyaya-Vaisesika, as a necessary condition, along with touch (sparsa) , of proper sense perception. The other five substances cannot be regarded as perceptible, since they cannot be seen or touched, and therefore are only inferred to exist. In the case of ether the apparent whiteness of it is due to particles of fire in it, while the darkness of night is not a substance, nor is it a quality ; if it were a quality it would be perceptible by day also, and therefore must be deemed to be merely absence of light. A variant of this doctrine in the school of Prabhakara declares darkness to be the absence of the knowledge of light. Kumarila claims darkness as a substance, because it is blue in colour and moves, these two facts being necessarily attributed to some
1 PrakaranapancikG, pp. 24, 54, 77, 84, 141 ff ; Manameyodaya, pp. 6ff, 66ff,'78ff ; Slokavarttika, p. 404 (v. 183); Tarkikaraksa, pp. 133, 134.
54 THE KARMA-MlMAMSA
substance, but the Nyaya denies these facts. Pointing out that a colour can be perceived only in light, and darkness is experienced when there is no light. Sridhara again suggests that darkness is the imposition of blue colour on something else. The necessity of inferring ether arises from the nature of sound, which must be provided with a substratum; unlike Kumarila, Prabhakara sees no sufficient ground to give to sound the rank of a distinct substance, a position which has obvious difficulties in a system which allots so pre-eminent a place to the word.
Air, in Prabhakara's view is neither hot nor cold, the apparent heat being due to fire particles, and the coolness to water particles diffused in it. Kumarila also regards it as perceptible, but does not claim that it has any colour; he rejects therefore the Nyaya view that it can only be inferred, colour being necessary to perception, and adopts the later Nyaya opinion which admits of direct perception through the sense of touch. In this and in many other details his school, if not the founder, clearly largely assimilated the Nyaya- Vaisesika physics, though it is clear that Kumarila himself was not prepared to accept the atomic theory as absolutely essential to his principles. Some of his followers went further, and claimed that ether, space and time were directly perceptible, but on these points the doctrine of both schools seems never to have been developed.
The account of qualities which inhere in substances, and are distinct from motion, given by both Prabhakara and Kumarila shows obvious obligations to the Vaisesika.1 Prabhakara gives as objects of perception the qualities of colour, taste, smell, and touch; number, dimension, indivi duality, conjunction, disjunction, priority and posteriority; pleasure, pain, desire, aversion, and effort, and, like the Vaisesika, distinguishes conjunctions and disjunctions ac cording as they are produced by the action of one or both of the things concerned, or rise mediately through another conjunction or disjunction. Kumarila, like Piasastapada, enumerates twenty-four qualities : colour, smell, taste,
1 Prakaranapancikd, pp. 54, 151 ; Mdnameyodaya, pp. 99-111 ; T6rkikaraksd,'p. 164.
THE WORLD OF REALITY 55
touch ; number, individuality, dimension, conjunction, disjunction, priority, posteriority; gravity, fluidity, viscidity; cognition, pleasure, pain, desire, aversion, effort, impression (covering velocity, elasticity, and mental impression), tone (dhvani), which is a quality of the air, revealing sound, manifestation (prdkatya), a quality common to all sub stances, perceptible and determining them, and potency. Potency is reckoned by Prabhakara as a distinct category; its existence is proved by inference: fire burns normally, but under the influence of some spell it ceases to have that effect ; there must, therefore, be something of special charac ter in the fire by virtue of which it burns. Words al^o have the potency to denote meanings, and so on ad infinitum. It is eternal in eternal things, but transient in transient things, coming into being with them and disappearing when they disappear, and thus differing from impression (sams- kara), which even in eternal things is evanescent. The Nyaya view is sensibly opposed to the recognition of any such conception, since, strictly speaking, the number of potencies in any object might be regarded as very numerous, negating the possibility of accepting potency as one quality or a distinct category of being. Number, which Prabhakara makes a separate category,1 in the list of Kumarila falls to the rank of a quality. The classification of qualities and their assignment to substances follows generally the classifi cation first given in infinite detail by Prasastapada. From his list Kumarila departs only in the substitution of tone for sound, and of manifestation and potency for merit and demerit. Unlike Prasastapada, he denies that cognition is the object of mental perception, though admitting this for the other special qualities of the self. From the school of Prabhakara that of Kumarila differs in asserting that indi viduality applies both to eternal things and to products, while the former asserts that it applies to eternal things alone. Priority and posteriority apply to both space and time ; the later Nyaya wisely rejects both as general quali ties, since they are essentially determinations of space and
1 In Prakaranapancikd, p. 54, it appears as a quality : impression in its various form is referred to (pp. 80, 81) as inferred only ; the full treatment occurred in the missing Prameyaparayana (seeibid. p. 111).
56 THE KARMA-MlMAMSA
time, or, as stated in the Mdnameyodaya, are special quali ties of these entities.
Action1 as a category covers only, as in the Nyaya- Vaisesika, the restricted field of motion, with its traditional five-fold divisions, as throwing up or down, drawing towards or expanding, and motions other than these. But Prabha- kara maintains that it is only an object of inference, while Kumarila holds that it is perceived. The argument of the former rests on the fact that, when we think we see motion, we only see conjunction and disjunction with points of space, these contacts subsisting only in outside space and not in the moving thing, in which the activity of motion must reside. The reply of Kumarila 's school is that it could only be inferred as the immaterial cause of the conjunction and disjunction of a thing with points in space, which would mean that it must subsist both in space and in the thing, whereas it exists in the thing only. We really see motion, which is in the thing and which brings about conjunction and disjunction in space, a doctrine which has now excellent modern support.
Generality both Prabhakara and Kumarila admit as real and as directly perceptible by the senses, and thus set them selves at variance with the Buddhist denial that there is any such thing as generality. The first Buddhist argument rests on the impossibility of the existence of any whole, which both schools of MImamsa deny. But further difficulties are raised. If generality is perceptible and is eternal, as claimed in the MImamsa, the absurdity arises of perpetual perception. Again, how is generality related to the indivi duals; is it present in its entirety in each?2 If so, then there are as many generalities as individuals, and there is mere duplication of names. If not, then it must exist in all collectively, and therefore be entirely unknown, since one can never know all the individuals which make up a generality. If it is eternal, and exists before the individuals,
1 Prakiranapancika, pp. 78-81 ; Mdnameyodaya, pp. 112, 113 ; a wider view is taken in £lokavdrttika, ,p. 707 (v. 74).
2 Prakaranapancikd, pp. 17-32; Slokavdrttika, pp. 545-65, 614- 39; Mdnameyodaya, pp. 95-99; cf. Nyayamanjarl, pp. 297-324; Aioka, Sdmdnyadusanadikprasdritd(Six Buddhist Nyaya Tracts, pp. 94-102) .
THE WORLD OF REALITY 57
it ought to be knowable by itself, which is plainly absurd; if it comes into being with the individual, how is it distinct from it? The reply given by both these schools is an appeal to consciousness; we have as an actual fact direct perception of generality, and we cannot be induced to dis believe it by any process of inference, which cannot have the validity of direct preception. To confute our belief it is necessary to adduce some defect in the organs of perception or a sublating cognition, and neither of these processes is possible. The Buddhists would ask us to disbelieve in the existence of colour, but we decline to do so, and equally we find no cogency in their request that we should refuse belief to the evidence of our senses regarding generality. Nor is it correct to ask whether generality is present in its entirety in each individual or collectively in all; these are concep tions which are applicable to individual things alone, not to the impartite generality, which is not to be compared either with a string which holds together a necklace of pearls, or the many parts which go to constitute a single living creature. It is not to be considered as any particular configuration or shape, but is a distinct entity sui generis. It must not be thought to be perceptible apart from the individual; such an existence of a separable character, if held by the Nyaya- Vaisesika, is definitely rejected by the Mimarhsa. In the ultimate essence, when we analyse our idea of generality, it rests on the fact that, despite differences in things, we recognise in them an essential identity; among cows of many colours and shapes, there is still the same nature. The relation of generality to the individual may be described as that of difference as by Prabhakara, or as different and non-different as in the school of Kumarila, but the view of the two schools is not materially different. Kumarila points out that in ordinary life we recognise things as individual or as generalities, according as we select the one or other of the two aspects which must always be present. The specification of a class is brought about by various causes, such as colour, etc., or time and place. Thus gold is distinguished by its colour from copper, melted butter from oil by its odour and taste, a jar from other articles by its shape, a horse by its neighing, a Brahman by origin,
58 THE KARMA-MlMAMSA
and in some cases by action also, in places where the duties of the castes are duly supervised by the king. Prabhakara, however, declines to admit of generalities such as Brahman- hood and Ksatriyahood, which Kumarila accepts.
Prabhakara also differs from Kumarila in his use of the category of inherence as a means of explaining the relation of the individual to the generality. When a new individual of a class comes into being, what is produced is not the existence of the generality, which is eternal, but of the relation of inherence between the individual and the class. Inherence differs from contact in that it does not presuppose the previous existence of the things affected by it, and, unlike the Nyaya-Vaisesika, Prabhakara does not hold that it is necessarily eternal. This affords an easy reply to the question of the fate of the class character on the destruction of an individual; it does not go away, as it has no motion; it does not subsist in the individual, which is no longer in being; it does not cease to exist, for it remains in other individuals, but the inherence between the class and the individual comes to an end. But Kumarila1 rejects in toto the idea of inherence as a true category; a relationship, he argues, can exist only between things which are established as distinct entities, and, as inherence is supposed to be a relation between things which, like the class and the individual, are inseparable, it is a contradiction in terms.
While Rumania's school admits, as usual, the existence of generalities of substance, quality, and action, Prabhakara declines to accept the last two or a summum genus of existence as a real generality, on the ground that, as each generality rests on the fact of actual perception, the genus existence must be disallowed, as we do not in fact perceive things as merely existing. The true sense of existence is merely the individuality of things (svarilpasattd)', it is not a true class character.
Similarity2 as a category is asserted by Prabhakara, who holds that its existence is proved by our consciousness in
1 Slokavdrltika, I, 1, 4, vv. 146-55 ; cf. ASoka's Avayavinira- karana (Six Buddhist Nyaya Tracts, pp. 78-86).
* Prakaranapuncika, pp. 110, 111 ; Slokavdrltika, pp. 438-41 (vv. 18-23), 565 (vv. 74-77) ; Tarkikaraksd, p. 164.
THE WORLD OF REALITY 59
precisely the same way as every other category. It cannot be held to be substance, for it exists in quality and motion as well as in substance. It cannot, in view of its relation to motion and to quality, be a quality; motion has no qualities, nor can a quality have a quality. It is not generality, for no comprehensive conception of it exists. It is quite other than the relation of inherence. It is not particularity, which in any case is not a true category, since it is no more than the quality of individuality. It must, therefore, be a distinct category, which is perceived in the apprehension of qualities, motions, or parts of two things as common to both. Kumarila's rejection of this category is based on the fact that similarity admits of degrees, e.g. the resemblance of a cow and a buffalo is considerable, that of a cow and a boar is slight; if there were a true category there could be no degrees. He agrees, however, with Prabhakara in regarding similarity as directly perceptible. It consists, in his view, in the fact of the possession by two objects of the same arrangement of parts, and he attributes the erection of a special class of similarity to a misunderstanding by the Vaisesikas of the doctrine of Vindhyavasin, which merely asserted that generality consisted in possession of unity of form (sdrupya), which was taken to mean likeness (sadrsya). The same author is elsewhere1 cited by Kumarila as denying the doctrine of the existence of the subtle transmigrating body, a view accepted from him by Kumarila, and as enunciating the principle of the genesis of inference, which is accepted also in the Slokavarttika. Who this author was is not apparent; he cannot, it is certain, be Isvarakrsna, nor is there any plausibility in identifying him with the Vindhyavasa who plays a part in the history of the Samkhya, whether or not he was really Isvarakrsna.2 He may, of course, have been an older teacher of the Mlmamsa school itself.
Cause is not reckoned by either school as a category, a fact significant of the curious failure of Indian
1 P_P. 704 (v.62), 393 (v. 143).
2 Samkhya System, pp. 62, 69. Gunaratna (Saddarfanasa'-nuccaya p. 104) cites a £loka of Vandhyavasin (!), who was clearly, in his view, not I&varakr§na, but it is hard to say of what value his evidence is, or to whom he refers.
60 THE KARMA-MlMAMSA
philosophers to find a due place for this issue, even when, as in the case of the Nyaya-Vaisesika, they by no means ignore its importance. But there seems no evidence that either Prabhakara or Kumarila contributed anything of novelty or value to the doctrine. In his discussion of perception, as we have seen, the former makes use of the doctrine of the division of causes into the material or inherent (samavayi- kdrana), and immaterial or non-inherent (asamavdyi) , a distinction, doubtless, taken from the Nyaya-Vaisesika.1 The denial by Kumarila of the conception of inherence would have precluded him from adopting such a distinction of causes.
Causation, however, affords Kumarila an argument in favour of his thesis of the reality of non-existence.2 That entity he classifies as prior, as the non-existence of curd in milk; subsequent or destruction, as the non-existence of milk in curd; mutual, as the non-existence of the horse in the cow and vice versa; and absolute, as the non-existence of a horn on the head of the hare. Without the recognition of the first two kinds, he contends, there could be no idea of causation : in its prior negation lies the character of the curd as effect, in its destruction that of the milk as the cause. Everything has two aspects: it regards its self, it exists, as regards anything else it is non-existent; and both these aspect? are real and necessary to each other. It is only through this fact that we can say, " There is no jar on the ground," or that we can ever differentiate things, which is possible only on the ground of a real existence of non-existence. It is impossible to perceive this entity, for perception must deal with the exis tent; the process of intellection is, therefore, purely mental; the ground is seen, the jar remembered, and then ensues the purely mental cognition styled negation, which must be distinguished from inference or any other form of know ledge.
1 It may be noted that Salikanatha commented on the PraSasta, p&dabhasya (Bodleian Catalogue, p. 244).
1 Slokaz-drttika, pp. 473-92 ; Manameyodaya, pp. 58-64, 114-18; cf. Nyayamaf.jari, pp. 49-63 ; Saddarsanasamuecaya, pp. 295-98.
IV GOD, THE SOUL, AND MATTER
THOUGH the MImamsa is so deeply concerned with the sacrifice, it has no belief in the doctrine that the rewards of offering are to be expected either from the deities to whom the offerings are directed to be made, or from a God as creator or apportioner of reward and punishment. The sacrifice generates an unseen potency, whence the goods desired by sacrificers are obtained ; the Veddnta Sutra (III, 2, 40) expressly negatives the idea that in Jaimini's view there was divine intervention in this regard, and the atheism of the true MImamsa is regarded with such unani mity as to render it impossible to explain it away.1 The full development, however, of the doctrine is, as usual, to be found in Prabhakara and Kumarila, who adopted from the Nyaya-Vaisesika the groundwork of their views of the world, but declined to follow that school in its speculations on the existence of a creator.2
The Nyaya-Vaisesika, accepting the doctrine of atoms on the one hand and of the periodical creation and destruc tion of the world on the other, had found it necessary to introduce the conception of a creator, in order to secure in some measure a mode of bringing about the renewal and destruction of the combinations of the atoms and their connection with souls. But Prabhakara and Kumarila alike deny absolutely the validity of the belief in the periodic creation and dissolution of all things ; they accept a con-
1 As does Max Miiller, Six Systems, pp. 275-79 ; cf. K. L. Sarkar, Tagore Law Lectures, 1905, p. 508.
2 Prakaranapancikd, pp. 137-40 ; Slokavarttika, pp. 639-80 ; Mdnameyodaya, pp. 70-74; cf. Ny&yamanjari, pp. 193-204; Sad- darSanatamuccaya, pp. 284 ff.
62 THE KARMA-MlMAMSA
stant process of becoming and passing away, but they find no ground for the systematisation of the process, so as to produce cycles of evolution and involution of souls. Experi ence, Prabhakara urges, shows us the bodies of all animals being produced by purely natural means; we can argue hence to the facts of the past and the future, and need invoke no extraneous aid. Moreover, the whole conception of God supervising the merits and demerits of men is idle; God cannot perceive merit or demerit by perception, since they are not perceptible, nor by the mind, which is confined to the body which it occupies. Supervision also is impossible, even had God the necessary knowledge ; it^must take the form either of contact, which is impossible as merit and demerit being qualities are not subject to contact, or inherence, and plainly a man's qualities cannot inhere in God. If the argument is adduced of the analogy of the carpenter, it may be replied that on this basis the creator would have to be an embodied spirit, and no embodied spirit can affect such subtle things as the atoms or merit and demerit. Nor is it conceivable that the atoms should themselves act under the will of God, for no parallel to such activity is known to us, and, if it were possible, it would follow from the eternity of the will of God that creation would be unceasing. The only true case of supervision known to us is that exercised by the soul over the body, which it occupies by virtue of its merit or demerit, and there is no need to hold that the world is more than an ever-changing sequence of things affected by the souls in it.
Rumania's treatment includes both an elaborate attack on the whole conception of creation and a special refutation of the Vaisesika views. He ridicules the idea of the exis tence of Prajapati before the creation of matter; without a a body, how could he feel desire ? If he possessed a body, then matter must have existed before his creative activity, and there is no reason to deny then the existence of other bodies. Nor is there any intelligible motive for creation ; granted that, when the world exists, conditions are regulated by merit and demerit, originally there was no merit or demerit, and the creation of a world full of misery was inexcusable, for it is idle to argue that a creator could only
GOD, THE SOUL, AND MATTER 63
produce a world in which there is sin and pain. Yet, if his action is conditioned, he cannot be omnipotent. If, again, it is alleged that the creation was for his amusement, this contradicts the theory that he is perfectly happy, and would involve him in much wearisome toil. Moreover, there is no possibility of establishing the reality of his creative activity. It could only rest on reports of the first of created men, and they could have no power to testify effectively to a state of things existing before they were brought into being. They could only rely on what they were told by the creator, and his assertions might be mere boasting. Nor is it at all satis factory to accept the belief in the creation of the Veda, which by no means enhances its value ; still less to hold that it resides with the creator during the periodic dissolu tions of the world, for which, again, there is not a shred of evidence.
Against the Vaisesika view of creation exception is justly taken to the difficulty involved in holding that in some manner the action of the Supreme Lord brings to a stand at one time the potencies of all the souls, and then awakens them all when a new creation is imminent. Against this view it is contended that the activity of men arising from their past deeds can never cease, and it is absurd needlessly to complicate matters by assuming both the force of men's deeds and the intervention of the desire of God. Moreover, it is impossible to explain why this desire should ever arise, and unintelligible to elucidate the mode in which the creator can act without a body or acquire a body.
Kumarila, however, does not content himself with refuting the Nyaya-Vaisesika doctrine; he attacks equally the Vedanta,1 on the simple ground that, if the absolute is, as it is asserted to be, absolutely pure, the world itself should be absolutely pure. Moreover, there could be no creation, for nescience is impossible in such an absolute. If, however, we assume that some other cause starts nescience to activity, then the unity of the absolute disappears. Again, if nescience is natural it is impossible to remove it, for that could be accomplished only by knowledge of the
1 Cf. Nyayamaftjari, pp. 525-31.
64 THE KARMA-MlMAMSA
self, which, on the theory of the natural character of nescience, is out of the question. Nor is the Sarhkhya doctrine of many selves and nature any more tenable as a theory of creation. The beginning of creation is held to be due to a disturbance in the equilibrium of the three constituents which make up nature. But how can such a disturbance take place at a first creation, when there are no potencies due to men's actions demanding fruition? Even at subsequent creations, how do latent potentialities by themselves become fruitful without any consciousness to direct them? And, if they do attain fruition, the Sarhkhya theory of liberation by knowledge is without value, since the potencies will remain able to come again into activity. Knowledge, it must be recognised, can never give freedom from bondage, which can be attained only by the exhaustion of action, for which the Sarhkhya metaphysics affords no adequate possibility, owing to the infinite potentiality of nature.
Though the existence of a creator is denied, the Mlmamsa accepts without reserve the doctrine of the existence of the self or soul,1 and Sabarasvamin elaborates the case for its existence; Prabhakara and Kumarila both develop the theme in close accordance with his views. The necessity of the existence of the self for the Mimamsa rests on its fundamental assumption that the sacrifices are performed to secure, in many cases, a reward not in this life. There must, therefore, be an eternal entity, distinct from the body, the sense organs, and cognitions, which is both the doer of actions and the reaper of their reward. It is not unnaturally objected that there is a strong presumption against claiming eternity for something which suffers change, but the more serious objection is made that men do not realise, when they reap results, the actions which brought these about, thus invalidating the value of the assumed continuity, and that there is nothing unnatural in a man determining to do an act which will lead to evil results in the future, secure in the knowledge that, when
Sutra, pp. 18-24 ; Prakaranapancika, pp. 141-60 ; Slokavarttika, pp. 689-728 ; Mdnanteyodaya,'pp. 78-84.
GOD, THE SOUL, AND MATTER 65
these results come to fruition, he will not remember their cause. To this Kumarila replies that remembrance has nothing whatever to do with the matter; the wise, who alone are worthy of sacrificing, realise in all their actions the law of retribution, even without remembering the facts of each case. Nor is it any argument against the eternity of the soul that it undergoes modifications; we see in actual experience abundant evidence of changes in man's condition in life without any cessation of the substantial identity, and we treat death as no more than a change through which the soul endures. The sea remains, despite the movements of its waves; the serpent uncoils, without change of essence.
To the theory of the substantial soul the Buddhist at once objects, and proposes instead the doctrine of the series of ideas, each of which gathers from its predecessor the impressions of its unending past. The performer, therefore, it is contended, is the same as the enjoyer, but this contention Kumarila rejects. It is impossible to accept this view, he argues, unless the first idea and the last in the series, from performance to result, have a common substratum. Apart from the fact that, if ideas are really momentary, there can be no interaction and no series, it is impossible on the series theory to find any rational basis for action, since the doer will clearly not reap what he did, and action without rational ground is out of the question among men. Moreover, the exact character of transmigration presents insuperable difficulties on the Buddhist theory. It is impossible for an immaterial idea to move about in a living body, much less to transfer itself from one body to another. The hypothesis of a subtle body which serves as an intermediary between one life and rebirth is denied by Vindhyavasin, and unsupported by any evidence, nor, if it existed, is it clear how an idea would pass with it. Nor, again, can the existence of an idea in the embryo be explained. The embryo has no sense organs and cannot have cognitions, and an idea is never known to exist save in the form of a cognition. Nor can it be supposed that the idea exists as a latent potentiality in the embryo without a substratum, while, if the sense organs are assumed to be the substratum,
5
66 THE KARMA-MIMAMSA
then they would possess intelligence, and rebirth would become impossible, since on their destruction intelligence would go also; further, the share played by the organs in developing the idea would contradict the Buddhist doctrine that the idea arises from a preceding idea only. Nor is there any evidence that the first cognition of the newborn child is due to a previous idea; we hold that it arises from the functioning of the sense organs. There must, therefore, be something which possesses the potentiality of ideas, is eternal, and capable of transmigration. This need is furnished by the soul, which is immaterial and omnipresent, and thus, without motion, is able to connect itself with one body after another.
The soul, then, is essentially active, for, unlike the Vaisesika school, the MImamsa does not, according to Kumarila, deem that motion is the only form of action, and it is through its superintending activity that the motions of the body are achieved. We must, therefore, conceive the soul engaged from time immemorial in the work of directing a body, the acts done in each life determining the character of the body attained in the next, a process which will cease only, if ever, when the soul ceases to obtain a bodily habitation.
Again, from another point of view the Buddhist conception of a series is imperfect. Granted that it is impossible to establish a soul merely on the ground of such attributes of the soul as pleasure, desire, or memory, adduced by the Vaisesika school as indications of the existence of the soul, since these might be explained on the theory of impressions, no such explanation is available to dispose of the cognition of the self. In the case of the two judgments, " I knew " and " I know," the theory of ideas breaks completely down. The first idea cannot, as past, know the later idea, nor can the later idea know the first. It is useless to appeal to a series, for the series was not present at the first cognition, nor is it present at the last. Nor is there any unity in the two cognitions, for the Buddhist refuses to recognise any classes. Nor can it be argued that similarity would suffice, for in cognitions of different objects, e.g. a horse and a cow, there is no similarity of cognition. The bare fact of each
GOD, THE SOUL, AND MATTER 67
being in one aspect a cogniser would at most give merely the bare recognition that there was a cogniser, but no personal identity. A true permanent substance is, therefore, essential, and such a substance explains far more effectively than any other hypothesis such phenomena as desire, memory, and pleasure and pain, while it is the indispensable basis of merit and demerit.
This permanent entity is quite distinct from the body, the senses, or cognition. The elements of the body are seen to be without intelligence, and the combination of such elements cannot produce intelligence. If, again, one element alone had this nature, the others could not coalesce with it to form a body. A dead body, which consists of precisely the same material as the living body, contains no intelligence. On the contrary, the fact that a body is an organised whole suggests irresistibly the fact that it serves the purpose of another which directs it, namely, the soul. Such phrases as " I am fat," or " I go," are merely natural transfers of use. On the other hand, the phrase " My body " shows clearly that the ego and the body are different. The same argument can be applied to the case of the sense organs, but others are also available; thus the fact that I feel with my hand what I see with my eyes shows that there is something beyond the sense organs. Again, a blind man remembers what he saw when his eyesight remained, which would be impossible if the organ were the self. More generally the analysis of any cognition reveals to us the fact that the " I " is not the body, nor the sense organs nor the cognition itself, but something over and beyond. Many people can have the same cognition as far as content is concerned, but each cognition has an individual refer ence, as is seen also with perfect clearness in the facts of memory; if there were no " I ", how could we have the fact that one, who has learned half a lesson at one time, can later on resume the task at the place at which he left off ? The objection, that the terms " My soul " indicate a differ ence between the " I " and the soul, is met by holding that in the word "soul" the meaning "cognition" is to be understood, cognition often being inaccurately described as the soul. The result can be confirmed by the evidence
68 THE KARMA-MlMAMSA
of the Samhitas and Brahmanas, the former of which implicitly, the latter explicitly, recognise the existence of the eternal soul.
There must, however, be something to mediate between the eternal and omnipresent soul and the world, else its knowledge would be eternal and omniscient, as emphati cally it is not. The mediator is furnished by mind, whose contact with the soul is the essential condition for its con sciousness in all its forms. For this contact it is necessary that mind should, in contrast to soul which is omnipresent, be atomic, and possess the capacity of extremely rapid motion, a fact which makes our experiences, even when truly successive as they are, appear on occasion to be simul taneous. Mind, however, can exist only in a body, which the soul must ensoul, and then through it the soul comes into contact with the outer world by means of the sense organs. Through the contact of external objects with the sense organs, mediated by the mind, the soul appreciates the outer world ; the mind directly conveys to it knowledge of plea sure, pain, desire, aversion and effort, which are among its qualities. It possesses further qualities : cognition, which is self cognised in the terminology of Prabhakara or, as Kumarila has it, inferred ; merit and demerit, which are inferred ; and impression (samskdra), which is produced by apprehension and results in memory, from whose opera tions it is inferred. The principle of impression, more over, really applies to merit and demerit, for these exist in the form of impressions of past activities, and can hardly be said to be separate qualities, since they merely sum up in terms of moral value the nature of the accumulated impressions ; hence, though they appear as distinct elements in the Nyaya-Vaisesika lists, one list of qualities attri buted to Kumarila more logically leaves out merit and demerit. Further, the soul possesses *he common qualities of number, namely, unity ; individuality; dimension as omnipresent as opposed to atomic, or of the same size as the body as held by the Jains; and conjunction and disjunction with mind. Nothing is more obscure than this relation between the soul and the mind. It is said to be brought about by merit and demerit, but it is obvious
GOD, THE SOUL, AND MATTER 69
that it is also affected by the activity of the soul, which is never regarded as merely passive in its attitude to mind. The impossibility of expressing the relationship intelligibly is inherent in the effort to bridge the gulf between the material and the immaterial worlds. But it is curious that, as in the Nyaya-Vaisesika, there is no real attempt in the Mlmamsa to explain in what way mind is active in the processes of reasoning. It is obvious that inference, and the other means of proof apart from sense perception, must be due to the activity of mind in contact with the soul, but insistence on the part of mind in the direct perception of pleasure, pain, desire, aversion, and effort has apparently resulted in obscuring the essential part which it must be deemed to play in the higher mental activities, if for no other reason than that they all rest, save verbal cognition and negation, on sense perception as an ultimate basis, and even verbal cognition and negation must be mediated to the soul by mind.
The soul, then, with the aid of the mind, is the enjoyer of all experience; the sense organs the instruments; the objects, external or internal, the world and the qualities of the soul; and the body is the abode of the sense organs and the mind, through whose instrumentality the soul has experi ence. Of bodies Prabhakara recognises three kinds only — womb-born, egg-born and sweat-born — omitting, with some Nyaya-Vaisesika authorities, the vegetable body, on the ground that its possession of sense organs is not established, despite the Jain views on this topic. None but earth bodies are accepted by Prabhakara, though the Nyaya-Vaisesika accepts the existence in other worlds of water bodies, fire bodies and air bodies; this excludes the Vedanta view, which finds in the body five or three elements or the variant which admits of four only. The body, however, in any event is essentially subservient to the soul, which acquires a body in accordance with its past deeds; in what manner this is accomplished neither Prabhakara or Kumarila tell us, for in truth the problem is incomprehensible.
So far the views of Prabhakara and Kumarila seem to be in general harmony, but there is a distinct discrepancy, if not a very important one, in their view of the manner in
70 THE KARMA-MlMAMSA
which the soul is cognised. In the view of Prabhakara the cognition is self-illumined, but this doctrine is not applicable to the soul. The Vedanta view, of course, insists on the doctrine of self-illumination in the case of the cognition and the soul as consciousness alike; Prabhakara objects that in this case the soul must be present in consciousness during the state of deep sleep no less than during the waking, dreaming and fourth states, and, as all our consciousness can be explained by hypothesis of the self-illumination of cognition, it is needless to assume any other self-luminous object. The Nyaya view, which makes the soul to be the object of direct perception, as opposed to the Vaisesika doctrine of the inferring of the soul, which is also found in older Nyaya, is rejected by Prabhakara on the ground that it serves to make the perceived also the perceiver, which is in his view absurd, a position for which there is clearly much better ground than in the cognate case of the denial of the mental perception of cognition. The theory which he adopts is, then, simply that in every cognition the soul enters into the cognition as a necessary element, and, therefore, in a sense the soul is cognised by the same means of valid cognition as the objects which it knows. But, while the soul is thus cognised, it is not cognised as a true object; it is cognised as the agent in cognition, just as a man who walks is the agent of walking, not the object. The soul, therefore, is the substratum of the self-illumined cognition, into which it enters in the element of " I," and this fact explains why in deep sleep there is no self-consciousness, since at that time there is no cognition, and the soul can be known only along with a cognition. But the fact that there is no cognition does not mean that there is no soul: consciousness is not, as in the Vedanta, the essence of the soul, but a mere quality of it, and in the state of liberation the soul remains eternally existent, though by ceasing to have cognitions it ceases to be cognised. While this view of the knowledge of the soul in self-consciousness is ingenious and not unhappy, laying as it does due stress on the necessary implication of the self in consciousness, it is a little difficult to see why Prabha kara did not admit that the soul was self-illumined, which is certainly the natural interpretation of the Sabarabhasya
GOD, THE SOUL, AND MATTER 71
(p. 22). That terra seems to apply more readily to the soul than to cognitions on his own theory, in which the cognition seems really to be inferred, as it actually is held to be by the school of Rumania.
How far Kumarila really differs from Prabhakara in these views is not clear. He certainly is credited by such texts as the Sdstradlpikd (p. 101) and the Sarvasiddhanta- samgraha (VIII, 37) with the view that the self is the object of direct perception by the mind, a view ascribed by the Nydyamanjarl (p. 429) to the Aupavarsas,1 and this is perhaps a legitimate deduction from the doctrine, which he certainly held, that the existence of the self is established through the notion of " I." The soul he holds to be the substratum of the " I " element in cognition, and this appears to be practically identical with Prabhakara's view that the soul is the substratum of the self-illumined cogni tion, and the " I " element in it. Kumarila, however, adopts in the Tantravarttika2 the doctrine that the soul is pure consciousness, though he distinguishes it from cogni tion, but this characteristic is hardly more than a verbal deviation from the view of Prabhakara, as far as practical results go.
Prabhakara and Kumarila are agreed as to the fact of there existing a multitude of separate souls, as is the neces sary supposition of the Sutra and the theme of the Bhdsya. The perception of another soul is obviously impossible, but one sees the activities of other bodies, and infers thence that they must be ensouled, just as one's own body is ensouled. Thus, if a pupil has learned half his task in one day, the fact that he continues to learn the next half the next day is a good ground for assuming that he possesses a soul. The same result can be arrived at from the fact that merit and demerit are infinitely various, and not one, as they must be if there were one soul only. The objection that pain is felt as localised, though there is but one soul in the body, is met by insisting that in reality the feeling is in the soul, and it is only the cause of the pain which can be said
1 Cf. Mdnameyodaya , p. 80. But in Slokavdrttika, p 525 (vv. 142, 143), he seems to accept self-illumination from the Bhdsya. » Trans., p. 516 ; so Slokavdrttika, p. 187 (v. 167),
72 THE KARMA-MIMAMSA
to be localised. The further Vedanta contention, that the sun, though one, appears by reflection in different substances to be endowed with diverse qualities, is also rebutted by the observation that the qualities, which appear different, do not really belong to the sun but to the reflecting medium. On this analogy the different qualities appearing in connection with the soul would belong to the bodies which are ensouled, and this conclusion is manifestly contrary to fact, since cognition, etc., are qualities, as we have seen, of soul, not of body. It is characteristic, however, of the tendency to import Vedanta conceptions into the Mimamsa that the Sarvasiddhanta- samgraha (VIII, 39) asserts categorically that there is one real supreme self, of which the individual selves are unreal differentiations.
Such being the nature of the soul of man, his normal lot is to continue in an unending cycle of lives, each determined from the outset by his actions in previous lives, unless he adopts the path which leads to freedom from this round of existence. The process of this liberation is sketched by Prabhakara ; first the man becomes disgusted by the troubles which attend this mortal life; then he realises that even the pleasures of this life are inseparable from pain, both in their attainment and in their disappearance; accordingly he devotes his mind to seeking final release from all worldly things. To this end he abstains from all prohibited acts, which lead to punishment hereafter, and also from all acts which are undertaken for the purpose of attaining some worldly or heavenly guerdon. He also exhausts the accumu lated store of his merit and demerit by undergoing the experiences which result thence. Finally he destroys the receptacle of experience by the knowledge of the soul, together with such concomitants as contentment, self-restraint and so forth, all things enjoined by the scriptures to prevent the return of the soul. When all this is accomplished, then the achievement of release is brought about. Prabhakara insists that, as the texts enjoin the know ledge of the soul for no ulterior purpose, it must be understood that the absence of rebirth is the reward of this knowledge. Liberation thus consists in the cessation of the operation of merit or demerit, and in total freedom from the body.
GOD, THE SOUL, AND MATTER 73
Liberation is purely negative in character; the soul exists as a mere existence without cognition, and without either pleasure or pain, both of which are essentially connected with the presence of the soul in the body. The existence thus achieved is thus like that attained in the view of the Vaisesika, the condition enjoyed by a stone as the author of the Sarvasiddhantasamgraha maliciously points out.
Rumania's views are largely in harmony with those of Pra- bhakara; thus he holds that liberation cannot be supposed to be bliss in any form, which is essentially transient and unreal. But, consistently with the main aim of the Mimamsa, he cannot admit that the Vedic texts, which seem to connect freedom from return with knowledge of the soul, mean that knowledge produces directly this result. This would con flict with the criticisms urged by him against the Samkhya that they erroneously hold that knowledge by itself can terminate bondage; it is, on the contrary, only possible to counteract the accumulated result of past deeds by working off the consequences; the function of knowledge, therefore, may prevent further accumulation of merit and demerit; it cannot annul what has been accumulated. So Kumarila attaches only an indirect value to knowledge of the self as a factor in sacrificial performance; in the first place, it serves to induce men to undertake sacrifices in cases where they would not else be willing to do so, for they understand the distinction of soul and body and the spiritual advantages of sacrificing. In the second place, through such knowledge men learn to perform the regular offerings, including not merely the fixed offerings but those for special occasions and penances, with the aim merely of avoiding the sin which arises from non-performance, and without any desire for the advantages accruing therefrom. It is agreed that men who do not desire the results of such sacrifices do not obtain them, a doctrine which, of course, is familiar in another application from the Bhagavadglta. Thus, then, by this means the actions of past lives are worked out, while no new action is accumulated, and the man becomes ready for final release in the shape of perpetual freedom from corporeal attachment. The final condition, then, of man will be the persistence of pure consciousness, but
74 THE KARMA-MlMAMSA
without cognition or feeling of any sort. This view, though in entire harmony with the Mimamsa, has suffered the usual fate at the hands of the later texts,1 in which it is asserted that the final condition of man is a state of constant bliss.
In what manner then does the performance of sacrifice operate as affecting the soul? The Mimamsa in both schools is confident that there is no question of rewards coming from the deity to whom the offerings are made; no deity is either eternal or omnipresent, and there could be no assurance of it ever receiving the numerous offerings made by diverse votaries, apart from the difficulty of the deity conferring rewards. There must, therefore, be a capacity, which does not exist prior to the sacrificial action, either in the principal performance or in the agent, but which is generated in the course of the performance. Before a man performs a sacrifice, which will lead to heaven, there is an incapacity in the offering and in the man himself to secure that result, but, when he has performed it, he becomes, as a result of the action, endowed with a potency, styled Apurva, which in the course of time will secure for him the end desired. The existence of this potency is testified to in the scriptures; its necessity is apparent by the means of proof known as presumption. We find in the Veda assertions that sacrifices produce certain results, and, as the operation of the sacrifice, as we see it, is transient, the truth of the scripture would be vitiated if we did not accept the theory of Apurva. Nor is there anything illogical in the doctrine; every action sets in force activities in substances or agents, and these come to fruition when the necessary auxiliaries are present. The action specified is called into existence by the injunction contained in the form of an optative in a sentence in the Veda.
From this doctrine Prabhakara dissents, elaborating instead a theory which is obviously a refinement on the simple view which Kumarila accepts from the older writers of the school and which best suits the Mimamsa Sutra.2 In his opinion the injunction rests in the sentence as a whole,
1 Manameyodaya, p. 88.
2 II, 1, 1 ff ; Prakaranapancika, pp. J85 ff ; Tantravarttika, II, 1, 1-$.
GOD, THE SOUL, AND MATTER 75
not in the optative verb, and he denies that from the action there arises directly the Apurva. On the contrary, the process is that the injunctive sentence lays down a mandate, Niyoga; this excites the man to exertion, and this exertion pertains to some form of action, indicated by the verb of the injunctive sentence. The exertion produces in the agent a result (kdrya) to which also the name of Niyoga is given by Prabhakara, on the ground that it is this which acts as an incentive to the agent to put forth exertion towards the performance of the action denoted by the verb of the in junctive clause. The Niyoga, however, is unable to produce its result, unless aided by something which Salikanatha styles fate, nor is it apparent that either in his terminology or in his view of the process Prabhakara's doctrine is any superior to that of Kumarila. It seems as if primarily it arose from nothing more important than the observation that the result produced in the agent was in one sense his motive to action as much as the sentence directing the action to be done, leading to a transfer of the term Niyoga, naturally applicable to the sentence, to the condition in the agent to which the more orthodox name of Apurva was usually applied.
In simple sacrifices there is only one Apurva produced, but in more complicated sacrifices there may be several, as a rule four. Thus in the new and full moon sacrifices, consisting of two sets of three oblations at new and full moon respec tively, there may be distinguished the Angapurva, pertaining to the minor acts of the several oblations; the Utpattyapurva, the result flowing from each of the three oblations in either set; the Samudayapurva, the result of each group of three; and the Phalapurva, the result of the whole performance re garded as a unit. But it is not every action which brings about an Apurva; those actions, which are devoted simply to some material result, though a part of the sacrifice, such as the appointment of priests or the threshing of corn, are not credited with any such effect, as they serve an immediate purpose and need no further explanation.
In the view of both schools there is a clear relation be tween the injunction and the action of the agent; the former possesses a verbal energy (sdbdl bhdvana} in its tendency to produce action by the agent, while the latter puts forth
76 THE KARMA-MlMAMSA
actual energy (drthl bhdvand} towards the end indicated in the injunction.1
It is significant of the theistic tendency of Indian thought that even the Miraarhsa was not exempt from transformation. Despite its emphatic denial of the existence of a Supreme Lord, the Sarvasiddhdntasamgraha (VIII, 40, 41) treats the end of man as to be obtained by meditation upon, and wor ship of, the Supreme_Spirit which is manifested in each man, and authors, such as Apade va and Laugaksi Bhaskara, declare that if the sacrifice is performed in honour of Govinda or the creator, Isvara, it leads to the highest good, basing this assertion on the authority of the Bhagavadgltd. Hence it is easy to explain the tendency of such works as the Sesvara- rnimdmsd of the polymath Venkatesa, where Vedanta tenets are grafted on the Mimamsa. Gunaratna, in his comment on the Saddarasanasamuccaya (p. 298) similarly attributes to Jaimini acceptance of the Maya doctrine.
The question, however, arises, how far, in accepting views of the future of the spirit, which are rejected by both Prabhakara and by Kumarila, and in imparting a theistic tinge to the doctrine, later texts relied on earlier authority, now lost to us. It must be remembered that in the Vedanta Siitra there are attributed to Jamini not merely views in entire harmony with his principles, such as insistence (IV, 1, 17) on the fact that works bear their due fruit without any divine intervention of any kind, but also opinions which show him in the unexpected light of a true Vedantin, though not of the orthodox doctrine of Samkara. Thus he is credited with the view that the order in which a man must pass through the various stages of life (dsramas) is fixed as from lower to higher, and never vice versa, and as explain ing away as metaphorical the assertion that the highest spirit is a span in size. More precise light is thrown on his doctrine by the fact that he adopted the view of the fate of the soul on departing, by which it is ultimately led by a spirit to Brahman, in the sense that the absolute Brahman is meant, though Samkara argued, apparently against the
1 Mimamsanyayaprakasa (ed. Benares) pp. 1-22, 118-35 ; Ml- niamsaparibhasd, pp. 25-30.
77
intention of the Vedanta Sutra (IV, 3, 7-14) that the refer ence is to the lower Brahman, the soul passing to the higher state only on the occurrence of the absorption of the lower Brahman. In its final condition the soul possesses, accord ing to Jaimini, all the qualities of the Brahman enumerated in the Chdndogya Upanisad (VIII, 7) , together with omnipo tence and omniscience, and further possesses a body and senses, having the power of assuming many diverse forms. Though this is not the view of Sarhkara(IV, 4, 5, 11) it can hardly be imagined that Jaimini really regarded this condition as pertaining to the soul merely preparatory to final absorption in the Brahman; we may rather suppose that on this topic his views were akin to those of Ramanuja, and perhaps of Badarayana himself.
If we were to hold that the Jaimini of the Karma- Mimamsa and the Jaimini of the Vedanta must be regarded as enunciating one body of doctrine, we would be forced to the admission that the later school of Mimamsa departed from the principles of the founder of the doctrine by ignor ing the fact that the Mimdmsd Sutra represented only one side of his thought. But to accept this would probably be to lay far too much stress on the traditional allocation of doctrines; it is far more plausible to assume that the views expressed in the Mimamsa do not represent one aspect of the thought of an individual sage, but are the expression of the doctrine of a school, which appealed to Jaimini and Badarayana only so far as it thought fit to adopt or discuss views of theirs. It would otherwise be altogether too remarkable that of two authorities, who covered much the same ground, we should have preserved the Sutra of one on the doctrine of action, and the Sutra of the other on the nature of the absolute, and in both cases the form of the Sutra is decidedly unfavourable to the view that it is the production of one definite author. The shadowy personalities of Jaimini and Badarayana can hardly claim much more effective reality than those of Gautama or Kanada, or even than Kapila himself.
If, on the one hand, there was a tendency to adapt the Mimamsa to theistic or pantheistic views, there was on the other a steady process of degradation of the deities to whom
78 THE KARMA-MlMAMSA
the offerings were supposed to be made. It can hardly be assumed that these deities were not believed to be real by the founders of the MImamsa. And there is nothing to show that Jaimini did not accept their existence. But the later doctrine, a_s evinced in such works as the Devatdsvarupavi- cdra of Apadeva, does not accept the validity of the descriptions of the deities given in the Puranas as showing the existence of such beings ; these passages rank as mere Arthavada ; the deity is merely that to which offering is made, and has no existence beyond the Mantras addressed to it.
THE RULES OF RITUAL INTERPRETATION
WE have seen that Prabhakara and Kumarila establish by their elaborate epistemological and metaphysical en quiries precisely the same results as were more simply accepted by Jaimini, the fact that duty or righteousness is inculcated by the Veda in the form of injunctions, which are to be carried out on the strength of the authority of that text as uncreated and eternal. The task of Jaimini, in all save the first Pada of Adhyaya I, is, therefore, to lay down the principles which will enable men rightly to perform the actions which the Veda enjoins, but which the vast extent of the Vedic literature renders it difficult to determine. The task falls essentially under two great heads; it is necessary to determine precisely to what texts and in what degree authority attaches, and it is requisite to classify systemati cally the various forms of injunction with reference to the actions which they enjoin. Both duties are performed, though occasionally in somewhat haphazard manner, in the Sutra; the more important one, the investigation of injunc tion, forms the main topic of many later works, while the compendia usually cover more or less adequately the whole field. The details of the discussions have necessarily little general value; they deal with incidents of sacrifices, which flourished only in the early days of the history of the Mlmarhsa, and in many cases the labour devoted to their investigation cannot but seem to us mis-spent. On the other hand, the principles of interpretation developed are often both valuable and interesting as examples of logical analysis.
Of the Vedic texts the Brahmanas afford the immediate material for the extraction of the injunctions which are the
80 THE KARMA-MlMAMSA
essential part of the Veda, but they contain also passages which cannot be treated as dealing with either positive or negative injunctions, and are classed either as explanatory matter, Arthavada, or name, Namadheya. The Arthavada (I, 2, 1-30) at first sight seems not to be entitled to authority, but Kumarila and Prabhakara alike defend its validity, the latter against the charge that such sentences are inexpressive, since they are not construed with injunctive verbal forms. The value of the Arthavada, both hold, lies in it either extolling desirable, or censuring forbidden, acts; it thus comes into immediate connection with injunction or prohi bition. Hence it follows that, in cases where it might be possible to extract from an Arthavada an injunction, it is needless to do so, the passage being adequately explained if it remains eulogistic of some action already enjoined. Arthavadas may be variously divided, but the simplest division is into three classes : the first is where in contradic tion of some other means of proof a quality is asserted to exist, as in " The post is the sun," which serves to extol the brilliance of the post. Or it may merely reiterate a truth known otherwise, as in " Agni is a protection from the cold." Or it merely may refer to something which has happened, neither contradicted by other means of proof nor already known.
The case of name (I, 4, 1-16) is far more obscure; dis cussions regarding it usually turn on one or other of the sentences udbhidd yajeta pasukdmah, " he who desires cattle should sacrifice with the udbhid " ; citrayd yajeta pastikdmah, "with the citrd "; agnihotram juholi, "he offers the Agnihotra"; and syenendbhicaran yajeta, "he who practices witchcraft should offer the syena sacrifice." It seems at least plausible to suppose that the subdivision owes its creation to the practical necessity of dealing with a limited number of obscure sacrificial terms, but was later extended into a wider area. In the developed theory,1 the justification of the classification of name is given as follows. Each word in the injunction must be brought into effective
Thibaut, Arthasathgraha, pp. xii, xiii ; MimamsanyayaprakaSa, pp. 85 ff.
RULES OF RITUAL INTERPRETATION 81
relation to the energy in the agent excited by the injunction. The energy requires a result to be achieved, an instrument wherewith to achieve it, and an indication of the procedure to be followed. The last requisite is furnished by various subsidiary injunctions, the instrument and the object are given by the verb yajeta and the qualification pasukdmah, which may be reduced to ydgena pasum bhdvayet, " he should effect or realise cattle by the use of the sacrifice." It remains to dispose of ndbhidd, and, various other sugges tions as to how it should be taken (e.g. as denoting a spade) being rejected, the conclusion is arrived at that it merely serves to limit the idea expressed by ydgena, is only a name, and thus deserves a separate place in the classification of texts.
Distinct from the Brahmana is the Mantra (I, 2, 31-53) of which no effective definition is attempted in the texts.1 The Mantras are divided into the Re, Saman and Yajus, according as they are recited, sung, or muttered, usually in a low tone, though some of the Yajus Mantras, the Nigadas, are said out aloud; the Yajus is usually held to be unmet- rical, though with small accuracy. Mantras do not lay down injunctions, but in the main they serve to denote something of value in connection with injunctions, especially the deity to whom offering is to be made. The tendency to find this characteristic is much overdone in the MImamsa, though even it is compelled to recognise now and then that a Mantra must be regarded as merely of a eulogistic character, or even that it is destined to have some supernatural fruit. The MImariisa position is, of course, an inevitable result of eliminating the goodwill of the deity as a real factor in the sacrifice ; the hymns with which their authors intended to confer pleasure on the gods become a somewhat cumbrous, and not altogether useful, part of the sacrificial apparatus.
Apropos of the Mantras of the Yajus type, however, the MImamsa develops some sensible rules of construction (II, 1,46-49), rendered necessary by the fact that, while the metre in the case of the Re and the song in the case of the Saman determined the extent of the Mantra, the Yajus
1 Mlmathsabalapr akdsa , pp. 58-70; Mtmdmsdparibhdsa, p. 40.
6
82 THE KARMA-MIMAMSA
Mantras were recorded in long paragraphs, with no obvious mechanical dividing marks. These principles are that of syntactical unity (ekavdkyatd) ; those words must be taken together which, when so united, form a single idea, or, as Prabhakara puts it, to suit his theory of injunction, express a single purpose, and which, taken apart, are not expressive of any idea or purpose. Secondly, there is admitted the principle of syntactical split (vakyabheda) , which permits us to break up what else might be taken as a single sentence into parts, each of which must contain a single idea. But this expedient is permissible only when there is a clear Vedic injunction to make the split, or when no other con struction is really possible, for otherwise the error is com mitted of multiplying Apurvas resulting from Mantras. None the less it is a necessary procedure in cases where it is made clear in any way that there are distinct acts to be accompanied by Mantras; "Pleasant I make this seat for thee ; sit upon it " (T.B. Ill, 7, 5, 2) would primd facie be one Mantra, but, as it is intended to serve the double purpose of accompanying the act of making the seat for the cake, and setting it down, it must be taken as two. Thirdly, there is the principle of extension (anusanga), which denotes that it is often necessary in the case of Mantras to supply with several sets of words a clause which follows the last of these sets only, and which might thus be deemed to belong to it alone.
Authority, however, is not confined to the Vedic Sarhhitas and the Brahmanas. It is extended to the Smrtis, in which term Kumarila1 includes primarily the Itihasas, Puranas, and the Smrti of Manu, these being works which claim universal application. The Itihasas and Puranas he deems to contain injunctions based on Vedic authority and much Arthavada, but he admits that there are also injunctions arising from mere worldly considerations, passages useful only to give pleasure, and other extraneous matter ; the hymns to deities serve to secure a transcendental result. They serve as wholes the useful purpose of instructing men
1 Tantravarttika, trans, pp. 25. 112 ff, 244. The Sutra (I, 3, 1-7, 11-14) never mentions the word Smrti and the commentators differ widely in their versions.
RULES OF RITUAL INTERPRETATION 83
of very various capacity and knowledge. In the case of the Smrtis proper only five alternatives are available: either they are completely erroneous, which is impossible, as these works are obviously useful and men are not so foolish as to believe nonsense ; or they are due to personal observation, which cannot be accepted; or they rest on tradition, which would give no assurance of validity; or they are deliberately intended to deceive, which is incredible for lack of motive and probability of success; or, finally, they represent lost Vedic tradition. For this decision there can be adduced a certain corroboration in the fact that for certain statements in Smrtis we can find confirmation in Vedic texts, whence we can assume that other statements were also derived from texts, now unhappily lost to us. Thus part of the Smrtis is derived from the Veda, part from ordinary motives of life, and the story material is Arthavada, as in the Itihasas and Puranas. The view of Prabhakara1 is not essentially different; he also accepts the inference of Vedic authority, but expressly negates it in the case of Smrtis which do not prescribe or prohibit any course of action, for example, statements that plants have souls, which contradicts his own denial of vegetable bodies. Both schools again agree (I, 3, 15, 16) in accepting as valid the Smrtis of such ^authors as Gautama, Vasistha, Baudhayana, Apastamba, Sankha and Harita, despite the occurrence in them of passages laying down certain customs as practised by persons in the east, and so on, a fact which apparently contradicts the universa lity of the Vedic injunction. The decision of the schools is that their injunctions are truly universal, and even the Veda lays down certain practices as to be performed by certain classes only, for instance, the Rajasuya is a sacrifice for kings only.
The Vedangas, or subsidiary treatises bearing on Vedic pronounciation, ritual, grammar, etymology and astronomy, are also admitted to rest in part on Vedic tradition, and the Mimamsa and philosophic treatises generally are permitted to share in Vedic authority ; thus Kumarila2 assures us that
1 Prakaranapancika, pp. 100, 101, 150, '' Slokav&rttika, p. 267 (v. 201).
84 THE KARMA-MlMAMSA
the Buddhist denial of the external world was not really meant to be taken as a serious contradiction of its reality, but to divert men's minds from undue attention to it, and so with other apparently erroneous tenets. In the case of the ritual Sutras the claim is made by some that they must be treated as true Veda themselves, but this is denied, for they have human authors, and are merely, like Smrtis, based on Vedic authority. In the case of grammar, however, a really interesting discussion is raised on the Sutra (I, 3, 4-30) by those who deny that it can be made out to rest on Vedic authority, and who go so far as to challenge the validity of the claims of the grammarians to be authoritative.
The argument of these unorthodox persons runs : words like gavi, in lieu of the grammarians' gauh, for cow are perfectly correct; they are fully expressive, they are percept ible by the ear, they are as eternal as any word is, and no beginning in time for them can be traced. The science of grammar, too, has no Vedic connection; it differs in no way from the process of explaining vernacular words for everyday use; it does not deal with actions which are the sole business of the Veda; it serves no useful purposes in relation to duty, as we do not need grammar to tell us the meanings of words; nor is grammar the source of usage, since, on the contrary, it rests upon and follows usage. The reply of the Mimarhsa is not convincing; it maintains that synonyms are not permissible, unless enjoined by Vedic use; as gauh expresses exactly the meaning " cow," any variant of it is wrong and undesirable, and has such power of denotation as it may actually possess, merely because of its similarity to gauh and not in its own right. The science of grammar also is essential to set out in orderly derivation the vast masses of words in the Vedas; moreover, not usage alone, but usage and grammar determine whether a word has the correct form to convey the meaning, and grammar in the last issue is more authoritative than mere usage. That correct words produce, when used, a transcendental result is proved by the fact that the Veda enjoins their use, and forbids the use of barbarous expressions; moreover, truth leads to supreme bliss, and the use of correct words is truth in speech. Fortunately this disquisition does not prevent either Prabha-
RULES OF RITUAL INTERPRETATION 85
kara or Kumarila (I, 3, 10) from agreeing that, when the Veda uses a barbarous word, it is to be interpreted in the sense barbarian usage accords to it, in lieu of attempting to foist upon it an etymological sense.1
The relation of Smrti authority to the Veda, in cases where there appears to be conflict, is the subject of divergence of opinion between Prabhakara and Kumarila. In the view of the former, which is apparently that of iheMimdmsd Sutra (I, 3,3-4), if a Smrti contradicts a Vedic passage, the former loses all authority, while, even in the case of Smrti passages which do not thus offend, it may be impossible to accord their injunctions any spiritual value, if they seem to be due to the avarice of the priests, as when the giving of the cloth from the sacrificial post to a priest is enjoined by Smrti authority. Kumarila, however, with his greater regard for tradition, reduces, as far as possible, cases of contradiction to mere instances where alternatives are permissible, and only holds that the Sutra recommends in the case of such alternatives the adoption of that which has direct, and not merely inferred/Vedic authority.
Below the Smrtis in value comes the practice of good men (I, 3, 8-9) or custom (I, 3, 15-23) on the simple ground that, while both must go back to Vedic authority to be valid, the former goes more directly to the fountainhead. In addition, however, to Smrti and practise must be reckon ed as sources of knowledge of duty the implications contained in Vedic texts, which may be deduced by us from them, even if not already set out in Smrtis or by tradition.
The essential function of all these sources is to give us knowledge of injunctions (vidhi), and injunctions are encite- ments to actions. Actions may be classified in various ways; there is a clear distinction between Vedic and worldly actions, with the former alone is the Mimamsa concerned. Actions of this type may be classed as positive, as negative (pratise- dha), or as partaking of both characters (paryuddsa), as in
1 Kumarila enlivens the discussion by giving a long list of errors in grammar committed even by grammarians, and similarly diversifies (I, 3, 7) his exposition of the practice of the good by an account of crimes attributed to gods and sages. On grammar, cf. Nyayamanjari, pp. 412-26.
86 THE KARMA-MlMAMSA
the case of a vow not to look at the sun taken by a student. Of actions1 the sacrificial are the most important, falling under the three main classes of Yaga, the offering to a deity of a substance; Homa, the offering of a substance in fire or water; and Dana, the waiver of one's ownership of an object in favour of a third party. Sacrifices, again, can be divided according as they serve as archetypes only, like the Agni- hotra, or as derivatives (vikrti), like the Masagnihotra, or as both, like the victim for Agni and Soma, based on the new and full moon sacrifice and itself a model for the offering at the Soma pressing; or as neither, like the Darvihoma, for special reasons given in the Sutra. More important is the division by purpose; the Nitya sacrifices must constantly be performed at the due seasons ; the Naimittika must be per formed on certain special occasions, as the Jyotis offering on the approach of spring, while the Kamya offerings are optional, being undertaken by a man who desires some special end, as in the case of the Karlristi performed to obtain rain.
The direction to perform an offering is laid down in an originating injunction (utpattividhi) or an injunction of application (viniyogavidhi), according as the matter concerned is a principal or a subordinate offering. In either case, it is frequently necessary to ascertain precisely how many actions are prescribed, and six rules for this purpose are laid down by the Mlmarhsa (II, 2 and 3). Difference in words is one clear indication, especially in the case of the verb, which is the most important part of a sentence of in junction; the repetition of the verb indicates distinct offer ings; the mention of a definite number of oblations is clear evidence; other sources are the difference of names, of materials to be used, and of context. In the case of the last item it is agreed that the occurrence of the same offering in two different recensions of one text, as in the case of the Kanva and Madhyarhdina texts of the Satapatha Brdhmana, does not constitute a diversity of context.
The originating injunctions are few in number, relating as they do to the principal Vedic offerings only, such as the
1 MimamstibalaprakaSa, pp. 81 ff ; Mitnaiiisdparibhasa, pp. 17 ft.
RULES OF RITUAL INTERPRETATION 87
Agnihotra, the new and full moon sacrifices, the Soma sacrifice, and so forth. It might have been expected that there would have been made some effort to systematise these offerings, but no trace of any attempt to effect this end is seen in the MImamsa, which accepts the sacrifices from the sacrificial tradition. It is true that there is a certain degree of order of progress from the simpler to the more complex, but this order is not absolute, being broken by the necessity of performing the Naimittika offerings on the occurrence of the special occasions which evoke them. Nor is there any principle discernible in the rewards attainable by these offerings ; they include such material things as wealth, usually in cattle, children, long life, rule, and, most frequently of all, heaven, which is held, on what is known as the Visvajit principle (IV, 3, 10-16), to be the reward promised in any case in which no specific boon is laid down.
The originating injunctions, however, do no more than excite in the mind of the hearer the desire to perform the action which they enjoin, generally in the form of a sacrifice ; it remains for other injunctions, those of appli cation, to denote the exact manner of procedure (itikar- tavyata), by specifying the numerous subsidiary actions requisite, and the materials and other necessaries for the performance. The discrimination between what is principal and what is subsidiary (sesa) occupies the greater part of the attention of the MImamsa, and it stands in a close relation to the motive for the performance of the various actions. Actions may be undertaken according to the Sutra (IV, 1, Iff), followed by Sabarasvamin, Prabhakara, and Kumarila, either for the sake of the agent (purusartha) , or for the sake of the offering (kratvartha) , while Partha- sarathi adds a third class of those which are neither for the one purpose or the other, giving as an instance" the Agnyadhana, or piling of the sacred fire. This innovation seems to be without warrant ; the original distinction corres ponds roughly to that between principal and subordinate actions ; the new and full moon offerings serve to benefit man by producing a due reward, while the fore-offerings, which form part of them, are merely subsidiary to the sacrifice ; materials normally are subsidiary to the sacrifice,
88 THE KARMA-MlMAMSA
any good results mentioned being treated as merely Artha- vada, though on occasions a thing like curds, which serves as an element in offerings, may be used to make efficient the sense organs of the sacrificer, and thus to serve for his benefit. The last instance shows that the correspondence between actions for the benefit of the agent and principal actions is by no means complete.
The question of what things can be subsidiary is the subject of an elaborate investigation; according to Badari's opinion, cited in the Mlmdmsd Sutra (III, 1,3), the only subsidiaries are substances, accessories, namely, the Mantras and the deities, and purificatory actions, such as the thresh ing of corn. To this list Jaimini adds actions or sacrifices generally, results, and agents. The distinction between the two sets, according to the Vrttikara, whom Sabarasvamin cites, is that the first three classes are essentially in their nature subsidiary, while the latter three are in one sense principal, in another subsidiary. Thus the sacrifices are principal with reference to the materials, but subsidiary to the result; the result is principal with reference to the sacrifice, but subsidiary towards the agent; the agent, again, is principal with regard to the result, but subsidiary to such acts as the measuring of the sacrificial post, which is to be related to his height. From another point of view the agent may be said to be subsidiary to the sacrifices, since it is to perform them that he acts.
Prabhakara1 divides the subsidiaries into four classes, according to the heads of class (jati), quality, substance, and actions, denoted by verbs (bhdvdrthdtmaka) . The last head he divides into those actions which are directly con ducive to the fulfilment of the sacrifice (samnipatyopakd- raka), and those which are more distantly conducive to this result (arddupakdraka). The former he classifies in four divisions; the bringing into existence (utpatti) of some object, as the production of dough by kneading the corn; the obtaining (prdpti) of. a substance already in existence, such as milk; the modification (vikrti) of a substance, as of
1 Prakaranapancikd, pp. 202 ff; cf. MimdmsanyayaprakaSa, pp. 62-67, where the division is twofold, siddha and kriya.
RULES OF RITUAL INTERPRETATION 89
ghee by boiling; and the purification (samskrti) of substance, such as the sprinkling of water over corn. These actions are all subservient to the sacrifice, and to the Apurva, which is produced by the sacrifice; they have no distinct Apurvas of their own. On the other hand, the indirect auxiliaries have Apurvas of their own, though some, as for instance the drink ing of milk by the sacrificer at the Jyotistoma, have also a seen result. In the case, however, of such a sentence as, "He offers to Tanunapat," no effect on the substance offered or the human agent is obvious, and we are bound to assume an Apurva appertaining to the action by itself, which is sub sidiary and auxiliary to the Apurva of the sacrifice as a whole. But Kumarila, who contents himself with a simpler division of all subsidiaries into the two classes of direct and indirect,1 goes much further in seeking to recognise subsidiary Apur vas. He postulates an Apurva for every injunction, instead of merely for injunctions of principal actions and such others as cannot be disposed of by any other means, and thus, while he does not assert that there is in the directly auxiliary acts themselves an Apurva, he holds that there is an Apurva in the fact that a choice has been made of the particular mode of action, e.g. in selecting the mode of thumping as the proper manner of cleansing the grain.
In dealing with injunctions of application there are six means by which the relations of subordination of actions, etc., may be determined (III, 2 and 3). The first and most important indication is express declaration; thus, if the injunction is given to use in honour of the Garhapatya fire a Mantra to Indra, this declaration (sruti) prevails over the normal conclusion that a verse to Indra must come in a ceremony in his honour. Second in importance is indirect implication (linga); thus it is a rule in the Soma sacrifice that the juice used is to be consumed; when, therefore, a Mantra (T.S. Ill, 2, 5, 1) is found which seems merely to refer to the drinking, it must be assumed that it covers by its reference to drinking all the operations connected with that action, such as the taking up of the potion, examining it, drinking it, and digesting it. Thirdly,
1 Sdstradipikd, pp. 202, 203 ; Mimdmsdparibhdsd, pp. 16 ff.
90 THE KARMA-MlMAMSA
syntactical connection is of value; thus, in one passage (£.B. IV, 4, 6, 16-18) we are able to decide that Re and Yajus mean the Rgveda and the Yajurveda, and not, as might else be thought, metrical and prose Mantras, because of the syntactical connection with the immediately preced ing words. Fourthly, context (prakarana) is of great importance; we have the general injunction that one should perform the new and full moon sacrifices, and the injunction to offer to Tanunapat; this principle enables us to find a purpose for the latter offering in connection with the former sacrifices; mere syntactical connection would not here help, as the sentences stand apart and are in themselves quite complete. Fifthly, order (krama) or position (sthana) is of service; thus in one passage (T.S. I, 6, 2, 4) occur three Mantras without indication of use; we can, however, by finding that elsewhere three offerings are enjoined in connection with these Mantras, assume that the order of the sacrifices and the Mantras is to correspond, one being used with each offering in order. Finally, names (samdkhyd) may supply information else wanting; thus Mantras, not otherwise identified, by being styled Hautra are known to fall within the sphere of the Hotr priest. Each of these means for adequate reasons is deemed to be of more value than the preceding, and in working out the principle in detail the Mimarhsa shows both skill and acumen, even when we admit that in many cases its reasonings were guided by the fact that a certain usage had become regular, and therefore that the sound conclusion was already given by customary practice.
While these injunctions of application determine the exact mode in which the ceremonies prescribed in the origi nating injunctions are to be performed, the order of the actions is prescribed by injunctions of performance (viniyoga- vidhi). On this point, however, there is a difference of view between Prabhakara and Kumarila (V, 1) ; the latter admits readily the existence of injunctions determining the order of performance by the process of extracting such directions from injunctions of application. Prabhakara, however, insists that an injunction of application cannot be deemed to deal with order, which is a matter of indifference,
RULES OF RITUAL INTERPRETATION 91
so long as an act is performed, but even he admits that a few cases occur in which the order of offerings is specially defined. As a rule, however, the order is left to be deter mined by minor indications. Thus it may be directly enjoined, or the order of the mention of the offerings may be decisive, or the order of the natural actions may be resorted to; thus the gruel must be cooked before the Agni- hotra is offered, although the text mentions the latter first. Again, the order of commencement is of importance; in the Vajapeya there are seventeen victims to be immolated; the offerer may begin with any one, but the different acts must l)e done to each following the initial order adopted. Posi tion, again, is of importance; thus in the Agnistoma there are three animal victims, one to Agni and Soma offered on the day before the sacrifice, the Savanlya on the day of the pressing of the Soma, and a barren cow on the final day. In the Sadyaskra, a modification of the Agnistoma, the three victims are to be offered on one day, that of the press ing; hence, as this is properly the day of the Savanlya victim, it is to be offered first, followed by that for Agni and Soma, and the barren cow. Lastly, the order of the principal actions prevails over that of subordinate actions; thus at the new moon sacrifice the preliminaries for the offering to Indra are performed before those of the offering to Agni, but the offering to Agni comes before that to Indra : accordingly, in the performance of subsequent rites, it is those connected with Agni that take precedence over those connected with Indra. In cases where none of these means give a clear result, any order may be resorted to, and so with offerings performed independently to obtain worldly goods. Nor is there any fixed order between the Soma sacrifices and the simpler rites known as Istis.
There remains the question of the right to perform sacrifices, which forms the subject of a set of injunctions relating to qualification (adhikaravidhi). Jaimini, it seems, took a generous view of the position of woman, con templating (VI, 1, 6-8) her as a performer of sacrifices, though, in the case of her being married, both she and her husband must co-operate in offering, and the Vedic Mantras would be recited by him only. Sabarasvamin already
92 THE KARMA-MIMAMSA
emphasises the disability of women arising from their ignorance of the Veda, which is not asserted by Jaimini, who doubtless reflects the older usage. Sudras are excluded (VI, 1, 26,33) from sacrificing for this very reason of ignor ance of the Veda, and in the later texts the admission of women even to a qualified share of the sacrifice is thus anomalous. Some small amount of means is also requisite in a sacrificer, and he must not be incapacitated by disease. Further details are given in the Srauta Sutras, which recog nise, like Jaimini, the case of certain classes who can take some part in sacrifice though not of the three higher classes, such as the Rathakara. In the case of Sattras only Brahmans of the Visvamitra family studying the same Kalpa Sutra are qualified to act; all act as sacrificers, and each individually obtains the whole benefit of the sacrifice, instead of it being shared collectively. Moreover, while the death of an ordinary sacrificer destroys the rite, in the case of a Sattra the place of any one incapacitated can be taken by another priest, who, however, obtains no share of the result. Only Brahmans again can eat the remnants of sacrifice, so that, if a Ksatriya has a Soma sacrifice performed for him, he must be given to drink a substitute for Soma remnants. On the other hand, the threefold duty of sacrifice to the gods, of Vedic study as payment of debts to the Rsis, and of the begetting of children as a debt to the Fathers, is incum bent on all these classes, not merely on those who may wish to attain the benefits of these actions (VI, 2, 31). Again he only may perform the Visvajit (VI, 7) who can afford a fee of 1,200 gold pieces, but, when he is bidden to give up all in it, that applies only to his riches, not to, e.g. his parents, and of his riches there are excluded lands, horses, and slaves in personal attendance, while the 1,000 years of performance is interpreted as so many days.
In addition to these divisions according to content injunctions can be classified on the basis of the knowledge already possessed by the agent of the mode of performance or actions possible.1 Thus an original injunction (apurva-
1 Kumarila on Mlmamsa Sutra, I, 2, 42 ; Arthasamgraha, pp. 17, 18 ; Mlmamsaparibhasa, pp. 10-12, 41.
RULES OF RITUAL INTERPRETATION 93
vidhi) is one which enjoins something otherwise unknown, as when a direction is given that grains are to be washed. A restrictive injunction (niyamavidhi} serves to fix as alone valid one out of several possible means of carrying out such an action, such as the husking of grain, which an injunction requires to be performed by pounding. An injunction of limitation or exclusion (parisamkhyavidhi) precludes one of several alternatives which otherwise might be resorted to; thus the injunction, "Five animals among animals with five nails may be eaten," precludes the eating of any animals not having that adornment. In this case the preclusion is implied, in other cases it may be explicit.
While an injunction directs a positive act, a prohibition (nisedha) serves to turn a man away from performing the action expressed in the verb and its object. The prohibition does not lead to any desirable result such as heaven; it serves none the less a useful purpose; the man, who obeys the direction not to eat the mysterious Kalanja, by observing this taboo escapes the hell which else had been his fate. In the technical phraseology of the Mimamsa the negative applies not to the sense of the verb, but to the optative affix; as an optative urges us to action, so a negatived optative turns away from it. In certain cases, however, this normal con dition of affairs is precluded, and the negative is immediately connected with the verbal sense. Of these the most important is the case (IV, 1, 3-6) of negative passages headed, " His vows are as follows." The Brahmacarin is under an obligation not to look at the sun as a vow; the force of this is not that he is to avoid the evil result of looking, for there is no such result, but, as the context indicates, he is to effect the destruction of evil by the resolution of not-looking at the sun. A similar mode of interpreting an apparent prohibition is seen in the case of the rule, " He should say, ' Ho, we sacrifice,' at the begin ning of all sacrificial verses, but not at the Anuyaja offerings." The reason for this procedure is that, if the rule were regarded as a prohibition proper, then it would necessarily follow that there was an antecedent rule enjoin ing the action, since a prohibition implies a previous rule to the opposite effect, and, this being so, the result of the
94 THE KARMA-MlMAMSA
prohibition would not be what was desired, since, owing to the equal validity of all Vedic sentences, the only result would be to make the action optional. Thus, instead of a prohibition, we have what is technically styled a Paryudasa, and the sense of the rule is that the words, " Ho, we sacrifice," which are uttered with the sacrificial verses, are to be uttered with those verses only which do not occur in Anuyajas.
These are the main topics, which, with numerous excur sions into subsidiary detail, fill Padas II-IV of Adhyaya I and Adhyayas II-VI of the Mlmdmsd Sutra, The next two Adhyayas deal with the transfer of details from the archetype to sacrifices whose form is derived from it, a discussion rendered necessary by the fact that in the Brahmanas there are many cases in which it is presumed that the details of one offering will be supplied from another, as in the often-quoted case of the Isu offering which is based on the Syena. The transference (atidesa) applies not merely to the mode of performance, but to materials and other details.1 It is regulated by- context (prakarana) or position; thus the Isu offering follows the Syena model, because they are enjoined in the same context. The rule of position again lays it down that the deity of the original offering is to take the same place in the transferred offering, and the offering material is also to be transferred. Transfer takes place by express injunction, as in the case of the Isu offering; by implied injunction, as in the case of the offering to Surya, which is based on the new and full moon offerings; by mention of the name of a sacrifice, as in the case of the Masagni- hotra, which is made in accordance with the Agnihotra; or by mention of the name of a purification (samskara), as when, the Avabhrtha being mentioned at the Varunapraghasa, it is performed like the Avabhrtha, or concluding bath, of the Agnistoma where the rite is purificatory.
The process of transfer, however, frequently involves modifications (iiha) in the Mantras used to accompany the rites, in order to adapt them to the change of circumstance.
1 PrakaranapancikS, p. 227 (v. 13).
RULES OF RITUAL INTERPRETATION 95
Elaborate rules are accordingly given in Adhyaya IX of the Sutra on this head. Nor only Mantras are altered, but also Samans ; thus at the Vaisyastoma the Kanvarathan- tara replaces the orginal Brhat and Rathantara Samans. In some cases purifications are modified; thus the wild rice (nivara) used at the Vajapeya offering, in place of the ordinary rice, is subjected to the processes of purification applicable to the latter. In other cases Mantras are not altered, but the number of times of their use is modified.
In other cases the transference must be accompanied by the annulment of details which are now inappropriate. The cases in which this occurs, enumerated in Adhyaya X, are numerous and complicated; thus an act may by change be rendered useless; in the Prajapatya rite, based on the new and full moon sacrifices, grains of gold replace rice grains, and the operations of husking and washing are therefore annulled. Again, if Yajus Mantras are given to be recited as Nigadas, which are invitations and therefore must be said aloud, the normal rule of muttering of Yajus Mantras is annulled in favour of the necessary loud utterance. Annulment, again, may be partial or complete, arid the later text books take special pleasure in developing the diverse forms in which it may appear. Opposed to annulment is combination (samuccaya) in which the new details of the derivative form are only added to the details of the original offering.
In Adhyaya XI the question is raised of th? relation of subsidiary to principal offerings as regards repetition of performance. In certain cases a single performance of subsidiaries gives effective aid to more than one principal action, as in the case of the Agnyadhana, which need only once be performed, the same consecrated fire serving for all subsequent sacrifices; this aid is styled Tantra. On the other hand, some subsidiaries must be repeated with each principal offering ; thus the subsidiaries of the rites performed at new and full moon respectively in those offer ings are nearly the same, but the lapse of time between the two rites renders the repetition of the subsidiaries essential ; this case is styled Avapa. But in some cases where a sub sidiary is merely performed for the purpose of aiding one
96 THE KARMA-MlMAMSA
principal operation, it may nonetheless aid also another principal operation; thus the fore-offerings for the victim to Agni and Soma at the Soma sacrifice serve for the cake offering also, and, if an altar has been made ready for a Soma sacrifice, the sacrificer may perform an Isti with it, if he will. This form is termed Prasanga.
Finally, in Adhyaya XII the topic of options (vikalpa) is disposed of ; options are of many kinds,1 indicated by reasoning, or by direct declaration, or depending on the wish of the agent ; nineteen subdivisions of each type are made, of which eight depend on the option furthering the performance of the rite, and eleven on its bringing about some benefit for the agent. By another principle of divi sion options are classed as limited or fixed (vy avast hit a) and unlimited (avyavasthita) , each class again being sub divided according as it rests on reasoning or declaration. But options as a rule are open to many objections, though this defect does not apply either to fixed options, or to those which depend on the will of the agent. The subjects to which options may apply are most varied, the use or non-use of certain Mantras, preference for one colour or another, the choice of kinds of grain, mode of action, and so forth.
1 MimamsabdlaprakaSa, p. 152 ; Mlmamsaparibhasa, pp. 41-44
VI
THE MIMAMSA AND HINDU LAW
THE fact that the MImamsa is an investigation of texts in order to evolve an orderly system for their interpretation as a harmonious whole brings it immediately into contact with legal interpretation, and the parallel is made the closer in that the chief object of the MImamsa is to determine injunctions, which are distinct from those of civil law mainly in the fact that they deal with sacrificial rather than civil obligations, and are enforced by spiritual rather than temporal penalties. Thus as early as the Dharma Sutras of Vasistha and Baudhayana we find that skill in the Mlmarhsa was con sidered a qualification for membership of a Parisad, to which the settlement of disputed questions of law was entrusted. The same rule reappears in Manu and in other Dharmasastras, and Kulluka at the outset of his commentary on Manu expressly states that he proposes to follow the MImamsa principles as the appropriate method of interpre tation. The first known commentator on Manu whose work has reached us is Medhatithi, who shows himself an adept in the Mimamsa principles, and whose date is most probably to be placed in the ninth or tenth century A.D. The close association of Mimamsa and law is shown in the works of Madhava who wrote an exposition of the Parasara Smrti, in those of Dinakara,1 brother of Kamalakara, who added to his legal treatises the Bhattadinakara on the Sastradipika; of his son Visvesvara or Gaga Bhatta, whose Bhattacintamani deals with the Sutra; of Kamalakara himself, and of Nllakantha who shows in his compilation, Bhagavanta- bhaskara, an excellent knowledge of the MImamsa topics
1 Cf. Mandlik, Vyavaharamayukha, p. LXXV.
98 THE KARMA-MIMAMSA
with which his father, Sarhkara Bhatta, had dealt. The parallelism, indeed, of the two enquiries only became the more salient as with the course of time the number of Smrtis and other texts claiming authority increased, and the ideal of reconciling their conflicting views was more and more strongly held. All the devices necessary for such an end existed in the Mlmarhsa, and we can understand from this reason why it was not thought necessary or desirable to develop a distinct science of legal interpretation.
Thus the essential doctrine of injunction in civil law is based on the principles adopted in the Mimamsa, and in the interpretation of the various kinds of injunction the civil law adapts to its own special needs the maxims of the sacred law. The distinction between injunction proper and a restrictive injunction (niyama) is applied in the sense that the latter is reduced to nothing more than a maxim or rule, which ought to be regarded, but which, if violated, does not render the action affected invalid; thus Manu's rule as to marrying an amiable and healthy girl is not an injunction, the violation of which renders void the marriage, but a counsel of prudence. The case of an injunction of limitation (parisamkhyd) raises difficulties, as there arises in regard to it the question whether or not it is to be deemed to imply a prohibition ; thus, when the injunction is laid down that the sons may divide the family property on the death of their parents, Jimutavahana puts the question whether it is to be inferred that they may do so only on the death of their parents, a view which he rejects.
Negative injunctions also raise a point of legal impor tance in the relation of prohibition proper, and a mere exception (paryudasa). Thus the general law of the succes sion of a son and other heirs is subject to the exclusion from succession of persons impotent, outcaste, lame, blind, and suffering from incurable diseases, who are entitled to maintenance merely. The negation in their case is essen tially to be treated as an exception to the general rule of succession ; it, therefore, applies only to persons so circum stanced at the moment when the succession would normally vest, and, therefore, if successors become so afflicted after becoming entitled to the succession, the rule does not in any
THE MIMAMSA AND HINDU LAW 99
way affect them, as, of course, it would do if it were a prohibition proper.1
In the interpretation of the Smrti injunctions the same principles are applicable as in the case of the interpretation of the injunctions of application in Vedic texts. Thus the express declaration of a text must be held to override any conclusion which might be deduced from it by suggestion (laksana), corresponding to liiiga in Jaimini. The decla ration of Manu (IX, 104) that "after the death of father and mother the sons should divide the paternal property, for they have no po;ver over it while their parents live," is an absolute declaration that they have no such power ; it is impossible to read the rule as forbidding partition during the parents' life, but acknowledging the power of the sons over the property. The power of suggestion, however, has also its own place ; thus Nanda Pandita in explaining how, although the word "substitute" was first applied specificially to five kinds of sons, it becomes applicable to all the twelve kinds legally recognised, adduces the Pranabhrt maxim (I, 4, 28) as his warrant. iTanabhrt originally denotes a Mantra used in consecrating a brick in the fire altar ; thence it passes to be the name of the brick, and from denoting the special bricks used applies more generally to any brick. The principle of syntactical connection (vakya) reappears, usually under the title Anvaya ; its superiority to context (prakarana) is illustrated by Raghunandana's discus sion of Manu's rule (XI, 209) that one who assaults a Brahman must undergo the Krcchra penance. If the context is invoked, this may seem merely to refer to the case of the new and full moon sacrifices, and therefore has no general or civil application, but the sound view is that it is to be treated as a single independent proposition.
The term, Arthavada, which plays so important a part in the Mimarhsa discussions is dropped in legal terminology, but the legal texts recognise the existence of such passages in the Smrtis and deal variously with them. One difficult problem is handled in the light of the maxim of Jaimini (I, 2, 19-25), dealing with declarations which have the
1 Tagore Law Lectures, 1905, pp. 332, 333.
100 THE KARMA-MlMAMSA
appearance of being injunctions but are not really so (vidhivannigadddhikarana) . Thus Jimutavahana x is enabled to hold that the text which provides that, " Though immovables or bipeds have been acquired by a man, no gift or sale of them without the assent of his sons," is to be completed by the words, " should be made," and not by the words " must be made." This interpretation reduces the sentence to a mere pious opinion, and avoids contradiction with the well-known injunction, which allows a man absolute power of disposal over property acquired by his own exertions, as opposed to ancestral possessions. The same maxim, however, has been interpreted as supporting the general rule that an injunction for which a reason is adduced is merely equivalent to an Arthavada, so that Vasistha's rule against the adoption of an only son is reduced to a pious expression of opinion, because it is followed by the explanation that a son is one who saves from hell his natural father, a fact which makes the adoption of an only son undesirable. The Mimamsa rule, however, goes no further than to hold that, if for a rule which has no known Vedic sanction a selfish motive can be seen, it is impossible to postulate for it the authority of a Vedic text, and 'the supposed rule of law is clearly too widely stated.
The obligation of law to the Mimamsa extends to every department of the topic, and it not merely in matters of interpretation that the legal writers borrow matter from the Mimamsa, but they show repeatedly traces of influence by the positive doctrines of that school in their bearing on the religious aspect of property and family rights. The doctrine of the three debts of man, sacrifice to the gods, study to the Rsis, and the begetting of a son for the Fathers, enunciated by Jaimini, affords three presumptions which, more or less effectively, are taken into account by the schools of law. In treating the principle of succession Jimutavahana uses, as a guiding principle in reconciling the conflicting statements of the Smrtis, the principle of securing as far as possible spiritual welfare, and in interpreting the rules regarding
1 Dayabhaga, trans. Colebrooke, II, 29, 30.
THE MIMAMSA AND HINDU LAW 101
self acquisition of property he tacitly favours the claims of Vedic study. Vijnanesvara, on the other hand, in his treatment of the right of succession accepts the guidance of the importance of maintaining the institution of the family. It is significant that in treating of the fundamental text of Gautama on ownership as derived from inheritance, pur chase, partition, seizure, or finding, both Jimutavahana and Vijnanesvara appeal to the Mimamsa in support of tenets which are essentially at variance. The view of the former accepts the doctrine that in all these modes of acquisition there is more than mere physical acquisition concerned ; the attaining possession must be in furtherance of duty in the widest sense. This view he supports by the doctrine that, when the priest becomes possessed of the remnant of the sacrifice, he does so not by the mere act of acceptance, but in virtue of the pious intention of the donor in dedicating the offerings to the deity. Thus for him succession becomes a matter of spiritual benefit, and the property comes to the heir not in virtue of his acquisition of it, but by a species of relinquishment by his ancestor, a principle upon which, it has ingeniously been suggested, the lawyers of Bengal might easily have built a doctrine permitting of the limited settlement of family property. Vijnanesvara,1 on the other hand, following a suggestion of Prabhakara, argues that Jaimini (IV, 1, 3-6) was of opinion that property was essentially a matter of popular recognition, and that the acquisition of property by an action in breach of law did not deprive the sacrifices made by means of it of full effi cacy. Hence Vijnanesvara's doctrine of succession rests on blood kinship, and heritage is defined by him as wealth that becomes the property of another solely by reason of relation ship to the owner. Similarly the Vyavahdramayukha (p. 32), in accepting the purely secular origin of property, nonetheless appeals to the Mimamsa treatment (VI, 7, 1, 2) of the Visvajit offering in order to show that one's children are not included in the term " property," for when at that offering the sacrificer is supposed to give away to the priests
1 On Yajnavalkya, II, 114; he cites Guru (p. 198, ed. Bombay, 1909), and the passage is found in the Brhati (Prabhakara School. P. 312).
102 THE KARMA-MIMAMSA
his all, his children are excluded from the gift. The same passage is also employed to express the limited character of the ownership of a king or a feudatory; his actual owner ship is restricted to whatever property he has acquired ; his position towards the territory is one of sovereignty or suzer- anity, entitling him to a maintenance but not to true owner ship; when a king is said to give a village, he does not trans fer the ownership of the land, which is not his to give, but assigns to the donee the right of drawing a maintenance from the village.
The Mimamsa also affords guidance to Vijnanesvara (II, 136) in a variety of details in connection with heritage and partition. The claim of woman to inherit is questioned on the ground that, as property is intended for sacrificial purposes and as save along with her husband a woman has no locus standi as a sacrificer, on the interpreta tion of the Mimamsa Sutra (VI, 1, 17-21) adopted in the commentators, there is no ground for her having the right of inheritance. This illiberal doctrine is disposed of by appeal to another passage of the Mimamsa Sutra (III, 4, 26) in which it refers to ornaments of gold worn by the priests and the sacrificer, though serving no sacrificial purpose. The exact share of a wife raises difficulties in view of the con flicting interpretation of the two main texts, the first of which provides that, if an owner divides property in his lifetime, he should make his wives have equal shares with his sons; and the second, that on partition after the death of the husband the wife should have a share equal to that of her sons. These passages are interpreted by some authorities to mean that, if the property is extensive, she is to have a mere subsistence from the estate, while, if it is small, she is to have an equal share. This view is rejected on the strength of the principle upheld by Jaimini (VII, 3, 19-25) that, so long as a text can yield a single coherent meaning, it is not right to treat it as broken into two incongruous parts. Simi larly it is on the Mimamsa rule (V, 1, 4-7) of following the order of things mentioned in a certain order that is based the claim that, when the parents of a childless son succeed to his property, the mother has a prior claim, because the term parents (pitarau) is explained in grammatical treatises
THE MIMAMSA AND HINDU LAW 103
as mother and father (matdpitarau), and not as father and mother. Another appeal to the Mimamsa doctrine is made in regard to Yajnavalkya's doctrine (II, 126) that a coparcener, who at the time of partition withholds part of the property, must give it up for division. The question arises whether the action is reprehensible or not, and is decided in the affirmative because in the Mimamsa (VI, 3, 20) it is ruled that a man who substitutes one form of meal for another, even if acting under a genuine misapprehension, still does wrong, so that, even if the coparcener had some right to the property and regarded it as his own, his conduct is censurable. Jimutavahana, as often, differs in part from Vijnanesvara, and extenuates the action. On a strict interpretation by Mimamsa principles again, it is not impossible to argue that Jimutavahana does not allow the disposal by will by a father of inherited property without provision being made for the maintenance of the sons; the conflicting view of the Privy Council is clearly hard to reconcile with the principles of Mimamsa.1
Adoption, like inheritance, affords a fruitful field^for the application of Mimamsa principles.2 The right of a Sudra to adopt, which is denied in the Suddhiviveka, on the ground that adoption must be accompanied by Vedic Mantras and an oblation which he cannot as a Sudra have performed, is vindicated on the ground of the occurrence of a certain offer ing for a Nisadasthapati (VI, 1, 51) in the Veda, although a Nisada is normally as a Sudra excluded from any Vedic rite. The Mantras can then be recited by an Aryan. A woman, again, can only adopt with the permission of her husband, as she cannot by herself perform Vedic rites and ceremonies (VI, 1,6). Again, a child when adopted cannot inherit his father's property or perform his Sraddha, accord ing to Manu; this rule, though restricted to these two facts, must be understood to apply generally on the analogy of terms like antarvedi in the Mimamsa (III, 7, 13, 14), which means not merely at the centre of the altar, but anywhere within it. By another maxim Nilakantha decides that
1 Tagore Law Lectures, 1905, pp. 405-11.
2 Vyavaharamayukha (ed. Bombay, 1880), pp. 40 ff.
104 THE KARMA-MlMAMSA
Saunaka's text, which asserts that the son of a daughter and the son of a sister are adopted by Sudras, is to be read to mean that these adoptions are generally permissible, and that they are specially so in the case of Sudras : the maxim used is the Maitravaruna, which rests on the interpretation of the two sentences, " He hands over a staff to the Maitravaruna priest; he initiates or invokes by means of the staff." The accepted opinion is that the handing over of the staff is a distinct injunction, the initiation or invocation subsidiary, and so here the part of the Sudra is only subsidiary to an established rule. Samkara Bhatta, his father, whom Nilakantha cites, expressly applies to the Sudra the duty of paying his debt to the Fathers, which is asserted of the Brahman as an instance in the Mimamsa Sutra.
Similar use of the Mimamsa is made in the same con nection by Nanda Pandita in the Dattakamlmamsa (c. 1600 A.D.). Thus on the analogy of the Vaisvadeva, which is a maxim (I, 4, 13-16) laying down that in the case of such a word as that the conventional sense is to be fol lowed in lieu of the etymological, he holds (VI, 27) that the term sapinda used of relationship is not to be restricted to the exact meaning suggested by the word as a compound. So also, in order to meet the objection of Medhatithi to an adopted son on the ground that the duty of man is fulfilled only by begetting a son, he adduces (I, 41) the maxim (VI, 3, 31) of the substitution of the Putika for the Soma plant. In determining the value of substitution the mode in which the substitute originated is unimportant, the question is whether it can serve its purpose adequately, and this an adopted son can easily do. Again, the objection to the rule that an adoptive father must perform the birth ceremony for an adopted child, though adoption is permitted up to the fifth year, is met by the use of the maxim (V, 4, 5-14) that, when a difficulty arises as to the order of performance of offerings, reason and necessity must be consulted, whence it follows that the performance of the birth ceremony is in order though tardy. The author of the Dattakacandrika similarly appeals to the Mimamsa doctrine(IV, 1, 22-24)of the relation of the principal and incidental aspects of an action, in order to support his view that, if one of two co-widows adopt, the
THE MIMAMSA AND HINDU LAW 105
child becomes the adopted child of the other as well. As the principal purpose of the action is to provide offspring for the dead husband, the result of doing so is that the son occupies the same position to either widow, this being a mere incidental matter. The same text elsewhere cites the Kapinjala maxim (XI, 1, 38-45) which indicates that, when the plural number is used in any injunction, in the absence of necessity requiring that a larger number should be deemed to be meant, the needs of the situation are fully met by restricting the number to three, the minimum indicated.
The sacrificial practice yields one obvious contribution to the law of partnership, such as those of trading companies, bodies of actors, and agricultural concerns. The rule is laid down that the returns are to be divided among the partners according to the amounts of their respective invest ments, on the analogy of the distribution of the sacrificial fees among the officiating priests (X, 3, 53-55). Thus, of one hundred cows twelve are given to each of the four principal priests, six to the next four, four each to the next four, and three each to the last group, the amounts being allocated in accordance with the comparative importance of their contri bution to the carrying out of the offering.
In the domain of quasi-criminal law Raghunandana solves a difficulty by the application of the analogy of the principle of Tantra, where a single performance of an action serves the purpose of more than one principal offering. The problem is raised in the case of the rule that the assailant of a Brahman must perform the Krcchra penance on pain of punishment; if, then, one offends against five Brahmans, is the penance to be performed five times ? The answer on the Tantra principle is in the negative. Medhatithi again has recourse to the Mimamsa maxim of Grahaikatva (III, 1, 13-15). which holds that in a general injunction the singular includes the plural and the masculine the feminine, in order to solve the doubt raised by the rule of Manu (V, 90) that a Brahman must not drink spirituous liquor, an ingenious objector having suggested that the text is restrict ed in its application to such action by a single male Brah man. The maxim, it will be seen, differs entirely in its effect from that given above, which in sacrifices and on
106 THE KARMA-MIMAMSA
other occasions reduces a plural to its bare meaning of three.
Even more interesting is a case in which the law of evidence is influenced by the Mimamsa doctrine of the self validity of cognitions. Yajnavalkya lays down (II, 80) that, if a man has brought forward witnesses, yet if at a later period he can produce more satisfactory testimony, the evidence already adduced is to be discredited. This pro cedure, at first sight drastic, is justified b> the adduction in the Mitaksard of the arguments adduced by the Vrttikara in support of the self-evidence of cognitions. Evidence is prima facie valid, unless it can be shown that the witness could not have known the facts, that his means of knowl edge were defective (kdranadosa) , or his evidence is displaced by other evidence, that is, the first cognition is sublated by a second cognition. Immediately after, Vijnanes- vara (II, 83) has recourse to the Mimamsa to provide a suitable penance for the witness whom he enjoins to with hold evidence or testimony, where the proof of the charge would result in the infliction of the capital penalty ; in these cases the usual punishment of a fine, or in the case of a Brahman banishment, is not in point ; still, to do away with the sin of the deviation from the truth the performance of a special offering, the Sarasvatesti, given in the Mimamsa, is prescribed.
As is natural, the obligations of the law books to the Mimamsa principles are still more marked in those parts of those treatises which deal, not with civil law (vyavahdra) in the narrower sense of the term, but with religious custom and penances. Even in the civil law, however, there is one point on which the law books differ in essentials from Jaimini ; it was necessary for the latter, in support of his doctrine of the eternity of the Veda, to maintain that its commands are universal, and thus he treats even Smrti texts which contain injunctions expressed as local practises as really laying down general principles. In the practical needs of the law, however, the utmost value is always attached to local customs, and the practice of good men, which thus in effect comes to outweigh maxims in Smrtis, if in any place these are not followed. Yet
THE MIMAMSA AND HINDU LAW 107
Jaimini's insistence on the supreme value of the Veda in all questions was not without effect; the tendency in the Smrtis is, in harmony no doubt with a common practice, to allot in the case of partition of property a larger share to the eldest son than to the others. But there is Vedic authority for the statement that Manu divided his property in equal shares among his sons, and this doctrine has finally prevailed in the law, despite the efforts of some of the compilers of digests to compromise the matter in order to obey the clear directions of the Smrtis.1 In the legal schools, again, it has been found necessary to assign relative weight to Puranas and Smrtis, a distinction which is not found in Kumarila, who accepts the Puranas on the same basis as the Smrtis. In the case of a divergence between Smrti and Purana the former should prevail in the view of Vyasa; the Purana represents no more than custom, while the Smrti is a step nearer to Sruti.2
While the Mimamsa thus stands in close relation with Indian law, in its enunciation of principles in the form of brief maxims (nydya), comparable with the headnotes of modern law reports, it stands in equally close relation with the popular vogue of maxims3 framed on the model whence the Mimamsa use is doubtless derived. Such popular maxims are freely cited by the text-books of the school, and it was presumbly on their analogy that the Adhikarana headings were derived; the remarkable divergence of the commentators4 in allotting Sutras to Adhikaranas indicates that the latter were not a primitive constituent of the Sutra text.
1 Cf. Mandlik, trans, of Vyavaharamayukha, p. 41, n. 1.
2 Tagore Law Lectures, 1905, pp. 234, 235 ; d., however, Mandlik, op. cit . p. xxx.
3 See Col. Jacob's Lankikany&ydnjali (2nd ed., 3 parts). * Above pp. 4, 13. So in the Veddnta Sutra.
INDEX
A CTION, 56, 66 •rA Actions, classification of,
85,86
Adoption, 103 , 104 Adhikarana, 4, 13, 107 Air, 53, 54 Aita§ayana, 4 Alayavijnana, 47 Analogy, 32, 33 Anandagiri, 11 Anantadeva, 13 Angapurva, 75 Annulment of rites, 95 Anubhuti, 17 Anumana, 30 Anusanga, 82 Anuvyavasaya, 22, 50 Anvaya, 99 Apastamba, 2 Apadeva, 13, 76, 78 Apoha, 40 Appayya Diksita, 12 Apprehension, validity of, 17-20 Aprama. 17
Apurva, 36, 73, 74,75, 89 Arthapatti, 33, 34 Arthasamgraha, 13 Arthavada, 80, 83, 88, 99, 100 Aryadeva, 7 Asanga, 46 A§vaghosa, 7 Atheism, 61
Atomic theory, 50, 54, 61, 62 Atreya, 4 Aupavarsas,'71 Avapa, 95 Aversion, 54, 55, 68
DADARAYANA, 4, 5, 35, 77 u Badari,4, 88 Baudhdyana Dharma Sutra, 97 Bhagavadgitd, 73, 76 Bhdmatl, 16 Bhartrhari, 11, 36 n. Bhartrmitra, 8 Bhasarvajna, 32
Bhatta Dinakara, 12, 97
Bhatta Samkara, 12, 98
Bhattabhdskara, 13
Bhdttacintamani, 12, 97
Bhattalamkara, 13
Bhdttarahasya, 13
Bhavadasa, 7
Bhavanatha Mi§ra, 12
Bhavana, 75, 76
Bodhayana, 8
Body, 67, 69 ; of creator, 62, 63
Brahman, 36 n., 39, 76, 77
Brahruanas, Mimamsa in, 1; re cognise soul, 68 ; contents of, 79-81
Brhatl, 9, 101 n.
Buddhist views, 6, 7, 14, 30, 40, 65, 84 ; and see Vijnanavada and !§unyavada.
/CATEGORIES, 52, 53
^ Cause, 23, 24, 59, 60
Chandogya Upanisad, 77
Cidananda, 28
Class signification of words, 39
Classification of actions, 85, 86
Cognition, validity of, 17-20 ; mode of apprehension of, 20- 22, 45, 46, 49, 50, 68, 70, 71 ; of soul, 70, 71; as quality of soul, 67
Colour, 54, 55
Conjunction, 54, 55 ; in percep tion, 23, 26, 35
Construction, rales of, 81, 82, 86, 87, 89, 99
Convention, as basis of lan guage, 36
Co-partners, 103
Creator, existence of denied, 36, 43, 61-64
Custom, value of, 85, 106, 107
r\ANA, 86
^ Darkness, 53, 54 Dattakacandrika, 104
INDEX
109
Dattakamimdmsa , 104 Deities, in sacrifice, 74, 77, 78, 81 Demerit, 55, 62, 67, 68, 71 Desire, 54, 55, 66, 67 Determinate perception, 24, 25 Devatasvarupavicara, 78 Dharma, 4, 35 Dharma Sutras, Mimamsa in, 2,
3, 97
Dharmakirti, 9 n., 14, 24 Dhvani, 55 Dignaga, 14, 24, 27, 28, 29, 30,
48, 52
Dimension, 54, 55, 68 Dinakara Bhatta, 12, 97 Disjunction, 54, 55 Dream cognition, 18, 44, 45
ETFFORT, 54, 55
*-* Ekavakyata, 82
Elasticity, 55
Error, 19, 20
Eternity, of word, 37-39 ; of
Veda, 42, 43, 106 Ether, 37, 38, 53,_54 Evidence. Mimamsa rules as
affecting law of, 106
|7ALLAClES,doctrineof, 30-32 r Fire, 53, 54 Fluidity, 55
(~* AGA, or VisveSvara, Bhatta, ^ 12, 97
GangeSa, 15
Gautama, 77
Generality, 26, 56-58 ; cf. 50, 51
Gesture, as a Pramana, 43
God, 36, 61-64, 76
Gopala Bhatta, 12
Govinda, 76"
Grahaikatva, 105
Grammar, value of, 84, 85
Gravity, 55
Gunaratna, 15, 76
Guru, i.e. Prabhakara, 9, 101 n.
LJARI, 8
1 * Haribhadra, 15 Hearing, 38, 39 Homa, 86
IDEAS, how known, 20-22,45, 46, 49, 50, 68, 70, 71
Imagination, function of in knowledge, 24, 25
Inmortality, 64, 65, 72
Import of words, 39, 40
Impression, 49, 55, 66, 68
Indeterminate perception, 24 25; alone valid, 52 n.
Individuality, 25, 54, 55, 68
Inference, 27-32
Infinite regress, 18, 22, 28
Inherence, 26, 54, 58
Injunction, 35, 74, 75, 85-94, 98
Intrinsic validity of apprehen sion. 17-20
Isu,94
ISvara, 76
IsVarakrsna, 59
Interpretation, rules of Vedic, 79_-96
Itihasa, authority of, 82, 83
TAIMINI, 4, 5, 35, 76, 77, 78, 79, J 88, 91, 99_, 106 Jaiminiyanyayamalavistara, 12 Jaiminisutrabhasya, 14 Jain views, 32 n., 34, 38, 68, 69 Jayanta_Bhatta, 14, 15 Jimutavahana, 100, 101, 103 Jivadeva, 13
K'ALANJA, 93
^ Kanada, 77
Kamalakara, 12, 97
Kapila, 77
Kapinjala, a Nyaya, 105
KaSika, 10
KeSava, 13
Khandadeva, 12, 13
Knowledge, theory of, 17-43 ; and liberation, 72, 73
Krcchra, 105
Krsna Diksita, 13
Kulluka, 97
Kumarila, 6, 7, 9, 10, 14, 15, 17-19, 21, 22, 25, 27, 28, 30, 34, 35, 38,40,41,46,49,52,53,61, 64, 69, 71, 73, 74, 75, 79, 80, 82, 83_, 85, 90
Kusumanjali, 14
110
THE KARMA-MIMAMSA
I ABUKAYANA, 4 *-• Laksana or Linga, 89, 99 Language, origin of, 36, 37 Laugaksi Bhaskara, 13, 76 Law, Mimamsa principles in,
97-107
Liberation, 72-74 Like, known by like, 23 Logic, 14, 17-43.
|V/[ADHAVA, 3, 4, 12, 15, 97
1V1 Madhyamikas, 46, 47
Mahdbhdrata, 5, 7
Mahdbhdsya, 5 n.
Mahadeva Vedantin, 12
Maitravaruna, a Nyaya, 104
Manadeva, 14
Manameyodaya, 14
Mandana MiSra, 11, 16
Mantra, 81, 82 ; Uha of, 94, 95
Manu, 82, 97, 98, 103
Matter, 50, 53, 54, 61, 67, 69
Mayukhamalikd, 12
Meaning, relation of to word, 35,36
Medhatithi, 98, 104
Memory, 19, 20, 25, 66, 67, 68
Mental perception, 20, 50 ; and see mind.
Merit, 55, 62, 67, 68, 71
Metaphysics, 44-78
Mimamsa, 1
Mimdmsd Sutra, 3-7, 37 n ; con tents, 79-96
MimdmsdbdlaprakdSa, 13
Mlmamsaka, 3, 5 n.
Mimdmsdkaustubha, 12
Mimdmsdnayaviveka, 12
Mlmdmsdnukramam, 12
MimdmsdnydyaprakdSa, 13
Mimdmsdmakaranda, 14
Mimdmsdpddukd, 14
Mimdmsdparibhdsd, 13
Mimamsdratna, 13
Mlmdmsdsdrasamgraha, 13
Mind, 23, 24, 68, 69, 70 ; and see Mental Perception.
Mitdksard, 9, 106
Mode of apprehension of cogni tion, 20-22, 45, 46, 49, 50, 68, 70, 71 ; of soul, 70, 71
Motion, 56, 66
MAGARJUNA, 6, 44
1 ^ Name, 80, 81
Nanda Pandita, 99, 104
Ndyakaratna, 12
Narayana, 14
Narayana, of Kerala, 14
Narayanatlrtha Muni, 13
Nescience, 63, 64
Nigadas, 95
Nihilism, 6, 7, 44, 45, 46-52
Nilakantha, 97, 103, 104
Nirnaya, 4
Nisadasthapati, 103
Nisedha, 93
Niyama, 93, 98
Niyoga, 75
Non-existence, 34, 35, 52, 53, 60
Non-perception, 34, 35, 60
Number, 52, 68
Nyaya, meaning of, 2, 107
Nyaya school, 14, 17-20, 22, 26,
27, 29, 30, 32, 33, 34, 36, 40,
55, 70 Nydya Siltra, 5, 14, 24, 29, 37 n.,
44, 46, 51 Nyaya-VaiSesika school, 14,
23, 50, 52, 53, 56, 57, 58, 60,
61, 63, 68, 69 Nydyakanikd, 11 Nydyamanjari, 14 Nydyaratndkara, 10 Nydyaratnamdld, 12 Nydyasudhd, 10 NydydvaUdldhiti, 12 Nydyavdrttika, 14
f^BJECTIVE idealism, 52 ^ Ontology, 44-60 Order of sacrifices, 90, 91
DAIN, 54, 67, 68, 73 * Pancatantra, 5 n. Panini, 5 n. Pafisamkhya, 93, 98 Parthasarathi MiSra, 10, 11, 12,
21,87
Particularity, 52, 59 Partnership, 105 Paryudasa, 85,86, 94, 98
INDEX
111
Perception, 22-27, 54, 63 ; of motion, 56 ; of similarity, 59
Phala, 27, 30
Phalapurva, 75
Pleasure, 54, 66, 67, 68, 73
Plurality of souls, 71, 72
Posteriority, 54, 55
Potency, 32, 55
Prabhacandra, 11
Prabhakara, 9, 10, 14, 15, 17, 19, 20, 22, 25, 27, 28, 30, 34, 35, 39, 40, 41, 52, 53, 61, 64, 69, 70, 71, 72, 74, 75, 79, 80, 83, 85, 90
Prakarana, 94, 99
Prakaranapancika, 16
Prakatya, 55
Prama\ 17
Pramana, 27 ; number of . 35, 43
Prameyapdrdyana, 9, 16, 55 n.
Pranabhrt, a Nyaya, 99
Prasanga, 96
Prasastapada, 15, 27, 29, 30, 54, 55, 60 n.
Pratisedha, 85, 86
Pratyabhijna, 17
Presumption, 33, 34 ; as proof of Apurva, 74
Priority, 54, 55
Purana, authority of, 82, 83, 107
Purvapaksa, 3
QUALITIES, 54, 55
DAGHAVANANDA Saras- *v vati, 12 Raghunandana, 99, 105 Raghunatha, 13 Ramakrsna, 12 Ramakrsna Udicya Bhatta-
carya, 13
Ramanuja, 12, 46, 77 RamesVara, 12 Rdnaka, 10 Rathakara,_92
Ratnakara§anti, idealism of, 48n. Recognition, 17 Remembrance, 17 ; and see
Memory.
Rumour, as a Pramana, 43 Rjuvimald, 9
CABARASVAMIN, 7, 8, 9, 10, j? 39, 44, 64, 87, 88, 91 Sabda, as form of absolute, 36
n. ; and see Word. Sacrifices, 22, 73, 74, 76 Saddarfanasamuccaya, 15 S"alikanatha, 9, 17, 60 n., 75 Saman, 81 Samantabhadra, 11 Sambhava, as a Pramana, 43 Sarbgati, 4
Samkara, 6,7, 9, 46, 76 S"arhkara Bhatta, 12, 98, 104 Samkaravijaya, 11 Sdmkhya Sutra, 5, 36, 38, 64, 73. SarhSaya, 3 Samudayapurva, 75 Sdrirakabhdsya, 10 SarvadarSanasamgraha, 15 Sarvasiddhdntasamgraha, 9, 15 ^astra, as a Pramana, 35, 41 Sdstradipikd, 11, 12 Sattras, 92 Saunaka, 104 Sautrantika, 49 Self -consciousness, 20-22 Self-evidence of cognitions, 17-
20, 106 ; and see Ideas. Senses, 23, 54, 55 Sense organs, 23, 67, 69 Series of ideas, 49, 65, 66 S"esa,87
Segvaramimamsa, 14, 76 Siddhanta, 4
Similarity, 32, 33, 52, 58, 59 Slokavdrttika, 10 Smell, 54, 55 Smrti, 82, 83, 107 SomeSvara, 10 Soul, 64-72
Sound, 37, 38, 39, 53, 54 Space, 53, 54, 55 Sphota, 33 Spirit ; see Soul. S*ridhara, 15, 54 Subodhini, 12 Substance, 53, 54
112
THE KARMA-MlMAMSA
Sucarita Misra, 10 Suddhiviveka, 103 Sudra, 103
Sunyavada, 6, 7, 44, 45, 46-52 SureSvara, 11 Svatahpramanya; 7 Syena, 94 Syllogism, 29, 30
MANTRA, 95, 105
•*• Tantravarttika, 10 Tdrkikaraksd , 14 Taste, 54, 55 Tattvabindu, 11 Tattvacintamani, 15 Theism, 76 ; and see God, Dei ties.
Time, 53. 54, 55 Touch, 53, 54, 55 Transcendental perception, 27,
28
Transfer of ceremonies, 94, 95 Transmigration, 65, 66 Tuptikd, 10
T 1DAYANA, 14, 32
~ Uddyotakara, 9 n., 14, 24 Una, 94, 95 Unseen principle, 20 Upakratnapardkrama, 13 Upamana, 33, 34 Upavarsa, 7, 8 Utpattyapurva, 75
WACASPATI MiSra, 11, 14, v 15 n., 24, 46 Vaidyanatha, 12 VaiSesika school, 32, 35, 36, 42-
66, 70, 73 Vaifesika Sutra, 5 VaiSvadeva, a Nyaya, 104 Vakyabheda, 82 Validity of knowledge, 17-20 ;
of percegtion, 25, 26 Vallabha Acarya, 13 Varadaraja, 14, 16 Vdrttikdbharana, 10 Varttikakara, 9 Vasubandhu, 46 Vatsyayana, 24, 46
Venkatadhvarin, 14
Venkafanatha, VenkateSa, 10, 76
VenkateSvara Diksita, 14
Veda, 36, 42, 43, 63
Vedanga, 83
Vedanta school, 36, 46, 63, 64, 69, 70, 72
Vedanta Sutra, 5, 6, 7, 44, 46, 61, 76,_77
Vijnanavada, 5, 6, 7, 20, 22, 46- 52 _
VijnaneSvara, 101
Vidhi ; see Injunction.
Vidhirasayana, 12 ; commen taries on, 12
Vidhitrayaparitrdna, 14
Vidhiviveka, 11
Vidyananda, 11
Vikalpa, 96
Vikrti, 2, 86
Vindhyavasin, 59, 65
Visaya, 3
Viscidity, 55
ViSvajit, a Nyaya, 87, 101
Vrttikara, 7, 8, 20, 22, 27, 29, 30, 32, 34, 35, 42, 88, 106
Vyapti, 27
Vyasa, 107
Vyavahdramayukha, 101, 103
Vyavasaya, 22, 50
\Y7ATER, 53, 54
vv Whole and part, 50, 51; cf. 56-58
Wife, share of husband's pro perty, 102
Woman, right to sacrifice, 91, 92, 102
Word, or verbal testimony, 35- 43
GA, 86
Yajnavalkya, 103, 106 Yajus, 81
Yuktisnehaprapuram, 12 Yoga Sutra, 5 Yogdcdras, 47
Yogins, perception of, 27 n., 48
-
University of Toronto library
DO NOT
REMOVE
THE
CARD
FROM
THIS
Acme Library Card Pocket
Under Pat. "Ref. Index File" Made by LIBRARY BUREAU
•