YALE UNIVERSITY LIBRARY HINDUISM AND BUDDHISM AN HISTORICAL SKETCH BY SIR CHARLES ELIOT 4 * • H.M. AMBASSADOR AT TOKYO IN THREE VOLUMES VOLUME II LONDON EDWARD AENOLD & Co. 1921 (AU rights resened) CONTENTS BOOK IV THE MAHAYANA OHAPTEE XVI. Main Featuees of the Mahayana XVII. BODHISATTVAS XVin. The Buddhas of Mahayanism XIX. Mahayanist Metaphysics XX. Mahayanist Sceiptuees XXI. Chronology of the Mahayana . xxn. Feom KLanishka to Vasubandhu XXIII. Indian Buddhism as seen by Pilgeims XXIV. Deoadbnoe of Buddhism in India THE Chinese page 37 2636_ 476376 90 107 BOOK V HINDUISM XXV. SrvA AND Vishnu 136 XXVI. Featuees of Hinduism: Ritual, Caste, Sect, Faith 166 XXVII. The Evolution of Hinduism. Bhagavatas and PISUPATAS 186 XXVIII. Sankaea. Sivaism in Southeen India. Kashmie. LingIyats 206 XXIX. Vishnuism in South India 228 XXX. Latee Vishnuism in Noeth India . . . 242 XXXI. Amalgamation of Hinduism and Islam. Kabie AND THE Sikhs 262 XXXII. glKTisM 274 XXXIII. Hindu Philosophy 291 BOOK IV THE MAHAYANA CHAPTER XVI MAIN FEATURES OE THE MAHAYANA The obscurest period in the history of Buddhism is that which foUows the reign of Asoka, but the enquirer cannot grope for long in these dark ages without stumbhng upon the word Mahayana. This is the name given to a movement which in its various phases may be regarded as a philosophical school, a sect and a church, and though it is not always easy to define its relationship to other schools and sects it certainly became a prominent aspect of Buddhism in India about the beginning of our era besides achieving enduring triumphs in the Ear East. The word^ signifies Great Vehicle or Carriage, that is a means of conveyance to salvation, and is contrasted with Hinayana, the Little Vehicle, a name bestowed on the more conservative party though not willingly accepted by them. The simplest description of the two Vehicles is that given by the Chinese traveUer I-Ching (635-713 A.D.) who sa;v7 them both as Hving reahties in India. He says^ "Those who worship Bodhisattvas and read Mahayana Sutras are called Mahayanists, while those who do not do this are caUed Hinayanists." In other words, the Mahayanists have scriptures of their own, not included in the Hinayanist Canon and adore superhuman beings in the stage of existence immediately below Buddhahood and practicaUy differing Httle from Indian deities. Many characteristics could be added to I-Ching's description but they might not prove universaUy true of the Mahayana nor entirely absent from the Hinayana, for however divergent the two Vehicles may have become when separated geographicaUy, for instance in Ceylon and Japan, it is clear that when they were in contact, as in ^ Sanskrit, Mahdydna; Chinese, Ta Ch'ing (pronounced Tai Sheng in many southern provinces); Japanese, Dai-jo; Tibetan, Theg-pa-chen-po; Mongolian, Takd-killgm; Sanskrit, Hiriaydna; Chinese, Hsiao-Ch'ing; Japanese, Sho-jo; Tibetan, Theg-dman; Mongolian t)tmMn-kiJlgS.n, In Sanskrit the synonyms agra- yana and uttama-yana are also found. ^ Record of Buddhist practices. Transl. Takakusu, 1896, p. 14. Hsuan Chuang seems to have thought that acceptance of the Yogacaryabhumi (Nanjio, 1170) was sssential for a Mahayanist. See his Ufe, transl. by Beal, p. 39, transl. by JuHen, p. 50. 1—2 4 THE MAHAYANA [ch India and China, the distinction was not always sharp. But ii general the Mahayana was more popular, not in the sense o being simpler, for parts of its teaching were exceedingly abstruse but in the sense of striving to invent or include doctrines agree able to the masses. It was less monastic than the olde: Buddhism, and more emotional; warmer in charity, more personal in devotion, more ornate in art, Hterature and ritual more disposed to evolution and development, whereas th( Hinayana was conservative and rigid, secluded in its cloisten and open to the plausible if unjust accusation of selfishness The two sections are sometimes described as northern and southern Buddhism, but except as a rough description of theii distribution at the present day, this distinction is not accurate, for the Mahayana penetrated to Java, while the Hinayana reached Central Asia and China. But it is true that the develop ment of the Mahayana was due to influences prevalent in northern India and not equaUy prevalent in the South. The terms PaH and Sanskrit Buddhism are convenient and as accurate as can be expected of any nomenclature covering so large a field. Though European writers usuaUy talk of two Yanas or Vehicles — the great and the Httle — ^and though this is clearly the important distinction for historical purposes, yet Indian and Chinese Buddhists frequently enumerate three. These are the iSrdvakaydna, the vehicle of the ordinary Bhikshu who hopes to become an Arhat, the Pratyekabuddhaydna for the rare beings who are able to become Buddhas but do not preach the law to others, and in contrast to both of these the Mahdydna or vehicle of Buddhas and Bodhisattvas. As a rule these three Vehicles are not regarded as hostile or even incompatible. Thus the Lotus sutra^, maintains that there is reaUy but one vehicle though by a wise concession to human weakness the Buddha lets it appear that there are three to suit divers tastes. And the Mahayana is not a single vehicle but rather a train comprising many carriages of different classes. It has an unfortunate but distinct later phase known in Sanskrit as Mantrayana and Vajrayana but generaUy described by Europeans as Tantrism, This phase took some of the worst features in Hinduism, such 1 Saddharma-Pundarika, chap. iii. For brevity, I usually cite this work by the title of The Lotus. XVI] MAIN FEATURES OF THE MAHAYANA 5 as speUs, charms, and the worship of goddesses, and with mis placed ingenuity fitted them into Buddhism. I shall treat of it in a subsequent chapter, for it is chronologically late. The silence of Hsiian Chuang and I-Ching impHes that in the seventh century it was not a noticeable aspect of Indian Buddhism. Although the record of the Mahayana in literature and art j is clear and even briUiant, it is not easy either to trace its rise 1 or connect its development with other events in India. Its . annals are an interminable Ust of names and doctrines, but j bring before us few Hving personalities and hence are duU. [ They are like a record of the Christian Church's fight against ( Allans, Monophysites and Nestorians with aU the great fi^gures .j of Byzantine history omitted or caUed in question. Hence I feaf (; that my readers (if I have any) may find these chapters repeUent, jja mist of hypotheses and a catalogue of ancient paradoxes. 1 1 can only urge that if the history of the Mahayana is uncertain, I its teaching fanciful and its scriptures tedious, yet it has been a force of the first magnitude in the secular history and art of :; China, Japan and Tibet and even to-day the most metaphysical of its sacred books, the Diamond Cutter, has probably more ¦ readers than Kant and Hegel. ,; Since the early history of the Mahayana is a matter for II argument rather than precise statement, it wiU perhaps be best J to begin with some account of its doctrines and literature and jproceed afterwards to chronology. I may, however, mention -that general tradition connects it with King Kanishka and j, asserts that the great doctors A^vaghosha Mid Nagarjuna Hved .'pin and immediately after his reign. The attitude of Kanishka ' and of the Council which he summoned towards the Mahayana is far from clear and I shall say something about this difficult .i^subject below. Unfortunately his date is not beyond dispute I for while a considerable consensus of opinion fixes his accession lat about 78 a.d., some scholars place it earher and others in the ¦.second century a.d.^ Apart from this, it appears estabUshed |that the Sukhavati-vyuha which is definitely Mahayanist was ^translated into Chinese between 147 and 186 a.d. We may U.ssume that it was then already well known and had been com- ^posed some time before, so that, whatever Kanishka's date may * The date 58 B.C. has probably few supporters among scholars now, especiaUy ifter Marshall's discoveries. 6 THE MAHAYANA [oh. xn have been, Mahayanist doctrines must have been in existence about the time of the Christian era, and perhaps considerably earHer. NaturaUy no one date Hke a reign or a council can be selected to mark the beginning of a great school. Such a body of doctrine must have existed piecemeal and unauthorized before it was collected and recognized and some tenets are older than others. Enlarging I-Ching's definition we may find in the Mahayana seven Hues of thought or practice. AU are not found in aU sects and some are shared with the Hinayana but probably none are found fuUy developed outside the Mahayana. Many of them have parallels in the contemporary phases of Hinduism. 1. A belief in Bodhisattvas and in the power of humai beings to become Bodhisattvas. 2. A code of altruistic ethics which teaches that everyoni must do good in the interest of the whole world and make ovei to others any merit he may acquire by his virtues. The aim o] the rehgious Hfe is to become a Bodhisattva, not to become ar Arhat, 3. A doctrine that Buddhas are supernatural beings, distri buted through infinite space and time, and innumerable. In the language of later theology a Buddha has three bodies and still later there is a group of five Buddhas. 4. Various systems of ideaUst metaphysics, which tend to regard the Buddha essence or Nirvana much as Brahman is regarded in the Vedanta. 5. A canon composed in Sanskrit and apparently later than the PaH Canon. 6. Habitual worship of images and elaboration of ritual. There is a dangerous tendency to rely on formulae and charms. 7. A special doctrine of salvation by faith in a Buddha, usually Amitabha, and invocation of his name. Mahayanism can exist without this doctrine but it is tolerated by most sects and considered essential by some. CHAPTER XVII BODHISATTVAS Let us now consider these doctrines and take first the worship of Bodhisattvas. This word means one whose essence is know ledge but is used in the technical sense of a being who is in process of obtaining but has not yet obtained Buddhahood. The PaH Canon shows Httle interest in the personaUty of Bodhisattvas and regards them simply as the preUminary or larval form of a Buddha, either Sakyamuni^ or some of his predecessors. It was incredible that a being so superior "to" ordinary humanity as a Buddha should be suddenly produced in a human family nor could he be regarded as an incarnation in the strict sense. But it was both logical and edifjnng to suppose that he was the product of a long evolution of virtue, of good deeds and noble resolutions extending through count less ages and culminating in a being superior to the Devas. Such a being awaited in the Tushita heaven the time fixed for his appearance on earth as a Buddha and his birth was accom panied by marvels. But though the PaU Canon thus recognizes the Bodhisattva as a type which, if rare, yet makes its appear ance at certain intervals, it leaves the matter there. It is not suggested that saints should try to become Bodhisattvas and Buddhas, or that Bodhisattvas can be helpers of mankind^. But both these trains of thought are natural developments of the older ideas and soon made themselves prominent. It is a characteristic doctrine of Mahayanism that men can try and should try to become Bodhisattvas. ' In dealing with the Mahayanists, I use the expression Sakyamuni in preference to Gotama. It is their own title for the teacher and it seems incongruous to use the purely human name of Gotama in describing doctrines which represent him as superhuman. " But Kings Hsin-byu-shin of Burma and Sri Suryavainsa Bama of Siam have left inscriptions recording their desire to become Buddhas. See my chapters on Burma and Siam below. Mahayanist ideas may easily have entered these countries Erom China, but even in Ceylon the idea of becoming a Buddha or Bodhisattva is lot unknown. See Manual of a Mystic (P.T.S. 1916), pp. xviii and 140 8 THE MAHAYANA [ch. In the PaH Canon we hear of Arhats, Pacceka Buddhas, and perfect Buddhas. For aU three the ultimate goal is the same, namely Nirvana, but a Pacceka Buddha is greater than an Arhat, because he has greater inteUectual powers though he is not omniscient, and a perfect Buddha is greater stiU, partly because he is omniscient and partly because he saves others. But if we admit that the career of the Buddha is better and nobler, and also that it is, as the Introduction to the Jataka recounts, simply the result of an earnest resolution to school himself and help others, kept firmly through the long chain of existences, there is nothing iUogical or presumptuous in making our goal not the quest of personal salvation, but the attainment of Bodhisattvaship, that is the state of those who may aspire to become Buddhas. In fact the Arhat, engrossed in his own salvation, is excused only by his humiUty and is open to the charge of selfish desire, since the passion for Nirvana is an ambition Hke any other and the quest for salvation can be best foUowed by devoting oneseU entirely to others. But though my object here is to render inteUigible the Mahayanist point of view including its objections to Hinayanism, I must defend the latter from the accusation of selfishness. The vigorous and authorita tive character of Gotama led him to regard all mankind as patients requiring treatment and to emphasize the truth that they could cure themselves if they would try. But the Buddhism of the Pali Canon does not ignore the duties of loving and instructing others^; it merely insists on man's power to save himself if properly instructed and bids him do it at once: "sell aU that thou hast and foUow me." And the Mahayana, if less self-centred, has also less seH-reHance, and seU-discipline. It is more human and charitable, but also more easygoing : it teaches the beUever to lean on external supports which if weU chosen may be a help, but if trusted without discrimination become paralyzing abuses. And if we look at the abuses of both systems the fossiUzed monk of the Hinayana wiU compare favourably ' E,g, in Itivuttakam 75, there is a description of the man who is like a drought and gives nothing, the man who is like rain in a certain district and the man who is Sabbabhutanukampako, compassionate to all creatures, and like rain falling everywhere. Similarly ib, 84, and elsewhere, we have descriptions of persons (ordinary disciples as well as Buddhas) who are bom for the welfare of gods and men bahujanahitaya, bahujanasukhaya, lokanukampaya, atthaya, hitaya, sukhaya devamanussanam. xvu] BODHISATTVAS 9 nth the tantric adept. It was to the corruptions of the Hahayana rather than of the Hinayana that the decay of Buddhism in India was due. The career of the Bodhisattva was early divided into stages bhumi) each marked by the acquisition of some virtue in his iriumphant course. The stages are variously reckoned as five, seven and ten. The Mahavastu^, which is the earUest work where ;he progress is described, enumerates ten without distinguishing ihem very clearly. Later writers commonly look at the Bodhi- sattva's task from the humbler point of view of the beginner tvho wishes to learn the initiatory stages. For them the Bodhisattva is primarily not a supernatural being or even a saint but simply a reUgious person who wishes to perform the iuties and enjoy the privileges of the Church to the fuU, much like a communicant in the language of contemporary Christianity. We have a manual for those who would foUow this path, in the Bodhicaryavatara of Santideva, which in its humiUty, sweetness and fervent piety has been rightly compared with the De [mitatione Christi. In many respects the virtues of the Bodhi sattva are those of the Arhat. His will must be strenuous and concentrated; he must cultivate the strictest morahty, patience, energy, meditation and knowledge. But he is also a devotee, a bhakta; he adores all the Buddhas of the past, present and future as weU as sundry superhuman Bodhisattvas, and he con fesses his sins, not after the fashion of the Patimokkha, but by accusing himseU before these heavenly Protectors and vowing to sin no more. Santideva Hved in the seventh century^ but tells us that he foUows the scriptures and has nothing new to say. This seems to be true for, though his book being a manual of devotion presents its subject-matter in a dogmatic form, its main ideas are stated and even elaborated in the Lotus. Not only are eminent figures in the Church, such as Sariputra and Ananda, there designated as future Buddhas, but the same dignity is predicted wholesale for five hundred and again for two thousand ' Ed. Senart, vol. i. p. 142. ^ The Bodhicaryavatara was edited by MinayefE, 1889 and also in the Journal of the Buddhist Text Society and the Bibliotheca Indica, De la VaU6e Poussin pubUshed parts of the text and commentary in his Bouddhisme and also a translation in 1907. 10 THE MAHAYANA [ch. monks while in Chapter x is sketched the course to be foUowed by "young men or young ladies of good family" who wish to become Bodhisattvas^. The chief difference is that the Bodhi caryavatara portrays a more spiritual Ufe, it speaks more of devotion, less of the milHon shapes that compose the heavenly host: more of love and wisdom, less of the merits of reading particular sutras. While rendering to it and the faith that produced it all honour, we must remember that it is typical of the Mahayana only in the sense that the De Imitatione Christi is typical of Roman CathoUeism, for both faiths have other sides. Santideva's Bodhisattva, when conceiving the thought of Bodhi or eventual supreme enlightenment to be obtained, it may be, only after numberless births, feels first a sympathetic joy in the good actions of aU Hving beings. He addresses to the Buddhas a prayer which is not a mere act of commemoration, but a request to preach the law and to defer their entrance into Nirvana. He then makes over to others whatever merit he may, possess or acquire and offers himself and all his possessions, moral and material, as a sacrifice for the salvation of all beings. This on the one hand does not much exceed the Umits of ddnam or the virtue of giving as practised by Sakyamuni in previous births according to the PaH scriptures, but on the other it contains in embryo the doctrine of vicarious merit and salvation through a saviour. The older tradition admits that the future Buddha {e.g. in the Vessantara birth-story) gives aU that is asked from him including Hfe, wife and children. To consider the surrender and transfer of merit (pattidana in PaH) as parallel is a natural though perhaps false analogy. But the transfer of Karma is not altogether foreign to Brahmanic thought, for it is held that a wife may share in her husband's Karma nor is it whoUy unknown to Sinhalese Buddhism^. After thus deUberately rejecting all personal success and selfish aims, the neophyte makes a vow (pranidhana) to acquire enUghten- ment for the good of all beings and not to swerve from the rules of life and faith requisite for this end. He is then a "son ^ The career of the Bodhisattva is also discussed in detaU in the Avatamsaka sutra and in works attributed to Nagarjuna and Sthiramati, the Lakshana-vimukta- hridaya-^astra and the Mahayana-dharma-dhatvavi^eshata-^astra. I only know of these works as quoted by Teitaro Suzuki. 2 See Childers, PaU Diet. s.v. Patti, Pattianuppadanam and Puniio. xvn] BODHISATTVAS 11 of Buddha," a phrase which is merely a natural metaphor for saying that he is one of the household of faith^ but still paves the way to later ideas which make the celestial Bodhisattva an emanation or spiritual son of a celestial Buddha. Asanga gives ^ a more technical and scholastic description of the ten bhumis or stages which mark the Bodhisattva's progress towards complete enUghtenment and culminate in a phase bearing the remarkable but ancient name of Dharmamegha known also to the Yoga phUosophy. The other stages are caUed: muditd (joyful) : virrtald (immaculate) : prabhdhari (Ught giving) : arcismati (radiant): durjaya (hard to gain): abhimukhi (facing, because it faces both transmigration and Nirvana) : duramgamd (far-going) : acald (immovable) : sddhumati (good minded). The incarnate Buddhas and Bodhisattvas of Tibet are a travesty of the Mahayana which on Indian soil adhered to the sound doctrine that saints are known by their achievements as men and cannot be selected among infant prodigies^. It was the general though not universal opinion that one who had entered on the career of a Bodhisattva could not fall so low as to be reborn in any state of punishment, but the spirit of humiUty and self-effacement which has always marked the Buddhist ideal tended to represent his triumph as incalculably distant. MeanwhUe, although in the whirl of births he was on the upward grade, he yet had his ups and downs and there is no evidence that Indian or Far Eastern Buddhists arrogated to themselves special claims and powers on the ground that they were weU advanced in the career of Buddhahood. The vow to suppress seU and foUow the Hght not only in this Ufe but in all future births contains an element of faith or fantasy, but has any reUgion formed a nobler or even equivalent picture of the soul's destiny or built a better staircase from the world of men to the immeasurable spheres of the superhuman? One aspect of the story of Sakyamuni and his antecedent births thus led to the idea that aU may become Buddhas. An ' It occurs in the PaU Canon, e.g. Itivuttakam 100. Tassa me tumhe putta orasa, mukhato jata, dhammaja. ' See Sylvain L6vi, Mahdydna-S'tUrdlankdra: introduction and passim. Por much additional information about the Bhumis see De la VaU^e Poussin's article "Bodhi sattva" in U.R.E. ' Eminent doctors such as Nagarjuna and Asanga are often described as Bodhi sattvas just as eminent Hindu teachers, e.g. Caitanya, are described as Avataras. 12 THE MAHAYANA [ch equally natural development in another direction created celestial and superhuman Bodhisattvas. The Hinayana held that Gotama, before his last birth, dwelt in the Tushita heaver enjoying the power and splendour of an Indian god and it looked forward to the advent of Maitreya. But it admitted no othei Bodhisattvas, a consequence apparently of the doctrine thai there can only be one Buddha at a time. But the luxuriant fancy of India, which loves to multiply divinities, soon broke through this restriction and fashioned for itself beautiful images of benevolent beings who refuse the bUss of Nirvana that they may aUeviate the sufferings of others^. So far as we can judge, the figures of these Bodhisattvas took shape just about the same time that the personaUties of Vishnu and Siva were acquiring consistency. The impulse in both cases is the same, namely the desire to express in a form accessible to human /'prayer and sympathetic to human emotion the forces which rule the universe. But in this work of portraiture the Buddhists laid more emphasis on moral and spiritual law than did the Brahmans : they isolated in personification quaUties not found isolated in nature. Siva is the law of change, of death and rebirth, with all the riot of slaughter and priapism which it entails : Vishnu is the protector and preserver, the type of good energy warring against evil, but the unity of the figure is smothered by mythology and broken up into various incarna tions. But Avalokita and Manjusri, though they had not such strong roots in Indian humanity as Siva and Vishnu, are genii of purer and brighter presence. They are the personifications of kindness and knowledge. Though manifold in shape, they have Uttle to do with mythology, and are analogous to the archangels of Christian and Jewish tradition and to the Amesha Spentas of Zoroastrianism. With these latter they may have some historical connection, for Persian ideas may well have influenced Buddhism about the time of the Christian era. How ever difficult it may be to prove the foreign origin of Boclhisat- tvas, few of them have a clear origin in Inclia and all of them 1 The idea that Arhats may postpone their entry into Nirvana for the good of the world is not unknown to the Pali Canon. According to the Maha Parin-Sutta the Buddha himself might have done so. Legends which cannot be caUed definitely Mahayanist relate how Pindola and others are to tarry until Maitreya oome and how Kasyapa in a less active r61e awaits him in a cave or tomb, ready to revive at his advent. See J.A. 1916, ii. pp. 196, 270. XVII] BODHISATTVAS 13 are much better known in Central Asia and China. But they are represented with the appearance and attributes of Indian Devas, as is natural, since even in the PaH Canon Devas form the Buddha's retinue. The early Buddhists considered that these spirits, whether caUed Bodhisattvas or Devas, had attained their high position in the same way as Sakyamuni himself, that is by the practice of moral and inteUectual virtues through countless existences, but subsequently they came to be regarded as emanations or sons of superhuman Buddhas. Thus the Karanda-vyuha relates how the original Adi-Buddha produced Avalokita by meditation and how he in his turn produced the universe with its gods. MiUions of unnamed Bodhisattvas are freely mentioned and even in the older books copious lists of names are found^, but two, Avalokita and Manjusri, tower above the rest, among whom only few have a definite personaUty. The tantric school counts eight of the first rank. Maitreya (who does not stand on the same footing as the others), Samantabhadra, Mahasthana- prapta and above aU Kshitigarbha, have some importance, especiaUy in China and Japan. Avalokita^ in many forms and in many ages has been one of the principal deities of Asia but his origin is obscure. His main attributes are plain. He is the personification of divine mercy and pity but even the meaning of his name is doubtful. In its fuU form it is Avalokitesvara, often rendered the Lord who looks down (from heaven). This is an appropriate title for the God of Mercy, but the obvious meaning of the participle avalokita in Sanskrit is passive, the Lord who is looked at. Kem^ thinks it may mean the Lord who is everywhere visible as a very present help in trouble, or else the Lord of View, Uke the epithet Drishtiguru appHed to Siva. Another form of the name is Lokesvara or Lord of the world and this suggests that avalokita may be a synonym of loka, meaning the visible uni verse. It has also been suggested that the name may refer to the smaU image of Amitabha which is set in his diadem and thus looks down on him. But such smaU images set in the head of a larger figure are not distinctive of Avalokita : they are found ' E.g. Lotus, chap. i. * De la VaUee Poussin's article "Avalokita" in E.E,E, may be consulted. 3 Lotus, 8.B.E. XXI. p. 407. 14 THE MAHAYANA [ch. in other Buddhist statues and paintings and also outside India, for instance at Palmyra. The Tibetan translation of the name^ means he who sees with bright eyes. Hsiian Chuang's rendering Kwan-tzii-tsai^ expresses the same idea, but the more usual Chinese translation Kuan-yin or Kuan-shih-yin, the deity who looks upon voices or the region of voices, seems to imply a verbal misunderstanding. For the use of Yin or voice makes us suspect that the translator identified the last part of Avalokite- ivara not with tSvara lord but with svara sound'. Avalokitesvara is unknown to the Pali Canon and the MiUnda Paiiha. So far as I can discover he is not mentioned in the Divyavadana, Jatakamala or any work attributed to Asvaghosha. His name does not occur in the LaUta-vistara but a list of Bodhisattvas in its introductory chapter includes Mahakarunacandin, suggesting Mahakaruna, the Great Com- I passionate, which is one of his epithets. In the Lotus* he is placed second in the introductory Ust of Bodhisattvas after Manjusri. But Chapter xxiv, which is probably a later addition, is dedicated to his praises as Samantamukha, he who looks ( every way or the omnipresent. In this section his character as the all-merciful saviour is fully developed. He saves those who icall on him from shipwreck, and execution, from robbers and 'all violence and distress. He saves too from moral evils, such 1 as passion, hatred and foUy. He grants children to women who j worship him. This power, which is commonly exercised by female deities, is worth remarking as a hint of his subsequent transformation into a goddess. For the better achievement of his merciful deeds, he assumes aU manner of forms, and appears in the guise of a Buddha, a Bodhisattva, a Hindu deity, a gobUn, or a Brahman and in fact in any shape. This chapter was translated into Chinese before 417 a.d. and therefore can hardly be later than 350. He is also mentioned in the Sukhavati-vyuha. 1 sPyan-ras-gzigs rendered in Mongol by Nidubar-udzakci. The other common Mongol name Ariobalo appears to be a corruption of Aryavalokita. 2 Meaning apparently the seeing and self-existent one. Cf. Ta-tzu-tsai as a name of &va. ' A maidservant in the drama Malatimadhava is caUed Avalokita. It is not clear whether it is a feminine form of the divine name or an adjective meaning looked-at, or admirable. ' 8.B.E. XXI. pp. 4 and 406 £E. It was translated in Chinese between a.d. 265 and 316 and chap, xxiv was separately translated between a.d. 384 and 417. See Nanjio, Catalogue Nos. 136, 137, 138. xvn] BODHISATTVAS 15 The records of the Chinese pUgrims Fa-Hsien andHsiian Chuang^ indicate that his worship prevailed in India from the fourth tiU the seventh century and we are perhaps justified in dating its beginnings at least two centuries earHer. But the absence of any mention of it in the writings of Asvaghosha is remarkable 2. Avalokita is connected with a mountain caUed Potala or Potalaka. The name is bome by the palace of the Grand Lama at Lhassa and by another Lamaistic estabUshment at Jehol in north China. It reappears in the sacred island of P'u-t'o near Ningpo. In aU these cases the name of Avalokita's Indian residence has been transferred to foreign shrines. In India there were at least two places caUed Potala or Potalaka — one at the mouth of the Indus and one in the south. No certain connection has been traced between the former and the Bodhisattva but in the seventh century the latter was regarded as his abode. Our information about it comes mainly from Hsiian Chuang ^ who describes it when speaking of the Malakuta country and as near the Mo-lo-ya (Malaya) mountain. But apparently he did not visit it and this makes it probable that it was not a rehgious centre but a mountain in the south of which Buddhists in the north wrote with Httle precision*. There is no evidence that Avalokita was first worshipped on this Potalaka, though he is often associated with mountains such as Kapota in Magadha and Valavati in Kataha ^. In fact the connection of Potala with Avalokita remains a mystery. Avalokita has, like most Bodhisattvas, many names. Among the principal are Mahakaruna, the Great Compassionate one, Lokanatha or Lokesvara, the Lord of the world, and Padmapani, or lotus-handed. This last refers to his appearance as portrayed in statues and miniatures. In the older works of art his figure ' Hsuan Chuang (Watters, n. 215, 224) relates how an Indian sage recited the Sui-hsin dharani before Kuan-tzu-tsai's image for three years. ^ As wiU be noticed from time to time in these pages, the sudden appearance of new deities in Indian hterature often seems strange. The fact is that until deities are generaUy recognized, standard works pay uo attention to them. ' Watters, vol. ii. pp. 228 ff. It is said that Potalaka is also mentioned in the Hwa-yen-ohing or Avatamsaka sutra. Tibetan tradition connects it with the Sakya family. See Csoma de Koros, Tibetan studies reprinted 1912, pp. 32-34. * Just as the Lankavatara sutra purports to have been deUvered at Laniapura- minudra-rnalaya-kikhara rendered in the Chinese translation as "in the city of Lanka on the summit of the Malaya mountain on the border of the sea." ' See Fouoher, Iconographie bouddhique, 1900, pp. 100, 102. 16 THE MAHAYANA [cs. is human, without redundant Umbs, and represents a youth in the costume of an Indian prince with a high jeweUed chignon, or 'sometimes a crown. The head-dress is usuaUy surmounted by a smaU figure of Amitabha. His right hand is extended in the position known as the gesture of charity^. In his left he carries a red lotus and he often stands on a larger blossom. His complexion is white or red. Sometimes he has four arms and in later images a great number. He then carries besides the lotus such objects as a book, a rosary and a jug of nectar^. ' The images with many eyes and arms seem an attempt to represent him as looking after the unhappy in all quarters aud stretching out his hands in help ^, It is doubtful if the Bodhisat tvas of the Gandhara sculptures, though approaching the type of Avalokita, represent him rather than any other, but nearly all the Buddhist sites of India contain representations of him which date from the early centuries of our era* and others are preserved in the miniatures of manuscripts^. He is not a mere adaptation of any one Hindu god. Some of his attributes are also those of Brahma. Though in some late texts he is said to have evolved the world from himseU, his characteristic function is not to create but, Hke Vishnu, to save and Hke Vishnu he holds a lotus. But also he has the title of Isvara, which is speciaUy applied to Siva. Thus he does not issue from any local cult and has no single mythological pedigree but is the idea of divine compassion represented with such materials as the art and mythology of the day offered. He is often accompanied by a female figure Tara®. In the tantric period she is recognized as his spouse and her images, common in northern India from the seventh century onwards, 1 Varamudra. ^ These as weU as the red colour are attributes of the Hindu deity Brahma. ' A temple on the north side of the lake in the Imperial City at Peking contains a gigantic image of him which has literaUy a thousand heads and a thousand hands, This monstrous figure is a warning against an attempt to represent metaphors literaUy. < WaddeU on the Cult of Avalokita, J,R,A,S, 1894, pp. 51 ff. thinks they are not earlier than the fifth century, ^ See especially Poucher, Iconographie Bouddhique, Paris, 1900. » See especiaUy de Blonay, Etudes pour servir a I'histoire de la ddesse bmiddhiqui Tdrd, Paris, 1895. Tara continued to be worshipped as a Hindu goddess after Buddhism had disappeared and several works were written in her honour. See Raj. Mitra, Search for Sk, MSS, iv. 168, 171, x. 67. xvn] BODHISATTVAS 17 show that she was adored as a female Bodhisattva. In Tibet T4r§, is an important deity who assumes many forms and even before the tantric influence had become prominent she seems to have been associated with Avalokita. In the Dharma- sangraha she is named as one of the four Devis, and she is mentioned twice under the name of To-lo Pu-sa by Hsiian Chuang, who saw a statue of her in Vaisali and another at Tiladhaka in Magadha. This last stood on the right of a gigantic figure of Buddha, Avalokita being on his left^. ¦ \ Hsiian Chuang distinguishes To-lo (Tara) and Kuan-tzii-tsai. The latter under the name of Kuan-yin or Kwannon has become the most popular goddess of China and Japan, but is apparently a form of Avalokita. The god in his desire to help mankind assumes many shapes and, among these, divine womanhood has by the suffrage of mUHons been judged the most appropriate. But Tara was not originally the same as Kuan-yin, though the fact that she accompanies Avalokita and shares his attributes imay have made it easier to think of him in female form^. 1 The circumstances in.-which Avalokita became a goddess are obscure. The Indian images of him are not feminine, although ihis sex is hardly noticed before the tantric period. He is not a i male deity like Krishna, but a strong, bright spirit and like the Christian archangels above sexual distinctions. No female form i of him is reported from Tibet and this confirms the idea that inone was known in India ^, and that the change was made in China. It was probably facilitated by the worship of Tara and lof Hariti, an ogress who was converted by the Buddha and is (frequently represented in her regenerate state caressing a child. i 1 About the time of Hsiian Chuang's travels Sarvajnamitra wrote a hymn to Tara which has been preserved and pubUshed by de Blonay, 1894. j 2 Chinese Buddhists say Tara and Kuan-Yin are the same but the difference jbetween them is this. Tara is an Indian and Lamaist goddess associated with ^ivalokita and in origin analogous to the Saktis of Tantrism. Kuan-yin is a female jform of Avalokita who can assume aU shapes. The original Kuan-yin was a male deity : male Kuan-yins are not unknown in China and are said to be the rule in (Korea. But Tara and Kuan-yin may justly be described as the same in so far as they are attempts to embody the idea of divine pity in a Madoima. ' But many scholars think that the formula Om manipadme hum, which is Jsupposed to be addressed to Avalokita, is reaUy an invocation to a form of Sakti iPaUed Manipadma. A Nepalese inscription says that "The Saktas call him Sakti" ,[E,B,E, vol. n. p. 260 and J,A, ix. 192), but this may be merely a way of saying that he is identical with the great gods of aU sects. E. n. 2 18 THE MAHAYANA [cH, She is mentioned by Hsiian Chuang and by I-Ching who adds that her image was aUeady known in China. The Chinese also worshipped a native goddess called T'ien-hou or T'ou-mu. Kuan- yin was also identified with an ancient Chinese heroine called; Miao-sheni. This is paraUel to the legend of Ti-tsang (Kshiti garbha) who, though a male Bodhisattva, was a virtuous! maiden in two of his previous existences. Evidently Chinese religious sentiment required a Madonna and it is not unnatur^ : if the god of mercy, who was reputed to assume many shapes j and to give sons to the childless, came to be thought of chiefly j in a feminine form. The artists of the T'ang dynasty usua]| represented Avalokita as a youth with a sUght moustache ani the evidence as to early female figures does not seem to mf strong 2, though a priori I see no reason for doubting their exis. tence. In 1102 a Chinese monk named P'u-ming pubHshedi romantic legend of Kuan-yin's earthly Hfe which helped to popularize her worship. In this and many other cases the latei developments of Buddhism are due to Chinese fancy and have no connection with Indian tradition. Tara is a goddess of north India, Nepal and the Lamaist Church and almost unknown in China and Japan. Her name means she who causes to cross, that is who saves, Hfe and ita troubles being by a common metaphor described as a sea. Tara also means a star and in Puranic mythology is the name given to the mother of Buddha, the planet Mercury. Whether the name was first used by Buddhists or Brahmans is unkno-wns but after the seventh century there was a decided tendency to i give TarS, the epithets bestowed on the Saktis of Siva and ! assimUate her to those goddesses. Thus in the Ust of her 108 I names ^ she is described among other more amiable attributes as ^ Harlez, Lime des esprits et des immortels, p. 195, and Dore, Becherches sur iei superstitions en Chine, pp. 94-138. 2 See FenoUosa, Epochs of Chinese and Japanese Art, i. pp. 105 and 124 ; Johnston, Buddhist China, 275 ff. Several Chinese deities appear to be of uncertain or varyinj sex. Thus Chun-ti is sometimes described as a deified Chinese General and sometime! identified with the Indian goddess Marioi. Yii-ti, generaUy masculine, is sometimei feminine. See Dor6, l.c, 212. StUl more strangely the Patriarch Asvaghosha (Ma Ming) is represented by a female figure. On the other hand the monk Ta ShetJ (c. 705 A.D.) is said to have been an incarnation of the female Kuan Yin. Mafijuin is said to be worshipped in Nepal sometimes as a male, sometimes as a female. See BendaU and Haraprasad, Nepalese MSS, p. Ixvii. ' de Blonay, l,c, pp. 48-57. xvn] BODHISATTVAS 19 terrible, furious, the slayer of evil beings, the destroyer, and KaH: also as carrying skulls and being the mother of the Vedas. Here we have if not the borrowing by Buddhists of a Saiva deity, at least the grafting of Saiva conceptions on a Bodhisattva. The second great Bodhisattva Maiijusri^ has other similar names, such as Manjunatha and Manjughosha, the word Manju meaning sweet or pleasant. He is also Vagisvara, the Lord of Speech, and Kumarabhiita, the Prince, which possibly impHes that he is the Buddha's eldest son, charged with the government under his direction. He has much the same literary history as Avalokita, not being mentioned in the PaH Canon nor in the earlier Sanskrit works such as the LaUta-vistara and Divyava- 'dana. But his name occurs in the Sukhavati-vyuha: he is the 'principal interlocutor in the Lankavatara sutra and is extolled in the Ratnakaranda-vyuha-sutra^. In the greater part of the 'Lotus he is the principal Bodhisattva and instructs Maitreya, 'because, though his youth is etemal, he has known many 'Buddhas through innumerable ages. The Lotus ^ also recounts how he visited the depths of the sea and converted the inhabi- -tants thereof and how the Lord taught him what are the duties ^of a Bodhisattva after the Buddha has entered finally into 'Nirvana. As a rule he has no consort and appears as a male Athene, aU inteUect and chastity, but sometimes Lakshmi or (Sarasvati or both are described as his consorts*. t His worship prevailed not only in India but in Nepal, Tibet, JChina, Japan and Java. Fa-Hsien states that he was honoured tin Central India, and Hsiian Chuang that there were stupas idedicated to him at Muttra^. He is also said to have been elncarnate in Ati^a, the Tibetan reformer, and in Vairocana who introduced Buddhism to Khotan, but, great as is his benevo lence, he is not so much the helper of human beings, which is Avalokita's special function, as the personification of thought, I.; [6 * Chinese, Man-chu-shih-U, or Wen-shu; Japanese, Monju; Tibetan, hjam-pahi- itibyans (pronounced Jam-yang). # Manju is good Sanskrit, but it must be confessed that the name has a Central- (jiUian ring. Iii) ^ Translated into Chinese 270 a.d. ' Chaps. XI and xill. $ * A special wojrk MafijuSrivikridita (Nanjio, 184, 185) translated into Chinese ilsil3 A.D. is quoted as describing ManjuM's transformations and exploits. " Hsuan Chuang also relates how he assisted a phUosopher caUed Ch'en-na = Dinnaga) and bade him study Mahayanist books. 2—2 20 THE MAHAYANA [c: knowledge, and meditation. It is for this that he has in l hands the sword of knowledge and a book. A beautiful figu from Java bearing these emblems is in the Berlin Museum Miniatures represent him as of a yeUow colour with the han( (when they do not carry emblems) set in the position known i , teaching the law^ Other signs which distinguish his images ai ! the blue lotus and the lion on which he sits. An interesting fact about Manjusri is his association wit China 3, not only in Chinese but in late Indian legends. Tl mountain Wu-t'ai-shan in the province of Shan-si is sacred t him and is covered with temples erected in his honour*. Th name (mountain of five terraces) is rendered in Sanskrit a Pancasirsha, or Pancasikha, and occurs both in the Svayamblil Purana and in the text appended to miniatures representiij Maiijusri. The principal temple is said to have been erectet between 471 and 500 a.d. I have not seen any statement tha the locaHty was sacred in pre-Buddhist times, but it wai probably regarded as the haunt of deities, one of whom- perhaps some spirit of divination — was identified with the wise Maiijusri. It is possible that during the various inroads oi Grseco-Bactrians, Yiieh-Chih, and other Central Asian tribej into India, Maiijusri was somehow imported into the pantheon of the Mahayana from China or Central Asia, and he has, especiaUy in the earHer descriptions, a certain pure and abstract quality which recaUs the Amesha-Spentas of Persia. But still his attributes are Indian, and there is Uttle positive evidence oi a foreign origin. I-Ching is the first to teU us that the Hindus beUeved he came from China ^. Hsiian Chuang does not mention this beUef , and probably did not hear of it, for it is an interesting detail which no one writing for a Chinese audience would have omitted. We may therefore suppose that the idea arose in India about 650 a.d. By that date the temples of Wu-t'ai-Shan would ' It is reproduced in Griinwedel's Buddhist Art in India, Translated by Gibson, 1901, p. 200. ^ Dharmacakramudra. ' For the Nepalese legends see S. Levi, Le Nepal, 1905-9. • For an account of this sacred mountain see Edkins, Religion in China, chap. XVTI to XIX. ^ See I-tsing, trans. Takakusu, 1896, p. 136. Por some further remarks on tt* possible foreign origin of Mafiju&i see below, chapter on Central Asia. The verm attributed to King Harsha (Nanjio, 1071) praise the reUquaries of China but without details. j xvn] BODHISATTVAS 21 have had time to become celebrated, and the visits paid to India by distinguished Chinese Buddhists would be Hkely to create the impression that China was a centre of the faith and frequented by Bodhisattvas^. We hear that Vajrabodhi (about 700) and Prajiia (782) both went to China to adore Manjusri. In 824 a Tibetan envoy arrived at the Chinese Court to ask for an image of Manjusri, and later the Grand Lamas officially recognized that he was incarnate in the Emperor^. Another legend relates that Maiijusri came from Wu-t'ai-Shan to adore a miraculous lotus * that appeared on the lake which then fiUed Nepal. With a blow of his sword he cleft the mountain barrier and thus drained the vaUey and introduced civilization. There may be hidden in this some tradition of the introduction of culture into Nepal but the Nepalese legends are late and in their I coUected form do not go back beyond the sixteenth century. After Avalokita and Manjusri the most important Bodhisat- , tva is Maitreya*, also caUed Ajita or unconquered, who is the , only one recognized by the PaU Canon^. This is because he does I not stand on the same footing as the others. They are super- . human in their origin as weU as in their career, whereas Maitreya , is simply a being who Hke Gotama has lived innumerable Hves and ultimately made himself worthy of Buddhahood which he awaits in heaven. There is no reason to doubt that Gotama , regarded himseU as one in a series of Buddhas : the PaH scriptures relate that he mentioned his predecessors by name, and also spoke of unnumbered Buddhas to come®. Nevertheless '. Maitreya or Metteyya is rarely mentioned in the Pali Canon''. 1 Some of the Tantras, e.g, the Mahacinakramacara, though they do not connect ; MaiijuW with China, represent some of their most surprising novelties as having t been brought thence by ancient sages Uke Vasishtha. « J.B.A.S. new series, xn. 522 and J.A.S.B. 1882, p. 41. The name Manchu S perhaps contributed to this beUef. j- ' It is described as a Svayambhu or spontaneous manifestation of the Adi- Buddha. •* * Sanskrit, Maitreya; PaU, Metteyya; Chinese, Mi-U; Japanese, Miroku; Mongol, Maidari; Tibetan, Byams-pa (pronounced Jampa). For the history of the Maitreya idea see especially Peri, B.E.F.E.O. 1911, pp. 439-457. » But a Siamese inscription of about 1361, possibly influenced by Chinese i'Mahayauism, speaks of the ten Bodhisattvas headed by Metteyya. See B.E.F.E.O. 1917, No. 2, pp. 30, 31. '' • E.g. in the Mahaparinibbana Sutra. «' ' Dig. Nik. XXVI. 25 and Buddhavamsa, xxvii. 19, and even this last verse is %aid to be an addition. 22 THE MAHAYANA [ch. He is, however, frequently aUuded to in the exegetical Pali Hterature, in the Anagata-vamsa and in the earHer Sanskrit works such as the LaHta-vistara, the Divyavadana and Maha- : vastu. In the Lotus he plays a prominent part, but stiU is subordinate to Manjusri. Ultimately he was ecHpsed by the two great Bodhisattvas but in the early centuries of our era he received much respect. His images are frequent in aU parts of the Buddhist world : he was beUeved to watch over the propaga tion of the Faith^, and to have made special revelations to AsaAga^. In paintings he is usuaUy of a golden colour: his statues, which are often gigantic, show him standing or sitting in the European fashion and not cross-legged. He appears to be represented in the earUest Gandharan sculptures and there was a famous image of him in Udyana of which Fa-Hsien (399-414 A.D.) speaks as if it were already ancient*. Hsiian Chuang describes it as weU as a stupa erected* to commemorate Sakyamuni's prediction that Maitreya would be his successor, On attaining Buddhahood he wiU become lord of a terrestrial paradise and hold three assembUes under a dragon fiower tree^, at which aU who have been good Buddhists in previous births wiU become Arhats. I-Ching speaks of meditating on the advent ! of Maitreya in language Uke that which Christian piety uses of the second coming of Christ and concludes a poem which ip incorporated in his work with the aspiration "Deep as the deptl of a lake be my pure and calm meditation. Let me look for the first meeting under the Tree of the Dragon Flower when I heai the deep rippling voice of the Buddha Maitreya*." But messianic ^ See e.g. Watters, Yuan Chwang, i. 239. 2 See Watters and Peri in B.E.F.E.O. 1911, 439. A teraple of Maitreya hai been found at Turfan in Central Asia with a Chinese inscription which speaks ol him as an active and benevolent deity manifesting himself in many forms. ' He has not fared weU in Chinese iconography which represents him as at enormously fat smiUng monk. In the Liang dynasty there was a monk called Pu-ta (Jap, Hotel) who was regarded as an incarnation of Maitreya and became a popula) subject for caricature. It would appear that the Bodhisattva himself has becomf superseded by this cheerful but undignified incarnation. * The stupa was apparently at Benares but Hsiian Chuang's narrative is noi clear and other versions make Rajagriha or Sravasti the scene of the prediction. ' Campa. This is his bodhi tree under which he wiU obtain enUghtenment ai Sakyamuni under the Ficus religiosa. Each Buddha has his own special kind ol bodhi tree. ' Record of the Buddhist religion, Trans. Takakusu, p. 213. See too AVatteU, Yuan Chwang, li. 57, 144, 210, 215. XVII] BODHISATTVAS 23 ideas were not much developed in either Buddhism or Hinduism and perhaps the figures of both Maitreya and Kalki owe some thing to Persian legends about Saoshyant the Saviour. The other Bodhisattvas, though lauded in special treatises, have left Httle impression on Indian Buddhism and have obtained in the Far East most of whatever importance they possess. The makers of images and miniatures assign to each his , proper shape and colour, but when we read about them we feel that we are deaUng not with the objects of real worship or even the products of a lively imagination, but with names and figures which have a value for picturesque but conventional art. Among the best known is Samantabhadra, the all gracious i, who is stiU a popular deity in Tibet and the patron saint of the sacred mountain Omei in China, with which he is associated as Manjusri with Wu-t'ai-shan. He is represented as green and riding on an elephant. In Indian Buddhism he has a moderately prominent position. He is mentioned in the Dharmasangraha and in one chapter of the Lotus he is charged with the special duty of protecting those who foUow the law. But the Chinese , pUgrims do not mention his worship. Mahasthamaprapta^ is a somewhat similar figure. A chapter of the Lotus (xix) is dedicated to him without however giving any clear idea of his personaUty and he is extoUed in several descriptions of Sukhavati or Paradise, especially in the Ami- tayurdhyana-sutra. Together with Amitabha and Avalokita he forms a triad who rule this Happy Land and are often repre sented by three images in Chinese temples. Vajrapani is mentioned in many Usts of Bodhisattvas [e.g. in the Dharmasangraha) but is of somewhat doubtful position as Hsiian Chuang caUs him a deva*. HistoricaUy his recognition as a Bodhisattva is interesting for he is merely Indra trans formed into a Buddhist. The mysterious personages caUed Vajradhara and Vajrasattva, who in later times are even ' Chinese P'u-hsien. See Johnston, From Peking to Mandalay, for an interesting account of Mt. Omei. * Or Mahasthana. Chinese, Tai-shih-ohih. He appears to be the Arhat Maud- galyayana deified. In China aud Japan there is a marked tendency to regard aU Bodhisattvas as ancient worthies who by their vows and virtues have risen to their present high position. But these euhemeristic explanations are common in the Far East and the real origin of the Bodhisattvas may be quite different. 3 E.g. Watters, i. p. 229, n. 215. 24 THE MAHAYANA [ch. identified with the original Buddha spirit, are further develop ments of Vajrapani. He owes his elevation to the fact that Vajra, originally meaning simply thunderbolt, came to be used as a mystical expression for the highest truth. More important than these is Kshitigarbha, Ti-tsang or Jizo^ who in China and Japan ranks second only to Kuan-yin. Visser has consecrated to him an interesting monograph ^ which shows what strange changes and chances may attend spirits and how ideal figures may alter as century after century they travel from land to land. We know Httle about the origin of Kshitigarbha. The name seems to mean Earth-womb and he has a shadowy counterpart in Akasagarbha, a simUar deity of the air, who it seems never had a hold on human hearts. The Earth is generaUy personified as a goddess* and Kshitigarbha has some sHght feminine traits, though on the whole decidedly masculine. The stories of his previous births relate how he was twice a woman : in Japan he was identified with the mountain goddess of Kamado, and he helps women in labour, a boon generaUy accorded by goddesses. In the pantheon of India he played an inconspicuous part*, though reckoned one of the eight great Bodhisattvas, but met with more general esteem in Turkestan, where he began to coUect the attributes afterwards defined in the Far East. It is there that his history and trans formations become clear. He is primarily a deity of the nether world, but Hke Amitabha and Avalokita he made a vow to help all Hving creatures and specially to deUver them from heU. The Taoists pictured hell as divided into ten departments ruled over by as many kings, and Chinese fancy made Ti-tsang the superintendent of these functionaries. He thus becomes not so much a Saviour as the kindly superintendent of a prison who preaches to the inmates and wilUngly procures their release. Then we hear of six Ti- tsangs, corresponding to the six worlds of sentient beings, the gracious spirit being supposed to multiply his personaUty in ' Kshitigarbha is translated into Chinese as Ti-tsang and Jizo is the Japanese pronunciation of the same two characters. 2 In Ostasiat. Ztsft. 1913-15. See too Johnston, Buddhist China, chap. vm. » The Earth goddess is known to the earUest Buddhist legends. The Buddha caUed her to witness when sitting under the Bo tree. * Three Sutras, analysed by Visser, treat of Kshitigarbha. They are Nanjio, Nos. 64, 65, 67. xvn] BODHISATTVAS 25 order to minister to the wants of aU. He is often represented as a monk, sta£f in hand and with shaven head. The origin of this guise is not clear and it perhaps refers to his previous births. But in the eighth century a monk of Chiu Hua^ was regarded as an incarnation of Ti-tsang and after death his body was gilded and enshrined as an object of worship. In later times the Bodhisattva was confused with the incarnation, in the same way as the portly figure of Pu-tai, commonly known as the laughing Buddha, has been substituted for Maitreya in Chinese iconography. In Japan the cult of the six Jizos became very popular. They were regarded as the deities of roads ^ and their effigies ultimately superseded the ancient phalUc gods of the crossways. In this martial country the Bodhisattva assumed yet another character as Shogun Jizo, a militant priest riding on horseback^ and wearing a helmet who became the patron saint of warriors and was even identified with the Japanese war god, Hachiman. Until the seventeenth century Jizo was worshipped principally by soldiers and priests, but subsequently his cult spread among aU classes and in aU districts. His benevolent activities as a guide and saviour were more and more emphasized: he heals sickness, he lengthens Ufe, he leads to heaven, he saves from hell: he even suffers as a substitute in hell and is the special protector of the souls of children amid the perils of the under world. Though this modern figure of Jizo is wrought with ancient materials, it is in the main a work of Japanese senti ment. ^ A celebrated monastery in the portion of An-hui which lies to the south of tho Yang-tse. See Johnston, Buddhist China, chaps, vni, ix and x. ' There is some reason to think that even in Turkestan Kshitigarbha was a god of roads. ' In Annam too Jizo is represented on horseback. CHAPTER XVIII THE BUDDHAS OF MAHAYANISM This mythology did not grow up around the Buddha without affecting the central figure. To understand the extraordinary changes of meaning both mythological and metaphysical which the word Buddha undergoes in Mahayanist theology we must keep in mind not the personaUty of Gotama but the idea that he is one of several successive Buddhas who for convenience may be counted as four, seven or twenty-four but who really form an infinite series extending without Hmit backwards into the past and forwards into the future i. This beUef in a series of Buddhas produced a plentiful crop of imaginary personaUties and also of speculations as to their connection with one another, with the phenomena of the world and with the human soul. In the PaU Canon the Buddhas antecedent to Gotama are introduced much Uke ancient kings as part of the legendary history of this world. But in the LaHta-vistara (Chap, xx) and the Lotus (Chap, vii) we hear of Buddhas, usuaUy described as Tathagatas, who apparently do not belong to this world at all, but rule various points of the compass, or regions described as Buddha-fields (Buddha-kshetra). Their names are not the same in the different accounts and we remain dazzled by an endless panorama of an infinity of universes with an infinity of shining Buddhas, illuminating infinite space. Somewhat later five of these unearthly Buddhas were formed into a pentad and described as Jinas^ or Dhyani Buddhas (Buddhas of contemplation), namely, Vairocana, Akshobhya, Ratnasambhava, Amit§,bha and Amoghasiddhi. In the fuUy developed form of this doctrine these five personages are ^ In Mahaparinib. Sut. i. 16 the Buddha is made to speak of aU the other Buddhas who have been in the long ages of the past and will be in the long ages of the future. " Though Dhyani Buddha is the title most frequently used in European works it would appear that Jina is more usual in Sanskrit works, and in fact Dhyani Buddha is hardly known outside Nepalese literature. Ratnasambhava and Amo ghasiddhi are rarely mentioned apart from the others. According to Getty (Oods oj Northern Buddhism, pp. 26, 27) a group of six, including the Idi-Buddha himself under the name of Vajrasattva, is sometimes worshipped. CH. XVIII] THE BUDDHAS OF MAHAYANISM 27 produced by contemplation from the Adi-Buddha or original Buddha spirit and themselves produce various reflexes, including Bodhisattvas, human Buddhas and goddesses like TarS,. The date when these beUefs first became part of the accepted Mahayana creed cannot be fixed but probably the symmetrical arrangement of five Buddhas is not anterior to the tantric period^ of Buddhism. The most important of the five are Vairocana and Amitabha. ; Akshobhya is mentioned in both the Lotus and SmaUer Sukha vati-vyuha as the chief Buddha of the eastern quarter, and a work purporting to be a description of his paradise stiU extant in Chinese^ is said to have been translated in the time of the Eastern Han dynasty. But even in the Far East he did not find many worshippers. More enduring has been the glory of Vairocana who is the chief deity of the Shingon sect in Japan and is represented by the gigantic image in the temple at Nara. In Java he seems to have been regarded as the principal and supreme Buddha. The name occurs in the Mahavastu as the designation of an otherwise unknown Buddha of luminous attributes and in the Lotus we hear of a distant Buddha-world caUed Vairocana-rasmi-prattmandita, embelHshed by the rays of the sun^. Vairocana is clearly a derivative of Virocana, a recognized title of the sun in Sanskrit, and is rendered in Chinese by Ta-jih meaning great Sun. How this solar deity first came to be regarded as a Buddha is not known but the connection between a Buddha and Ught has always been recognized. Even the PaH texts represent Gotama as being luminous on some occasions and in the Mahayanist scriptures Buddhas are radiant and Ught-giving beings, surrounded by halos of prodigious extent and emitting flashes which illuminate the depths of space. The visions of innumerable paradises in aU quarters containing jeweUed stupas and lighted by refulgent Buddhas which are frequent in these works seem founded on astronomy vaporized under the influence of the idea that there are milHons of universes all equaUy transitory and imsubstantial. There is no reason, so ' About the same period Siva and Vishnu were worshipped in five forms. See below. Book v. chap. in. sec. 3 ad fin. 2 Nanjio, Cat. No. 28. ' Virocana also occurs in the Chandogya Up. vm. 7 and 8 as the name of an Asura who misunderstood the teaching of Prajapati. Verooana is the name of an Aeura in Sam. Nik. i. xi. 1. 8. 28 THE MAHAYANA [CH. far as I see, to regard Gotama as a mythical solar hero, but the celestial Buddhas ^ clearly have many solar attributes. This is natural. Solar deities are so abundant tn Vedic mythology that it is hardly possible to be a benevolent god without having something of the character of the sun. The stream Of foreign reUgions which flowed into India from Bactria and Persia about the time of the Christian era brought new aspects of sun worship such as Mithra, HeUos and ApoUo and strengthened the tendency to connect divinity and Ught. And this connection was pecuHarly appropriate and obvious in the case of a Buddha, for Buddhas are clearly revealers and light-givers, conquerors of darkness and dispeUers of ignorance. Amitabha (or the Buddha of measurelessH^it^, rising suddenly from an obscure origin, has like Avalokita and Vishnu becoine one of the great gods of Asia. He is also known as Amitayiis or measureless Hfe, and is therefore a god of Hght and immortahty". According to both the Lotus and the SmaUer Sukhavati-vyiiha he is the lord of the western quarter but he is unknown to the LaHta-vistara. It gives the ruler of the west a lengthy title^, which suggests a land of gardens. Now Paradise, which has bibUcal authority as a name for the place of departed spirits, , appears to mean in Persian a park or enclosed garden and the Avesta speaks of four heavens, the good thought Paradise, the goodr^dfdT'aradise, the good deed Paradise and the^JEndless Lights^. This last expression bears a remarkable resemblance I to the name of Amitabhaand we can understand that heshould rule the west, because it_^ the home to which the sun and deputed ,spints_go. Amitabha's Paradise is caUed Sukhavati or Happy Land. In the Puranas the city of Varuna (who is suspected of having a non-Indian origin) is said to be situated in the west and is caUed Sukha (Linga P. and Vayu P.) or Mukhya (so Vishnu P. and others). The name AmitS,bha also occurs in the Vishnu Purana as the name of a class of gods and it is curious that they are in one place* associated with other ^ The names of many of these Buddhas, perhaps the majority, contain some word expressive of Ught suoh as Aditya, prabha or tejas. * Chap. XX. Pushpavalivanarajikusumitabhijna. » E.g. Yashts. xxn. and xxrv. S.B.E. vol. xxni. pp. 317 and 344. The title Pure Land (Chinese Ch'ing-t'u, Japanese Jo-do) has also a Persian ring about it, See further in the chapter on Central Asia. * Vishnu P., Book in. chap. n. xvni] THE BUDDHAS OF MAHAYANISM 29 deities caUed the Mukhyas. The worship of Amitabha, so far as its history can be traced, goes back to Saraha, the teacher of Nagarjuna. He is said to have been a Sudra and his name seems un-Indian. This supports the theory that this worship was foreign and imported into India^. This worship and the doctrine on which it is based are an almost complete contradiction of Gotama's teaching, for they amount to this, that rehgion consists in faith in Amitabha and prayer to him, in return for which he wiU receive his foUowers after death in his paradise. Yet this is not a late travesty of Buddhism but a relatively early development which must have begun about the Christian era. The principal works in which it is preached are the Greater Sukhavati-vyuha or Description of the Happy Land, tranalaited-inta Chinese betweeirT47"and 186 A.D., the lesser work of the same name translated in 402 a.d. and the' Sutra of meditation on Amitayus^ translated in 424. The first of these works purports to be a discourse of Sakyamuni himself, deUvered on the Vulture's Peak in answer to the questions of Ananda. He relates how innumerable ages ago there was a monk called Dharmakara who, with the help of the Buddha of that period, made a vow or vows^ to become a Buddha but on_conditiqns7 That_^s to say he rejected the Bu33hahood"to which^^he might become entitled unless Uis inerits obtained certain advantages for others, and_ having obtained Buddhahood on these conditions he can now cause them tb"be~fliffiUed. In other words he can apportion his.yast sfore of accumulated merit to such jiera^^and^in such manner as he chooses. The gist of the conditions is that he should when he obtained^BuddKahdod be lord of a paradise whose inhabitants Uve In "linbroken Happiness uhtil they obtain Nirvana. All who have thought of this paradise ten times are to be admitted therein, unless they have committed grievous sin, and Amitabha will appear to them at the moment of death so that their thoughts may not be troubled. The Buddha shows Ananda a * See below : Section on Central Asia, and Griinwedel, Mythologie, 31, 36 and notes : Taranatha (Shiefner), p. 93 and notes. ^ Amitayur-dhyana-sutra. AU three works are translated in S.B.E, vol. xlix. ' Pranidhana. Not only Amitabha butaU Bodhisattvas (especially Avalokita and Kshitigarbha) are supposed to have made such vows. This idea is very common in China and Japan but goes back to Indian sources. See e.g. Lotus, xxiv. verse 3. 30 THE MAHAYANA [ca miraculous vision of this paradise and its joys are described u language recalling the account of the New Jerusalem in th( book of Revelation and, though coarser pleasures are excluded! aU the deUghts of the eye and ear, such as jewels, gardens, flowers, rivers and the songs of birds await the faithful. The smaUer Sukhavati-vyuha, represented as preached by Sakyamuni at Sravasti, is occupied almost entirely with a description of the paradise. It marks a new departure in definitely preaching salvation by faith only, not by works, whereas the previous treatise, though dweUing on the efficacy of faith, also makes merit a requisite for Hfe in heaven. But the shorter discourse says dogmaticaUy "Beings are not bom in that Buddha country as a reward and result of good works performed in this present life. No, aU men or women who hear and bear in mind for one, two, three, four, five, six or seven nights the name of Amitayus, when they come to die, Amitayus wiU stand before them in the hour of death, they wiU depart this Hfe with quiet minds and after death they wiU be born in Paradise." The Amit^yur-dhyana-sutra also purports to be the teaching of Sakyamuni and has an historical introduction connecting it with Queen Vaidehi and King Bimbisara. In theology it is more advanced than the other treatises : it is famiUar with the doctrine of Dharma-kaya (which wiU be discussed below) and it represents'' the rulers of paradise as a triad, Amitayus being assisted by Avalokita and Mahasthamaprapta^. Admission to the paradise can be obtained in various ways, but the method recommended is the practice of a series of meditations which are described in detail. The system is comprehensive, for salvation can be ob tained by mere virtue with Httle or no prayer but also by a single invocation of Amitayus, which suffices to free from deadly sins. Strange as such doctrines appear when set beside the Pali texts, it is clear that in their origin and even in the form which they assume in the larger Sukhavati-vyiiha they are simply an exaggeration of ordinary Mahayanist teaching 2. Amit§,bha is ^ These Bodhisattvas are also mentioned but without much emphasis in the Greater Sukhavati-vyuha. ^ Even in Hinayanist works suoh as the Nidanakatha Sumedha's resolution to become a Buddha, formed as he Ues on the ground before Dipankara, has a resem blance to Amida's vow. He resolves to attain the truth, to enable mankind to cross the sea of the world and only then to attain Nirvana. XVIII] THE BUDDHAS OF MAHAYANISM 31 merely a monk who devotes himself to the reUgious Ufe, namely seeking bodhi for the good of others. He differs from every day devotees only in the degree of sanctity and success obtained by his exertions. The operations which he performs are nothing but examples on a stupendous scale of parinamana or the assignment of one's own merits to others. His paradise, though in popular esteem equivalent to the Persian or Christian heaven, is not really so : strictly speaking it is not an ultimate ideal but a blessed region in which Nirvana may be obtained without toil ofcafe^ "" " -.— - - Though this teaching had brilHant success in China and Japan, where it still flourishes, the worship of Amitabha was never predominant in India. In Nepal and Tibet he is one among many deities : the Chinese pilgrims hardly mention him : his figure is not particularly frequent in Indian iconography^ and, except in the works composed speciaUy in his honour, he appears as an incidental rather than as a necessary figure. The whole doctrine is hardly strenuous enough for Indians. To pray to the Buddha at the end of a sinful Hfe, enter his paradise and obtain ultimate Nirvana in comfort is not only open to the same charge of egoism as the Hinayana scheme of salvation but is much easier and may lead to the abandonment of reUgious effort. And the Hindu, who above aU things Ukes to busy himself with his own salvation, does not take kindly to these expedients. Numerous deities promise a long speU of heaven as a reward for the mere utterance of their names ^, yet the beUever continues to labour earnestly in ceremonies or meditation. It would be interesting to know whether this doctrine of salvation by the utterance of a single name or prayer originated among Buddhists or Brahmans. In any case it is closely related to old ideas about the^ mamcjpower ojryedic, veraesl The five Jinas and other supernatural personages are often regarded as manifestations of a single Buddha-force and at last this force is personified as Adi-Buddha^. This admittedly 1 ' See Fouoher, Iconographie Bouddhique dans I'Inde. ^ The Bhagavad-glta states quite clearly the doctrine of the death-bed prayer (vm. ad init.). "He who leaves this body and departs remembering me in his last moments comes to my essence. Whatever form (of deity) he remembers when he finaUy leaves this body, to that he goes having been used to ponder on it." ' ^ See art. Adi-Buddha in E.R.E. Asanga in the Sutralankara (ix. 77) condemns ihe doctrine of Adi-Buddha, showing that the term was known then, even if it \ 32 THE MAHAYANA [ch theistic form of Buddhism is late and is recorded from Nepal Tibet (in the Kalacakra system) and Java, a distribution whicl impHes that it was exported from Bengali But another form in which the Buddha-force is impersonal and analogous to the Parabrahma of the Vedanta is much older. Yet when this philosophic idea is expressed in popular language it comes very near to Theism. As Kern has pointed out, Buddha is not called Deva or Isvara in the Lotus simply because he is above such beings. He declares that he has existed and will exist for incalculable ages and has preached and wiU preach in innumer able milUons of worlds. His birth here and his nirvana are Ulusory, kindly devices which may help weak disciples but do not mark the real beginning and end of his activity. This imphes a view of Buddha's personaUty which is more precisely defined in the doctrine known as Trikaya or the three bodies^ and expounded in the Mahayana-sutralankara, the Awakening of Faith, the Suvarna-prabhasa sutra* and many other works. It may be stated dogmaticaUy as foUows, but it assumes somewhat divergent forms according as it is treated theologicaUy or meta- physicaUy. "^ A Buddha has three bodies or forms of existence. The first is the Dharma-kaya, which is the essence of aU Buddhas. It is true knowledge or Bodhi. It may also be described as Nirvana and also as the one permanent reality underlying aU phenomena and all individuals. The second is the Sambhoga-kaya, or body had not the precise dogmatic sense which it acquired later. His argument is that no one can become a, Buddha without an equipment (Sambhara) of merit and knowledge. Such an equipment can only be obtained from a previous Buddha and therefore the series of Buddhas must extend infinitely backwards. ^ For the prevalence of the doctrine in mediaeval Bengal see B. K. Sarkar, l Folklore Element in Hindu Culture, which is however sparing of precise references, The Dharma or Nirafijana of the Sunya Purana seems to be equivalent to Adi- i Buddha. Sometimes the Adi-Buddha is identified with Vajrasattva or Samantabhadra, although these beings are otherwise classified as Bodhisattvas. This appean analogous to the procedure common in Hinduism by which a devotee declares that his special deity is aU the gods and the supreme spirit. 2 It would appear that some of the Tantras treat of five bodies, adding to tie three here given others suoh as the Anaudakaya, Vajrakaya and Svabhavakaya. For this doctrine see especiaUy De la VaUee Poussin, J.B.A.S. 1906, pp. 943-99' and Musion, 1913, pp. 257 ff. Jigs-med nam-mk4, the historian of Tibetan Buddiif^ describes four. See Huth, Oes. d. Bud. in d. Mongolei, vot. ii. pp. 83-89. Hinduism also assigns to Uving beings three bodies, the Karana-^arira, Unga^. and sthulai ' Translated into Chinese by Dharmaraksha between 397 and 439 a.d. XVIII] THE BUDDHAS OF MAHAYANISM 33 of enjoyment, that is to say the radiant and superhuman form in which Buddhas appear in their paradises or when otherwise manifesting themselves in celestial splendour. The third is the Nirmana-kaya, or the body of transformation, that is to say the human form wom by Sakyamuni or any other Buddha and regarded as a transformation of his true nature and almost a i distortion, because it is so partial and inadequate an expression 1] of it. Later theology regards Amitabha, Amitayus and SakyaT" muni as a series corresponding to the three bodies. Amitabha does not reaUy express the whole Dharma-kaya, which is incapable of personification, but when he is accurately dis tinguished from Amitayus (and frequently they are regarded as synonyms) he is made the more remote and ethereal of the two. Amitayus with his rich ornaments and his flask containing the water of etemal Hfe is the ideal of a splendidly beneficent saviour and represents the Sambhoga-kaya^. Sakyamuni is the same beneficent being shrunk into human form. But this is only one aspect, and not the most important, of the doctrine of the three bodies. We can easily understand the Sambhoga-kaya and Nirmana-kaya : they correspond to a deity such as Vishnu and his incarnation Krishna, and they are puzzUng in Buddhism simply because we think naturaUy of the older view (not entirely discarded by the Mahayana) which makes the human Buddha the crown and apex of a series of Hves that find in him their fulfilment. But it is less easy to understand the Dharma-kaya. The word should perhaps be translated as body of the law and the thought originaUy underlying it may have been that the essential nature of a Buddha, that \^hich makes him a Buddha, is the law which he preaches. As we might say, the teacher Hves in his teaching : while it survives, he is active and not dead. The change from metaphor to theology is iUustrated by ' Hsiian Chuang when he states ^ (no doubt quoting from his , edition of the Pitakas) that Gotama when dying said to those around him "Say not that the Tathagata is undergoing final ' The prototype of the Sambhoga-kaya is found in the PaU Canon, for the Buddha says (Mahaparinib. Sut. in. 22) that when he appears among the different classes of s gods his form and voice are simUar to theirs. ^ Watters, vol. II. p. 38. "Spiritual essence" is Fa-shen in Chinese, i.e. Dharma- , kaya. Another passage is quoted to the effect that "henceforth the observances of all my disciples constitute the Tathagata's Fa-shen, etemal and imperishable." E. n. 3 34 THE MAHAYANA [ch extinction: his spiritual presence abides for ever unchangeable.' This apparently corresponds to the passage in the PaU Canoni which runs "It may be that in some of you the thought may arise, the word of the Master is ended: we have no more a teacher. But it is not thus that you should regard it. The truths and the rules which I have set forth, let them, after I am gone, be the Teacher to you." But in Buddhist writings, including the oldest PaH texts, Dharma or Dhamma has another important meaning. It signifies phenomenon or mental state (the two being identical for an ideaUstic philosophy) and comprises both the external and the internal world. Now the Dharma-kaya is emphaticaUy not a phenomenon but it may be regarded as the substratum or totaUty of phenomena or as that which gives phenomena whatever reaUty they possess and the double use of the word dharma rendered such divagations of meaning easier 2. Hindus have a tendency to identify being and know ledge. According to the Vedanta phUosophy he who knows Brahman, knows that he himself is Brahman and therefore he actuaUy is Brahman. In the same way the true body of the Buddha is prajna or knowledge^. By this is meant a knowledge which transcends the distinction between subject and object and which sees that neither animate beings nor inanimate thmgs have individuaUty or separate existence. Thus the Dharma- kaya being an intelligence which sees the iUusory quaUty of the world and also how the iUusion originates* may be regarded as the origin and ground of all phenomena. As such it is also called Tathagata-garbha and Dharma-dhatu, the matrix or storehouse of all phenomena. On the other hand, inasmuch as it is beyond them and impHes their unreaUty, it may also be regarded as the annihilation of aU phenomena, in other words as Nirvana. In fact the Dharma-kaya (or Bhuta-tathata) is sometimes^ defined in words similar to those which the PaU Canon makes the Buddha use when asked if the Perfect Saint exists after death— ¦ "it is neither that which is existence nor that which is non- ¦^ Mahaparinib. Sut. VI. i. * Something similar might happen in English if think and thing were pro nounced in the same way and a thing were beUeved to be that which we can thini. ' See Ash^asahasrika Prajna-paramita, chap, rv, near beginning. • It is in this last point that no inferior InteUigence can foUow the thought of a Buddha. ' The Awakming of Faith, Tsitaro Suzuki, p. 59. I xvin] THE BUDDHAS OF MAHAYANISM 35 existence, nor that which is at once existence and non-existence nor that which is neither existence nor non-existence." In more theological language it may be said that according to the general opinion of the Mahayanists a Buddha attains to Nirvana by the very act of becoming a Buddha and is therefore beyond everything which we caU existence. Yet the compassion which he feels for mankind and the good Karma which he has accumu lated cause a human image of him (Nirmana-kaya) to appear among men for their instruction and a superhuman image, perceptible yet not material, to appear in Paradise. -•--- CHAPTER XIX MAHAYANIST METAPHYSICS Thxts the theory of the three bodies, especially of the Dharma- kaya, is bound up with a theory of ontology. Metaphysics became a passion among the traveUers of the Great Vehicle as psychology had been in earUer times. They may indeed he reproached with being bad Buddhists since they insisted on speculating on those questions which Gotama had declared to be unprofitable and incapable of an answer in human language. He refused to pronounce on the whence, the whither and the nature of things, but bade his disciples walk in the eightfold path and analyse the human mind, because such analysis con duces to spiritual progress. India was the last country in the world where such restrictions were Hkely to be observed. Much Mahayanist Hterature is not reUgious at aU but simply meta physics treated in an authoritative and ecclesiastical manner, The nature and origin of the world are discussed as freely as in the Vedanta and with simUar results : the old ethics and psy chology receive scant attention. Yet the difference is less than might be supposed. Anyone who reads these treatises and notices the number of apparently eternal beings and the talk about the universal mind is Hkely to think the old doctrine that nothing has an atman or soul, has been forgotten. But this impression is not correct ; the doctrine of Nairdtmyam is asserted so uncompromisingly that from one point of view it may be said that even Buddhas do not exist. The meaning of this doctrine is that no being or object contains an unchangeable permanent self, which Hves unaltered in the same or in different bodies. On the contrary individual existences consist of nothing but a coUection of skandhas or a santdna, a succession or_series of mental phenomena. In the PaH books this doctrine.isj.pplied^ chieflyto the soul and psychological enquiries. The Mahayana- applied it to the external world and proved by ingenious argu ments' that nothing at_. all exists. Similarly the doctrine of iCarma is maintained, though it is seriously modified by the CH. XIX] MAHAYANIST METAPHYSICS 37 admission that merit can be transferred from one personality to another. The Mahayana continued to teach that an act once performed affects a particular series of mental states until its effect is exhausted, or in popular language that an individual enjoys or suffers through a series of births the consequences of previous acts. Even the instance of Amitabha's paradise, though it strains the doctrine of Karma to the utmost, does not repudiate it. For the believer performs an act — to wit, the invocation of Amitabha — to which has been attached the wonderful result that the performer is reborn in a blessed state. This is not essentiaUy different from the idea found in the Pali Canon that attentions paid to a BuddUa may be rewarded by a happy rebirth In Leaven^ Mahayanist metaphysics, Hke aU other departments of this theology, are beset by the difficulty that the authorities who treat of them are not always in accord and do not pretend to be in accord. The idea that variety is permissible in beUef and conduct is deeply rooted in later Buddhism: there are many vehicles, some better than others no doubt and some very ramshackle, but all are capable of conveying their passengers to salvation. NominaUy the Mahayana was divided into only two schools of philosophy : practicaUy every important treatise propounds a system with features of its own. The two schools are the Yogacaras and Madhyamikas^- Both are idealists and deny the reaUty of the external world, but whereas the Yogacaras (also caUed Vijnanavadins) admit that Vijiiana or consciousness and the series of states of which it consists are real, the Madhya- mikas refuse the title of reahty to both the subjective and the objective world and hence gained a reputation of being complete nihilists. Probably the M§;dhyamikas are the older school. Both schools attach importance to the distinction between relative and absolute knowledge. Relative knowledge is true for human beings living in the world : that is to say it is not more false than the world of appearance in which they Hve. The Hinayanist doctrines are true in this sense. Absolute knowledge I ' E.g. in Mahaparinib. Sut. iv. 57, the Buddha says "There has been laid up by Cnnda the smith (who had given him his last meal) a karma redounding to length ¦ of Ufe, to good fortune, to good fame, to the inheritance of heaven, and of sovereign power." i * Strictly speaking Madhyamaka is the name of the school Madhyamika of its I adherents. Both forms are used, e.g. Madhyamakakarikas and Madhyamikasfltra. 38 THE MAHAYANA [ch. rises above the world of appearance and is altogether true but difficult to express in words. The Yogacara makes three divisions, dividing the inferior knowledge into two. It dis tinguishes first iUusory knowledge {parikalpita) such as mistaking a piece of rope for a snake or belief in the existence of individual souls. Secondly knowledge which depends on the relations of things {paratantra) and which though not absolutely wrong is necessarily Hmited, such as belief in the real existence of ropes and snakes. And thirdly absolute knowledge {parinishpanna), which understands all things as the manifestation of an under lying principle. The Madhyamikas more simply divide knowledge into samvriti-satya and paramdrtha-satya, that is the truth of everyday Ufe and transcendental truth. The world and ordinary reUgion with its doctrines and injunctions about good works are real and true as samvriti but in absolute truth {paramdrtham} we attain Nirvana and then the world with its human Buddhas and its gods exists no more. The word Mnyam or iunyatd, that is void, is often used as the equivalent of paramdrtham. Void must be understood as meaning not an abyss of nothingness but that which is found to be devoid of all the attributes which we try to ascribe to it. The world of ordinary experience is not void, for a great number of statements can be made about it, but absolute truth is void, because nothing whatever can be predicated of it. Yet even this colourless designation is not perfectly accurate^, because neither being nor not-being can be predicated of absolute truth. It is for this reason, namely that they admit neither being nor not-being but something between the two, that the foUowers of Nagarjuna are known as the MMhyamikas or school of the middle doctrine, though the European reader is tempted to say that their theories are extreme to the point of being a reductio ad absurdum of the whole system. Yet though much of their logic seems late and useless sophistry, its affinity to early Buddhism cannot be denied. The fourfold proposition that the answer to certain questions cannot be any of the statements "is," "is not," "both is and is not," "neither is nor is not," is part of the earUest known stratum of Buddhism. The Buddha himself is represented as 1 Nagarjuna says Siinyam iti na vaktavyam a^unyam iti va bhavet Ubhayam nobhayam ceti prajiiaptyartham tu kathyate, "It cannot be caUed void or not void or both or neither but in order to somehow indicate it, it is called Sunyata." XIX] MAHAYANIST METAPHYSICS 39 sayingi that most people hold either to a beUef in being or to a beUef in not being. But neither belief is possible for one who considers the question with full knowledge. "That things have being is one extreme: that things have no being is the other extreme. These extremes have been avoided by the Tathagata and it is a middle doctrine that he teaches, " namely, dependent origination as explained in the chain of twelve Hnks. The Madhyamika theory that objects have no absolute and inde pendent existence but appear to exist in virtue of their relations is a restatement of this ancient dictum. The Mahayanist doctors find an ethical meaning in their negations. If things possessed svabhdva, real, absolute, self- determined existence, then the four truths and especially the cessation of suffering and attainment of sanctity would be impossible. For if things were due not to causation but to their own self-determining nature (and the Hindus always seem to imderstand real existence in this sense) cessation of evU and attainment of the good would be alike impossible: the four Noble Truths imply a world which is in a state of constant becoming~"tEaFis"a worU which is not really existent. But for aU that the doctrine of Sunyatd as stated in the Madhyamika aphorisms ascribed to Nagarjuna leaves an im pression of audacious and ingenious sophistry. After laying down that every object in the world exists only in relation to every other object and has no self -existence, the treatise pro ceeds to prove that rest and motion are alike impossible. We speak about the path along which we are passing but there is reaUy no such thing, for if we divide the path accurately, it always proves separable into the part which has been passed over and the part which wiU be passed over. There is no part which is being passed over. This of course amounts to a denial of the existence of present time. Time consists of past and future separated by an indivisible and immeasurable instant. The minimum of time which has any meaning for us impHes a change, and two elements, a former and a subsequent. The present minute or the present hour are faUacious expressions^. 1 Sam. Nik. xxn. 90. 16. " Gotama, the founder of the Nyaya phUosophy, also admitted the force of tho arguments against the existence of present time but regarded them as a reductio ad absurdnim. Shadworth Hodgson in his Philosophy of Reflection, vol. i. p. 253 also treats of the question. 40 THE MAHAYANA . [ch. Therefore no one ever is passing along a path. Again you cannot logically say that the passer is passing, for the sentence is redundant: the verb adds nothing to the noun and vice versa: but on the other hand you clearly cannot say that the non- passer is passing. Again if you say that the passer and the passing are identical, you overlook the distinction between the agent and the act and both become unreal. But you cannot maintain that the passer is different from the passing, for a passer as distinct from passing and passing as distinct from a passer have no meaning. "But how can two entities exist at aU, if they exist neither as identical with one another nor as different from one another?" The above, though much abridged, gives an idea of the logic of these siitras. They proceed to show that aU maimer of things, such as the five skandhas, the elements, contact, attachment, - fire and fuel, origination, continuation and extinction have no real existence. Similar reasoning is then appHed to reUgious topics: the world of transmigration as weU as bondage and Uberation are declared non-existent. In_reaHt^ riQ, souLis^in bondage and none isjreleased,^. Similarly Karma, the Buddha himself, the four truths. Nirvana and the twelve Hnks in the chain of causation are aU unreal. This is not a declaration of scepticism. It means that the Buddha as a human or celestial being and Nirvana as a state attainable in this world are con ceivable only in connection with this world and therefore, Hke the world, unreal. No rehgious idea can enter into the unreal (that is the practical) life of the world unless it is itseU unreal. This sounds a topsy turvy argument but it is really the same as the Advaita doctrine. The Vedanta is on the one hand a scheme of salvation for Hberating souls which transmigrate unceasingly in a world ruled by a personal God. But when true knowledge is attained, the soul sees that it is identical with the Highest Brahman and that souls which are in bondage and God who rules the world are iUusions Hke the world itself. But the Advaita has at least a verbal superiority over the Madhyamika philosophy, for in its terminology Brahman is the real and the existent con trasted with the world of iUusion. The result of giving to what the Advaita caUs the real and existent the name of siinyatS. or 1 The Sankhya phUosophy makes a simUar statement, though for different XIX] MAHAYANIST METAPHYSICS 41 void is disconcerting. To say that everything without distinction is non-existent is much the same as saying that everything is existent. It only means that a wrong sense is habituaUy given to the word exist, as if it meant to be self-contained and without relation to other objects. Unless we can make a verbal contrast and assert that there is something which does exist, it seems futile to insist on the unreaUty of the world. Yet this mode of thought is not confined to text-books on logic. It invades the scriptures, and appears (for instance) in the Diamond Cutter^ which is stiU one of the most venerated booki of devotion in China and Japan. In this_wo A the Buddha explains .that a Bodhisattva must resolve to deUver_aU living beings and. yet niusFundBTStSSdTThat^aHerT&ie has thus„deliyered innumerable beings, no one has been_deliyergd. And why ? Because no one is~to"be^caIIed"a"Bodhisattva for whom there exists the idea of a being, or person. Siniilarly a saint does not think that he is a saint, for if he did so think, fie* would beUeve in a self, and a person. There occur continuaUy in this work phrases cast in the foUowing form : " what was preached as a store of merit, that was preached as no store of merit^ by the Tathagata and there fore it is caUed a store of merit. If there existed a store of merit, the Tathagata would not have preached a store of merit." That is to say, if I understand this dark language rightly, accumulated merit is part of the world of iUusion which we live in and by speaking of it as he did the Buddha impHed that it, Hke every thing else in the world, is reaUy non-existent. Did it belong to the sphere of absolute truth, he would not Have spoken of it as if it were one of the things commonly but erroneously supposed to exist. FinaUy we are told of the highest knowledge "Even the smaUest thing is not known or perceived there ; therefore it is caUed the highest perfect knowledge." That is to say perfect knowledge transcends aU distinctions ; it recognises the iUusory nature of all individuaUty and the truth of sameness, the never- changing one behind the ever-changing many. In this sense it is said to perceive nothing and know nothing. One might expect that a philosophy thus prone to use the ' Vajracchedika. See S.B.E. vol. xlix. It was translated into Chinese by Kumarajiva (384^17 a.d.). 2 Or in other repetitions of the same formula, beings, ideas, good things, signs, etc., etc. 42 THE MAHAYANA [ce, language of extreme nihiUsm would sUp into a destructive, or at least negative system. But Mahayanism was puUed equally strongly in the opposite direction by the popular and mytho logical elements which it contained and was on the whole incUned to theism and even polytheism quite as much as to atheism and acosmism. A modern Japanese writer^, says that Dharma-kaya "may be considered to be equivalent to the Christian conception of the Godhead." This is excessive as a historical statement of the view current in India during the early centuries of our era, but it does seem true that Dharma-kaya was made the equivalent of the Hindu conception of Param Brahma and also that it is very nearly equivalent to the Chinese Tao^. The work called Awakening of Faith^ and ascribed to Asvaghosha is not extant in Sanskrit but was translated into Chinese in 553 a.d. Its doctrine is practicaUy that of the Yogacara school and this makes the ascription doubtful, but it is a most important treatise. It is regarded a's authoritative in China and Japan at the present day and it Ulustrates the triple tendency of the Mahayana towards metaphysics, mythology, and devotional piety. It declares that faith has four aspects. Three of these are the three Jewels, or Buddha, the Law and the Church, and cover between them the whole field of rehgion and morahty as generally understood. The exposition is tinged with a fine unselfish emotion and teUs the beUever that though he should strive not for his own emancipation but for the salvation of others yet he himseU receives unselfish and super natural assistance. He is remembered and guarded by Buddhas and Bodhisattvas in aU quarters of the Universe who are etemaUy trying to Hberate mankind by various expedients (upaya). By expedient is meant a modified presentment of the truth, which is easier of comprehension and, if not the goal, at least on the road to it, such as the Paradise of Amitabha* ' Soyen Shaku, Sermons of a Buddhist Abbot, p. 47. * See for a simple and persuasive statement of these abstruse doctrines a charming little book caUed Wu-Wei by H. Borel. * Translated from the Chinese by Teitaro Suzuki, 1900. The translation must be used with care, as its frequent use of the word soul may lead to misunderstanding, * Asanga's work Mahdydna-sHlrdlankdra (edited and translated by S. Levi) which covers much of the same ground is extant in Sanskrit as well as in Chinese and Tibetan translations. It is a lucid and authoritative treatise but does not appear to have ever been popular, or to be read now in the Far East. For Yogacara see also Musion, 1904, p. 370. XIX] MAHAYANIST METAPHYSICS 43 But the remaining aspect of faith, which is the one that the author puts first in his enumeration, and treats at great length, is "to beUeve in the fundamental truth, that is to think joyfuUy of suchness." By suchness (in Sanskrit bhuta-tathatd, in Chinese CMnju) is meant absolute truth as contrasted with the relative truth of ordinary experience^. The word is not illuminating nor likely to excite rehgious emotion and the most that can be said for it is that it is less dreary than the void of NagS.rjuna. Another and more positive synonym is dharma-dhdtu, the aU- embracing totaUty of things. It is only through our ignorance and subjectivity that things appear distinct and individuate. Could we transcend this subjectivity, isolated objects would cease to exist. Things in their fundamental nature cannot be named or explained : they are beyond the range of language and perception : they have no signs of distinction but possess absolute sameness (samata). From this totaUty of things nothing can be excluded and to it nothing can be added. Yet it is also sunyata, negation or_thejroid,JbecauseXt„ cannot be said to possess any of _the ^t^butes pf the world we Uve in : neither existence nor non-existence, nor unity nor pluraUty can be predicted of it. According to the celebrated formula of Nagarjuna known as the eight Nos there is in it "neither production {utpdda) nor destruction {uccheda) nor annihilation {nirodha) nor persistence {sasvatd) nor unity {ekdrtha) nor plurality {ndndrtha) nor coming in {dgamana) nor going out {nirgama)." But when we perceive that both subject and object are unreal we also see that suchness is the one reality and from that point of view it may be regarded as the Dharma-kaya of aU Buddhas. It is also called Tathagata- garbha, the womb or store-house of the Buddha, from which all indiAridual existences are evolved under the law of causation, but this aspect of it is already affected by ignorance, for in Bhiita-tathata as known in the Ught of the highest truth there is neither causation nor production. The Yogacara employs the word Sunyatd (void), though not so much as its sister school, but it makes special use of the term dlaya-vijndna, the receptacle or store of consciousness. This in so far as it is superindividual is an aspect of suchness, but when it affirms and particularises itseU it becomes citta, that is the human mind, or to be more 1 The discussion of tathaid in Kathavatthu, xix. 5 seems to record an early phase of these speculations. 44 THE MAHAYANA [ch. accurate the substratum of the human mind from which is developed manas, or the principle of wiU, self -consciousness and self-affirmation. Similarly the Vedanta philosophy, though it has no term corresponding to dlaya-vijndna, is familiar with the idea that Brahman is in one aspect immeasurable and all- embracing but in another is infinitesimal and dweUs in the human heart : or that Brahman after creating the world entered into it. Again another aspect of suchness is enUghtenment {bodhi), that is absolute knowledge free from the limitations of subject and object. This "is the universal Dharma-kaya of tite Tathagatas" and on account of this aU Tathagatas are spoken of as abiding in enUghtenment a priori. This enUghtenment may be negative (as iunydta) in the sense that it transcends all relations but it may also be affirmative and then "it transforms and unfolds itself, whenever conditions are favourable, in the form of a Tathagata or some other form in order that aU beings may be induced to bring their store of merit to maturity^." It will be seen from the above that the absolute truth of the Mahayanists varies from a severely metaphysical conception, the indescribable thing in itself, to something very Hke an all- pervading benevolent essence which from time to time takes shape in a Buddha. And here we see how easy is the transition from the old Buddhism to a form of pantheism. For if we admit that the Buddha is a superhuman InteUigence appearing from time to time according to a certain law, we add Httle to this statement by saying that the essence or spirit of the cosmos manifests itseU from time to time as a Buddha. Only, such words as essence or spirit are not reaUy correct. The world of individuals is the same as the highest truth, the same as the Dharma-kaya, the same as Nirvana. It is only through ignorance that it appears to be different and particularized. Ignorance, the essence of which consists in beUeving in the distinction between subject and object, is also caUed defilement and the highest truth passes through various stages of defilement ending with that where under the influence of egoism and passion the external world of particulars is beUeved to be everything. But the various stages may influence one another^ so that under a higher influence the mind which is involved in subjectivity 1 Awakening of Faith, Teitaro Suzuki, pp. 62 and 70. 2 The process is generally oaUed Vasana or perfuming. XIX] MAHAYANIST METAPHYSICS 45 begins to long for Nirvana. Yet Nirvana is not something different from or beyond the world of experience; it does not reaUy involve annihUation of the skandhas. Just as in the Advaita he who has the true knowledge sees that he himself and everything else^J^ahman, jQ.fpj_the Mahayanist aU things are seen icHSi^irvana, to be the pharma-kaya. It is sometiines^ said that there are four kinds of Nirvana (a) absolute Nirvana, which is a synonym of the Dharma-kaya and in that sense universally present in all beings, (6) upadhisesha-nirvana, the state of enUghtenment which can be attained during Hfe, while the body with its Hmitations still remains, (c) anupadhisesha- nirvana, a higher degree of the same state attained after death when the hindrances of the body are removed, {d) Nirvana without abode or apratishthita-nirvana. Those who attain to this understand that there is no real antithesis between Samsara and Nirvana 2: they do not seek for rest or emancipation but devote themselves to beneficent activity and to leading their feUows to salvation. Although these statements that Nirvana and Samsara are the same are not at all in the manner of the older Buddhism, yet this ideal of disinterested activity combined with Nirvana is not inconsistent with the portrait of Gotama preserved in the PaH Canon. The Mahayanist Buddhism of the Far East makes free use of such phrases as the Buddha in the heart, the Buddha mind and the Buddha nature. These seem to represent such Sanskrit terms as Buddhatva and Bodhicitta which can receive either an ethical or a metaphysical emphasis. The former Une of thought is weU shown in Santideva^ who treats Bodhicitta as the initial impulse and motive power of the rehgious Ufe, com bining inteUectual iUumination and unselfish devotion to the good of others. Thus regarded it is a guiding and stimulating principle somewhat analogous to the Holy Spirit in Christianity. But the Bodhicitta is also the essential quaUty of a Buddha (and the Holy Spirit too is a member of the Trinity) and in so far as a man has the Bodhicitta he is one with aU Buddhas. 1 Vijiianamatra Sastra. Chinese version quoted by Teitaro Suzuki, Outlines of Mahdydna Buddhism, p. 343. Apparently both upadhi and upadhi are used in Buddhist Sanskrit. Upadi is the Pali form. ^ So the Madhyamika Sastra (xxv. 19) states that there is no difference between Samsara and Nirvana. Cf. Rabindranath Tagore, Sadhana, pp. 160-164. ' E,g, Bodhicaryavatara, chap. I, caUed praise of the Bodhicitta. 46 THE MAHAYANA [oh. xix This conception is perhaps secondary in Buddhism but it is also as old as the Upanishads and only another form of the doctrine that the spirit in every man (antaryamin) is identical with the Supreme Spirit. It is developed in many works stiU popular in the Far East^ and was the fundamental thesis of Bodhidharma, the foimder of the Zen school. But the practical character of the Chinese and Japanese has led them to attach more import ance to the moral and intellectual side of this doctrine than to the metaphysical and pantheistic side. i E.g. the P'u-t'i-hsin-U-hsiang-lun (Nanjio, 1304), translated from Nagarjuna, and the Ta-Ch'eng-fa-chieh-wu-ch'a-pieh-lun, translated from Sthiramati (Nanjio, 1258). CHAPTER XX MAHAYANIST SCRIPTURES In a previous chapter I have discussed the PaU Canon and I shaU subsequently have something to say about the Chinese and Tibetan Canons, which are libraries of reUgious and edifying works rather than sacred books similar to the Vedas or the Bible. My present object is to speak of the Sanskrit Hterature, chiefly sutras, which appeared contemporaneously with the rise of Mahayanism in India. The Mahayanist scriptures are the largest body of sacred writings extant in the world, but it is not easy either to define the Umits of the Canon or to say when it was put together. According to a common tradition Kanishka played for the Church of the Great Vehicle much the same part as Asoka for the Theravadins and summoned a Council which wrote com mentaries on the Tripitaka. This may be reasonably held to include a recension of the text commented on but we do not know what that text was, and the brief and perplexing accounts of the CouncU which we possess indicate not that it gave its imprimatur to Mahayanist sutras but that it was specially concemed with the Abhidharma works of the Sarvastivadin school. In any case no Canon formed in the time of Kanishka can have been equivalent to the coUections of writings accepted to day in China and Tibet, for they contain works later than any date which can be assigned to his reign, as do also the nine sacred books revered in Nepal. It was agreed among Indian Buddhists that the scriptures were divided among the three Pitakas or baskets, but we may surmise that there was no unanimity as to the precise contents of each basket. In India the need for unanimity in such matters is not felt. The Brah mans always recognized that the most holy and most jealously preserved scriptures could exist in various recensions and the Mahabharata shows how generations of respectful and un critical hearers may aUow adventitious matter of all sorts to 48 THE MAHAYANA [ch. be incorporated in a work. Something of the same kind happened with the Pitakas. We know that the PaU recension which we possess was not the only one, for fragments of a Sanskrit version have been discovered. There was probably a large floating Hterature of sutras, often presenting several recensions of the same document worked up in different ways. Just as additions were made to the Ust of Upanishads up to the middle ages, although the character of the later works was different from that of the earHer, so new sutras, modern in date and in tone, were received in the capacious basket. And just as the Puranas were accepted as sacred books without undermining the authority of the Vedas, so new Buddhist scriptures superseded without condemning the old ones. Various Mahayanist schools had their own versions of the Vinaya which apparently contain the same rules as the PaU text but also much additional narrative, and Asanga quotes from works corresponding to the PaH Nikayas, though his doctrine belongs to another age^. The Abhidharma section of the PaH Canon seems however to have been pecuUar to the Theravada school. The Sarvastivadin Pitaka of the same name was entirely different and, judging from the Chinese Canon, the Mahayanists gave the title to philosophic works by such authors as Asanga and Vasubandhu, some of which were described as revelations from Maitreya. SpeciaUy characteristic of Mahayanist Buddhism are the Vaipulya^ sutras, that is sutras of great extension or develop ment. These works, of which the Lotus is an example, foUow the same scheme as the older sutras but are of wider scope and on a much larger scale, for they often consist of twenty or more chapters. They usuaUy attempt to give a general exposition of the whole Dharma, or at least of some aspect of it which is 1 In the Mahayana-sutralankara he quotes frequently from the Samyukta, and Ekottara Agamas, corresponding to the Samyiitta and Anguttara Nikayas of the PaU. 2 A reading Vaitulya has also been found in some manuscripts of the Lotus discovered at Kashgar and it is suggested that the word may refer to the sect of VetuUas or Vetulyakas mentioned in the Commentary on the Kathavatthu as holding that the Buddha reaUy remained in the Tushita heaven and sent a phantom to represent him in the world and that it was Ananda, not the Budiia, who preached the law. See Kern, Vers, en Med. der K. Ak. v. Wetenschappen, Letterk,,' R, 4 D. vm. pp. 312-9, Amsterdam, 1907, and De la VaUde Poussin's notice of this article in J,B,A,S. 1907, pp. 434-6. But this interpretation does not seem very probable. XX] MAHAYANIST SCRIPTURES 49 extoUed as sufficient for the right conduct of Hfe. The chief speaker is usuaUy the Buddha, who is introduced as teaching on the Vulture Peak, or some other well-known locality, and surrounded by a great assemblage many of whom are super human beings. The occasion of the discourse is commonly signaUzed by his sending forth rays of light which iUuminate the universe until the scene includes other worlds. As early as the Anguttara Nikaya^ we find references to the danger of a taste for ornate and poetic sutras and these compositions seem to be the outcome of that taste. The Uterary ideas and methods which produced them are iUustrated by the Sutralankara of Asva ghosha, a coUection of edifying tales, many of which use the materials suppHed by the PaH Nikayas and Vinaya but present them in a more effective and artistic form. It was thought a pious task to ampUfy and embellish the simple narratives handed down by tradition. The Mahayanist scriptures are composed in Sanskrit not in PaH, but it is only rarely — for instance in the works of Asvaghosha — that Buddhist Sanskrit conforms to the rules of the classical language. UsuaUy the words deviate from this standard both in form and meaning and often suggest that the text as we have it is a sanskritized version of an older work in some popular dialect, brought into partial conformity with Uterary usage. In the poetical portions, this process of sanskritization encountered greater difficulties than in prose, because metre and prosody often refused to admit the changes required by grammar, so that this poetical dialect cannot be caUed either Sanskrit, PaH or Magadhi but remains a mixture of learned and popular speech. But Sanskrit did not become a sacred language for the Mahayanists like Latin for Roman CathoUcs. It is rather PaU which has assumed this position among the Hinayanists, for Burmese and Sinhalese translations of the Pitakas acquired no authority. But in the north the principle ^ that every man might read the Buddha's word in his own vernacular was usually respected : and the populations of Central Asia, the Chinese, the Tibetans, and the Mongols translated the scriptures into their 1 » IV. 160. 5. '' See CuUavagga, v. 33. The meaning evidently is that the Buddha's words are , not to be enshrined in an artificial Uterary form which wiU prevent them from ( being popular. £. n 4 50 THE MAHAYANA [ch. own languages without attaching any superstitious importance to the original words, unless they were Dharanis or spells. About the time of the Christian era or perhaps rather earUer, greater use began to be made of writing for reUgious purposes. The old practice of reciting the scriptures was not discontinued but no objection was made to preserving and reading them in written copies. According to tradition, the PaU scriptures were committed to writing in Ceylon during the reign of Vattagamani, that is according to the most recent chronology about 20 B.C., and Kanishka caused to be engraved on copper plates the com mentaries composed by the council which he summoned. In Asvaghosha! we find the story of a Brahman who casuaUy taking up a book to pass the time Ughts on a copy of the Sutra of the Twelve Causes and is converted. But though the Buddhists remained on the whole true to the old view that the important thing was to understand and disseminate the substance of the Master's teaching and not merely to preserve the text as if it were a sacred formula, stiU we see growing up in Mahayanist works ideas about the sanctity and efficacy of scripture which are foreign to the PaH Canon. Many sutras (for instance the Diamond Cutter) extol themselves as aU -sufficient for salvation: the Prajna-paramita commences with a salutation addressed not as usual to the Buddha but to the work itseff, as if it were a deity, and Hodgson states that the Buddhists of Nepal worship their nine sacred books. Nor was the idea excluded that certain words, especiaUy formulae or speUs caUed Dharani, have in themselves a mysterious efficacy and potency^. Some of these are cited and recommended in the Lotus ^. In so far as the repetition of sacred words or speUs is regarded as an integral part of the religious life, the doctrine has no warrant in the earHer teaching. It obviously becomes more and more pro minent in later works. But the idea itself is old, for it is clearly the same that produced a beUef in the Brahmanic mantras, particularly the mantras of the Atharva Veda, and early Buddhism did not reject mantras in their proper place. Thus* the deities present themselves to the Buddha and offer to teach him a formula which will protect his disciples from the attacks of evil spirits. Hsuan Chuang even states that the council which i ' Sutralankara, I. 2. 2 See WaddeU, "The Dharani cult" in Ostasiat. Ztsft, 1912, pp. 155 ff. ^ Chap. XXI, which is however a later addition. < Dig. Nik. 32. XX] MAHAYANIST SCRIPTURES 51 sat at Rajagriha after the Buddha's death compiled five Pitakas, one of which consisted of Dharanis i, and it may be that the coUection of such texts was begun as early as the coUection of discourses and rules. But for many centuries there is no evidence that they were in any way confounded with the Dharma. The Mahayanist scriptures are so voluminous that not even the clergy were expected to master any considerable part of them^. Indeed they make no claim to be a connected whole. The theory was rather that there were many vehicles plying on the road to salvation and many guide books. No traveUer thought of taking the whole library but only a few volumes which suited him. Most of the Chinese and Japanese sects avowedly base themselves upon three sutras, selected according to the taste of each school from the hundreds quoted in cata logues. Thus the T'ien-t'ai sect has for its scriptures the Lotus, the Nirvana-siitra and the Prajna-paramita, while the Shin-shu sect admits only the three Amidist sutras. The foUowing are the names of some of the principal Mahayanist scriptures. Comparatively few of them have been published in Europe and some exist only in Chinese or Japanese translations. 1. PrajM-paramita or transcendental knowledge ^ is a generic name given to a whole Hterature consisting of treatises on the doctrine of sunyata, which vary greatly in length. They are classed as sutras, being described as discourses delivered by the Buddha on the Vulture Peak. At least ten are known, besides excerpts which are sometimes described as substantive works. The great coUection translated into Chinese by Hsiian Chuang is said to consist of 200,000 verses and to comprise sixteen different sutras* The earUest translation of one of these treatises into Chinese (Nanjio, 5) was made about 170 a.d. and • Watters, rfiaw Chwang, ii. p. 160. ^ The Mahavyutpatti (65) gives a list of 105 sutras. ' The word param-ita means as an adjective gone to the further shore or trans cendent. As a feminine substantive it means a transcendent virtue or perfection. * See WaUeser, Prajndpdramitd in Quellen der Religionsgeechichte, pp. 15 ff. S,B,E, XLIX. Nanjio, Catalogue Nos. 1-20 and Rajendralala Mitra's Nepalese Buddhist Literature, pp. 177 ff. Versions are mentioned consisting of 125,000 verses, 100,000 verses, 25,000 verses, 10,000 verses and 8000 verses respectively. (Similarly at the beginning of the Mahabharata we are told that the Epic consists of 8800 verses, of 24,000 and of 100,bOO.) Of these the last or Ashtasahasrika has been pubUshed in the Bibliotheca Indica and the second or Satasahasrikfi is in process 4—2 52 THE MAHAYANA [ch. everything indicates that portions of the Prajiia-paramita are among the earliest Mahayanist works and date from about the first century of our era. Prajna not only means knowledge of the absolute truth, that is to say of .siinyata or the void, but is regarded as an ontological principle synonymous with Bodhi and Dharma-kaya. Thus Buddhas not only possess this knowledge in the ordinary sense but they are the knowledge manifest in human form, and Prajna is often personified as a goddess. AU these works lay great stress on the doctrine of sunyata, and the non-existence of the world of experience. The longest re cension is said to contain a polemic against the Hinayana. The Diamond Cutter is one of the best known of these trans cendental treatises and the two short works caUed Heart of the Prajiiaparamita, which are widely read in Japan, appear to be brief abstracts of the essence of this teaching. 2. The Saddharma-pundarika, or Lotus of the Good Law^, is one of the best known Mahayanist sutras and is highly esteemed in China and Japan. It purports to be a discourse deUvered by Sakyamuni on the Vulture Peak to an assemblage . of Bodhisattvas. The Lotus clearly affirms the multipHcity of vehicles, or various ways of teaching the law, and also the eternity of the Buddha, but it does not emphasize, although it mentions, the doctrine of sunyata. The work consists of two parts of which the second (chaps, xxi-xxvi) is a later addition. This second part contains speUs and many mythological narratives, including one of an ancient Bodhisattva who bumt himself aUve in honour of a former Buddha. Portions of the Lotus were translated into Chinese under the Western Tsia Djmasty 265-316 a.d. and it is quoted in the Maha-prajfia- paramita-sastra ascribed to Nagarjuna^. The first part is of publication. It is in prose, so that the expression "verses" appears not to mean that the works are Gathas. A Khotanese version of the Vajracchedika is edited in Hoernle's Manuscript Remains by Sten Konow. The Sanskrit text was edited by Max MiiUer in Anecdota Oxoniensia, ' The Sanskrit text has been edited by Kern and Nanjio in Bibliotheca Buddhica; translated by Bumouf {Le Lotus de la bonne Loi), 1852 and by Kern (Saddharma- pundarika) in S.B.E, vol. xxi. ^ There appears to have been an earUer Chinese version of 255 A.D. but it has been lost. See Nanjio, p. 390. One of the later Chinese versions aUudes to the existence of two recensions (Nanjio, No. 139). See B.E.F.E.O. 1911, p. 453. Frag ments of a shorter and apparently earUer recension of the Lotus have been discovered in E. Turkestan. See J,R,A,S, 1916, pp. 269-277. XX] MAHAYANIST SCRIPTURES 53 probably not later than the first century a.d. The Lotus is unfortunately accessible to EngUsh readers only in a most unpoetic translation by the late Professor Kern, but it is a great reUgious poem which starting from humanity regards rehgion as cosmic and universal, rather than something mainly con cerned with our earth. The discourses of Sakyamuni are accompanied in it by stupendous miracles culminating in a grand cosmic phantasmagojia in which is evoked the stupa containing the body of a departed Buddha, that is a shrine containing the eternal truth. 3. The LaHta-vistara^ is a Hfe of Sakyamuni up to the com mencement of his mission. Though the setting of the story is miraculous and Buddhas and Bodhisattvas innumerable are freely spoken of, yet the work does not enunciate the character istic Mahayanist doctrines so definitely as the other treatises here enumerated. It is said to have originally belonged to the school of the Sarvastivadins and to have been subsequently accepted by the Mahayanists, and though it is not an epic but a coUection of baUads and legends, yet it often reads as if it were a preliminary study for Asvaghosha's Buddhacarita. It contains Sanskrit versions of old legends, which are almost verbal renderings of the PaU text, but also new material and seems to be conscious of relating novelties which may arouse scepticism for it interrupts the narrative to anathematize those who do not beheve in the miracles of the Nativity and to extol the merits of faith {iraddhd not bhakti). It is probably coeval with the earHer Gandharan art but there are no facts to fix its date 2. 4. The Lankavatara^ gives an account of the revelation of the good Law by Sakyamuni when visiting Lanka. It is pre sumably subsequent to the period when Ceylon had become a 1 Edited by Rajendralala Mitra in the Bibliotheca Indica and partially translated in the same series. A later critical edition by Lefmann, 1902-8. * The early Chinese translations seem doubtful. One said to have been made under the later Han has been lost. See Nanjio, No. 159. ' See Bumouf, Introduction, pp. 458 ff. and J,R.A.S. 1905, pp. 831 ff. Rajen dralala Mitra, Nepalese Buddhist Literature, p. 113. A brief analysis is given in J,A,S,B, June, 1905 according to which the sutra professes to be the work of a human author, Jina of the clan of Katyayana born at Campa. An edition of the Sanskrit text pubUshed by the Buddhist Text Society is cited but I have not seen it. Chinese translations were made ia 443 and 515 but the first is incomplete and does not correspond with our Sanskrit text. 54 THE MAHAYANA [ch. centre of Buddhism, but the story is pure fancy and unconnected with history or with older legends. It relates how the Buddha aUghted on Mt Malaya in Lanka. Ravana came to pay his respects and asked for definitions of virtue and vice which were given. The Bodhisattva Mahamati (apparently Manjusri) pro ceeded to propound a series of more abstruse questions which are answered at considerable length. The Lankavatara repre sents a mature phase of speculation and not only criticizes the Sankhya, Pasupata and other Hindu schools, but is conscious of the growing resemblance of Mahayanism to Brahmanic phUosophy and tries to explain it. It contains a prophecy about Nagarjuna and another which mentions the Guptas, and it appears to aUude to the domination of the Huns. This aUusion would make its date as late as the sixth century but a translation into Chinese which is said to correspond with the Sanskrit text was made in 513. If so the barbarians referred to cannot be the Huns. An earHer translation made in 443 does not agree with our Sanskrit text and perhaps the work existed in several recensions. 5. The Suvarna-prabhasa or GHtter of Gold^ is a Vaipulya sutra in many ways resembling the Lotus. It insists on the supernatural character of the Buddha. He was never reaUy bom nor entered into Nirvana but is the Dharma-kaya. The scene is laid at Rajagriha and many Brahmanic deities are among the interlocutors. It was translated into Chinese about 420 a.d. and fragments of a translation into Uigur have been discovered in Turkestan^. The contents comprise phUosophy, legends and speUs. 6. Ganda-vjruha^ or the Structure of the World, which is compared to a bubble. The name is not found in the catalogue of the Chinese Tripitaka but the work is said to be the same as the Avatamsaka sutra which is popular in the Far East under the name of Hua-yen in China or Ke-gon in Japan. The identity of the two books could not have been guessed from the extracts and analyses which have been pubUshed but is guaranteed by 1 Abstract by Rajendralala Mitra, Nepalese Buddhist Lit. p. 241. » See Nanjio, No. 127 and F. W. K. MuUer in Abhandl. der K, Preuss, Ahd. der Wissenschaften, 1908. The Uigur text is published in Bibliotheca Buddhica, ' 1914. Fragments of the Sanskrit text have also been found in Turkestan. » Abstract by Raj. Mitra, Nepalese Buddhist Lit. pp. 90 ff. The ^ikshasamuooaya cites the Ga^Kja-vyiiha several times and does not mention the Avataipsaka. XX] MAHAYANIST SCRIPTURES 55 high authorities^. It is possible however that the Ganda-vyuha is only a portion of the larger work caUed Avatarpsaka. So far as can be judged from the extracts, this text preaches in a fuUy developed form, the doctrines of Sunyata, Dharma-kaya, the omnipresence of the Buddha and the redemption of the world by the exertions of Bodhisattvas. Yet it seems to be early, for a portion of it was translated into Chinese about 170 a.d. (Nanjio, 102) and about 405 Kumarajiva translated a com mentary on it ascribed to Nagarjuna (Nanjio, 1180). 7. Tathagata-guhyaka. This work is known by the analysis of Rajendralala Mitra from which it appears to be a Tantra of the worst class and probably late. Its proper title is said to be Sriguhyasamaja. Watanabe states that the work catalogued by Nanjio under No. 1027 and translated into Chinese about 1000 A.D. is an expurgated version of it. The Sikshasamuccaya cites the Tathagata-guhya-siitra several times. The relations of these works to one another are not quite clear. 8. Samadhiraja^ is a Vyakarana or narrative describing different forms of meditation of which the Samadhiraja is the greatest and best. The scene is laid on the Vulture's Peak and the principal interlocutors are Sakyamuni and Candraprabha, a rich man of Rajagriha. It appears to be the same as the Candrapradipa-sutra and is a complete and copious treatise, which not only expounds the topic from which it takes its name but incidentaUy enumerates the chief principles of Mahayanism. Watanabe* states that it is the Yiiefi-teng-san-mei-ching (Nanjio, 191) translated about 450 and again in 557 a.d. 9. Dasabhiimisvara*. An account of the ten stages in the career of a Bodhisattva before he can attain to Buddhahood. The scene is laid in the paradise of Indra where Sakyamuni was temporarUy sojourning and the principal interlocutor is a Bodhi sattva named Vajragarbha. It is said to be the same as the Dasabhumika-siitra first translated into Chinese about 300 a.d. ^ The statement was first made on the authority of Takakusu quoted by Wintemitz in Oes, Ind, Lit, ii. i. p. 242. Watanabe in J,B,A,S, 1911, 663 makes an equaUy definite statement as to the identity of the two works. The identity is confirmed by PeUiot in J,A. 1914, ii. pp. 118-121. * Abstract by Raj. Mitra, Nepalese Buddhist Lit. pp. 81 ff. Quoted in Santideva's Bodhicaryavatara, vrti. 106. » See J,R.A,S, 1911, 663. • Abstract by Raj. Mitra, Nepalese Buddhist Lit, pp. 81 ff. 56 THE MAHAYANA [ch. (Nanjio, 105 and 110) but this work appears to be merely a portion of the Ganda-vyuha or Avatamsaka mentioned above. These nine works are all extant in Sanskrit and are known in Nepal as the nine Dharmas, the word Dharma being an abbreviation for Dharmaparydya, revolution or exposition of the law, a term frequently used in the works themselves to describe a comprehensive discourse deUvered by the Buddfia. They are aU quoted in the Sikshasamuccaya, supposed to have been written about 650 a.d. No simUar coUection of nine seems to be known in Tibet or the Far East and the origin of the selection is obscure. As however the Ust does not include the Svayambhu Purana, the principal indigenous scripture of Nepal, it may go back to an Indian source and represent an old tradition. Besides the nine Dharmas, numerous other sutras exist in Sanskrit, Chinese, Tibetan and the languages of Central Asia. Few have been edited or translated and even when something is known of their character detailed information as to their contents is usuaUy wanting. Among the better known are the foUowing. 10. One of the sutras most read in China and admired because its style has a Uterary quaUty unusual in Buddhist works is commonly known as the Leng-yen-ching. The fuU title is Shou-leng-yen-san-mei-ching which is the Chinese trans- Utteration of Surangama Samadhi^. This sutra is quoted by name in the Sikshasamuccaya and fragments of the Sanskrit text have been found in Turkestan^. The Siirangama-Samadhi Sutra has been conjectured to be the same as the Samadhiraja, but the accounts of Rajendralala Mitra and Beal do not support this theory. Beal's translation leaves the impression that it resembles a PaH sutta. The scene is laid in the Jetavana with few miraculous accessories. The Buddha discusses with Ananda the location of the soul and after confuting his theories expounds the doctrine of the Dharma-kaya. The fragments found in Turkestan recommend a particular form of meditation. 1 1 . Taranatha informs us that among the many Mahayanist works which appeared in the reign of Kanishka's son was the ' Translated in part by Beal, Catena of Buddhist Scriptures, pp. 286-369. See also Teitaro Suzuki, Outlines of Mahdydna, p. 157. For notices of the text see Nanjio, Nos. 399, 446, 1588. Fa-Hsien, chap. xxix. For the equivalence of Shou- leng-yen and Surangama see Nanjio's note to No. 399 and JuUen, Methode, 1007 and Vasilief, p. 175. * See Sikshas, ed. BendaU, pp. 8, 91 and Hoemle, Manuscript remains, I. pp. 125 ff. XX] MAHAYANIST SCRIPTURES 57 Ratna-kuta-dharma-paryaya in 1000 sections and the Ratnakuta is cited not only by the Sikshasamuccaya but by Asanga^. The Tibetan and Chinese canons contain sections with this name comprising forty-eight or forty-nine items among which are the three important treatises about Amitabha's paradise and many dialogues called Paripriccha, that is, questions put by some personage, human or superhuman, and furnished with appro priate repUes^ The Chinese Ratnakiita is said to have been compiled by Bodhiruchi (693-713 a.d.) but of course he is responsible only for the selection not for the composition of the works included. Section 14 of this Ratnakuta is said to be identical with chapters 11 and 12 of the Mulasarvastivadin Vinaya *- 12. The Guna-karanda-vyuha and Karanda-vyuha are said to be two recensions of the same work, the first in verse the second in prose. Both are devoted to the praise of Avalokita who is represented as the presiding deity of the universe. He has refused to enter Buddhahood himself until all living creatures attain to true knowledge and is speciaUy occupied in procuring the release of those who suffer in hell. The Guna- karanda-vyuha contains a remarkable account of the origin of the world which is said to be absent from the prose version. The primeval Buddha spirit, Adi-Buddha or Svayambhu, pro duces Avalokita by meditation, and Avalokita produces the material world and the gods of Hinduism from his body, Siva from his forehead, Narayana from his heart and so on. As such doctrines are not known to have appeared in Indian Buddhism before the tenth century it seems probable that the versified edition is late. But a work with the title Ratna-karandaka- vyuha-siitra was translated into Chinese in 270 and the Karanda- vyuha is said to have been the first work translated into Tibetan*. 1 Mahayana-sutralankara, xix. 29. ' E,g. the Rashtra-pala-paripriccha edited in Sanskrit by Finot, Biblioth, Buddhica,. 1901. The Sanskrit text seems to agree with the Chinese version. The real number of sutras in the Ratnakuta seems to be 48, two being practicaUy the same but represented as uttered on different occasions. ' There is another somewhat simUar coUection of sutras in the Chinese Canon caUed Ta Tsi or Mahasannipata but unlike the Ratnakuta it seems to contain few weU-known or popular works. ' I know of these works only by Raj. Mitra's abstracts, Nepal, Bud, Lit, pp. 95 and 101. The prose text is said to have been published in Sanskrit at Calcutta, 1873. 58 THE MAHAYANA [ch. 13. The Karuna-pundarikai or Lotus of Compassion is mainly occupied with the description of an imaginary continent called Padmadhatu, its Buddha and its many splendours. It exists in Sanskrit and was translated into Chinese about 400 a.d. (Nanjio, No. 142). 14. The Mahavairocanabhisambhodhi caUed in Chinese Ta- jih-ching or Great Sun sutra should perhaps be mentioned as it is the principal scripture of the Chen-yen (Japanese Shingon) school. It is a late work of unknown origin. It was translated into Chinese in 724 a.d. but the Sanskrit text has not been found. There are a great number of other sutras which are important for the history of Hterature, although Httle attention is paid to them by Buddhists at the present day. Such are the Mahayanist version of the Mahaparinirvana recounting the death and burial of the Buddha and the Mahasannipata-sutra, which apparently includes the Suryagarbha and Candragarbha sutras. AU these works were translated into Chinese about 420 a.d. and must therefore be of respectable antiquity. Besides the sutras, there are many compositions styled Avadanas or pious legends^. These, though recognized by Mahayanists, do not as a rule contain expositions of the Sunyata and Dharma-kaya and are not sharply distinguished from the more imaginative of the Hinayanist scriptures*. But they introduce a multipHcity of Buddhas and Bodhisattvas and represent Sakyamuni as a superhuman worker of miracles. They correspond in many respects to the PaH Vinaya but teach right conduct not so much by precept as by edifying stories and, Hke most Mahayanist works they lay less stress upon monastic discipUne than on unselfish virtue exercised throughout successive existences. There are a dozen or more coUections of Avadanas of which the most important are the Mahavastu and the Divyavadana. The former* is an encyclopaedic work which contains inter alia a life of Sakyamuni. It describes itseU as 1 Raj. Mitra, Nepalese Buddhist Lit, pp. 285 ff. The Sanskrit text was published for the Buddhist Text Society, Calcutta, 1898. 2 Avadana is primarUy a great and glorious act : hence an account of such an act. ' The Avadana-^ataka (Peer, Annates du Music Chiimet, xvm) seems to be entirely Hinayanist. * Edited by Senart, 3 vols. 1882-1897. Windisch, Die Komposition des Mahd- vastu, 1909. Article "Mahavastu" in E,R,E, XX] MAHAYANIST SCRIPTURES 59 belonging to the Lokottaravadins, a section of the Aryamaha- sanghikas. The Lokottaravadins were an ancient sect, pre cursors of the Mahayana rather than a branch of it, and much of the Mahavastu is parallel to the PaU Canon and may have been composed a century or two before our era. But other parts seem to belong to the Gandharan period and the mention of Chinese and Himnish writing points to a much later date^. If it was originaUy a Vinaya treatise, it has been distended out of aU recognition by the addition of legends and anecdotes but it stiU retains a certain amount of matter found also in the Pali and Tibetan Vinayas. There were probably several recensions tn which successive additions were made to the original nucleus. One interpolation is the lengthy and important section called Dasabhiimika, describing the career of a Bodhisattva. It is the only part of the Mahavastu which can be caUed definitely Mahayanist. The rest of the work marks a transitional stage in doctrine, just as its language is neither Prakrit or Sanskrit but some ancient vernacular brought into partial conformity with Sanskrit grammar. No Chinese translation is known. The Divyavadana^ is a coUection of legends, part of which is known as the Asokavadana and gives an edifying Ufe of that pious monarch. This portion was translated into Chinese a.d. 317-420 and the work probably dates from the third century of our era. It is loosely constructed: considerable portions of it seem to be identical with the Vinaya of the Sarvastivadins and others with passages in the works of Asvaghosha. The Avadanas He on the borderland between scripture and pious Hterature which uses human argument and refers to scripture for its authority. Of this Hterature the Mahayanist church has a goodly coUection and the works ascribed to such doctors as Asvaghosha, Nagarjuna, Asanga and Vasubandhu hold a high place in general esteem. The Chinese Canon, places many of them in the Pitakas (especiaUy in the Abhidharma Pitaka) and not among the works of misceUaneous writers. The Mahayanist scriptures are stiU a Hving force. In Nepal the nine Dharmas receive superstitious homage rather than * So too do the words Horapathaka (astrologer), Ujjhebhaka ( ? Uzbek), Peli- yaksha (? FeUx). The word Yogacara (i. 120) may refer simply to the practice of Yoga and not to the school which bore this name. * Edited by CoweU and NeU, 1886. See Nanjio, 1344. 60 THE MAHAYANA [CH. intelligent study, but in Tibet and the Far East the Prajna- paramita, the Lotus and the sutras about Amitabha are in daUy use for pubhc worship and private reading. I have heard the first-named work as weU as the Leng-yen-ching expounded, that is, read aloud with an extempore paraphrase, to lay congrega tions in China, and the section of it caUed the Diamond Cutter is the book which is most commonly in the hands of religious Tibetans. The Lotus is the special scripture of the Nichiren sect in Japan but is universally respected. The twenty-fourth chapter which contains the praises of Avalokita is often printed separately. The Amitabha sutras take the place of the New Testament for the Jodo and Shin sects and copies of them may also be found in almost every monastery throughout China and Annam. The Suvarnaprabhasa is said to be speciaUy popular among the Mongols. I know Chinese Buddhists who read the Hua-yen (Avatamsaka) every day. Modem Japanese writers quote frequently from the Lankavatara and Kasyapa-parivarta but I have not met with any instance of these works being in popular use. I have mentioned already the obscurity surrounding the history of the Mahayanist Canon in India and it may seem to throw doubt on the authenticity of these scriptures. Unauthentic they certainly are in the sense that European criticism is not Ukely to accept as historical the discourses which they attribute to the Buddha and others, but there is no reason to doubt that they are treatises composed in India early in our era and repre senting the doctrines then prevalent. The reUgious pubhc of India has never felt any difficulty in accepting works of merit — and often only very moderate merit — as revelations, whether caUed Upanishads, Puranas, Sutras or what not. Only rarely have such works received any formal approbation, such as recognition by a council. Indeed it is rather in Ceylon, Burma, Tibet and China than in India itself that authoritative Hsts of scriptures have been compiled. The natural instinct of the Hindus was not to close the Canon but to leave it open for any additions which might be vouchsafed. Two sketches of an elastic Mahayanist Canon of this kind are preserved, one in the Sikshasamuccaya^ attributed to Santideva, who probably flourished in the seventh century, and ^ Edited by BendaU in Bibl. Buddhica. XX] MAHAYANIST SCRIPTURES 61 the other in a Uttle work caUed the Duration of the Law, reporting a discourse by an otherwise unknown Nandimitra, said to have Hved in Ceylon 800 years after the Buddha's death^. The former is a compendium of doctrine iUustrated by quotations from what the author regarded as scripture. He cites about a hundred Mahayanist sutras, refers to the Vinaya and Divya vadana but not apparently to the Abhidharma. He mentions no Tantras^ and not many Dharanis. The second work was translated by Hsiian Chuang and was therefore probably written before 600 a.d.^ Otherwise there is no external evidence for fixing its date. It represents Nandi mitra as explaining on his deathbed the steps taken by the Buddha to protect the True Law and in what works that Law is to be found. Like the Chinese Tripitaka it recognizes both Mahayanist and Hinayanist works, but evidently prefers the former and styles them coUectively Bodhisattva-Pitaka. It enumerates about fifty sutras by name, beginning with the Prajna-paramita, the Lotus and other weU-known texts. Then comes a Ust of works with titles ending in Samadhi, foUowed by others caUed Paripficcha* or questions. A new category seems to be formed by the Buddhavatamsaka-sutra with which the sutras about Amitabha's Paradise are associated. Then comes the Mahasannipata-sutra associated with works which may correspond to the Ratnakuta division of the Chinese Canon^. The writer adds that there are "hundreds of myriads of similar sutras classified in groups and categories." He mentions the Vinaya and Abhidharma without further particulars, whereas in describing the Hinayanist versions of these two Pitakas he gives many detaUs. The importance of this list Ues in the fact that it is Indian rather than in its date, for the earUest catalogue of the Chinese Tripitaka compiled about^ 510 is perhaps older and certainly ' Nanjio, No. 1466. For a leamed discussion of this work see Levi and Chavannes in J,A, 1916, Nos. i and n. ^ It is not Ukely that the Tathagatha-guhya-sutra which it quotes is the same as the Tantra with a similar name analysed by Rajendralal Mitra. » Watters, J,E,A.S. 1888, p. 331 says there seems to have been an earlier translation. * Many works with this title wUl be found in Nanjio. ^ But the Chinese title seems rather to represent Ratnarasi. ° See Nanjio, pp. xiU-xvii. 62 THE MAHAYANA [CH. xx ampler. But if the catalogue stood alone, it might be hard to say how far the selection of works in it was due to Chinese taste. But taking the Indian and Chinese evidence together, it is clear that in the sixth century Indian Mahayanists {a) tolerated Hinayanist scriptures while preferring their own, (6) made Uttle use of the Vinaya or Abhidharma for argument or edification, though the former was very important as a code, (c) recognized extremely numerous sutras, grouped in various classes such as Mahasannipata and Buddhavatamsaka, {d) and did not use works called Tantras. Probably much the same is true of the fourth century and even earlier, for Asanga in one work^ quotes both Maha- and Hinayanist scriptures and among the former cites by name seventeen sutras, including one caUed Paripriccha or questions. ^ MahaySna-sfitralankara. See Levi's introduction, p. 14. The "Questions" sutra is Brahma-paripriccha. CHAPTER XXI CHRONOLOGY OF THE MAHAYANA In the previous chapters I have enumerated some features of Mahayanism, such as the worship of Bodhisattvas leading to mythology, the deification of Buddhas, entaiUng a theology as compHcated as the Christian creeds, the combination of meta physics with rehgion, and the rise of new scriptures consecrating all these innovations. I will now essay the more difficult task of arranging these phenomena in some sort of chronological setting. The voluminous Chinese Hterature concerning Buddhism offers valuable assistance, for the Chinese, unlike the Hindus, have a natural disposition to write simple narratives recording facts and dates. But they are diarists and chroniclers rather than historians. The Chinese pilgrims to India give a good account of their itinerary and experiences, but they have Httle idea of investigating and arranging past events and merely recount traditions connected with the places which they visited. In spite of this their statements have considerable historical value and on the whole harmonize with the literary and archaeological data furnished by India. The Tibetan Lama Taranatha who completed his History of Indian Buddhism^ in 1608 is a less satisfactory authority. He merits attention but also scepticism and caution. His work is a compilation but is not to be despised on that ground, for the Tibetan translations of Sanskrit works offer a rich mine of information about the history of the Mahayana. Unfortunately few of these works take the historical point of view and Tara- natha's own method is as uncritical as his materials. Dire confusion prevails as to chronology and even as to names ^, so » Translated by Schiefner, 1869. Taranatha informs us (p. 281) that his chief authorities were the history of Kshemendrabhadra, the Buddhapurana of Indra- datta and Bhataghati's history of the succession of the Aoaryas. * The Tibetans generaUy translate instead of transUterating Indian names. It is as it an EngUsh history of Greece were to speak of Leader of the People instead of Agesilaus. 64 THE MAHAYANA [ch that the work is almost useless as a connected account, though it contains many interesting details. Two epochs are of special importance for the development of later Indian Buddhism, that of Kanishka and that of Vasu bandhu and his brother Asanga. The reader may expect me to discuss at length the date of Kanishka's accession, but I do not propose to do so for it may be hoped that in the next few years archaeological research in India or Central Asia wiU fix the chronology of the Kushans and meanwhile it is waste of time to argue about probabiUties or at any rate it can be done profitably only in special articles. At present the majority of scholars place his accession at about 78 a.d., others put it back to 58 B.C. and arrange the Kushan kings in a different order^, while stiU others ^ think that he did not come to the throne until the second century was weU advanced. The evidence of art, particidarly of numismatics, indicates that Kanishka reigned towards the end of his dynasty rather than at the beginning, but the use of Greek on his coins and his traditional connection with the beginnings of the Mahayana are arguments against a very late date. If the date 78 a.d. is accepted, the conversion of the Yiieh-chih to Buddhism and its diffusion in Central Asia cannot have been the work of Kanishka, for Buddhism began to reach China by land about the time of the Christian era^. There is however no reason to assume that they were his work. Kanishka, like Constantino, probably favoured a winning cause, and Buddhism may have been graduaUy making its way among the Kushans and their neighbours for a couple of centuries before his time. In any case, however important his reign may 1 They place Kanishka, Vasishka, Huvishka and Vasudeva before Kadphises I and Kadphises II. ^ E,g, Stael Holstein who also thinks that Kanishka's tribe should be caUed Kusha not Kushan. Vincent Smith in his latest work (Oxford History of India, p. 130) gives 120 a.d^ as the most probable date. ' My chief difficulty in accepting 78-123 a.d. as the reign of Kanishka is that the Chinese Annals record the doings of Pan Ch'ao between 73 and 102 in Central Asia, with which region Kanishka is beUeved to have had relations, and yet do not mention his name. This silence makes it primd facie probable that he Uved either before or after Pan Ch'ao's career. The catalogues of the Chinese Tripitaka state that An-Shih-Kao (148-170 a.d.) translated the Margabhumi-sutra of Sangharaksha, who was the chaplain of Kanishka. But this unfortunately proves nothing except that Kanishka cannot have been very late. The work is not a scripture for whose recognition some lapse of time must be postulated. An-Shih-kao, who came from the west, may very well have translated a recent and popular treatise. XXI] CHRONOLOGY OF THE MAHAYANA 65 have been for the Buddhist Church, I do not thinlt that the history of the Mahayana should be made to depend on his date. Chinese translations, supported by other evidence, indicate that the Mahayanist movement had begun about the time of our era. If it is proved that Kanishka Uved considerably later, we should not argue that Mahayanism is later than was supposed but rather that his relation towards it has been misunderstood^. The date of Vasubandhu has also been much discussed and scholars have generaUy placed him in the fourth or fifth century but Peri^ appears to have proved that he Hved from about 280 to 360 A.D. and I shaU adopt this view. This chronology makes a reasonable setting for the development of Buddhism. If Kanishka reigned from about 78 to 123 a.d. or even later, there is no difficulty in supposing that Asvaghosha flourished in his reign and was foUowed by Nagarjuna. The coUapse of the Kushan Empire was probably accompanied by raids from Iranian tribes, for Persian influence appears to have been strong in India during the confused interval between the Kushans and Guptas (225-320). The latter inaugurated the revival of Hinduism but stiU showed favour to individual Buddhists, and we know from Fa-Hsien that Buddhism was fairly flourishing during his visit to India (399-415). There is nothing improbable in supposing that Vasubandhu, who is stated to have Hved at Court, was patronized by the early Guptas. The blank in Buddhist history which follows his career can be explained first by the progress of Hinduism at the expense of Buddhism and secondly by the invasions of the Huns. The Chinese pilgrim Sung-Yiin has left us an account of India in this distressful period and for the seventh century the works of Hsiian Chuang and I-Ching give copious information. In investigating the beginnings of the Mahayana we may start from the epoch of Asoka, who is regarded by tradition as the patron and consoHdator of the Hinayanist Church. And the tradition seems on the whole correct: the united evidence of ' In this connection we may remember Taranatha's statement that Kanishka's CounoU put an end to dissentions which had lasted about a century. But he also states that it was after the CouncU that Mahayanist texts began to appear. If Kanishka flourished about 50 a.d. this would fit in with Taranatha's statements and what we know of the history of Buddhism. * B.E.F.E.O. 1911, 339-390. Satifchandra Vidyabhushana arrived at the same conclusion in J.A.S.B. 1905, p. 227. «. rr. 5 66 THE MAHAYANA [ch. texts and inscriptions goes to show that the Buddhists of Asoka's time held the chief doctrines subsequently professed by the Sinhalese Church and did not hold the other set of doctrines known as Mahayanist. That these latter are posterior in time is practicaUy admitted by the books that teach them, for they are constantly described as the crown and completion of a pro gressive revelation. Thus the Lotus ^ illustrates the evolution of doctrine by a story which curiously resembles the parable of the prodigal son except that the returned penitent does not recognize his father, who proceeds to reveal graduaUy his name and position, keeping back the fuU truth to the last. SimUarly it is held in the Far East that there were five periods in Sakya muni's teaching which after passing through the stage of the Hinayana culminated in the Prajna-paramita and Amitabha sutras shortly before his death. Such statements admit the historical priority of the Hinayana: it is rudimentary (that is early) truth which needs completion and expansion. Many critics demur to the assumption that primitive Buddhism was a system of ethics purged of superstition and mythology. And in a way they are right. Could we get hold of a primitive Buddhist, we should probably find that miracles, magic, and superhuman beings played a large part in his mind and that the Buddha did not appear to him as what we call a human teacher. In that sense the germs of the Mahayana existed in the Ufetime of Gotama. But the difference between early and later Buddhism lies in this, that the deities who surround the Buddha in the PaH Pitakas are mere accessories: his teaching would not be affected if they were aU removed. But the Bodhi sattvas in the Lotus or the Sutra of the Happy Land have a doctrinal significance. Though in India old ideas persist with unusual vitaUty, stUl even there they can Hve only if they either develop or gather round them new accretions. As one of the religions of India, Buddhism was sensitive to the general movement of Indian thought, or rather it was a part of that movement. We see as clearly in Buddhist as in non-Buddhist India that there was a tendency to construct philosophic systems and another tendency to create deities satisfying to the emotions as weU as to the inteUect and yet another tendency to compose new scriptures. But apart ^ Chap. IV. XXI] CHRONOLOGY OF THE MAHAYANA 67 from this parallel development, it becomes clear after the Christian era that Buddhism is becoming surrounded by Hindu ism. The influence is not indeed one-sided: there is interdepen dence and interpenetration but the net result is that the general Indian features of each reUgious period overpower the speciaUy Buddhist features and in the end we find that while Hinduism has only been profoundly modified Buddhism has vanished. If we examine the PaU Pitakas, including the heresies mentioned in the Kathavatthu, we find that they contain the germs of many Mahayanist ideas. Thus side by side with the human portrait of the Buddha there is the doctrine that he is one in a series of supernatural teachers, each with the same Ufe-history, and this Ufe is connected with the whole course of nature, as is shown by the sympathetic earthquakes which mark its crises. His birth is supernatural and had he wiUed it he could have Hved untU the end of the present Kalpa^- So, too, the nature of a Buddha when he is released from form, that is after death, is deep and unfathomable as the ocean^. The Katha vatthu condemns the ideas (thus showing that they existed) that Buddhas are bom in aU quarters of the universe, that the Buddha was superhuman in the ordinary affairs of Ufe, that he was not reaUy bom in the world of men and that he did not preach the Law himseU. These last two heresies are attributed by the commentary to the Vetulyakas who are said to have beUeved that he remained in the Tushita heaven and sent a phantom to preach on earth. Here we have the rudiments of the doctrine afterwards systematized under the name of the three bodies of Buddha. SimUarly though Nirvana is regarded as primarily an ethical state, the PaU Canon contains the expression Nirvana- dhatu and the idea* that Nirvana is a sphere or realm {dyatanam) which transcends the transitory world and in which such antitheses are coming and going, birth and death, cease to exist. This foreshadows the doctrine of Bhuta-tathata and we seem to hear a prelude to the dialectic of Nagarjuna when the Katha vatthu discusses whether Sunnata or the void is predicable of the Skandhas and when it condemns the views that anything now existing existed in the past: and that knowledge of the present is possible (whereas the moment anything is known it ' Mahaparinib. Sut. ni. ^ Majj. NUr. 72. ' Udana, vni. 1^. 6—2 68 THE MAHAYANA [ch. is reaUy past). The Kathavatthu also condemns the proposition that a Bodhisattva can be reborn in realms of woe or faU into error, and this proposition hints that the career of a Bodhisattva was considered of general interest. The Mahayana grows out of the Hinayana and in many respects the Hinayana passes into it and is preserved unchanged. It is true that in reading the Lotus we wonder how this marvel lous cosmic vision can represent itseUas the teaching of Gotama, but the Buddhacarita of Asvaghosha, though embelHshed with literarymythology, hardly advances in doctrine beyond the Pali sutras describing the marvels of the Buddha's nativity ^ and the greater part of Nagarjuna's Friendly Epistle, which purports to contain an epitome of the faith, is in phraseology as weU as thought perfectly in harmony with the PaU Canon. Whence comes this difference of tone in works accepted by the same school? One difficulty of the historian who essays to account for the later phases of Buddhism is to apportion duly the influence of Indian and foreign elements. On the one hand, the Mahayana, whether we call it a development or perversion, is a product of Indian thought. To explain its trinities, its saviours, its doctrine of self sacrifice it is not necessary to seek abroad. New schools, anxious to claim continuity and antiquity, gladly retained as much of the old doctrine as they could. But on the other hand, Indian Buddhism came into contact with foreign, especiaUy Iranian, ideas and undoubtedly assimilated some of them. From time to time I have drawn attention to such cases in this work, but as a rule the foreign ideas are so thoroughly mastered and indianized that they cease to be obvious. They merely open up to Indian thought a new path wherein it can move in its own way. In the period foUowing Asoka's death Buddhism suffered a temporary ecUpse. Pushyamitra who in 184 B.C. overthrew the Mauryas and estabUshed the Sunga dynasty was a patron of the Brahmans. Taranatha describes him^ as a ferocious persecutor, and the Divyavadana supports the story. But the persecution, if it reaUy occurred, was probably local and did not seriously check the spread of Buddhism, which before the time of Kanishka had extended northwards to Bactria and Kashmir. The latter territory became the special home of the Sarvastivadins. It was ' Aooariyabbhutasuttam. Majj. Nik. 123. ^ Chap. xvi. XXI] CHRONOLOGY OF THE MAHAYANA 69 in the reign of Pushyamitra that the Graeco-Bactrian king Menander or MiUnda invaded India (155-3 B.C.) and there were many other invasions and settlements of tribes coming from the north-west and variously described as Sakas, Pahlavas, Parthians and Yavanas, culminating in the conquests of the Kushans. The whole period was disturbed and confused but some general statements can be made with considerable confidence. From about 300 B.C. to 100 a.d. we find inscriptions, buildings and statues testifying to the piety of Buddhist and Jain donors but hardly any indications of a similar liberaUty to Brahmans. In the second and third centuries a.d. grants of land to Brahmans and their temples begin to be recorded and in the fourth century (that is with the rise of the Gupta Dynasty) such grants become frequent. These facts can hardly be inter preted otherwise than as meaning that from 300 B.C. to 100 a.d. the upper classes of India favoured Buddhism and Jainism and did not favour the Brahmans in the same way or to the same extent. But it must be remembered that the rehgion of the Brahmans continued throughout this period and produced a copious Hterature, and also that the absence of works of art may be due to the fact that their worship was performed in sacrfficial enclosures and that they had not yet begun to use temples and statues. After the first century a.d. we have first a gradual and then a rapid rise in Brahmanic influence. Inscrip tions as well as books indicate that a Hnguistic change occurred in the same period. At first popular dialects were regarded as sufficiently dignified and current to be the medium for both scripture and official records. Sanskrit remained a thing apart — the pecuhar possession of the Brahman literati. Then the popular language was sanskritized, the rules of Sanskrit grammar being accepted as the standard to which it ought to conform, though perfect conformity was impracticable. In much the same way the modem Greeks try to bring Romaic into Une with classical Greek. FinaUy Sanskrit was recognized as the proper language for Hterature, govemment and rehgion. The earUest inscriptions^ in correct Sanskrit seem to date from the second century a.d. Further, the invaders who entered India from the ' That of Rudradaman at Gimar, dated 72 in the Saka Era, has hitherto been considered the oldest, but it is now said that one discovered at Isapur near Muttra is older. See J.B.A.S 1912, p. 114. 70 THE MAHAYANA [ch. north-west favoured Buddhism on the whole. Coins indicate that some of them worshipped Siva^ but the number and beauty of Buddhist monuments erected under their rule can hardly be interpreted except as a sign of their patronage. And their con version was natural for they had no strong reUgious convictions of their own and the Brahmans with their pride of caste shrank from foreigners. But Buddhism had no prejudice of race or class : it was animated by a missionary spirit and it was probably the stronger creed at this period. It not only met the invaders on their entry into India but it sent missionaries to them in Bactria and Afghanistan, so that to some extent they brought Buddhism with them. But it was a Buddhism combined with the most varied elements. HeUenic art and reUgion had made the figures of ApoUo, Herakles and HeUos famiUar in Bactria, and both Bactria and northern India were in touch with Zoroastrians. The mixed cults of these borderlands readily professed allegiance to the Buddha but, not understanding Indian ideas, simply made him into a deity and having done this were not Hkely to repudiate other Indian deities. Thus tn its outward form the Buddhism of the invaders tended to be a compound of Indian, Greek and Persian ideas in which Sun worship played a large part, for not only Indian myths, but ApoUo and HeUos and the Persian Mithra aU entered into it. Persian infiuence in art is discernible as early as the architecture of Asoka : in doctrine it has something to do with such figures as Vairocana and Amitabha. Graeco-Roman influence also was powerful in art and through art affected rehgion. In Asoka's time Hkenesses of the Buddha were unknown and the adoration of images, if not entirely due to the art of Gandhara, was at least encouraged by it. But though coins and sculpture bring clearly before us a medley of deities corresponding to a medley of human races, they do not help us much in tracing the growth of thought, phases of which are preserved in a Hterature sufficiently copious though the record sometimes fails at the points of transition where it would be of most interest. It is natural that sacred books should record accepted results rather than tentative innovations and even disguise the latter. But we can fix a few dates which enable us to judge what shape Buddhism was taking E.g. Kadphises II and Vasudeva. XXI] CHRONOLOGY OF THE MAHAYANA 71 about the time of the Christian era. The Tibetan historian Taranatha is not of much help, for his chronology is most confused, but still he definitely connects the appearance of Mahayanist texts with the reign of Kanishka and the period immediately foUowing it^ and regards them as a new pheno menon. Greater assistance is furnished by the Chinese translators, whose dates are known with some exactitude. Thus the earUest Buddhist work rendered into Chinese is said to be the sutra of forty -two sections, translated by Kasyapa Matanga in 67 a.d. It consists of extracts or resumes of the Buddha's teaching mostly prefaced by the words "The Buddha said," doubtless in imitation of the Confucian Analects where the introductory formula "The master said" plays a similar part. Its ideas and precepts are Hinayanist^: the Arhat is held up as the ideal and in a remarkable passage* where the degrees of sanctity are graded and compared no mention is made of ^odhisattvas. This first translation was foUowed by a long series of others, principaUy from the Siitra-Pitaka, for very Uttle of the Vinaya was translated before the fifth century. A great number of Hinayanist sutras were translated before 300 a.d. but very few after 450. On the other hand portions of the sutra about Amida's Paradise, of the Prajna-paramita, and of the Avatamsaka were translated about 150 a.d. and translations of the Lotus and LaHta-vistara appeared about 300. Great caution is necessary in using these data and the circumstances of China as well as of India must be taken into account. If translations of the Vinaya and complete coUections of sutras are late in appearing, it does not follow that the corresponding Indian texts are late, for the need of the Vinaya was not felt until monasteries began to spring up. Most of the translations made before the fifth century are extracts and of indifferent workmanship. Some are retained in the Chinese Tripitaka but are superseded by later versions. But however inaccurate and incomplete these older translations may be, if any of them can be identified with a part of an extant Sanskrit * Chaps, xii, xin. ' The last section (42) as translated by Teitaro Suzuki in the Sermons of a Buddhist Abbot may seem an exception, for it contains suoh statements as " I con sider the doctrine of sameness as the absolute ground of reaUty." But the transla tion seems to me doubtful. s Sec. 11. 72 THE MAHAYANA [ch. work it foUows that at least that part of the work and the doctrines contained in it were current in India or Central Asia some time before the translation was made. Applying this principle we may conclude that the Hinayana and Mahayana were flourishing side by side in India and Central Asia in the first century a.d. and that the Happy Land sutras and portions of the Prajiia-paramita aUeady existed. From that time on wards Mahayanist Hterature as represented by Chinese transla tions steadily increases, and after 400 a.d. Hinayanist Hterature decHnes, with two exceptions, the Vinaya and the Abhidharma books of the Sarvastivadins. The Vinaya was evidently regarded as a rule of Hfe independent of theology, but it is remarkable that Hsiian Chuang after his return from India in 645 should have thought it worth whUe to translate the philosophy of the Sarvastivadins. Other considerations render this chronology probable. Two conspicuous features of the Mahayana are the worship of Bodhisattvas and idealist phUosophy. These are obviously parallel to the worship of Siva and Vishnu, and to the rise of the Vedanta. Now the worship of these deities was probably not prevalent before 300 B.C., for they are almost unknown to the PaU Pitakas, and it was fuUy developed about the time of the Bhagavad-gita which perhaps assumed its present form a Httle before the Christian era. Not only is the combination of devotion and metaphysics found in this work similar to the tone of many Mahayanist sutras but the manifestation of Krishna in his divine form is Uke the transformation scenes of the Lotus^. The chief moral principle of the Bhagavad-gita is substantially the same as that prescribed for Bodhisattvas. It teaches that action is superior to inaction, but that action should be whoUy disinterested and not directed to any seffish object. This is precisely the attitude of the Bodhisattva who avoids the inaction of those who are engrossed in self -culture as much as the pursuit of wealth or pleasure. Both the Gita and Mahayanist treatises lay stress on faith. He who thinks on Krishna when dying goes to Krishna^ just as he who thinks on Amitabha goes 1 Just as all gods and worlds are seen within Krishna's body, so we are told in the Karanda-vyuha (which is however a later work) that in the pores of Avalokita's skin are woods and mountains where dwell saints and gods. 2 Bhag. G. VIU. 5. XXI] CHRONOLOGY OF THE MAHAYANA 73 to the Happy Land and the" idea is not unknown to the Pali texts, for it finds complete expression in the story of Mattha- kundaU^. The idea of a benevolent deity to be worshipped with devo tion and faith and not with ceremonies is strange to old Buddhism and old Brahmanism aUke. It was a popular idea which became so strong that neither priests nor Bhikshus could ignore it and in its ultimate result it is hard to say whether Buddhist or Brahmanic elements are more prominent. Both Avalokita and Krishna are Devas. The former has the beauty of hoUness and the strength which it gives, but also the weakness of a somewhat abstract figure: the latter is very personal and springs from the heart of India but to those who are not Hindus seems wanting in purity and simphcity. The divine character of both figures is due to Brahmanism rather than Buddhism, but the new form of worship which laid stress on a frame of mind rather than on ceremonial and the idea of Avataras or the periodic appearance of superhuman saviours and teachers indi cate the influence of Buddhism on Brahmanism. There is a similar parallel between the newer Buddhist phUosophy and the Vedantist school represented by Sankara, and Indian critics detected it. Sankara was called a Pracchanna- bauddha or crjrpto-buddhist by his theological opponents^ and the resemblance between the two systems in thought, if not in word, is striking. Both distinguish relative and absolute truth : for both the relative truth is practically theism, for both absolute truth is beyond description and whether it is called Brahman, Dharma-kaya or Sunyata is not equivalent to God in the Christian or Mohammedan sense. Just as for the Vedantist there exist in the light of the highest knowledge neither a personal God nor an individual soul, so the Madhyamika Sutra can declare that the Buddha does not reaUy exist. The Maha yanist philosophers do not use the word Maya but they state the same theory in a more subjective form by ascribing the appearance of the phenomenal world to ignorance, a nomen- ' Commentary on Dhammapada, P.T.S. edition, pp. 25 ff. especiaUy p. 33. * See Ramanuja, Sribhashya, li. 2, 27 and Padma-Purana uttarakanda 43 (quoted by Snhtankar in Vienna Oriental Journ, vol. xxii. 1908). Mayavadam asacchastram praechannam bauddham ucyate. The Madhvas were speciaUy bitter in their denunciation of Sankara. 74 THE MAHAYANA [ch. clature which is derived from the Buddha's phrase, "From ignorance come the Sankharas." Here, as elsewhere, Buddhist and Brahmanic ideas acted and reacted in such complex interrelations that it is hard to say which has borrowed from the other. As to dates, the older Upanishads which contain the foundations but not the complete edifice of Vedantism, seem a Httle earHer than the Buddha. Now we know that within the Vedantist school there were divergences of opinion which later received classic expression in the hands of Sankara and Ramanuja. The latter rejected the doctrines of Maya and of the difference between relative and absolute truth. The germs of both schools are to be found in the Upanishads but it seems probable that the ideas of Sankara were originally worked out among Buddhists rather than among Brahmans and were rightly described by their opponents as disguised Buddhism. As early as 520 a.d. Bodhi dharma preached in China a doctrine which is practicaUy the same as the Advaita. The earUest known work in which the theory of Maya and the Advaita philosophy are clearly formulated is the metrical treatise known as the Karika of Gaudapada. This name was borne by the teacher of Sankara's teacher, who must have lived about 700 A.D., but the high position accorded to the work, which is usuaUy printed with the Mandukya Upanishad and is practically regarded as^ a part of it, make an earHer date probable. Both in language and thought it bears a striking resemblance to Buddhist writings of the Madhyamika school and also contains many ideas and similes which reappear in the works of Sankara^. On the other hand the Lankavatara Siitra which was translated into Chinese in 513 and therefore can hardly have been composed later than 450, is conscious that its doctrines resemble Brahmanic philosophy, for an interlocutor ' Or as itseU forming four separate Upanishads. For other arguments in favour of an early date see WaUeser, Alterer Veddnta, pp. 14 ff. He states that the Karika is quoted in the Tibetan translations of Bhavaviveka's Tarkajvdla, Bhavaviveka was certainly anterior to the travels of Hsiian Chuang and perhaps was much earUer. But if he died about 600 a.d. a work quoted by him can hardly have been later than 550 and may be much earUer. But see also Jacobi iu J.A,O.S, April, 1913, p. 51. " For the resemblances to Nagarjuna see J.R.A.S. 1910, pp. 136 ff. EspeciaUy remarkable are ii. 32 na nirodho na cotpattir, etc., and iv. 59 and the whole argu ment that causation is impossible. Noticeable too is the use of Buddhist terms Uke upaya, nirvaija, buddha and adibuddha, though not always in the Buddhist sense. XXI] CHRONOLOGY OF THE MAHAYANA 75 objects that the language used in it by the Buddha about the Tathagatha-garbha is very Hke the Brahmanic doctrine of the Atman. To which the Buddha repHes that his language is a concession to those who cannot stomach the doctrine of the negation of reahty in aU its austerity. Some of the best known verses of Gaudapada compare the world of appearance to the apparent circle of fire produced by whirhng a Ughted torch. This striking image occurs first in the Maitrayana Upanishad (VT. 24), which shows other indications of an acquaintance with Buddhism, and also in the Lankavatara Siitra. A real affinity unites the doctrine of Sankara to the teaching of Gotama himseU. That teaching as presented in the PaH Pitakas is marked by its negative and deUberately circum scribed character. Its rule is silence when strict accuracy of expression is impossible, whereas later philosophy does not shrink from phrases which are suggestive, if not exact. Gotama refuses to admit that the human soul is a fixed entity or Atman, but he does not condemn (though he also does not discuss) the idea that the whole world of change and becoming, including human souls, is the expression or disguise of some one ineffable principle. He teaches too that the human mind can grow until it develops new faculties and powers and becomes the Buddha mind, which sees the whole chain of births, the order of the world, and the reahty of emancipation. As the object of the whole system is practical. Nirvana is always regarded as a terminus ad quem or an escape (nissaranam) from this transitory world, and this view is more accurate as well as more edifying than the view which treats Brahman or Sunyata as the origin of the universe. When the Vedanta teaches that this changing troubled world is merely the disguise of that unchanging and untroubled state into which saints can pass, it is, I beheve, foUowing Gotama's thought, but giving it an expression which he would have considered imperfect. CHAPTER XXII FROM KANISHKA TO VASUBANDHU Tradition, as mentioned above, connects the rise of the Mahayana with the reign of Kanishka. Materials for forming a picture of Indian Ufe under his rule are not plentiful but it was clearly an age of fusion. His hereditary dominions were ample and he had no need to spend his reign in conquests, but he probably subdued Kashmir as well as Khotan, Yarkand and Kashgar^. Hostages from one of these states were sent to reside in India and all accounts agree that they were treated with generosity and that their sojourn improved the relations of Kanishka with the northern tribes. His capital was Purushapura or Peshawar, and the locaUty, Hke many other features of his reign, indicates a tendency to amalgamate India with Persia and Central Asia. It was embeUished with masterpieces of Gandharan sculpture and its chief ornament was a great stupa built by the king for the reception of the reUcs of the Buddha which he coUected. This building is described by several Chinese pilgrims^ and its proportions, though variously stated, were sufficient to render it celebrated in all the Buddhist world. It is said to have been several times bumt, and rebuilt, but so soUd a structure can hardly have been totally destroyed by fire and the greater part of the monument discovered in 1908 probably dates from the time of Kanishka. The base is a square measuring 285 feet on each side, with massive towers at the corners, and on each of the four faces projections bearing stair- * The uncertainty as to the date of Kanishka naturaUy makes it uncertain whether he was the hero of these conquests. Kashmir was certainly included in the dominions of the Kushans and was a favourite residence of Kanishka. About 90 A.D. a Kushan king attacked Central Asia but was repulsed by the Chinese general Pan-Ch'ao. Later, after the death of Pan-Ch'ao (perhaps about 103 a.d.), he renewed the attempt and conquered Kashgar, Yarkand and Khotan. See Vincent Smith, Early History of India, Srd ed. pp. 253 ff. 2 See Fa-Hsien, ed. Legge, p. 33, B.E.F.E.O. 1903 (Sung Yiin), pp. 420 ff. Watters, Yiian Chwang, i. pp. 204 ff. J.R.A.S. 1909, p. 1056, 1912, p. 114. For the general structure of these stupas see Foucher, L'art Grico-Bouddhique du Gandhara, pp. 45 ff. CH. xxn] FROM KANISHKA TO VASUBANDHU 77 cases. The sides were ornamented with stucco figures of the Buddha and according to the Chinese pilgrims the super structure was crowned with an iron piUar on which were set t-rt^enty-five gilded disks. Inside was found a metal casket, still containing the sacred bones, and bearing an inscription which presents two points of great interest. Firstly it mentions "Agisala the overseer of works at Kanishka's vihara," that is, probably Agesilaus, a foreigner in the king's service. Secondly it states that the casket was made "for the acceptance of the teachers of the Sarvastivadin sect^," and the idea that Kanishka was the special patron of the Mahayana must be reconsidered in the Hght of this statement. Legends ascribe Kanishka's fervour for the Buddhist faith not to education but to conversion. His coinage, of which abundant specimens have been preserved, confirms this for it presents images of Greek, Persian, Indian and perhaps Baby lonian deities showing how varied was the mythology which may have mingled with Gandharan Buddhism. The coins bearing figures of the Buddha are not numerous and, as he . undoubtedly left behind him the reputation of a pious Buddhist, it is probable that they were struck late in his reign and represent his last rehgious phase 2. Hsiian Chuang* repeats some legends which relate that he was originaUy anti-Buddhist, and that after his conversion he summoned a council and built a stupa. The substance of these legends is probable. Kanishka as a barbarian but docile conqueror was Hkely to adopt Buddhism if he wished to keep abreast of the thought and civiUsation of his subjects, for at that time it undoubtedly inspired the inteUect and art of north-western India. Both as a statesman and as an enquirer after truth he would wish to promote harmony and stop sectarian squabbles. His action resembles that of Con stantino who after his conversion to Christianity proceeded to summon the Council of Nicaea in order to stop the dissensions of the Church and settle what were the tenets of the rehgion which he had embraced, a point about which both he and 1 J.R.A.S. 1909, p. 1058. "Acaryanam Sarvastivadinam pratigrahe." " SimUarly Harsha became a Buddhist late in lite. ' Watters, vol. i. p. 203. He places Kanishka's accession 400 years after the death of the Buddha, which is one of the arguments for supposing Kanishka to have reigned about 50 B.C., but in another passage (Watters, i. 222, 224) he appears to place it 500 years after the death. 78 THE MAHAYANA [ch. Kanishka seem to ha\'e felt some uncertainty. Our knowledge of Kanishka's Council depends chiefly on the traditions reported by Hsiian Chuang ^ which present many difficulties. He teUs us that the king, acting in consultation with Parsva, issued summonses to aU the learned doctors of his realm. They came in such crowds that a severe test was imposed and only 499 Arhats were selected. There was some discussion as to the place of meeting but finaUy Kashmir ^ was selected and the king built a monastery for the Brethren. When the Council met, there arose a question as to whether Vasumitra (who is not further described) should be admitted seeing that he was not an Arhat but aspired to the career of a Bodhisattva. But owing to the interposition of spirits he was not only admitted but made president. The texts of the Tripitaka were coUected and the Council "composed 100,000 stanzas of Upadesa Sastras explanatory of the canonical siitras, 100,000 stanzas of Vinaya-vibhasha Sastras explanatory of the Vinaya and 100,000 of Abhidharma-vibhasha Sastras explanatory of the Abhidharma. For this exposition of the Tripitaka all learning from remote antiquity was thoroughly examined; the general sense and the terse language (of the Buddhist scriptures) was again and again made clear and dis tinct, and learning was widely diffused for the safe-guiding of disciples. King Kanishka caused the treatises when finished to be written out on copper plates and enclosed these in stone boxes which he deposited in a tope made for the purpose. He then ordered spirits to keep and guard the texts and not to aUow any to be taken out of the country by heretics ; those who wished to study them could do so in the country. When leaving to return to his own country, Kanishka renewed Asoka's gift of all Kashmir to the Buddhist Church*." Paramartha (499-569 A.D.) in his Life of Vasubandhu^ gives an account of a council generally considered to be the same as 1 Watters, vol. i. 270-1. 2 But Taranatha says some authorities held that it met at Jalandhara. Some Chinese works say it was held at Kandahar. ' Watters, l.c. ' Translated by Takakusu in T'oung Pao, 1904, pp. 269 ff. Paramartha was a native of Ujjain who arrived at Nanking in 548 and made many translations, but it is quite possible that this Ufe of Vasubandhu is not a translation but original notes of his own. xxn] FROM KANISHKA TO VASUBANDHU 79 that described by Hsuan Chuang, though the differences in the two versions are considerable. He says that about five hundred yearsi after the Buddha's death {i.e. between 87 b.c. and 13 a.d. if the Buddha died 487 b.c.) an Indian Arhat called Katyayani- putra, who was a monk of the Sarvastivadin school, went to Kipin or Kashmir. There with 500 other Arhats and 500 Bodhi sattvas he coUected the Abhidharma of the Sarvastivadins and arranged it in eight books caUed Ka-lan-ta (Sanskrit Grantha) or Kan-tu (PaH Gantho). This compilation was also called Jfiana-prasthana. He then made a proclamation inviting all who had heard the Buddha preach to commrmicate what they remembered. Many spirits responded and contributed their reminiscences which were examined by the Council and, when they did not contradict the sutras and the Vinaya, were accepted, but otherwise were rejected. The selected pieces were grouped according to their subject-matter. Those about wisdom formed the Prajiia Grantha, and those about meditation the Dhyana Grantha and so on. After finishing the eight books they pro ceeded to the composition of a commentary or Vibhasha and invited the assistance of Asvaghosha. When he came to Kashmir, Katyayani-putra expounded the eight books to him and Asvaghosha put them into Uterary form. At the end of twelve years the composition of the commentary was finished. It consisted of 1,000,000 verses.... Katyayani-putra set up a stone inscribed with this proclamation. "Those who hereafter leam this law must not go out of Kashmir. No sentence of the eight books, or of the Vibhasha must pass out of the land, lest other schools or the Mahayana should corrupt the true law." This proclamation was reported to the king who approved it. The sages of Kashmir had power over demons and set them to guard the entrance to the country, but we are told that anyone desirous of learning the law could come to Kashmir and was in no way interrupted. There foUows a story teUing how, despite this prohibition, a native of Ayodhya succeeded in learning the law in Kashmir 1 Chinese expressions like "in the five hundred years after the Buddha's death" probably mean the period 400-500 of the era commencing with the Buddha's death and not the period 500-600. The period 1-100 is "the one hundred years," 1 01-200 "the two hundred years" and so on. See B.E.F.E.O. 1911, 356. But it must be remembered that the date of the Buddha's death is not yet certain. The latest theory (Vincent Smith, 1919) places it in 554 e.o. 80 THE MAHAYANA [CH. and subsequently teaching it in his native land. Paramartha's account seems exaggerated, whereas the prohibition described by Hsiian Chuang is inteUigible. It was forbidden to take the official copies of the law out of Kashmir, lest heretics should tamper with them. Taranatha^ gives a singularly confused account of the meeting, which he expressly calls the third councU, but makes some important statements about it. He says that it put an end to the dissensions which had been distracting the Buddhist Church for nearly a century and that it recognized all the eighteen sects as holding the true doctrine: that it put the Vinaya in writing as well as such parts of the Sutrapitaka and Abhidharma as were still unwritten and corrected those which already existed as written texts : that all kinds of Mahayanist writings appeared at this time but that the Sravakas raised no opposition. It is hard to say how much history can be extracted from these vague and discrepant stories. They seem to refer to one assembly regarded (at least in Tibet) as the third council of the Church and held under Kanishka four or five hundred years ^ after the Buddha's death. As to what happened at the council tradition seems to justify the following deductions, though as the tradition is certainly jumbled it may also be incorrect in detaUs. {a) The councU is recognized only by the northern Church and is unknown to the Churches of Ceylon, Burma and Siam. It seems to have regarded Kashmir as sacred land outside which the true doctrine was exposed to danger, {b) But it was not a speciaUy Mahayanist meeting but rather a conference of peace and compromise. Taranatha says this clearly : in Hsiian Chuang's accqunt an assembly of Arhats (which at this time must have meant Hinayanists) elect a president who was not an Arhat and according to Paramartha the assembly consisted of 500 Arhats and 500 Bodhisattvas who were convened by a leader of the Sarvastivadin school and ended by requesting Asvaghosha to revise their work, (c) The Uterary result of the council was the ' Chap. XII. ' See Watters, r. pp. 222, 224 and 270. It is worth noting that Hsiian Chuang says Asoka lived one hundred years after the Buddha's death. See Watters, i. p. 267. See also the note of S. L6vi in J.R.A.S. 1914, pp. 1016-1019, citing traditions to the effect that there were 300 years between Upagupta, the teacher of Asoka, and Kanishka, who is thus raade to reign about 31 a.d. On the other hand Kanishka'g chaplain Sangharaksha is said to have lived 700 years after the Buddha. XXII] FROM KANISHKA TO VASUBANDHU 81 composition of commentaries on the three Pitakas. One of these, the Abhidharma-mahavibhasha-sastra, translated into Chinese in 437-9 and stiU extant, is said to be a work of encyclopaedic character, hardly a commentary in the strict sense. Paramartha perhaps made a confusion in saying that the Jnana-prasthana itseU was composed at the council. The traditions indicate that the council to some extent sifted and revised the Tripitaka and perhaps it accepted the seven Abhidharma books of the Sarvasti vadins'. But it is not stated or impHed that it composed or sanctioned Mahayanist books. Taranatha merely says that such books appeared at this time and that the Hinayanists raised no active objection. But if the above is the gist of the traditions, the position described is not clear. The council is recognized by Mahayanists yet it appears to have resulted in the composition of a Sarvasti vadin treatise, and the tradition connecting the Sarvastivadins with the council is not likely to be wrong, for they are recognized in the inscription on Kanishka's casket, and Gandhara and Kashmir were their headquarters. The decisions of councils are often poUtic rather than logical and it may be that the doctors summoned by Kanishka, while compiling Sarvastivadin treatises, admitted the principle that there is more than one vehicle which can take mankind to salvation. Perhaps some compromise based on geography was arranged, such as that Kashmir should be left to the Sarvastivadin school which had long flourished there, but that no opposition should be offered to the Mahayanists elsewhere . The relations of the Sarvastivadins to Mahayanism are exceedingly difficult to define and there are hardly sufficient materials for a connected account of this once important sect, but I wiU state some facts about it which seem certain. It is ancient, for the Kathavatthu aUudes to its doctrines^. It flourished in Gandhara, Kashmir and Central Asia, and Kanishka's casket shows that he patronized it*. But it appears * See Takakusu in J.P.T.S. 1905, pp. 67 ff. For the Sarvastivadin Canon, see my chapter on the Chinese Tripitaka. ' See above, vol. i. p. 262. For an account of the doctrines see also Vasilief, 245 ff. RockhUl, Life of the Buddha, pp. 190 ff. ' Its connection with Gandhara and Kashmir is plainly indicated in its own scriptures. See Przyluski's article on " Le Nord-Ouest de I'Inde dans le Vinaya des Mulasarvastivadins," J.A. 1914, ii. pp. 493 ff. This Vinaya must have received con siderable additions as time went on and in its present form is posterior to Kanishka. E. n. 6 82 THE MAHAYANA [ch. to have been hardly known in Ceylon or Southern India. It was the principal northern form of Hinayanism, just as the Theravada was the southern form. I-Ching however says that it prevaUed in the Malay Archipelago. Its doctrines, so far as known, were Hinayanist but it was distinguished from cognate schools by holding that the external world can be said to exist and is not merely a continual process of becoming. It had its own version of the Abhidharma and of the Vinaya. In the time of Fa-Hsien the latter was stiU pre served oraUy and was not written. The adherents of this school were also caUed Vaibhashikas, and Vibhasha was a name given to their exegetical Hterature. But the association of the Sarvastivadins with Mahayanists is clear from the council of Kanishka onwards. Many eminent Buddhists began by being Sarvastivadins and became Mahayan ists, their earHer beUef being regarded as preUminary rather than erroneous. Hsuan Chuang translated the Sarvastivadin scriptures in his old age and I-Ching belonged to the Mula sarvastivadin school'; yet both authors write as if they were devout Mahayanists. The Tibetan Church is generaUy regarded as an extreme form of Mahayanism but its Vinaya is that of the Sarvastivadins. Though the Sarvastivadins can hardly have accepted ideaHst metaphysics, yet the evidence of art and their own version of the Vinaya make it probable that they tolerated a moderate amount of mythology, and the Mahayanists, who Hke aU philosophers were obUged to admit the provisional vaUdity of the external world, may also have admitted their analysis of the same as provisionally vaUd. The strength of the Hinayanist schools lay in the Vinaya. The Mahayanists showed a tendency to replace it by legends and vague if noble aspirations. But a code of discipline was necessary for large monasteries and the code of the Sarvastivadins enjoyed general esteem in CJentral Asia and China. Three stages in the history of Indian Buddhism are marked by the names of Asvaghosha, Nagarjuna and the two brothers 1 The distinction between Sarvastivadin and Mulasarvastivadin is not clear to me. I can only suggest that when a section of the school accepted the Mahavibhasha and were known as Vaibhashikas others who approved of the school chiefly on , account of its excellent Vinaya caUed themselves Primitive Sarvastivadins. j XXII] FROM KANISHKA TO VASUBANDHU 83 Asanga and Vasubandhu. It would be easier to give a precise description of its development if we were sure which of the works ascribed to these worthies are authentic, but it seems that Asvaghosha represents an ornate and transitional phase of the older schools leading to Mahayanism, whereas Nagarjuna is connected with the Prajiia-paramita and the nihilistic philosophy described in the preceding chapter. Asanga was the founder of the later and more scholastic system caUed Yogacara and is also associated with a series of revelations said to have been made by Maitreya. As mentioned above, tradition makes Asvaghosha', one of the most brilHant among Sanskrit writers, Uve at the court of Kanishka^ and according to some accounts he was given to the Kushans as part of a war indemnity. The tradition* is confirmed by the style and contents of his poems and it has been noted by Foucher that his treatment of legends is in remarkable accord with their artistic presentment in the Gandharan sculptures. Also fragmentary manuscripts of his dramas discovered in Central Asia appear to date from the Kushan epoch. Asva ghosha's rank as a poet depends chiefiy on his Buddhacarita, or Hfe of the Buddha up to the time of his enUghtenment. It is the earhest example of a Kavya, usuaUy translated as artificial epic, but here Uterary skill is subservient to the theme and does not, as too often in later works, overwhelm it. The Buddha is its hero, as Rama of the Ramayana, and it sings the events of his earHer Ufe in a fine flow of elaborate but impassioned language. Another of his poems*, discovered only a few years ago, treats of the conversion of Nanda, the Buddha's half-brother. ' See Sylvain Levi, J.A. 1908, xn. 57 ff., and Wintemitz, Oes. Ind. Lit. Ii. i. pp. 201 ff. 2 The only reason for doubting it is that two stories (Nos. 14 and 31) in the Sutralankara (which appears to be a genuine work) refer to Kanishka as if he had reigned in the past. This may be a poetic artifice or it may be that the stories are interpolations. See for the traditions Watters on Yuan Chwang, n, 102-4 and Takakusu in J.R.A,8, 1905, p. 53 who quotes the Chinese Samyukta-ratna-pitaka- sutra and the Record of Indian Patriarchs. The Chinese Ust of Patriarchs is com patible with the view that Asvaghosha was aUve about 125 a.d. for he was the twelfth Patriarch and Bodhidharma the twenty-eighth visited China in 520. This gives about 400 years for sixteen Patriarchs, which is possible, for these worthies were long-Uved. But the list has Uttle authority. ^ The traditions are conveniently coUected in the introduction to Teitaro Suzuki's translation of The Awakening of Faith. * The Saundaranandakavya. 6—2 84 THE MAHAYANA [ch. Various other works are ascribed to Asvaghosha and for the history of Buddhism it is of great interest to decide whether he was really the author of The Awakening of Faith, This skiUul exposition of a difficult theme is worthy of the writer of the Buddhacarita but other reasons make his authorship doubtful, for the theology of the work may be described as the fuU-blown flower of Mahayanism untainted by Tantrism. It includes the doctrines of Bhiita-tathata, Alayavijnana, Tathagatagarbha and the three bodies of Buddha. It would be dangerous to say that these ideas did not exist in the time of Kanishka, but what is known of the development of doctrine leads us to expect their fuU expression not then but a century or two later and other circumstances raise suspicions as to Asvaghosha's authorship. His undoubted works were translated into Chinese about 400 A.D. but The Awakening of Faith a century and a haU later'. Yet if this concise and authoritative compendium had existed in 400, it is strange that the earHer translators neglected it. It is also stated that an old Chinese catalogue of the Tripitaka does not name Asvaghosha as the author^. The undoubted works of Asvaghosha treat the Buddha with ornate but grave rhetoric as the hero of an epic. His progress is attended by miracles such as Indian taste demands, but they hardly exceed the marvels recounted in the PaH scriptures and there is no sign that the hero is identified, as in the Ramayana of Tulsi Das or the Gospel according to St John, with the divine spirit. The poet clearly feels personal devotion to a Saviour. He dwells on the duty of teaching others and not seffishly seeking one's own salvation, but he does not formulate dogmas. The name most definitely connected with the early pro mulgation of Mahayanism is Nagarjuna*. A preponderance of 1 See Nanjio, Nos. 1182, 1351, 1250, 1299. It is noticeable that the translator Paramartha shows a special interest in the life and works of Asanga and Vasubandhu. ^ See Winternitz, Ges. Ind. Lit. ii. i. p. 211. It is also noticeable that The Awakening of Faith appears to quote the Lankavatara sutra which is not generaUy regarded as an early Mahayanist work. ' Nagarjuna cannot have been the founder of the Mahayana for in his Maha- prajiia-pararaita-Sastra (Nanjio, 1169, translation by Kumarajiva) he cites inter alia the Lotus, the Vimalakirti-sutra, anda work caUed Mahayana-Sastra. SeeB.E.F.E.O. 1911, p. 453. For Nagarjuna see especiaUy Griinwedel, Mythologie, pp. 29 ff. and the bibUography given in the notes. Jour. Budd. Text. Soe. v. part iv. pp. 7 ff. Watters, Yuan Chwang, pp. 200 ff. TaranStha, chap, xv and Wintemitz, Oes, Ind, Lit, n, i. pp. 250 ff. XXII] FROM KANISHKA TO VASUBANDHU 85 Chinese tradition makes him the second patriarch after Asva ghosha' and this agrees with the Kashmir chronicle which impUes that he Uved soon after Kanishka^. He probably flourished in the latter half of the second century. But his biographies extant in Chinese and Tibetan are almost whoUy mythical, even crediting him with a Ufe of several centuries, and the most that can be hoped is to extract a few grains of history from them. He is said to have been by birth a Brahman of Vidarbha (Berar) and to have had as teacher a Sudra named Saraha or Rahulabhadra. When the legend states that he visited the Nagas in the depths of the sea and obtained books from them, it seems to admit that he preached new doctrines. It is noticeable that he is represented not only as a philosopher but as a great magician, builder, physician, and maker of images. Many works are attributed to him but they have not the same authenticity as the poems of Asvaghosha. Some schools make him the author of the Prajiia-paramita but it is more usuaUy regarded as a revelation. The commentary on it known as Maha-prajiia-paramita-sastra is generally accepted as his work. A consensus of tradition makes him the author of the Madhyamika* aphorisms of which some account has been given above. It is the principal authority of its school and is provided with a commentary attributed to the author himself and with a later one by Candrakirti*. There is also ascribed to him a work caUed the SuhriUekha or friendly letter, a compendium of Buddhist doctrines, addressed to an Indian king^. This work 1 He is omitted from the list of Buddhabhadra, giving the succession according to the Sarvastivadins, to which school he did not belong. I-Ching classes him with Asvaghosha and Aryadeva as belonging to the early period. ' Rajatarangini, i. 173, 177. ' Edited in the Bibliotheca Buddhica by De la VaUee Poussin and (in part) in the Journal of the Buddhist Text Soe. See too WaUeser, Die Mittlere Lehre des Ndgdrjuna nach der Tibetischen Version iibertragen, 1911 : nach der Chinesischen Version iibertragen, 1912. • The ascription of these works to Nagarjuna is probably correct for they were translated by Kumarajiva who was sufficiently near him in date to be in touch with good tradition. ' The name of this king, variously given as Udayana, Jetaka and Satavahana, has not been identified with certainty from the various transcriptions and transla tions in the Chinese and Tibetan versions. See J. Pali Text Soe. for 1886^and I-Ching Records of the Buddhist Religion (trans. Takakusu), pp. 158 ff. The Andhra kings who reigned from about 240 B.C. to 225 a.d. all claimed to belong to the Satavahana dynasty. The stupa of Amaravati in the Andhra territory is surrounded by a stone raUing ascribed to the period 160-200 a.d. and Nagarjuna may have addressed a pious king Uving about that time. 86 THE MAHAYANA [ch. is old for it was translated into Chinese in 434 a.d. and is a homily for laymen. It says nothing of the Madhyamika philosophy and most of it deals with the need of good conduct and the terrors of future punishment, quite in the manner of the Hinayana. But it also commends the use of images and incense in worship, it mentions Avalokita and Amitabha and it holds up the ideal of attaining Buddhahood. Nagarjuna's author ship is not beyond dispute but these ideas may weU represent a type of popular Buddhism sUghtly posterior to Asvaghosha'. In most Usts of patriarchs Nagarjuna is followed by Deva, also called Aryadeva, Kanadeva or Nilanetra. I-Ching mentions him among the older teachers and a commentary on his principal work, the Satasastra, is attributed to Vasubandhu^. Little is known of his special teaching but he is regarded as an important doctor and his pupil Dharmatrata is also important if not as an author at least as a compiler, for Sanskrit coUections of verses corresponding to the PaU Dhammapada are ascribed to him. Aryadeva was a native of southern India*. The next epocH in the history of Buddhism is marked by the names of Asanga and Vasubandhu. The interval between them and Deva produced no teacher of importance, but Kumar- alabdha, the founder of the Sautrantika school and perhaps identical with Kumarata the eighteenth Patriarch of the Chinese lists, may be mentioned. Hsiian Chuang says* that he was carried off in captivity by a king who reigned somewhere in the east of the Pamirs and that he, Asvaghosha, Nagarjuna and Deva were styled the four shining suns. Asanga and Vasubandhu were brothers, sons of a Brahman -who lived at Peshawar. They were both converted from the Sarvastivadin school to Mahayanism, but the third brother 1 For other works attributed to Nagarjuna see Nanjio, Nos. 1169, 1179, 1180, 1186 and Walleser's introduction to Mittlere Lehre nach der Chinesischen Version The Dharmasangraha, a Sanskrit theological glossary, is also attributed to Nagar juna as weU as the tantric work Pancakrama. But it is not Ukely that the latter dates from his epoch. « Nanjio, No. 1188. ' The very confused legends about him suggest a comparison with the Dravidian legend of a devotee who tore out one of his eyes and offered it to Siva. See Griin wedel, Mythologie, p. 34 and notes. Polemics against various Hinayanist sects are ascribed to him. See Nanjio, Nos. 1259, 1260. * Watters, Yuan Chwang, ii. p. 286. Hsiian Chuang does not say that the four were contemporary but that in the time of Kumaralabdha they were caUed the four Suns. xxn] FROM KANISHKA TO VASUBANDHU 87 Virincivatsa never changed his convictions. Tradition connects their career with Ayodhya as well as with Peshawar and Vasubandhu enjoyed the confidence of the reigning monarch, who was probably Candragupta I. This identification depends on the hypothesis that Vasubandhu Uved from about 280 to 360 A.D. which, as already mentioned, seems to me to have been proved by M. Peri'. The earUer Gupta kings though not Buddhists were tolerant, as is shown by the fact that the king of Ceylon 2 was aUowed to erect a magnificent monastery at Naianda in the reign of Samudragupta (c. 330-375 a.d.). Asanga founded the school known as Yogacara and many authorities ascribe to him the introduction of magical practices and Tantrism. But though he is a considerable figure in the history of Buddhism, I doubt if his importance or culpabiUty is so great as this. For if tradition can be trusted, earHer teachers especiaUy Nagarjuna dealt in spells and invocations and the works of Asanga* known to us are characterized by a somewhat scholastic piety and are chiefly occupied in defining and des cribing the various stages in the spiritual development of a BodMsattva. It is true that he admits the use of magical formulae* as an aid in this evolution but they form only a slight part of his system and it does not appear that the Chen-yen or Shingon sect of the Far East (the Sanskrit Mantrayana) traced its lineage back to him. Our estimate of his position in the history of Buddhism must depend on our opinion as to the authorship of The Awakening of Faith, If this treatise was composed by Asvaghosha then doctrines respecting the three bodies of Buddha, the Tathagata- garbha and the Alaya-vijfiana were not only known but scientificaUy formulated considerably before Asanga. The con clusion cannot be rejected as absurd — ^for Asvaghosha might speak differently in poems and in philosophical treatises — but ' For Asanga and Vasubandhu see Peri in B,E,F,E,0, 1911, pp. 339-390. Vincent Smith in Early History of India, third edition, pp. 328-334. Wintemitz, Oes, Ind, Lit, ii. i. p. 256. Watters, YiiMn Chwang, i. pp. 210, 355-359. Taranatha, chap. xxn. Griinwedel, Mythologie, p. 35. * Meghavarman. See V. Smith, l,c, 287. ^ Two have been preserved in Sanskrit : the Mahayana-siitralankara (Ed. v. Transl., S. Levi, 1907-1911) and the Bodhisattva-bhumi (Enghsh summary in Musion, 1905-6). A brief analysis of the Uterature of the Yogacara school according to Tibetan authorities is given by Stcherbatskoi in Musion, 1905, pp. 144^155. * Mahayana-sutral. xvili. 71-73. The ominous word maithuna also occurs in this work, xvirt. 46. 88 • THE MAHAYANA [ch. it is surprising, and it is probable that the treatise is not his. If so, Asanga may have been the first to elaborate systematicaUy (though not to originate) the idea that thought is the one and only reahty. Nagarjuna's nihiUsm was probably the older theory. It sounds late and elaborate but still it follows easUy if the dialectic of Gotama is appUed uncompromisingly not only to our mental processes but to the external world. Yet even in India the result was felt to-be fantastic and sophistical and it is not surprising if after the lapse of a few generations a new system of ideahsm became fashionable which, although none too inteUigible, was abstruse rather than paradoxical. Asanga was alleged to have received revelations from Maitreya and five of his works are attributed to this Bodhisattva who enjoyed considerable honour at this period. It may be that the veneration for the Buddha of the future, the Messiah who would reign over his saints in a pure land, owed something to Persian influence which was strong in India during the decadence of the Kushans'. Both Mithraism and Manichaeism classified their adepts in various ranks, and the Yogacara doctors who dehght in grading the progress of the Bodhisattva may have borrowed something from them 2. Asanga's doctrine of defile ment (klesa) and purification may also owe something to Mani, as suggested by S. Levi. In spite of his Uterary merits Asanga remains a doctor rather than a saint or poet*. His speculations have Httle to do with either Gotama or Amitabha and he was thus not in Uving touch with either the old or new schools. His brother Vasubandhu had perhaps a greater position. He is reckoned as the twentieth Patriarch and Tibetan tradition connects him with the worship of Amitabha*. Paramartha's Hfe of Vasubandhu represents him as having frequented the court of Vikramaditya (to be identified with Candragupta I), who at first favoured the Sankhya philosophy I Vincent Smith, l,c. p. 275. " But there are of course abundant Indian precedents, Brahmanical as weU as Buddhist, for describing various degrees of sanctity or knowledge. ' The wooden statues of Asanga and Vasubandhu preserved in the Kofukaji at Nara are masterpieces of art but can hardly claim to be other than works of imagina tion. They date from about 800 a.d. See for an exceUent reproduction Tajima's Select Relics, ii. x. * See Eitel and Griinwedel, but I do not know in what texts this tradition is found. It is remarkable that Paramartha's Ufe (T'oung Pao, 1904, pp. 269-296) does not say either that he was twentieth patriarch or that he worshipped Amida. xxn] FROM KANISHKA TO VASUBANDHU 89 but accorded some patronage to Buddhism. During this period Vasubandhu was a Sarvastivadin but of Uberal views' and while in this phase wrote the Abhidharma-kosa, a general exposition of the Abhidharma, mainly according to the views of the Vaibhashikas but not without criticism. This celebrated work is not weU known in Europe^ but is still a text-book amongst Japanese Buddhist students. It gained the esteem of all schools and we are given to understand that it presupposed the philosophy of the Vibhasha and of the Jfiana-prasthana. According to Paramartha the original work consisted of 600 aphorisms in verse which were sent by the author to the monks of Kashmir. They approved of the composition but, as the aphorisms were concise, asked for fuller explanations. Vasubandhu then expanded his verses into a prose commentary, but meanwhile his views had undergone a change and when he disapproved of any Vaibhashika doctrine, he criticized it. This enlarged edition by no means pleased the brethren of Kashmir and caUed forth polemics. He also wrote a controversial work against the Sankhya philosophy. Late in Hfe Vasubandhu, moved by the entreaties of his brother Asanga, became a devout Mahayanist and wrote in his old age Mahayanist treatises and commentaries*. - ^ On receiving a large donation ho bruit three monasteries, one for Hinayanists, one for Mahayanists and one for nuns. ' The work consists of 600 verses (Karika) with a lengthy prose commentary (Bhashya) by the author. The Sanskrit original is lost but translations have been preserved in Chinese (Nanjio, Nos. 1267, 1269, 1270) and Tibetan (see Cordier, Cat, du Fonds tibetain de la Bib, Nat. 1914, pp. 394, 499). But the commentary on the Bhashya caUed Abhidharma-kosa- vyakhya, or Sphutartha, by YAsomitra has been preserved in Sanskrit in Nepal and frequently cites the verses as well as ths Bhashya in the original Sansltrit. A number of European savants are at present occupied with this Uterature and Sir Denison Ross (to whom I am indebted for much information) contemplates the publication of an Uighur text of Book i found in Central Asia. At present (1920), so far as T know, the only portion of the Abhidharma-k6sa in print is De la VaUee Poussin's edition and translation of Book m, containing the Tibetan and Sanskrit texts but not the Chinese (De la VaUee Poussin — Vasubandhu ei Yaiomitra, London, 1914^18). This chapter deals with such topics as the structure of the universe, the manner and place of rebirth, the chain of causation, the geography of the world, the duration and characteristics of Kalpas, and the appearance of Buddhas and Cakravartins. ' See Nanjio, pp. 371-2, for a list of his works translated into Chinese. Hsiian Chuang's account differs from the above (which is taken from Paramartha) in details. He also tells a curious story that Vasubandhu promised to appear to his friends after death and ultimately did so, though he forgot his promise untU people began to say he had gone to hell. CHAPTER XXIII INDIAN BUDDHISM AS SEEN BY THE CHINESE PILGRIMS About the time of Vasubandhu there existed four schools of Indian Buddhism called Vaibhashika, Sautrantika, Madhyamika and Yoga or Yogacara'. They were speciaUy concemed with philosophy and apparently cut across the older division into eighteen sects, which at this period seem to have differed mainly on points of discipUne. Though not of great practical import ance, they long continued to play a certain part in controversial works both Buddhist and Brahmanic. The first two which were the older seem to have belonged to the Hinayana and the other two even more definitely to the Mahayana. I-Ching^ is quite clear as to this. "There are but two kinds of the so-caUed Mahayana" he says, "first the Madhyamika, second the Yoga.... These two systems are perfectly in accordance with the noble doctrine. Can we say which of the two is right? Both equaUy conform to truth and lead us to Nirvana" and so on. But he does not say that the other two systems are also aspects of the truth. This is the more remarkable because he himseU foUowed the Mula-sarvastivadins. Apparently Sarvastivadin and Vai bhashika were different names for the same school, the latter being appHed to them because they identified themselves with the commentary (Vibhasha) already mentioned whereas the former and older designation came to be used chiefly with reference to their disciplinary rules. Also there were two groups of Sarvastivadins, those of Gandhara and those of Kashmir. The name of Vaibhashika was appHed chiefly to the latter who, if we may find a kernel of truth in legends which are certainly exaggerated, endeavoured to make Kashmir a holy land with a monopoly of the pure doctrine. Vasubandhu and Asanga appear to have broken up this isolation for they first preached 1 See Vasilief, Le Bouddhisme, Troisi^me supplement, pp. 262 ff. Koppen, Rel. des Buddha, i. 151. Takakusu in J. Pali Te^t Society, 1905, pp. 67-146. " Records, translated by Takakusu, p. 15. CH. XXIII] THE CHINESE PILGRIMS 91 the Vaibhashika doctrines in a Uberal and eclectic form outside Kashmir and then by a natural transition and development went over to the Mahayana. But the Vaibhashikas did not disappear and were in existence even in the fourteenth century'. Their chief tenet was the real existence of external objects. In matters of doctrine they regarded their own Abhidharma as the highest authority^. They also held that Gotama had an ordinary human body and passed first into a preliminary form of Nirvana when he attained Buddhahood and secondly into complete Nirvana at his death. He was superhuman only in the sense that he had intuitive knowledge and no need to learn. Their contempt for sutras may have been due to the fact that many of them discountenance the Vaibhashika views and also to a knowledge that new ones were continuaUy being composed. I-Ching, who ends his work by asserting that all his state ments are according to the Arya-miila-sarvastivada-nikaya and no other, gives an interesting summary of doctrine. "Again I say: the most important are only one or two out of eighty thousand doctrines of the Buddha : one should conform to the worldly path but inwardly strive to secure true wisdom. Now what is the worldly path? It is obeying prohibitive laws and avoiding any crime. What is the true wisdom? It is to obliterate the distinction between subject and object, to follow the exceUent truth and to free oneseU from worldly attachments: to do away with the trammels of the chain of causaUty : further to obtain merit by accumulating good works and finally to realize the excellent meaning of perfect reality." Such a statement enables us to understand the remark which he makes elsewhere that the same school may belong to the Hinayana and Mahayana in different places, for, whatever may be meant by wisdom which aims at obhterating the difference between subject and object, it is clearly not out of sympathy with Yogacara doctrines. In another place where he describes the curriculum followed by monks he says that they learn the Yogacarya-^astra first and then eight compositions of Asanga and Vasubandhu. Among the works prescribed for logic is the Nyayadvara-sastra attributed to Nagarjuna. The monk * They are mentioned in the Sarva-dar^ana-sangraha. 2 Kern (Indian Buddhism, p. 126) says they rejected the authority of the Sutras altogether but gives no reference. 92 THE MAY AH AN A [ch. should learn not only the Abhidharma of the Sarvastivadins but also the Agamas, equivalent to the Sutra-pitaka. So the study of the siitras and the works of Asanga and Vasubandhu is approved by a Sarvastivadin. The Sautrantikas', though accounted Hinayanists, mark a step in the direction of the Mahayana. The founder of the school was Kumaralabdha, mentioned above. In their estimation of scripture they reversed the views of the Vaibhashikas, for they rejected the Abhidharma and accepted only the siitras, arguing that the Abhidharma was practicaUy an extract from them. As Uterary criticism this is correct, if it means that the more ancient siitras are older than the oldest Abhidharma books. But the indiscriminate acceptance of siitras led to a creed in which the supernatural played a larger part. The Sautrantikas not only ascribed superhuman powers to the Buddha, but beheved in the doctrine of three bodies. In phUosophy, though they were reaUsts, they held that external objects are not per ceived directly but that their existence is inferred^. Something has aUeady been said of the two other schools, both of which denied the reahty of the external world. The differences between them were concerned with metaphysics rather than theology and led to no popular controversies. Up to this point the history of Indian Buddhism has proved singularly nebulous. The most important dates are a matter of argument, the chief personages haU mythical. But when the records of the Chinese pilgrims commence we are in touch with something more soUd. They record dates and facts, though we must regret that they only repeat what they heard and make no attempt to criticize Indian traditions or even to weave them into a connected chronicle. Fa-Hsien, the first of these interesting men, left China in 399 and resided in India from 405 to 411, spending three years at Pataliputra and two at TamraUpti. He visited the Panjab, Hindustan and Bengal and his narrative leaves the impression that all these were in the main Buddhist countries : of the Deccan which he did not visit he heard that its inhabitants were barbarous and not Buddhists, though it contained some 1 See VasUief, pp. 301 ff. and various notices in Hsuan Chuang and Watters. Also de la Vallte Poussin's article in E.R.E. ' Hsiian Chuang informs us that when he was in Srughna he studied the Vibhasha of the Sautrantikas, but the precise significance of this term is not plain. xxm] THE CHINESE PILGRIMS 93 Buddhist shrines. Of the Middle Kingdom (which according to his reckoning begins -with Muttra) he says that the people are free and happy and neither kill any Uving creature nor drink intoxicating liquor'. He does not hint at persecution though he once or twice mentions that the Brahmans were jealous of the Buddhists. Neither does he indicate that any strong animosity prevailed between Maha- and Hinayanists. But the two parties were distinct and he notes which prevailed in each locaHty. He left China by land and found the Hinayana pre valent at Shen-shen and Wu-i (apparently locaHties not far from Lob-Nor) but the Mahayana at Khotan. Nearer India, in countries apparently corresponding to parts of Kashmir and Gilgit, the monks were numerous and all Hinayanist. The same was the case in Udyana, and in Gandhara the Hinayanists were stiU in the majority. In the Panjab both schools were prevalent but the Hinayana evidently strong. In the district of Muttra the Law was stUl more flourishing, monasteries and topes were numerous and ample alms were given to the monks. He states that the professors of the Abhidharma and Vinaya made offerings to those works, and the Mahayanists to the book Prajaa-paramita, as weU as to Manjusri and Kwan-shih-yin. He found the country in which are the sacred sites of Sravasti, Kapilavastu and Kusinara sparsely inhabited and desolate, but this seems to have been due to general causes, not specially to the decay of rehgion. He mentions that ninety-six ^ varieties of erroneous views are found among the Buddhists, which points to the existence of numerous but not acutely hostile sects and says that there still existed, apparently in Kosala, followers of Devadatta who recognized three previous Buddhas but not Sakyamuni. He visited the birth-places of these three Buddhas which contained topes erected in their honour. He found Magadha prosperous and pious. Of its capital, Patna, he says "by the side of the topes of Asoka has been made a Mahayana monastery very grand and beautiful, there is also a Hinayana one, the two together containing 600 or 700 monks." It is probable that this was tj^pical of the religious condition of Magadha and Bengal. Both schools existed but the • Pa-Hsien's Travels, chap. xvi. " This figure is probably deduced from some artificial calculation of possible heresies like the 62 wrong views enumerated in the Brahma- Jala siitra. 94 THE MAHAYANA [cu. Mahayana was the more flourishing. Many of the old sites, such as Rajagriha and Gaya, were deserted but there were new towns near them and Bodh Gaya was a place of pilgrimage with three monasteries. In the district of TamraUpti (Tamluk) on the coast of Bengal were 22 monasteries. As his principal object was to obtain copies of the Vinaya, he stayed three years in Patna seeking and copying manuscripts. In this he found some difficulty, for the various schools of the Vinaya, which he says were divided by trivial differences only, handed down their respective versions orally. He found in the Mahayanist monastery one manuscript of the Mahasanghika rules and considered it the most complete, but also took down the Sarvastivadin rules. After the death of Vasubandhu few names of even moderate magnitude stand out in the history of Indian Buddhism. The changes which occurred were great but gradual and due not to the initiative of innovators but to the assimilative power of Hinduism and to the attractions of magical and emotional rites. But this tendency, though it doubtless existed, did not become conspicuous until about 700 a.d. The accounts of the Chinese pilgrims and the Hterature which has been preserved suggest that in the intervening centuries the monks were chiefly occupied with scholastic and exegetical work. The most distinguished successors of Asanga were logicians, among whom Dinnaga was pre-eminent. Sthiramati' andGunamati appear to have belonged to the same school and perhaps Bhavaviveka^ too. The state ments as to his date are inconsistent but the interesting fact is recorded that he utiUzed the terminology of the Sankhya for the purposes of the Mahayana. Throughout the middle ages the study of logic was pursued but Buddhists and Jains rather than by Brahmans*. Vasu bandhu composed some treatises dealing exclusively with logic but it was his disciple Dinnaga who separated it definitely from philosophy and theology. As in ideaHst philosophy, so in pure logic there was a paraUel movement in the Buddhist and Brahmanic schools, but if we may trust the statements of * He must have Uved in the fourth century as one of his works (Nanjio, 1243) was translated between 397 and 439. ^ Watters, Yilan Chwang, ii. 221-224. Nanjio, 1237. The works of Gunamati also are said to show a deep knowledge of the Sankhya philosophy. ' For the history of logic in India, see Vidyabhusana's interesting work Mediaeval School of Indian Logic, 1909. But I cannot accept aU his dates. xxm] THE CHINESE PILGRIMS 95 vacaspatimisra (about 1100 a.d.) Dinnaga interpreted the aphorisms of the Nyaya philosophy in a heterodox or Buddhist sense. This traces the beginnings of Indian logic to a Brahmanic source but subsequently it flourished greatly in the hands of Buddhists, especially Dinnaga and Dharmakirti. The former appears to have been a native of Conjevaram and a contemporary of KaUdasa. Both the logician and the poet were probably alive in the reign of Kumaragupta (413-455). Dinnaga spent much time in Naianda, and though the Sanskrit originals of his works are lost the Tibetan translations' are preserved. The Buddhist schools of logic continued for many centuries. One flourished in Kashmir and another, founded by Candrago- min, in Bengal. Both lasted almost until the Mohammedan conquest of the two countries. From about 470 to 530 a.d. northern India groaned under the tyranny of the Huns. Their King Mihiragula is represented as a determined enemy of Buddhism and a systematic destroyer of monasteries. He is said to have been a worshipper of Siva but his fury was probably inspired less by religious animosity than by love of pillage and slaughter. About 530 a.d. he was defeated by a coaUtion of Indian princes and died ten years later amid storms and portents which were beheved to signify the descent of his wicked soul into hell. It must have been about this time that Bodhidharma left India for he arrived in Canton about 520. According to the Chinese he was the son of a king of a country called Hsiang-Chih in southern India^ and the twenty -eighth patriarch and he became an important figure in the rehgion and art of the Far East. But no aUusion to him or to any of the Patriarchs after Vasu bandhu has been found in Indian literature nor in the works of Hsiian Chuang and I-Ching. The inference is that he was of no importance in India and that his reputation in China was not great before the eighth century: also that the Chinese Hsts of patriarchs do not represent the traditions of northern India. ' Diimaga's principal works are the Pramana-samuccaya andthe Nyaya-prave^a. Hsiian Chuang calls him Ch'en-na. See Watters, n. 209. See Stcherbatskoi in Musion, 1904, pp. 129-171 for Diimaga's influence on the development of the Naiyayika and Vaifeshika schools. » His personal name is said to have been P'u-ti-to-lo and his surname Ch'a-ti-li. The latter is probably a corruption of Kshatriya. Hsiang-Chih possibly represents a name beginning with Gandha, but I can neither flnd nor suggest any identification. 96 THE MAHAYANA [ch. ReUgious feeHng often ran high in southern India. Buddhists, Jains and Hindus engaged in violent disputes, and persecution was more frequent than in the north. It is easy to suppose that Bodhidharma being the head of some heretical sect had to fly and followed the example of many monks in going to China. But if so, no record of his school is forthcoming from his native land, though the possibihty that he was more than an individual thinker and represented some movement unknown to us cannot be denied. We might suppose too that since Nagarjuna and Aryadeva were southerners, their pecuhar doctrines were coloured by Dravidian ideas. But our available documents indicate that the Buddhism of southern India was almost entirely Hinayanist, analogous to that of Ceylon and not very sympathetic to the Tamils. The pilgrims Sung-Yiin and Hui-Sheng' visited Udyana and Gandhara during the time of the Hun domination (518-521). They found the king of the former a pious Buddhist but the latter was governed by an EphthaUte chieftain, perhaps Mihiragula himseU, who was a worshipper of demons. Of the Yetha or EphthaHtes they make the general observation that "their rules of poHteness are very defective." But they also say that the population of Gandhara had a great respect for Bud dhism and as they took back to China 170 volumes, " aU standard works belonging to the Great Vehicle," the EphthaUte persecu tion cannot have destroyed the faith in north-western India. But the evil days of decay were beginning. Henceforward we have no more pictures of untroubled piety and prosperity. At best Buddhism receives royal patronage in company with other rehgions ; sectarian conflicts increase and sometimes we hear of persecution. About 600 a.d. a king of Central Bengal named Sa^anka who worshipped Siva attempted to extirpate Buddhism in his dominions and destroyed the Bo tree at Bodh Gaya ^ . On the other hand we hear of the pious Piirnavarman, king of Magadha, who made amends for these sacrileges, and of Siiaditya, king of the country caUed Mo-lo-po by the Chinese, who was so careful of animal Ufe, that he even strained the water drunk by his horses and elephants, lest they should consume minute insects. 1 See B.E.F.E.O. 1903, pp. 379 ff. 2 His evil deeds are several times mentioned by Hsiian Chuang. It required a miracle to restore the Bo tree. xxm] THE CHINESE PILGRIMS 97 We know more of Indian Buddhism in the seventh century than in the periods which precede or follow it. The epoch was marked by the reign of the great king, or rather emperor, Harsha-Vardhana (606-648 a.d.), and the works written by Baiia, Bhartrihari and others who frequented his court have come down to us. Also we are fortunate in possessing the copious narrative of Hsiian Chuang, the greatest of the Chinese pUgrims, who spent sixteen years (629-645) in India as well as the work known as the "Record of the Buddhist reUgion as practised in India and the Malay Archipelago," composed by I-Ching who traveUed in those countries from 671 to 695. I-Ching also wrote the Hves of sixty Chinese pilgrims who visited India during the seventh century and probably there were many others of whom we have no record. The reign of Harsha is thus illustrated by a number of contemporary dateable works unusual in India. The king himseU wrote some Buddhist hymns', and three dramas are ascribed to him but were probably composed by some of the Uterary men whom he patronized. For aU that, the religious ideas which they contain must have had his approval. The RatnavaH and Priyadarsika are secular pieces and so far as they have any rehgious atmosphere it is Brahmanic, but the Naga- nanda is a Buddhist reUgious drama which opens with an invocation of the Buddha and has a jataka story for its plot^. Bana was himseU a devout Brahman but his historical romance Harshacarita and his novel caUed Kadambari both describe a mixture of rehgions founded on observation of contemporary life. In an interesting passage* he recounts the king's visit to a Buddhist ascetic. The influence of the holy man causes the more inteUigent animals in his neighbourhood, such as parrots, to devote themselves to Buddhist lore, but he is surrounded by devotees of the most diverse sects, Jains, Bhagavatas, Panca- ratras, Lokayatikas with foUowers of Kapila, Kanada and many 1 See Ettinghausen, Earshavardhana, Appendix m. * The appearance of Gauri as a dea ex machina at the end hardly shows that Harsha's Buddhism had a Saktist tinge but it does show that Buddhists of that period turned naturaUy to Sivaite mythology. 3 Harshacarita, chap. vm. The parrots were expounding Vasubandhu's Abhi- dharmako^a. Bana frequently describes troops of holy men apparently Uving in harmony but including foUowers of most diverse sects. See Kadambari, 193 and 394 : Harshacar. 67. B.n. 7 9'8 THE MAHAYANA [ch. other teachers. Mayura, another Uterary protege of Harsha's, was Uke Bana a Brahman, and Subandhu, who flourished a Httle before them, ignores Buddhism in his romance called Vasa- vadatta. But Bhartrihari, the stiU popular gnomic poet, was a Buddhist. It is true that he osciUated between the court and the cloister no less than seven times, but this vacillation seems to have been due to the weakness of the flesh, not to any change of convictions. For our purpose the gist of this Uterature is that Hinduism in many forms, some of them very unorthodox, was becoming the normal religion of India but that there were stiU many eminent Buddhists and that Buddhism had sufficient prestige to attract Harsha and sufficient Hfe to respond to his patronage. About 600 A.D. India was exhausted by her struggle with the Huns. After it there remained only a multitude of smaU states and obscure dynasties, but there was evidently a readiness to accept any form of unifjdng and tranquilUzing rule and for nearly half a century this was provided by Harsha. He con quered northern India from the Panjab to Bengal but failed to subdue the Deccan. Though a great part of his reign was spent in war, learning and education flourished. Hsiian Chuang, who was his honoured guest, gives a good account of his adminis tration but also makes it plain that brigandage prevaUed and that traveUing was dangerous. After 643 Harsha, who was growing elderly, devoted much attention to reUgion and may be said to have become a Buddhist, while aUowing himself a certain eclectic freedom. Several creeds were represented among his immediate relatives. Devotion to Siva was traditional in the family : his father had been a zealous worshipper of the Sun and his brother and sister were Buddhists of the Sammitiya sect. Harsha by no means disowned Brahmanic worship, but in his latter years his procUvity to Buddhism became more marked and he endeavoured to emulate the piety of Asoka. He founded rest houses and hospitals, as well as monasteries and thousands of stupas. He prohibited the taking of Hfe and the use of animal food, and of the three periods into which his day was divided two were devoted to rehgion and one to business. He also exercised a surveiUance over the whole Buddhist order and advanced meritorious members. Hsiian Chuang has left an interesting account of the reUgious xxm] THE CHINESE PILGRIMS 99 fetes and spectacles organized by Harsha. At Kanauj he attended a great assembly during which a solemn procession took place every day. A golden image of Buddha was borne on an elephant and Harsha, dressed as Indra, held a canopy over it, while his ally Raja Kumara', dressed as Brahma, waved a fly-whisk. It was subsequently washed by the king's own hands and in the evening his Majesty, who like Akbar had a taste for religious discussion, Ustened to the arguments of his Chinese guest. But the royal instructions that no one was to speak against the Master of the Law were so peremptory that even his biographer admits there was no real discussion. These edifying pageants were interrupted by disagreeable incidents which show that Harsha's tolerance had not produced complete harmony. A temporary monastery erected for the fetes caught fire and a fanatic attempted to stab the king. He confessed under examination that he had been instigated to the crime by Brahmans who were jealous of the favours which the Buddhists received. It was also estabUshed that the incendiaries were Brahmans and, after the ringleaders had been punished, five hundred were exiled. Harsha then proceeded to Allahabad to superintend a quinquennial distribution of alms. It was his custom to let treasure accumulate for five years and then to divide it among holy men and the poor. The proceedings lasted seventy -five days and the concourse which coUected to gaze and receive must have resembled the fair still held on the same spot. Buddhists, Brahmans and Jains aU partook of the royal bounty and the images of Buddha, Surya and Siva were worshipped on successive days, though greater honour was shown to the Buddha. The king gave away everything that he had, even his robes and jewels, and finaUy, arrayed in clothes borrowed from his sister, rejoiced saying "all I have has entered into incorruptible and imperishable treasuries." After this, adds Hsiian Chuang, the king's vassals offered him jewels and robes so that the treasury was replenished. This was the sixth quinquennial distribution which Harsha had held and the last, for he died in 648. He at first favoured the Hinayana but subsequently went over to the Mahayana, being moved in part by the exhortations of Hsiian Chuang. ' It is curious that Baria (Harshacarita, vn.) says of this prince that from child hood he resolved never to worship anyone but Siva. 7—2 100 THE MAHAYANA [ch. Yet the substance of Hsiian Chuang's account is that though Buddhism was prospering in the Far East it was decaying in India. Against this can be set instances of royal piety Hke those described, the fame enjoyed by the shrines and schools of Magadha and the conversion of the king of Tibet in 638 a.d. This event was due to Chinese as weU as Indian influence, but would hardly have occurred unless in north-eastern India Buddhism had been esteemed the reUgion of civiHzation. StiU Hsiian Chuang's long catalogue of deserted monasteries' has an unmistakable significance. The decay was most pronounced in the north-west and south. In Gandhara there were only a few Buddhists : more than a thousand monasteries stood untenanted and the Buddha's sacred bowl had vanished. In TakshasUa the monasteries were numerous but desolate : in Kashmir the people foUowed a mixed faith. Only in Udyana was Buddhism held in high esteem. In Sind the monks were numerous but indolent. No doubt this desolation was largely due to the depredations of Mihiragula. In the Deccan and the extreme south there was also a special cause, namely the prevalence of Jainism, which somewhat later became the state rehgion in several kingdoms. In Kahnga, Andhra and the kingdom of the Colas the pilgrim reports that Jains were very numerous but counts Buddhist monasteries only by tens and twenties. In Dravida there were also 10,000 monks x)f the Sthavira school but in Malakuta among many ruined monasteries only a few were stUl inhabited and here again Jains were numerous. For aU Central India and Bengal the pUgrim's statistics teU the same tale, namely that though Buddhism was represented both by monasteries and monks, the Deva-temples and un beUevers were also numerous. The most favourable accounts are those given of Kanauj, Ayodhya and Magadha where the sacred sites naturally caused the devout to congregate. The statistics which he gives as to sects are interesting *. The total number of monks amounted to about 183,000. Of these only 32,000 belonged definitely to the Mahayana: more 1 The Rashtrapalaparipriccha (Ed. Finot, pp. ix-xi, 28-33) inveighs against the moral degeneration of the Buddliist clergy. This work was translated into Chinese between 589 and 618, so that demoraUsation must have begun in the sixth century. » See Rhys Davids in J,R,A.8, 1891, pp. 418 ff. xxm] THE CHINESE PILGRIMS 101 than 96,000 to the Hinayana, and 54,500 studied both systems or at any rate resided in monasteries which tolerated either course of study. Some writers speak as if after our era Mahayan ism was predominant in India and the Hinayana banished to its extreme confines such as Ceylon and Kashmir. Yet about a.d. 640 this zealous Mahayanist' states that haU the monks of India were definitely Hinayanist whUe less than a fifth had equaUy definite Mahayanist convictions. The Mahayana laid less stress on monasticism than the Hinayana and therefore its strength may have lain among the laity, but even so the admitted strength of the Hinayana is remarkable. Three Hinayanist schools are frequently mentioned, the Sthaviras, Sarvastivadins and Sam- mitiyas. The first are the weU-known Sinhalese sect and were found chiefly in the south (Conjeevaram) and in East Bengal, besides the monks of the Sinhalese monastery at Gaya. The Sarvastivadins were found, as their history would lead us to expect, chiefly in the north and beyond the frontiers of India proper. But both were outnumbered by the Sammitlyas, who amounted to nearly 44,000 monks. The chief doctrine^ of this sect is said to have been that individuals (puggalo) exist as such in the truest sense. This doctrine was supported by reference to the sutra known as the Burden and the Burden bearer*. It does not assert that there is a permanent and unchangeable soul (atta) but it emphasizes the reaUty and importance of that personahty which aU accept as true for practical purposes. It is probable that in practice this beUef differed Uttle from the ordinary Brahmanic doctrine of metempsychosis and this may be one reason for the prevalence of the sect. I-Ching, though he does not furnish statistics, gives a clear conspectus of Buddhist sects as they existed in his time. He starts from the ancient eighteen sects but divides them into four groups or Nikayas. (a) The Arya-Mahasanghika-nikaya. This comprised seven subdivisions but was apparently the least influential school as it was not predominant anywhere, though 1 Hsiian Chuang was not disposed to underrate the numbers of the Mahayana for he says that the monks of Ceylon were Mahayanists. ^ See the beginning of the Kathavatthu. The doctrine is formulated in the words Puggalo npalabbhati saccikatthaparamatthenati, and there foUows a dis cussion between a member of the orthodox school and a Puggalavadin, that is one who beUeves in the existence of a person, soul or entity which transmigrates from this world to another. ' Sam. Nik. xxn. 221. 102 THE MAHAYANA [ch. it coexisted with other schools in most parts. The Lokottara vadins mentioned by Hsiian Chuang as existing at Bamiyan belonged to it. They held that the Buddha was not subject to the laws of nature. (6) Arya-Sthavira-nikaya. This is the school to which our PaH Canon belongs. It was predominant in southern India and Ceylon and was also found in eastern Bengal. (c) The Arya-Mula-sarvastivada-nikaya with four subdivisions. Almost aU belonged to this school in northern India and it was flourishing in Magadha. {d) The Arya-Sammitiya-nikaya- with four subdivisions flourished in Lata and Sindhu. Thus the last three schools were preponderant in southern, northern and western India respectively. AU were foUowed in Magadha, no doubt because the holy places and the University of Naianda attracted all shades of opinion, and Bengal seems to have been similarly catholic. This is substantially the same as Hsiian Chuang's statement except that I-Ching takes a more favourable view of the position of the Sarvastivada, either because it was his own school or because its position had reaUy improved. It would seem that in the estimation of both pilgrims the Maha- and Hinayana are not schools but modes in which any school can be studied. The Nikaya' or school appears to have been chiefly, though not exclusively, concemed with the rule of disciphne which naturally had more importance for Buddhist monks than it has for European scholars. The observances of each Nikaya were laid down in its own recension of the scriptures which was sometimes oral and sometimes in writing. Probably aU the eighteen schools had separate Vinayas, and to some extent they had different editions of the other Pitakas, for the Sarvastivadins had an Abhidharma of their own. But there was no objection to combining the study of Sarvastivadin literature with the reading of treatises by Asanga and Vasu- ' This use of Nikaya must not be confused with its other use to denote a division of the Sutra-Pitaka. It means a group or coUection and hence can be used to denote either a body of men or a coUection of treatises. These Nikayas are also not the same as the four schools (Vaibhashikas, etc.), mentioned above, which were speculative. Similarly in Europe a Presbyterian may be a Calvinist, but Presby terianism has reference to Church government aud Calvinism to doctrine. There were in India at this time (1) two vehicles, Maha- and Hinayana, (2) four speculative schools, Vaibhashikas, etc., (3) four disciplinary schools, Mula-sarv&sti- vadins, etc. These three classes are obviously not mutuaUy exclusive. Thus I-Ching approved of (a) the Mahayana, (6) the Madhyamika and Yogacara, which he did not consider inconsistent and (c) the Mula-sarvastivada. xxm] THE CHINESE PILGRIMS 103 bandhu' or sutras such as the Lotus, which I-Ching's master read once a day for sixty years. I-Ching himseU seems to regard the two Vehicles as alternative forms of rehgion, both exceUent in their way, much as a CathoUc theologian might impartially explain the respective advantages of the active and contemplative Hves. "With resolutions rightly formed" he says "we should look forward to meeting the coming Buddha Maitreya. If we wish to gain the lesser fruition (of the Hinayana) we may pursue it through the eight grades of sanctification. But if we learn to foUow the course of the greater fruition (of the Mahayana) we must try to accompUsh our work through long ages 2." I-Ching observes that both Vehicles agree in prescribing the same disciphne, in prohibiting the same offences and enjoining the practice of the noble truths. His views, which are sub- stantiaUy those of Hsuan Chuang*, must be those current in the seventh century when the Hinayana was allowing the Mahayana to overgrow it without resistance, but the relations of the two creeds are sometimes stated differently. For instance the AnguUmaUya sutra*, known only in a Tibetan translation, states that whereas for the Hinayana such formulae as the four truths and the eightfold path are of cardinal importance, the Mahayana does not recognize them, and it is undoubtedly true that the VaipiUya sutras frequently ignore the famiUar doctrines of early Buddhism and hint that they belong to a rudimentary stage of instruction. I-Ching makes no mention of persecution but he deplores the decay of the faith. "The teaching of the Buddha is becoming less prevalent in the world from day to day" he says. "When I compare what I have witnessed in my younger days and what I see to-day in my old age, the state is altogether different and we are bearing witness to this and it is hoped we shall be more 1 I-Ching, transl. Takakusu, p. 186. " Three Asankhya Kalpas. I-Ching, Takakusu's transl. pp. 196-7. He seems to regard the Mahayana as the better way. He quotes Nagarjuna's aUusions to Avalokita and Amitayus with apparent approval; he tells us how one of his teachers worshipped Amitayus and strove to prepare himself for Sukhavati and how the Lotus was the favourite scripture of another. He further teUs us that the Mad hyamika and the Yoga systems are both perfectly correct. ^ Hsiian Chuang speaks of Mahayanists belonging to the Sthavira school. « Quoted by Bockhill, Life of the Buddha, pp. 196 ff. 104 THE MAHAYANA [ch. attentive in future." Though he speaks regretfully of lax or incorrect disciphne, he does not complain of the corruption of the faith by Tantrism and magical practices. He does however deprecate in an exceedingly curious passage the prevalence of rehgious suicide'. Except for progressive decay, the condition of Indian Buddhism as described by the two pUgrims is much the same. Meals were suppHed to monks in the monasteries and it was no longer usual to beg for food in the streets, since the practice is mentioned by I-Ching as exceptional. On Upavasatha days it was the custom for the pious laity to entertain the monks and the meal was sometimes preceded by a rehgious service per formed before an image and accompanied by music. I-Ching describes the musical services with devout enthusiasm. "The priests perform the ordinary service late in the afternoon or in the evening twilight. They come out of the monastery and walk three times round a stupa, offering incense and flowers. Then they all kneel down and one of them who sings well begins to chant hymns describing the virtues of the great Teacher and continues to sing ten or twenty ^lokas. They then return to the place in the monastery where they usuaUy assemble and, when all have sat down, a reciter mounting the Hon-seat (which is near the head priest) reads a short sutra. Among the scriptures for such an occasion the ' Service in three parts ' is often used. This is a selection of Asvaghosha. The first part contains ten slokas of a hymn. The second part is a selection from some scripture consisting of the Buddha's words. Then there is an additional hymn as the third part of the service, of more than ten slokas, being prayers that express the wish to bring one's merits to maturity. After the singing the assembled Bhikshus exclaim Subhashita or Sadhu, that is weU-said or bravo. The reader descends and the Bhikshus in order salute the Hon-seat, the seats of Bodhisattvas and Arhats, and the superior of the monastery^." 1 Chaps, xxxvm and xxxix. He seems to say that it is right for the laity to make an offering of their bodies by burning but not for Bhikshus. The practice is recognized and commended in the Lotus, chap, xxn, which however is a later addition to the original work. 2 I-Ching, transl. Takakusu, pp. 153-4 somewhat abridged. I-Ching (pp. 156-7) speaks of Matricheta as the principal hymn writer and does not identity him with A^vagho.sha. xxm] THE CHINESE PILGRIMS 105 I-Ching also teUs us of the ceremonial bathing of images and prefaces his description by the remark that "the meaning of the Truths is so profound that it is a matter beyond the com prehension of vulgar minds while the ablution of the holy images is practicable for aU. Though the Great Teacher has entered Nirvana yet his image exists and we should worship it with zeal as though in his presence. Those who constantly offer incense and fiowers to it are enabled to purify their thoughts and those who perpetually bathe his image are enabled to over come the sins that involve them in darkness'." He appears to contemplate chiefly the veneration of images of Sakyamuni but figures of Bodhisattvas were also conspicuous features in temples, as we know not only from archaeology but from the biography of Hsuan Chuang, where it is said that worshippers used to throw flowers and silk scarves at the image of Avalokita and draw auguries from the way they feU. Monasteries were HberaUy decorated with statues, carvings and pictures^- They often comprised several courts and temples. Hsiian Chuang says that a monastery in Magadha which he caUs Ti-lo-shi-ka had "four courts with three storeyed halls, lofty terraces and a succession of open passages.... At the head of the road through the middle gate were three temples with disks on the roof and hung with small bells; the bases were surrounded by balustrades, and doors, windows, beams, walls, and stairs were ornamented with gilt work in relief." In the three temples were large images representing the Buddha, Tara and Avalokita. The great centres of Buddhist learning and monastic Hfe, mentioned by both pilgrims, were Valabhi or Balabhi in Gujarat and Naianda. The former was a district rather than a single locality and contained 100 monasteries with 6000 monks of the Sammitiya school. Naianda was in Magadha not far from Gaya. The date of its foundation is unknown but a great temple (though apparently not the first) was built about 485 a.d.* * I beUeve the golden image in the Arakan Pagoda at Mandalay is stiU washed with a ceremonial resembUng that described by I-Ching. " I-Ching says that monasteries commonly had a statue of Mahakala as a guardian deity. ' By the Gupta king, Narasinha Gupta Baladitya. Much information about Naianda wUl be found in Satis Chandra Vidyabhusana's Mediaeval School of Indian Logic, pp. 145-147. Hsiian Chuang (Life, transl. Beal, p. Ill) says that it was 106 THE MAHAYANA [ch. xxm Fa-Hsien mentions a viUage caUed Nala but without indicating that it was a seat of learning. Hence it is probable that the University was not then in existence or at least not celebrated. Hsiian Chuang describes it as containing six monasteries built by various kings and surrounded by an enclosing waU in which there was only one gate. I-Ching writing later says that the estabhshment owned 200 viUages and contained eight haUs with more than 3000 monks. In the neighbourhood of the monastery were a hundred sacred spots, several marked by temples and topes. It was a resort for Buddhists from aU countries and an educational as weU as a reUgious centre. I-Ching says that students spent two or three years there in learning and disputing after which they went to the king's court in search of a govem ment appointment. Successful merit was rewarded not only by rank but by grants of land. Both pUgrims mention the names of several celebrities connected with Naianda. But the worthies of the seventh century did not attain to more than scholastic eminence. The most important Hterary figure of the age is Santideva of whose Hfe nothing is known. His writings however prove that the Buddhism of this period was not a corrupt superstition, but could inspire and nourish some of the most beautiful thoughts which the creed has produced. built 700 years before his time, that is, in the first century B.C. He dwells on the beauty of the buUdings, ponds and flowers. CHAPTER XXIV DECADENCE OF BUDDHISM IN INDIA The theme of this chapter is sad for it is the decadence, degra dation and ultimate disappearance of Buddhism in India. The other great reUgions offer no precise paraUel to this phenomenon but they also do not offer a paraUel to the circumstances of Buddhism at the time when it flourished in its native land. Mohammedanism has been able to maintain itself in comparative isolation: up to the present day Moslims and Christians share the same cities rather than the same thoughts, especially when (as often) they belong to different races. European Christianity after a few centuries of existence had to contend with no rival of approximately equal strength, for the struggle with Moham medanism was chiefly military and hardly concerned the merits of the faiths. But Buddhism never had a similarly paramount and unchaUenged position. It never attempted to extirpate its rivals. It coexisted with a mass of popular superstition which it only gently reprobated and with a powerful hereditary priest hood, both inteUectual and pUant, tenacious of their own ideas and yet ready to countenance almost any other ideas as the price of ruHng. Neither Islam nor Christianity had such an adversary, and both of them and even Judaism resemble Buddhism in having won greater success outside their native lands than in them. Jerusalem is not an altogether satisfactory spectacle to either Christians or Jews'. StiU aU this does not completely explain the disappearance of Buddhism from India. Before attempting to assign reasons, we shaU do weU to review some facts and dates relating to the period of decadence. If we take aU India into consideration the period is long, but in many, indeed in most, districts the process of decay was rapid. In the preceding chapter I have mentioned the accounts of Indian Buddhism which we owe to the Chinese traveUers, Hsiian Chuang and I-Ching. The latter frankly deplores the decay of ^ Written before the war. 108 THE MAHAYANA [oh. the faith which he had witnessed in his own Hfe {i.e. about 650-700 A.D.) but his travels in India were of relatively smaU extent and he gives less local information than previous pilgrims. Hsiian Chuang describing India in 629-645 a.d. is unwiUing to admit the decay but his truthful narrative lets it be seen. It is only of Bengal and the present United Provinces that he can be said to give a favourable account, and the prosperity of Buddhism there was largely due to the personal influence of Harsha'. In central and southern India, he tells us of little but deserted monasteries. It is clear that Buddhism was dying out but it is not so clear that it had ever been the real reUgion of this region. In many parts it did not conquer the population but so to speak built fortresses and left garrisons. It is probable that the Buddhism of Andhra, Kalinga and the south was represented by Httle more than such outposts. They included Amaravati, where portions of the ruins seem assignable to about 150 A.D., and Ajanta, where some of the cave paintings are thought to be as late as the sixth century. But of neither site can we give any continuous history. In southern India the introduction of Buddhism took place under the auspices of Asoka himself, though his inscriptions have as yet been found only in northern Mysore and not in the Tamil country. The Tamil poems Manimegalei and Silappadigaram, especiaUy the former, represent it as prevalent and still preserving much of its ancient simphcity. Even in later times when it had almost completely disappeared from southern India, occasional Buddhist temples were founded. Rajaraja endowed one at Negapatam about 1000 A.D. In 1055 a monastery was erected at Belgami in Mysore and a Buddhist town named Kalavati is mentioned as existing in that state in 1533^. But in spite of such survivals, even in the sixth century Buddhism could not compete in southern India with either Jainism or Hinduism and there are no traces of its existence in the Deccan after 1150. For the Konkan, Maharashtra and Gujarat, Hsiian Chuang's statistics are fairly satisfactory. But in all this region the Sammitiya sect which apparently was nearer to Hinduism than the others was the most important. In Ujjain Buddhism was ' Even at Kanauj , the scene of Harsha' s pious festivities, there were 100 Buddhist monasteries but 200 Deva temples. ^ Rice, Mysore and Coorg from the Inscriptions, p. 203. XXIV] DECADENCE OF BUDDHISM IN INDIA 109 almost extinct but in many of the western states it Hngered on, perhaps only in isolated monasteries, until the twelfth century. Inscriptions found at Kanheri (843 and 851 a.d.), Dambal (1095 A.D.) and in Miraj (1110 a.d.) testify that grants were made to monasteries at these late dates'. But further north the faith had to endure the violence of strangers. Sind was conquered by the Arabs in 712; Gujarat and the surrounding country were invaded by northern tribes and such invasions were always inimical to the prosperity of monasteries. This is even more true of the Panjab, the frontier provinces and Kashmir. The older invaders such as the Yiieh-chih had been favourably disposed to Buddhism, but those who came later, such as the Huns, were predaceous barbarians with Httle rehgion of any sort. In Hsiian Chuang's time it was only in Udyana that Buddhism could be said to be the reUgion of the people and the torrent of Mohammedan invasion which swept continuously through these countries during the middle ages overwhelmed all earHer reUgions, and even Hinduism had to yield. In Kashmir Buddhism soon became corrupt and according to the Rajatarangini 2 the monks began to marry as early as the sixth century. King LaUtaditya (733-769) is credited with having built monasteries as weU as temples to the Sun, but his successors were Sivaites. Bengal, especiaUy western Bengal and Bihar, was the strong hold of decadent Buddhism, though even here hostile influences were not absent. But about 730 a.d. a pious Buddhist named Gopaia founded the Paia dynasty and extended his power over Magadha. The Paias ruled for about 450 years and suppHed a long and devout Une of defenders of the faith. But to the east of their dominions lay the principaUty of Kanauj, a state of varying size and fortunes and from the eighth century onwards a stronghold of Brahmanic learning. The revolution in Hinduism which definitely defeated, though it did not annihilate, Buddhism is generaUy connected with the names of Kumarila Bhatta (c. 750) and Sankara (c. 800). We know the doctrines of these teachers, for many of their works have come down to us, but when we enquire what was their poHtical importance, or the scope and extent of the ' See the note by Biihler in Journ. Pali Text Soe. 1896, p. 108. 2 Rajatarangini, in. 12. 110 THE MAHAYANA [ch. movement which they championed we are conscious (as so often) of the extraordinary vagueness of Indian records even when the subject might appeal to religious and philosophic minds'. Kumarila is said to have been a Brahman of Bihar who abjured Buddhism for Hinduism and raged with the ardour of a proselyte against his ancient faith. Tradition ^ represents him as instigating King Sudhanvan to exterminate the Buddhists. But nothing is known of this king and he cannot have had the extensive empire with which he is credited. Sankara was a Brahman of the south who in a short Hfe found time to write numerous works, to wander over India, to found a monastic order and build four monasteries. In doctrine and discipUne he was more pUant than Kumarila and he assimilated many strong points of Buddhism. Both these teachers are depicted as the successful heroes of pubhc disputa tions in which the interest at stake was considerable. The vanquished had to become a disciple of the vanquisher or to forfeit his Hfe and, if he was the head of an institution, to surrender its property. These accounts, though exaggerated, are probably a florid version of what occurred and we may surmise that the popular faith of the day was generaUy victorious. What violence the rising tide of Hinduism may have wrought, it is hard to say. There is no evidence of any general persecution of Buddhism in the sense in which one Christian sect persecuted another in Europe. But at a rather later date we hear that Jains were persecuted and tortured by Saiva princes both in southern India and Gujarat, and if there were any detailed account, epigraphic or Hterary, of such persecutions' in the eighth and ninth centuries, there would be no reason for doubting it. But no details are forthcoming. Without resorting to massacre, an anti-Buddhist king had in his power many effective methods of hostiUty. He might confiscate or transfer monastic property, or forbid his subjects to support monks. Con sidering the state of Buddhism as represented by Hsiian Chuang and I-Ching it is probable that such measures would suffice to ensure the triumph of the Brahmans in most parts of India. > See for the supposed persecution of Buddhism in India, J.P.T.S. 1896, pp. 87-92 and 107-111 and J.R.A.S. 1898, pp. 208-9. 2 As contained in the Sankara-dig-vijaya ascribed to Madhava and the Sankara- vijaya ascribed to Anandagiri. XXIV] DECADENCE OF BUDDHISM IN INDIA 111 After the epoch of Sankara, the history of Indian Buddhism is confined to the Pala kingdom. Elsewhere we hear only of isolated grants to monasteries and similar acts of piety, often striking but hardly worthy of mention in comparison with the enormous number of Brahmanic inscriptions. But in the Paia kingdom' Buddhism, though corrupt, was fiourishing so far as the number of its adherents and royal favour were concerned. Gopaia founded the monastery of Odontapuri or Udandapura, which according to some authorities was in the town of Bihar. Dharmapaia the second king of the dynasty (c. 800 a.d.) built on the north bank of the Ganges the even more celebrated University of VikramasUa^, where many commentaries were composed. It was a centre not only of tantric learning but of logic and grammar, and is interesting as showing the connection between Bengal and Tibet. Tibetans studied there and Sanskrit books were translated into Tibetan within its cloisters. Dharma paia is said to have reigned sixty -four years and to have held his court at Patna, which had fallen into decay but now began to revive. According to Taranatha his successor Devapaia built Somapuri, conquered Orissa and waged war with the unbeUevers who had become numerous, no doubt as a result of the preaching of Saidiara. But as a rule the Paias, though they favoured Buddhism, did not actively discourage Hinduism. They even gave grants to Hindu temples and their prime ministers were generaUy Brahmans who* used to erect non-Buddhist images in Buddhist shrines. The dynasty continued through the eleventh century and in this period some information as to the condition of Indian Buddhism is afforded by the relations between Bengal and Tibet. After the persecution of the tenth century Tibetan Buddhism was revived by the preaching of monks from Bengal. Mahipala then occupied the throne (c. 978- 1030) and during his reign various learned men accepted invita- 1 Taranatha in his twenty-eighth and foUowing chapters gives an account, unfortunately very confused, of the condition of Buddhism under the Pala dynasty. See also B. K. Sarkar, Folklore Element in Hindu Culture, chap, xn, in which there are many interesting statements but not sufficient references. ' See Vidyabhusana's Mediaeval School of Indian Logic, p. 150, for an account of this monastery which was perhaps at the modem Parthaghata. I have found no account of what happened to Naianda in this period but it seems to have dis appeared as a seat of learning. ^ See Taranatha, chap, xxvni. 112 THE MAHAYANA [ch. tions to Tibet. More celebrated is the mission of Atisa, a monk of the Vikramasila monastery, which took place about 1038. That these two missions should have been invited and despatched shows that in the eleventh century Bengal was a centre of Buddhist learning. Probably the numerous Sanskrit works preserved in Tibetan translations then existed in its monasteries. But about the same time the power of the Paia dynasty, and with it the infiuence of Buddhism, were curtailed by the estab lishment of the rival Sena dynasty in the eastern provinces. StiU, under Ramapaia, who reigned about 1100, the great teacher Abhayakara was an ornament of the Mahayana. Taranatha' says that he corrected the text of the scriptures and that in his time there were many Pandits and resident Bhikshus in the monasteries of Vikramasila, Bodh-Gaya and Odontapuri. There is thus every reason to suppose that in the tweUth century Buddhism still flourished in Bihar, that its clergy numbered several thousands and its learning was held in esteem. The blow which destroyed its power was struck by a Mohammedan invasion in 1 193. In that year Ikhtiyar-ud-Din Muhammad^, a general of Kutb-ud-Din, invaded Bihar with a band of only two hundred men and with amazing audacity seized the capital, which, consisting chiefiy of palaces and monasteries, collapsed without a blow. The monks were massacred to a man, and when the victors, who appear not to have understood what manner of place they had captured, asked the meaning of the Hbraries which they saw, no one was found capable of reading the books*. It was in 1193 also that Benares was conquered by the Mohammedans. I have found no record of the sack of the monastery at Sarnath but the ruins are said to show traces of fire and other indications that it was overwhelmed by some sudden disaster. The Mohammedans had no special animus against Buddhism. They were iconoclasts who saw merit in the destruction of images and the slaughter of idolaters. But whereas Hinduism was spread over the country. Buddhism was concentrated in 1 Chap. XXXVI. It is interesting to notice that even at this late period he speaks of Hinayanists in Bengal. 2 Often caUed Muhammad Bakhtyar but Bakhtyar seems to have been reaUy his father's name. ' Raverty, Tabat-i-Nasiri, p. 552. "It was discovered that the whole of that fortress and city was a coUege and in the Hindi tongue they caU a coUege Bihar," XXIV] DECADENCE OF BUDDHISM IN INDIA 113 the great monasteries and when these were destroyed there remained nothing outside them capable of withstanding either the violence of the MosHms or the assimilative influence of the Brahmans. Hence Buddhism suffered far more from these invasions than Hinduism but still vestiges of it lingered long' and exist even now in Orissa. Taranatha says that the immediate result of the MosHm conquest was the dispersal of the surviving teachers and this may explain the sporadic occurrence of late Buddhist inscriptions in other parts of India. He also tells us that a king named Cangalaraja restored the ruined Buddhist temples of Bengal about 1450. Elsewhere ^ he gives a not discouraging picture of Buddhism in the Deccan, Gujarat and Rajputana after the MosHm conquest of Magadha but adds that the study of magic became more and more prevalent. In the Hfe of Caitanya it is stated that when travelUng in southern India (about 1510 a.d.) he argued with Buddhists and confuted them, apparently somewhere in Arcot*. Manuscripts preserved in Nepal indicate that as late as the fifteenth or sixteenth century BengaU copyists wrote out Buddhist works, and there is evidence that Bodh-Gaya continued to be a place of pilgrimage. In 1585 it was visited by a Nepalese named Abhaya Raja who on his retum erected in Patau a monastery imitated from what he had seen in Bengal, and in 1777 the Tashi Lama sent an embassy. But such instances prove Uttle as to the rehgion of the surrounding Hindu population, for at the present day numerous Buddhist pilgrims, especiaUy Burmese, frequent the shrine. The control of the temple passed into the hands of the Brahmans and for the ordinary BengaH Buddha became a member of India's numerous pantheon. Pandit Haraprasad Sastri mentions a singular poem caUed Buddhacaritra, completed in 1711 and celebrating an incarnation of Buddha which appar ently commenced in 1699 and was to end in the reappearance of the golden age. But the being called Buddha is a form of Vishnu and the work is as strange a jumble of rehgion as it is ( ' Many of them have been coUected by Pandit Haraprasad Sastri in Jour, As, Soe, Bengal, 1895, pp. 55 ff. and in his Discovery of living Buddhism in, Bengal, Calcutta, 1897. ^ Chap. XL ad fin. Is the Ramacandra whom he mentions the last Yadava '^King (about 1314) ? Taranatha speaks of his son. ' Caitanya-oarit-amrita, chap, vn, transl. by Jadunath Sarkar, p. 85. This biography was written in 1582 by Krishnadas. Caitanya died in 1533. B. n. 8 114 THE MAHAYANA [ch. of languages, being written in "a curious medley of bad Sanskrit, bad Hindi and bad Bihari." It is chiefly in Orissa that traces of Buddhism can stiU be found within the Hmits of India proper. The Saraks of Baramba, Tigaria and the adjoining parts of Cuttack describe themselves as Buddhists'. Their name is the modern equivalent of Sravaka and they apparently represent an ancient Buddhist community which has become a sectarian caste. They have little knowledge of their religion but meet once a year in the cave temples of Khandagiri, to worship a deity called Buddhadeva or Caturbhuja. AU their ceremonies commence with the formula Ahimsd parama Aharrrwt and they respect the temple of Puri, which is suspected of having a Buddhist origin. Nagendranath Vasu has published some interesting detaUs as to the survival of Buddhist ideas in Orissa^. He traces the origin of this hardy though degraded form of Mahayanism to Ramai Pandit*, a tantric Acarya of Magadha who wrote a work called Sunya Purana which became popular. Orissa was one of the regions which offered the longest resistance to Islam, for it did not succumb until 1568. A period of Sivaism in the tenth and eleventh centuries is indicated by the temples of Bhubanesvar and other monuments. But in the twelfth and thirteenth centuries the reigning dynasty were worshippers of Vishnu and built the great temples at Puri and Konarak, dedicated to Jagannatha and Siirya-narayana respectively. We do not however hear that they persecuted Buddhism and there are reasons for thinking that Jagannatha is a form of the Buddha* and that the temple at Puri was originaUy a Buddhist site. It ' Census of India, 1901 : vol. vi. Bengal, pp. 427-430. " The Archaeological Survey of Mayurdbhanj (no date? 1911), vol. I. pp. cv-cckui.- The part containing an account of Buddhism in Orissa is also printed separately with the title Modern Buddhism, 1911. , ' For Ramai Pandit see Dinesh Chandra Sen, Hist, Bengali Language and Lit, pp. 30-37, and also B. K. Sarkar, Folklore Element in Hindu Culture, p. 192, and elsewhere. He appears to have been born at the end of the tenth century aud though the Siinya Puraija has been re-edited and interpolated parts of it are said to be in very old BengaU. * Nagendranath Vasu quotes a couplet from the Mahabharata of the poet ' Saraladasa : "I pay my humble respects to the incarnation of Buddha who in the form of Buddha dwells in the Nilaeala, i.e. Puri." The Imperial Gazetteer of ¦ India (s.v. Puri Town) states that in modem representations of Vishnu's ten ' avataras, the ninth, or Buddliavatara, is sometimes represented by Jagannatha. I xxiv] decadence of BUDDHISM IN INDIA 115 is said that it contains a gigantic statue of the Buddha before which a waU has been built and also that the image of Jagannatha, which is Httle more than a log of wood, is really a case enclos ing a Buddhist reUc. King Prataparudra (f 1529) persecuted Buddhism, which impHes that at this late date its adherents were sufficiently numerous to attract attention. Either at the beginning of his reign or before it there flourished a group of six poets of whom the principal were Acyutananda Dasa and Caitanya Dasa'. Their works are nominally devoted to the celebration of Krishna's praises and form the chief vernacular scripture of the Vaishnavas in Orissa but in them Krishna, or the highest form of the deity by whatever name he is called, is constantly identified with Sunya or the Void, that favourite term of Mahayanist philosophy. Passages from them are also quoted stating that in the Kali age the followers of the Buddha must disguise themselves ; that there are 3000 crypto-buddhists hidden in various parts of Orissa, that Hari has been incarnate in many Buddhas and that the Buddha wiU appear again on earth. The phrase "I take refuge in the Buddha, in Mata Adisakti (= Dharma) and in the Sangha" is also quoted from these works and Caitanya Dasa describes five Vishnus, who are apparently identical with the five Dhyani Buddhas^. Taranatha states that the last king of Orissa, Mukunda Deva, who was overthrown by the Mohammedans in 1568, was a Buddhist and founded some temples and monasteries. In the seventeenth century, there flourished a Buddhist poet named Mahadevadasa*, and the Tibetan pilgrim Buddhagupta visited among other sites the old capital of Mayurabhanja and saw a stupa there. It is claimed that the tribe known as Bathuris or Bauris have always been crypto-buddhists and have preserved their ancient customs. They are however no credit to their rehgion, for one of their principal ceremonies is hook-swinging*. The doctrine of the Bathuris is called Mahima Dharma and experienced an interesting revival in 1875 ^. A bUnd man named Blnma Bhoi had a vision of the Buddha who restored his sight 1 I give the dates or the authority of Narandra Nath while thinking that they may be somewhat too early. The two authors named wrote the 6unya Samhita , and Nirgu^ia Mahatmya respectively. " l.c. clxxvi ff., ccxix-ccxxiii, oexxxi. " Author of a poem oaUed Dharmagita. * l.c. cxvi ff. and ocxxxU. ' l.c. ccxxxiv ff. 8—2 116 THE MAHAYANA [ch. and bade him preach the law. He attracted some thousands of adherents and led a band to Puri proclaiming that his mission was to bring to Hght the statue of Buddha concealed in the temple. The Raja resisted the attempt and the foUowers of Bhima Bhoi were worsted in a sanguinary encounter. Since that time they have retired to the more remote districts of Orissa and are said to hold that the Buddha wiU appear again in a new incarnation. They are also called Kumbhipatias and according to the last census of India (1911) are hostile to Brahmans and probably number about 25,000. Traces of Buddhism also survive in the worship of a deity caUed Dharma-Raja or Dharma-Thakur which stUl prevaUs in western and southern Bengal'. Priests of this worship are usually not Brahmans but of low caste, and Haraprasad thinks that the laity who foUow it may number "several milHons." Though Dharma has come to be associated with the goddess of small-pox and is beUeved even by his adorers to be a form of Vishnu or of Siva, yet Dhyana, or meditation, forms a part of his worship and the prayers and literature of the sect retain some traces of his origin. Thus he is said to be highly honoured in Ceylon and receives the epithet Sunyamurti. A corrupt form of Buddhism still exists in Nepal^. This country when first heard of was in the hands of the Nevars who have preserved some traditions of a migration from the north and are akin to the Tibetans in race and language, though Hke many non-Aryan tribes they have endeavoured to invent for themselves a Hindu pedigree. Buddhism was introduced under Asoka. As Indian influence was strong and communica tion with Tirhut and Bengal easy, it is probable that Buddhism in Nepal reflected the phases which it underwent in Bengal. A Nepalese inscription of the seventh century gives a Hst of shrines of which seven are Sivaite, six Buddhist and four Vishnuite*. After that date it was more successful in main- ^ See Haraprasad Sastri, l.c. He gives a curious account of one of his temples in Calcutta. See also B. K. Sarkar, Folklore Element in Hindu Culture for the decadence of Buddhism in Bengal aud its survival in degenerate forms. ' See B. H. Hodgson, Essays on the languages, literature and religion of Nepal and Tibet, 1874. For the religion of Nepal see also Wright, History of Nepal, 1877; C. BendaU, Journal of Literary and Archaeological Research in Nepal, 1886; Rajen dralal Mitra, Sanskrit Buddhist literature of Nepal; and especiaUy S. L^vi, Le Nipal, 3 vols. 1905-8. ' S. L^vi in J.A. n. 1904, p. 225. He gives the date as 627. XXIV] DECADENCE. OF BUDDHISM IN INDIA 117 tainmg itseff, for it did not suffer from Mohammedan attacks and was less exposed to the assimilative influence of Brahman ism. That influence however, though operating in a foreign country and on people not bred among Brahmanic traditions, was nevertheless strong. In 1324 the king of Tirhut, being expeUed thence by Mohammedans, seized the throne of Nepal and brought with him many learned Brahmans. His dynasty was not permanent but later in the fourteenth century a subsequent ruler, Jayasthiti, organized society and rehgion in consultation with the Brahman immigrants. The followers of the two reUgions were arranged in paraUel divisions, a group of Buddhists classified according to occupation corresponding to each Hindu caste, and appropriate rules and ceremonies were prescribed for the different sections. The code then established is stUl in force in essentials and Nepal, being intellectually the pupU of India, has continued to receive such new ideas as appeared in the plains of Bengal. When these ascended to the mountain vaUeys they were adopted, with free modification of old and new material alike, by both Buddhists and Hindus, but as both sects were geographicaUy isolated, each tended to resemble the other more than either resembled normal Buddhism or Hindmsm. NaturaUy the new ideas were mainly Brahmanic and Buddhism had no chance of being fortified by an importa tion of even moderately orthodox doctrine. In the fourteenth century arose the community of wandering ascetics called Nathas who were reverenced by Hindus and Buddhists ahke. They rejected the observances of both creeds but often com bined their doctrines and, though disavowed by the Brahmans, exercised a considerable influence among the lower castes. Some of the pecuhar deities of Nepal, such as Matsyendranath, have attributes traceable to these wanderers. In 1769 Nepal was conquered by the Gurkhas. This tribe seems related to the Tibetan stock, as are the Nevars, but it had long been hinduized and claimed a Rajput ancestry. Thus Gurkha rule has favoured and accelerated the hinduizing of Nepalese Buddhism. Since the time of Hodgson the worship of the Adi-Buddha, or an original divine Buddha practicaUy equivalent to God, has been often described as characteristic of Nepalese reUgion and such a worship undoubtedly exists. But recent accounts indicate that it is not prominent and also that it can hardly be con- 118 THE MAHAYANA [ch. sidered a distinct type of monotheistic Buddhism. The idea that the five Dhyani-Buddhas are emanations or manifestations of a single primordial Buddha-spirit is a natural development of Mahayanist ideas, but no definite statement of it earHer than the Kalacakra Hterature is forthcoming, though many earlier works point towards it'. In modern Nepal the chief temple of the Adi-Buddha is on the hill of Svayambhu (the self -existent) near Katmandu. According to a legend preserved in the Svayambhii Purana, a special divine manifestation occurred in ancient times on an adjoining lake; a miraculous lotus arose on its surface, bearing an image, over which a Caitya was subse quently erected. The shrine is greatly venerated but this Adi-Buddha, or Svayambhii, does not differ essentially from other miraculous images in India which are said not to consist of ordinary matter but to embody in some special way the nature of a deity. The rehgion of Nepal is less remarkable for new developments of Buddhism than for the singular fusion of Buddhism with Hinduism which it presents and which helps us to understand what must have been the last phase in Bengal. The Nepalese Brahmans tolerate Buddhism. The Nepaia- mahatmya says that to worship Buddha is to worship Siva, and the Svayambhii Purana returns the compUment by recom mending the worship of Pasupati^. The official itinerary of the Hindu pilgrim includes Svayambhii, where he adores Buddha under that name. More often the two religions adore the same image under different names: what is Avalokita to the one is Mahakala to the other. Durga is explained as being the incarna tion of the Prajna-paramita and she is even identified with the Adi-Buddha. The Nepalese pantheon Hke the Tibetan contains three elements, often united in modern legends : firstly aboriginal deities, such as Nagas and other nature spirits: secondly definitely Buddhist deities or Bodhisattvas of whom Manjusri receives the most honour : thirdly Hindu deities such as Ganesa and Krishna. The popular deity Matsyendranath appears to combine aU three elements in his own person. Modern accounts of Nepal leave the impression that even ' The doctrine of the Adi-Buddha is fully stated in the metrical version of the Kftran^a-vyuha which appears to be a later paraphrase of the prose edition. See Wintemitz, Gesch. Ind. Lit. ii. i. 238. * Compare the fusion of Sivaism and Buddhism in Java. XXIV] DECADENCE OF BUDDHISM IN INDIA 119 corrupt Buddhism is in a bad way, yet the number of reUgious estabUshments is considerable. Celibacy is not observed by their inmates, who are caUed banras (bandyas). On entering the order the novice takes the ancient vows but after four days he returns to his tutor, confesses that they are too hard for him and is absolved from his obhgations. The classes known as Bhikshus and Gubharjus officiate as priests, the latter being the higher order. The principal ceremony is the offering of melted butter. The more learned Gubharjus receive the title of Vajracarya' and have the sole right of officiating at marriages and funerals. There is Httle learning. The oldest 'scriptures in use are the so-caUed nine Dharmas 2. Hodgson describes these works as much venerated and Rajendralal Mitra has analysed them, but Sylvain Levi heard Httle of them in 1898, though he mentions the recitation of the Prajna-paramita. The Svayambhii Purana is an account of the manifestation of the Adi-Buddha written in the style of those portions of the Brahmanic Puranas which treat of the glories of some sacred place. In its present form it can hardly be earUer than the sixteenth century a.d. The Nepaia-mahatmya is a similar work which, though of Brahmanic origin, puts Buddha, Vishnu and Siva on the same footing and identifies the first with Krishna. The Vagvati-mahatmya* on the other haijd is strictly Sivaite and ignores Buddha's claims to worship. The Vamsavah, or Chronicle of Nepal, written in the Gurkha language (Parbatiya) is also largely occupied with an account of sacred sites and buildings and exists in two versions, one Buddhist, the other Brahmanical. But let us return to the decadence of Buddhism in India. It is plain that persecution was not its main cause nor even very important among the accessory causes. The available records contain clearer statements about the persecution of Jainism than of Buddhism but no doubt the latter came in for some rough handUng, though not enough to annihilate a vigorous sect. Great numbers of monasteries in the north were demoHshed by the Huns and a similar catastrophe brought 1 Or Vajracarya-arhat-bhikshu-buddha, which in itseU shows what a medley Nepalese Buddhism has become. 2 See above chap. xx. for some account of these works. ' Dedicated to the sacred river Vagvati or Bagmati. 120 THE MAHAYANA [ch. about the coUapse of the Church in Bihar. But this last incident cannot be called rehgious persecution, for Muhammad did not even know what he was destroying. Buddhism did not arouse more animosity than other Indian rehgions: the significant feature is that when its temples and monasteries were demoUshed it did not Hve on in the hearts of the people, as did Hinduism with all its faults. The relation between the laity and the Church in Buddhism is curious and has had serious consequences for both good and evil. The layman "takes refuge" in the Buddha, his law and his church but does not swear exclusive aUegiance: to foUow supplementary observances is not treasonable, provided they are not in themselves objectionable. The Buddha prescribed no ceremonies for births, deaths and marriages and apparently expected the laity to continue in the observance of such rites as were in use. To-day in China and Japan the good layman is little more than one who pays more attention to Buddhism than to other faiths. This charitable pUancy had much to do with the victories of Buddhism in the Far East, where it had to struggle against strong prejudices and could hardly have made its way if it had been intolerant of local deities. But in India we see the disadvantages of the omission to make the laity members of a special corporation and the survival of the Jains, who do form such a corporation, is a clear object lesson. Social Hfe in India tends to combine men in castes or in com munities which if not castes in the technical sense have much the same character. Such communities have great vitahty so long as they maintain their pecuhar usages, but when they cease to do so they soon disintegrate and are reabsorbed. Buddhism from the first never took the form of a corporation. The special community which it instituted was the sangha or body of monks. Otherwise, it aimed not at founding a sect but at including aU the world as lay believers on easy terms. This principle worked well so long as the faith was in the ascendent but its effect was disastrous when decline began. The Une dividing Buddhist laymen from ordinary Hindus became less and less marked: distinctive teaching was found only in the monasteries: these became poorly recruited and as they were graduaUy deserted or destroyed by Mohammedans the religion of the Buddha disappeared from his native land. XXIV] DECADENCE OF BUDDHISM IN INDIA 121 Even in the monasteries the doctrine taught bore a closer resemblance to Hinduism than to the preaching of Gotama and it is this absence of the protestant spirit, this pHant adaptability to the ideas of each age, which caused Indian Buddhism to lose its individuahty and separate existence. In some locaHties its disappearance and absorption were preceded by a monstrous phase, known as Tantrism or Saktism, in which the worst elements of Hinduism, those which would have been most repulsive to Gotama, made an unnatural alhance with his church. I treat of Tantrism and Saktism in another chapter. The original meaning of Tantra as appUed to Uterary compositions is a simpUfied manual'. Thus we hear of Vishnuite Tantras and in this sense there is a real similarity between Buddhist and tantric teaching, for both set aside Brahmanic tradition as needlessly compHcated and both profess to preach a simple and practical road to salvation. But in Hinduism and Buddhism aUke such words as Tantra and tantric acquire a special sense and imply the worship of the divine energy in a female form called by many names such as KaU in the former, Tara in the latter. This worship which in my opinion should be called Saktism rather than Tantrism combines many elements: ancient, savage superstitions as well as ingenious but fanciful speculation, but its essence is always magic. It attempts to attain by magical or sacramental formulae and acts not only prosperity and power but salvation, nirvana and union with the supreme spirit. Some of its sects practise secret immoral rites. It is sad to confess that degenerate Buddhism did not remain uncorrupted by such abuses. It is always a difficult and speculative task to trace the early stages of new movements in Indian rehgion, but it is clear that by the eighth century and perhaps earHer the Buddhism of Bihar and Bengal had fallen a prey to this influence. Apparently the pubUc ritual in the Viharas remained unchanged and the usual language about nirvdna Sind SUnyatd was not discarded, but it * Hardly any Buddhist Tantras have been edited in Europe. See BendaU, Subhdehita-sangraha for a coUection .of extracts (also pubUshed in Musion, 1905), and De la Vallee Poussin, Bouddhisme, Mudes et Matiriaux. Id. Pancakrama, 1896. While this book was going through the press I received the Tibetan Tantra caUed Shrichakrasambhara (Avalon's Tantric Texts, vol. vn) with introduction by A. Avalou, but have not been able to make use of it. 122 THE MAHAYANA [ch. was taught that those who followed a certain curriculum could obtain salvation by magical methods. To enter this curriculum it was necessary to have a qualified teacher and to receive from him initiation or baptism (abhisheka). Of the subsequent rites the most important is to evoke one of the many Buddhas or Bodhisattvas recognized by the Mahayana and identify oneself with him'. He who wishes to do this is often caUed a sadhaka or magician but his achievements, Hke many Indian miracles, are due to seU-hypnotization. He is directed to repair to a lonely place and offer worship there with flowers and prayers. To this office succeed prolonged exercises in meditation which do not depart much from the ancient canon since they include the four Brahma-viharas. Their object is to suppress thought and leave the mind empty. Then the sadhaka fiUs this void with the image of some Bodhisattva, for instance Avalokita. This he does by uttering mystic syUables called bija or seed, because they are supposed to germinate and grow into the figures which he wishes to produce. In this way he imagines that he sees the emblems of the Bodhisattva spring up round him one by one and finaUy he himseU assumes the shape of Avalokita and becomes one with him. Something similar stiU exists in Tibet where every Lama chooses a tutelary deity or Yi-dam whom He summons in visible form after meditation and fasting^. Though this procedure when set forth methodicaUy in a mediaeval manual seems an absurd travesty of Buddhism, yet it has Hnks with the early faith. It is admitted in the Pitakas that certain forms of meditation* lead to union with Brahma and it is no great change to make them lead to union with other supernatural beings. Still we are not here breathing the atmosphere of the Pitakas. The object is not to share Brahma's heaven but to become temporarily identified with a deity, and this is not a byway of religion but the high road. But there is a further stage of degradation. I have already mentioned that various Bodhisattvas are represented as accom panied by a female deity, particularly Avalokita by Tara. The 1 See Fouoher, Iconographie bouddhique, pp. 8 ff. De la Valine Poussin, Boud dhisme, StudeJi et Matiriaux, pp. 213 ff. For Japanese tantric ceremonies see the Si-Do-In-Dzon in the Annates du Musie Guimet, vol. vm. " In ancient Egypt also the Kher heb or magician-priest claimed the power of becoming various gods. See Budge, Osiris, ii. 170 and Wiedemann, Magic im alten Aegypten, 13 ff. ' The BrahmS-viharas. E.g. Dig. Nik. xm. XXIV] DECADENCE OF BUDDHISM IN INDIA 123 mythological and metaphysical ideas which have grown up round Siva and Durga also attached themselves to these couples. The Buddha or Bodhisattva is represented as enjoying nirvana because he is united to his spouse, and to the three bodies already enumerated is added a fourth, the body of perfect bUss'. Sometimes this idea merely leads to further developments of the practices described above. Thus the devotee may imagine that he enters into Tara as an embryo and is born of her as a Buddha^. More often the argument is that since the bhss of the Buddha consists in union with Tara, nirvana can be obtained by sexual union here, and we find many of the tantric wizards represented as accompanied by female companions. The adept should avoid all action but he is beyond good and evil and the dangerous doctrine that he can do evil with impunity, which the more respectable sects repudiate, is expressly taught. The sage is not defiled by passion but conquers passion by passion; he should commit every infamy: he should rob, lie and kill Buddhas*. These crazy precepts are probably little more than a speculative appUcation to the moral sphere of the doctrine that aU things are non-existent and hence equivalent. But though tantrists did not go about robbing and murdering so freely as their principles aUowed, there is some evidence that in the period of decadence the morality of the Bhikshus had faUen into great discredit. Thus in the aUegorical Vishnuite drama caUed Prabodhacandrodaya and written at Kalanjar near the end of the eleventh century Buddhists and Jains are repre sented as succumbing to the temptations of inebriety and voluptuousness. It is necessary to mention this phase of decadence but no good purpose would be served by dweUing further on the absurd and often disgusting prescriptions of such works as the Tathagata-guhyaka. If the European reader is incUned to condemn unreservedly a reUgion which even in decrepitude could find place for such monstrosities, he should remember that the aberrations of Indian religion are due not to its ' Mahasukhakaya or vajrakaya. " De la VaUde Poussin, Bouddhisme, Studes et Matiriaux, p. 153. ' See Suhhd.shita-sa'hgraha edited by BendaU. Part n. pp. 29 ff. especiaUy p. 41. Parasvahara^iam karyam paradaranishevanam Vaktavyam canritam nityam sarvabuddhaifi^ca ghatayet. See also Tathagata-guhyaka in Rajendralal Mitra's Sanskrit Literature in Nepal, pp. 261-264. 124 THE MAHAYANA [ch. inherent depravity, but to its universaUty. In Europe those who foUow disreputable occupations rarely suppose that they have anything to do with the Church. In India, robbers, murderers, gamblers, prostitutes, and maniacs all have their appropriate gods, and had the Marquis de Sade been a Hindu he would probably have founded a new tantric sect. But though the details of Saktism are an unprofitable study, it is of some importance to ascertain when it first invaded Buddhism and to what extent it superseded older ideas. Some critics' seem to imply — for their statements are not very expHcit — that Saktism formed part if not of the teaching of the Buddha, at least of the medley of beUefs held by his disciples. But I see no proof that Saktist beliefs — that is to say erotic mysticism founded on the worship of goddesses — were prevalent in Magadha or Kosala before the Christian era. Although Siri, the goddess of luck, is mentioned in the Pitakas, the popular deities whom they bring on the scene are almost exclusively masculine 2. And though in the older Brahmanic books there are passages which might easily become tantric, yet the transition is not made and the important truths of rehgion are kept distinct from unclean rites and thoughts. The Bfihad-aranyaka contains a chapter which hardly admits of translation but the object of the practices inculcated is simply to ensure the birth of a son. The same work (not without analogies in the ecstatic utterances of Christian saints) boldly compares union with the Atman to the bhss of one who is embraced by a beloved wife, but this is a mere illustration and there is no hint of the doctrine that the goal of the rehgious Hfe is obtainable by maithuna. StiU such passages, though innocent in themselves, make it easy to see how degrading superstitions found an easy entrance into the noblest edifices of Indian thought and possibly some heresies condemned in the Kathavatthu* indicate that even at this early date the Buddhist Church was contaminated by erotic fancies. But, if so, there is no evidence that such malpractices were widespread. The * For instance De la Vallee Poussin in his Bouddhisme, Etudes et Matiriaux, 1896. In his later work, Bouddhisme, Opinions sur I'histoire de la dogmatique, he modifies his earUer views. ° See Dig. Nik. xx. and xxxn. ' Kathav. xxin. 1 and 2. XXIV] DECADENCE OF BUDDHISM IN INDIA 125 appendices to the Lotus' show that the worship of a many- named goddess, invoked as a defender of the faith, was beginning to be a recognized feature of Buddhism. But they contain no indications of left-handed Tantrism and the best proof that it did not become prevalent until much later is afforded by the narratives of the three Chinese pilgrims who all describe the condition of rehgion in India and notice anything which they thought singular or reprehensible. Fa-Hsien does not mention the worship of any female deity 2, nor does the Life of Vasu bandhu, but Asanga appears to aUude to Saktism in one passage *. Hsiian Chuang mentions images of Tara but without hinting at tantric ritual, nor does I-Ching aUude to it, nor does the evidence of art and inscriptions attest its existence. It may have been known as a form of popular superstition and even have been practised by individual Bhikshus, but the silence of I-Ching makes it improbable that it was then countenanced in the schools of Magadha. He complains* of those who neglect the Vinaya and "devote their whole attention to the doctrine of nothingness," but he says not a word about tantric abuses"- The change probably occurred in the next haU century* for Padma-Sambhava, the founder of Lamaism who is said to have resided in Gaya and Naianda and to have arrived in Tibet in 747 A.D., is represented by tradition as a tantric wizard, and about the same time translations of Tantras begin to appear in Chinese. The translations of the sixth and seventh centuries, including those of I-Ching, comprise a considerable though not preponderant number of Dharanis. After the seventh century * These appendices are later additions to the original text but they were trans lated into Chinese in the third century. Among the oldest Sanskrit MSS. from Japan is the Ushnisha-vijaya-dharani and there is a goddess with a simUar name. But the Dharani is not Saktist. See text in Anec. Oxon. Aryan series. ^ He speaks of Kwan-shih-yin but this is probably the male Avalokita. ' Mahayana-siitralaukara, ix. 46. Of course there may be many other aUusions in yet unedited works of Asanga but it is noticeable that this aUusion to maithuna is only raade in passing and is not connected with the essence of his teaching. * Transl. Takakusu, p. 51. ° Taranatha, chap, xxn seems also to assign a late origin to the Tantras though his remarks are neither clear nor consistent with what he says in other passages. He is doubtless right in suggesting that tantric rites were practised surreptitiously before they were recognized openly. « It is about this time too that we hear of Tantrism in Hinduisra. In the drama Malati and Madhava (c. 730 a.d.) the heroine is kidnapped and is about to be sacrificed to the goddess Cauda when she is rescued. 126 THE MAHAYANA [ch. these became very numerous and several Tantras were also translated'. The inference seems to be that early in the eighth century Indian Buddhists officially recognized Tantrism. Tantric Buddhism was due to the mixture of Mahayanist teaching with aboriginal superstitions absorbed through the medium of Hinduism, though in some cases there may have been direct contact and mutual influence between Mahayanism and aboriginal beUefs. But as a rule what happened was that aboriginal deities were identified with Hindu deities and Bud dhism had not sufficient independence to keep its own pantheon distinct, so that Vairocana and Tara received most of the attributes, brahmanic or barbarous, given to Siva or KaU. The worship of the goddesses, described in their hinduized form as Durga, Kali, etc., though found in most parts of India was specially prevalent in the sub -Himalayan districts both east and west. Now Padma-Sambhava was a native of Udyana or Swat and Taranatha represents the chief Tantrists ^ as coming from there or visiting it. Hsiian Chuang* teUs us that the inhabitants were devout Mahayanists but speciaUy expert in magic and exorcism. He also describes no less than four sacred places in it where the Buddha in previous births gave his flesh, blood or bones for the good of others. Have we here in a Buddhist form some ancient legend of dismemberment Hke that told of Sati in Assam? Of Kashmir he says that its rehgion was a mixture of Buddhism with other beUefs*. These are precisely the con ditions most favourable to the growth of Tantrism and though 1 See the latter part of Appendix ii in Nanjio's Catalogue. ' E.g. Lalitavajra, LUavajra, Buddha^auti, Ratnavajra. Taranatha also (tr. Schiefner, p. 264) speaks of Tantras "Welche aus Udyana gebracht und nie in Indien gewesen sind." It is also noticeable, as Griinwedel has pointed out, that many of the siddhas or sorcerers bear names which have no meaning in Aryan languages : Bir-va-pa, Na-ro-pa, Lui-pa, etc. A curious late tradition represents Saktism as coming from China. See a quotation from the Mahacinatantra in tha Archaeological Survey of Mayurabhanj, p. xiv. Either China is here used loosely for some country north of the Himalayas or the story is pure fancy, for with rare exceptions (for instance the Lamaism of the Yiian dynasty) the Chinese seem to have rejected Saktist works or even to have expurgated them, e.g. the Tathagata- guhyaka. ' His account of Udyana and Kashmir wUl be found in Watters, chapters vn and vm. ' Traces of Buddhism stiU exist, for according to Biihler the Nilaraata Purana orders the image of Buddha to be worshipped on Vaisakha 15 to the accompani ment of recitations by Buddhist ascetics. XXIV] DECADENCE OF BUDDHISM IN INDIA 127 the bulk of the population are now Mohammedans, witchcraft and sorcery are stiU rampant. Among the Hindu Kashmiris' the most prevalent reUgion has always been the worship of Siva, especiaUy in the form representing him as haU male, half female. This cult is not far from Saktism and many allusions ^ in the Rajatarangini indicate that left-hand worship was known, though the author satirizes it as a corruption. He also several times mentions* Matri-cakras, that is circles sacred to the Mothers or tantric goddesses. In Nepal and Tibet tantric Buddhism is fuUy developed but these countries have received so much from India that they exhibit not a parallel growth, but late Indian Tantrism as imported ready-made from Bengal. It is here that we come nearest to the origins of Tantrism, for though the same behefs may have flourished in Udyana and Kashmir they did not spread much in the Panjab or Hindustan, where their progress was hindered at first by a healthy and vigorous Hinduism and subsequently by Mohammedan in vasions. But from 700 to 1197 a.d. Bengal was remote alike from the main currents of Indian rehgion and from foreign raids: Httle Aryan thought or learning leavened the local superstitions which were infecting and stifling decadent Bud dhism. Hsiian Chuang informs us that Bhaskaravarma king of Kamariipa* attended the fetes celebrated by Harsha in 644 a.d. and inscriptions found at Tezpur indicate that kings with Hindu names reigned in Assam about 800 a.d. This is agreeable to the supposition that an amalgamation of Sivaism and aboriginal rehgion may have been in formation about 700 a.d. and have influenced Buddhism. In Bihar from the eighth century onwards the influence of Tantrism was powerful and disastrous. The best information about this epoch is still to be found in Taranatha, in spite of his defects. He makes the interesting statement that in the reign of Gopaia who was a Buddhist, although his ministers were not (730-740 A.D.), the Buddhists wished their religious buUdings to 1 For notices of Kashmirian reUgion see Stein's translation of the Rajatarangini and Biihler, Tour in Search of Sanskrit manuscripts. J. Bomb. A.S. 1877. » VI. 11-13, vn. 278-280, 295, 523. « I. 122, 335, 348 : in. 99, v. 55. ' Alst) caUed Kumara. 128 THE MAHAYANA [ch. be kept separate from Hindu temples but that, in spite of protests, Hfe-sized images of Hindu deities were erected in them'- The ritual too was affected, for we hear several times of burnt offerings 2 and how Bodhibhadra, one of the later pro fessors of Vikramasila, was learned in the mystic lore of both Buddhists and Brahmans, Naianda and the other viharas continued to be seats of learning and not merely monasteries, and for some time there was a regular succession of teachers. Taranatha gives us to understand that there were many students and authors but that sorcery occupied an increasingly important position. Of most teachers we are told that they saw some deity, such as Avalokita or Tara. The deity was summoned by the rites akeady described* and the object of the performer was to obtain magical powers or siddhi. The successful sorcerer was known as siddha, and we hear of 84 mahasiddhas, stiU celebrated in Tibet, who extend from Rahulabhadra Nagarjuna to the thirteenth century. Many of them bear names which appear not to be Indian. The topics treated of in the Tantras are divided into Kxiya (ritual), Carya (apparently corresponding to Vinaya), Yoga, and Anuttara-yoga. Sometimes the first three are contrasted with the fourth and sometimes the first two are described as lower, the third and fourth as higher. But the Anuttara-yoga is always considered the highest and most mysterious*. Taranatha says* that the Tantras began to appear simultaneously with the Mahayana siitras but adds that the Anuttara-yoga tantras appeared gradually^. He also observes that the Acarya Ananda- garbha'' did much to spread them in Magadha. It is not until 1 Similarly statues of Mahadevi are found in Jain temples now, i.e. in Gujarat. " Thia very unbuddhist practice seems to have penetrated even to Japan. Burnt offerings form part of the ritual in the temple of Narita. ' See for instance the account of how Kamalarakshita summoned Yamari. ' So too the Sarphitas of the Vaishnavas and the Agaraas of the Saivas are said to consist of four quarters teaching Jnana, Yoga, Kriya and Carya respectively. See Sohrader, Introd. to Pancaratra, p. 22. Sometimes five classes of Tantras are enumerated which are perhaps all subdivisions of the Anuttara-yoga, namely Guhyasamaja, Mayajala, Buddhasararaayoga, CandraguhyatUaka, Manju&ikrodha. See Tfiranatha (Schiefner), p. 221. ' Chap. XLin. But this seems hardly consistent with his other statements. ' The Lamas in Tibet have a simUar theory of progressive tantric revelation. See WaddeU, Buddhism of Tibet, pp. 56, 57. ' In the reigu of Mahipala, 978-1030 a.d. XXIV] DECADENCE OF BUDDHISM IN INDIA 129 a late period of the Paia dynasty that he mentions the Kalacakra which is the most extravagant form of Buddhist Tantrism. This accords with other statements to the effect that the Kalacakra tantra was introduced in 965 A.D. from Sambhala, a mysterious country in Central Asia. This system is said to be Vishnuite rather than Sivaite. It speciaUy patronizes the cult of the mystic Buddhas such as Kalacakra and Heruka, all of whom appear to be regarded as forms of Adi-Buddha or the primordial Buddha essence. The Siddha named Pito is also described as the author of this doctrine', which had less import ance in India than in Tibet. On the other hand Taranatha gives us the names of several doctors of the Vinaya who flourished under the Pala dynasty. Even as late as the reign of Ramapaia (? 1080-1120) we hear that the Hinayanists were numerous. In the reign of Dharma paia (c. 800 A.D.) some of them broke up the great silver image of Heruka at Bodh-Gaya and burnt the books of Mantras ^- These instances show that the older Buddhism was not entirely overwhelmed by Tantrism* though perhaps it was kept ahve more by pUgrims than by local sentiment. Thus the Chinese inscriptions of Bodh-Gaya though they speak at length of the three bodies of Buddha show no signs of Tantrism. It would appear that the worship celebrated in the holy places of Magadha preserved a respectable side until the end. In the same way although Tantrism is strong in the Uterature of the Lamas, none of the many descriptions of Tibet indicate that there is anything scandalous in the externals of religion. Probably in Tibet, Nepal and mediaeval Magadha aUke the existence of disgraceful tantric Uterature does not indicate such widespread depravity as might be supposed. But of its putrefying influence in corrupting the minds of those who ought to have preserved ^ Taranatha, p. 275. For the whole subject see Griinwedel, Mythologie des Buddhismus, pp. 41-2 and my chapters on Tibet below. ^ Schiefner (transl. Taranatha, p. 221) describes these Sravakas or Hinayanists as " Saindhavas welche Qravakas aus Simhala u.s.w. waren." They are apparently the same as the Saindhava-9ravakas often mentioned by Taranatha. Are they Hinayanists from Sindh where the Sammitiya school was prevalent ? See also Pag Sam Jon Zang, pp. cxix, 114 and 134 where Sarat Chandra Das explains Sendha-pa as a brahmanical sect. ' The curious story (Taranatha, p. 206) in which a Buddhist at first refuses on reUgious grounds to take part in the evocation of a demon seems also to hint at a disapproval of magic. E. n. 9 130 THE MAHAYANA [ch. the pure faith there can be no doubt. More than any other form of mixed beUef it obUterated essential differences, for Buddhist Tantrism and Sivaite Tantrism are merely two varieties of Tantrism. What is happening at Bodh-Gaya at present' illustrates how Buddhism disappeared from India. The abbot of a neighbouring Sivaite monastery who claims the temple and grounds does not wish, as a Mohammedan might, to destroy the building or even to efface Buddhist emblems. He wishes to supervise the whole estabhshment and the visits of pilgrims, as weU as to place on the images of Buddha Hindu sectarian marks and other orna ments. Hindu pilgrims are stiU taken by their guides to venerate the Bodhi tree and, but for the presence of foreign pilgrims, no casual observer would suppose the spot to be anything but a Hindu temple of unusual construction. The same process went a step further in many shrines which had not the same celebrity and effaced all traces and memory of Buddhism. At the present day the Buddha is recognized by the Brahmans as an incarnation of Vishnu^, though the recognition is often quaUfied by the statement that Vishnu assumed this form in order to mislead the wicked who threatened to become too powerful if they knew the true method of attaining superhuman powers. But he is rarely worshipped in proprid persojid^. As a rule Buddhist images and emblems are ascribed to Vishnu or Siva, according to sectarian preferences, but in spite of fusion some Hngering sense of original animosity prevents Gotama from receiving even such respect as is accorded to incarnations Hke Parasurama. At Bodh-Gaya I have been told that Hindu pilgrims are taken by their guides to venerate the Bodhi-tree but not the images of Buddha. Yet in reviewing the disappearance of Buddhism from India we must remember that it was absorbed not expeUed. The result of the mixture is justly called Hinduism, yet both in "• This passage was written about 1910. In the curious temple at Gaya called Bishnupad the chief object of veneration is a foot-like mark. Such impressions are venerated in many parts of the world as Buddha's feet and it seems probable, considering the locality, that this footprint was attributed to Buddha before it was transferred to Vishnu. * There are no very early references to this Avatara. Tt is mentioned in some of the Puranas (e.g. Bhagavata and Agni) and by Kshemendra. » But see the instances quoted above from Kashmir and Nepal. XXIV] DECADENCE OF BUDDHISM IN INDIA 131 usages and beUefs it has taken over much that is Buddhist and without Buddhism it would never have assumed its present shape. To Buddhist influence are due for instance the rejection by most sects of animal sacrifices : the doctrine of the sanctity of animal Hfe: monastic institutions and the ecclesiastical discipUne found in the Dravidian regions. We may trace the same influence with more or less certainty in the philosophy of Sankara and outside the purely rehgious sphere in the develop ment of Indian logic. These and similar points are dealt with in more detail in other parts of this work and I need not dwell on them here. 9—2 BOOK V HINDUISM BOOK y The present book deals with Hinduism and includes the period just treated in Book rv. In many epochs the same mythological and metaphysical ideas appear in a double form, Brahmanic and Buddhist, and it is hard to say which form is the earlier. Any work which Hke the present adopts a geographical and historical treatment is bound to make Buddhism seem more important than Hinduism and rightly, for the conversion and transformation of China, Japan and many other countries are a series of exploits of great moment for the history not merely of rehgion but of civUization. Yet when I think of the antiquity, variety and vitaUty of Hinduism in India — no small sphere — the nine chapters which foUow seem very inadequate. I can only urge that though it would be easy to fill an encyclopaedia with accounts of Indian beUefs and practices, yet there is often great simUarity under superficial differences : the main Hues of thought are less numerous than they seem to be at first sight and they tend to converge. CHAPTER XXV SIVA AND VISHNU The striking difference between the earUer and later phases of Indian religious belief, between the Vedic hymns, Brahmanas, Upanishads and their accessory treatises on the one hand, and the epics, Puranas, Tantras and later Uterature on the other, is due chiefly to the predominance in the latter of the great gods Siva and Vishnu, with the attendant features of sectarian wor ship and personal devotion to a particular deity. The difference is not wholly chronological, for late writers sometimes take the Vedic standpoint and ignore the worship of these deities, but stiU their prominence in Hterature, and probably in popular mythology, is posterior to the Vedic period. The change created by their appearance is not merely the addition of two imposing figures to an already ample pantheon ; it is a revolution which might be described as the introduction of a new rehgion, except that it does not come as the enemy or destroyer of the old. The worship of the new deities grows up peacefully in the midst of the ancient rites ; they receive the homage of the same popula tion and the ministrations of the same priests. The transition is obscured but also was f aciUtated by the strength of Buddhism during the period when it occurred. The Brahmans, confronted by this formidable adversary, were disposed to favour any popular religious movement which they could adapt to their interests. When the Hindu revival sets in under the Guptas, and Buddhism begins to decline, we find that a change has taken place which must have begun several centuries before, though our imperfect chronology does not permit us to date it. Whereas the Vedic sacrificers propitiated all the gods impartially and regarded ritual as a sacred science giving power over nature, the worshipper of the later deities is generally sectarian and often emotional. He selects one for his adoration, and this selected deity becomes not merely a great god among others CH. XXV] ISIVA AND VISHNU 137 but a gigantic cosmical figure in whom centre the phUosophy, poetry and passion of his devotees. He is almost God in the European sense, but stUl Indian deities, though they may have a monopoly of adoration in their own sects, are never entirely similar to Jehovah or AUah. They are at once more mythical, more human and more philosophical, since they are conceived of not as creators and riders external to the world, but as forces manifesting themselves in nature. An exuberant mythology bestows on them monstrous forms, celestial residences, wives and offspring : they make occasional appearances in this world as men and animals; they act under the influence of passions which if titanic, are but human feelings magnified. The philoso pher accommodates them to his system by saying that Vishnu or Siva is the form which the Supreme Spirit assumes as Lord of the visible universe, a form which is real only in the same sense that the visible world itseff is real. Vishnu and Rudra are known even to the Rig Veda but as deities of no special eminence. It is only after the Vedic age that they became, each for his own worshippers, undisputed Lords of the Universe. A Umiting date to the antiquity of Sivaism and Vishnuism, as their cults may be caUed, is furnished by Buddhist Uterature, at any rate for north-eastern India. The PaH Pitakas frequently' introduce popular deities, but give no prominence to Vishnu and Siva. They are apparently mentioned under the names of Venhu and Isana, but are not differentiated from a host of spirits now forgotten. The Pitakas have no pre judices in the matter of deities and their object is to represent the most powerful of them as admitting their inferiority to the Buddha. If Siva and Vishnu are not put forward in the same way as Brahma and Indra, the inference seems clear: it had not occurred to anyone that they were particularly important. The suttas of the Digha Nikaya in which these lists of deities occur were perhaps composed before 300 b.c.^ About that date Megasthenes, the Greek envoy at Patahputra, describes two Indian deities under the names of Dionysus and Herakles. They are generally identified with Eo-ishna and Siva. It might be difficult to deduce this identity from an analysis of each ^ See especiaUy Dig. Nik. xx. and xxxn. ^ But the Usts raay be pieces of folk-lore older than the suttas in which they are incorporated. 138 HINDUISM [CH. description and different authorities have identified both Siva and Krishna with Dionysus, butthe fact remains that a somewhat superficial foreign observer was impressed with the idea that the Hindus worshipped two great gods. He would hardly have derived this idea from the Vedic pantheon, and it is not clear to what gods he can refer if not to Siva and Vishnu. It thus seems probable that these two cults took shape about the fourth century B.C. Their apparently sudden appearance is due to their popular character and to the absence of any record in art. The statuary and carving of the Asokan period and immediately succeeding centuries is exclusively Buddhist. No temples or images remain to illustrate the first growth of Hinduism (as the later form of Indian religion is commonly styled) out of the earlier Brahmanism. Literature (on which we are dependent for our information) takes Httle account of the early career of popular gods before they win the recognition of the priesthood and aristocracy, but when that recognition is once obtained they appear in aU their majesty and without any hint that their honours are recent. As already mentioned, we have evidence that in the fifth or sixth century before Christ the Vedic or Brahmanic rehgion was not the only form of worship and philosophy in India. There were popular deities and rites to which the Brahmans were not opposed and which they countenanced when it suited them. What takes place in India to-day took place then. When some aboriginal deity becomes important owing to the prosperity of the tribe or locaUty with which he is connected, he is recognized by the Brahmans and admitted to their pantheon, perhaps as the son or incarnation of some personage more generaUy accepted as divine. The prestige of the Brahmans is sufficient to make such recognition an honour, but it is also their interest and millennial habit to secure control of every important rehgious movement and to incorporate rather than suppress. And this incorporation is more than mere recognition : the parvenu god borrows something from the maimers and attributes of the olympian society to which he is introduced. The greater he grows, the more considerable is the process of fusion and borrowing. Hindu phUosophy ever seeks for the one amongst the many and popular thought, in a more confused way, pursues the same goal. It combines and identifies its xxv] Siva and vishnu 139 deities, feeling dimly that taken singly they are too partial to be truly divine, or it pUes attributes upon them striving to make each an adequate divine whole. Ainong the processes which have contributed to form Vishnu and Siva we must reckon the invasions which entered India from the north-west'. In Bactria and Sogdiana there met and were combined the art and rehgious ideas of Greece and Persia, and whatever elements were imported by the Yiieh-chih and other tribes who came from the Chinese frontier. The person aUties of Vishnu and Siva need not be ascribed to foreign influence. The ruder invaders took kindly to the worship of Siva, but there is no proof that they introduced it. But Persian and Graeco-Bactrian influence favoured the creation of more definite deities, more personal and more pictorial. The gods of the Vedic hymns are vague and indistinct : the Supreme Being of the Upanishads altogether impersonal, but Mithra and ApoUo, though divine in their majesty, are human in their persons and in the appeal they make to humanity. The influence of these foreign conceptions and especiaUy of their representation in art is best seen in Indian Buddhism. Hinduism has not so ancient an artistic record and therefore the Graeco-Bactrian influence on it is less obvious, for the sculpture of the Gupta period does not seem due to this inspiration. Neither in outward form nor in character do Vishnu and Siva show much more resemblance to ApoUo and Mithra than to the Vedic gods. Their exuberant, fantastic shapes, their many heads and arms, are a symbol of their complex and multiple attributes. They are not restricted by the limits of personaUty but are great polymorphic forces, not to be indicated by the limits of one human shape^. ' The Dionysus of Megasthenes is a deity who comes from the west with an army that suffers from the heat of the plains. H we conld be certain that he meant Siva by Dionysus this would be valuable evidence. But he clearly misunderstood many things in Indian reUgion. Greek legends connected Dionysus with India and the East. ' MacdoneU seems to me correct in saying (J.R..i.S. 1915, p. 125) that one reason why Indian deities have many arms is that they may be able to carry the various symbols by which they tire characterized. Another reason is that worship is DsuaUy accompanied by dhyana, that is forming a mental image of the deity as described in a particul.ir text. E.g. the worshipper repeats a mantra which describes a deity in laognage which was originaUy metaphorical as having many heads and arms and at the same time he ought to make a mental image of such a figure. 140 HINDUISM [CH. Though alike in their grandeur and multipHcity, Vishnu and I Siva are not otherwise similar. In their completely developed forms they represent two ways of looking at the world. The main ideas of the Vaishnavas are human and emotional. The deity saves and loves : he asks for a worship of love. He appears in human incarnations and is known as weU or better by these incarnations than in his original form. But in Sivaism the main current of thought is scientific and philosophic rather than emotional'. This statement may seem strange if one thinks of the wild rites and legends connected with Siva and his spouse. Nevertheless the fundamental conception of Sivaism, the cosmic force which changes and in changing both destroys and repro duces, is strictly scientific and contrasts with the human, pathetic, loving sentiments of Vishnuism. And scandalous as the worship of the generative principle may become, the potency of this impulse in the world scheme cannot be denied. Agreeably to his character of a force rather than an emotion Siva does not become incarnate^ as a popular hero and saviour Uke Rama or Krishna, but he assumes various supernatural forms for special purposes. Both worships, despite their differences, show charac teristics which are common to most phases of Indian rehgion. Both seek for deUverance from transmigration and are penetrated with a sense of the sorrow inherent in human and animal life : both develop or adopt phUosophical doctrines which rise high above the level usuaUy attained by popular beUefs, and both ' But some forms of Sivaism in southern India come even nearer to emotional Christianity than does Vishnuism. 2 I cannot discover that any aUeged avatara of Siva has now or has had formerly any importance, but the Vay u.Linga and Kurma Purana give lists of such incarnations, as does also the Catechism of the Shaiva religion translated by Foulkes. But Indian sects have a strong tendency to ascribe aU possible achievements and attributes to their gods. The mere fact that Vishnu becomes incarnate incites the ardent Sivaite to say that his god can do the same. A curious instance of this rivalry is found in the story that Siva manifested himself as Sarabha-murti in order to curb the ferocity of Vishnu when incarnate in the Man Lion (see Gopiuatha Rao, Hindu Icon. p. 45). Siva often appears in a special form, not necessarUy human, for a special purpose (e.g. Virabhadra) and some tantric Buddhas seem to be imitations of these appari tions. There is a strong element of Sivaism borrowed from Bengal in the mythology of Tibet and MongoUa, where such personages as Hevajra, Saipvara, and Mahakala have a considerable importance under the strange title of Buddhas. xxv] Siva and vishnu ui have erotic aspects in which they faU below the standard of morality usually professed by important sects whether in Asia or Europe. The name Siva is euphemistic. It means propitious and, like Eumenides, is used as a deprecating and complimentary title for the god of terrors. It is not his earliest designation and does not occur as a proper name in the Rig Veda where he is known as Rudra, a word of disputed derivation, but probably meaning the roarer. Comparatively few hymns are addressed to Rudra, but he is clearly distinguished from the other Vedic gods. Whereas they are cheerful and benevolent figures, he is maleficent and terrible : they are gods of the heaven but he is a god of the earth. He is the "man-slayer" and the sender of disease, but if he restrains these activities he can give safety and health. "Slay us not, for thou art gracious," and so the Destroyer comes to be the Gracious One'. It has been suggested that the name Siva is connected with the Tamil word ^ivappu red and also that Rudra means not the roarer but the red or shining one. These etymologies seem to me possible but not proved. But Rudra is different in character from the other gods of the Rig Veda. It would be rash to say that the Aryan invaders of India brought with them no god of this sort but it is probable that this element in their pantheon increased as they gradually united in blood and ideas with the Dravidian population. But we know nothing of the beUefs of the Dravidians at this remote period. We only know that in later ages emotional rehgion, finding expression as so-caUed devil-dancing in its lower and as mystical poetry in its higher phases, was prevalent among them. The White Yajur Veda^ contains a celebrated prayer known as the Satarudriya addressed to Rudra or the Rudras, for the power invoked seems to be now many and now one. This deity, who is described by a long string of epithets, receives the name of Sankara (afterwards a well-known epithet of ^iva) and is blue-necked. He is begged to be Siva or propitious, but the word is an epithet, not a proper name. He haunts mountains and deserted, uncanny places : he is the patron of violent and lawless men, of soldiers and robbers (the two are evidently ^ The passage from one epithet to the other is very plain in R.V, I. 114. * Book XVI. 142 HINDUISM [ch. considered much the same), of thieves, cheats and piUerers', but also of craftsmen and huntsmen and is himseU "an observant merchant": he is the lord of hosts of spirits, "Ul-formed and of aU forms." But he is also a great cosmic force who "dweUs in flowing streams and in biUows and in tranquU waters and in rivers and on islands.. .and at the roots of trees..." : who "exists in incantations, in punishments, in prosperity, in the soU, in the threshing-floor. . .in the woods and in the bushes, in sound and in echo... in young grass and in foam... in gravel and in streams... in green things and in dry things... Reverence to the leaf and to him who is in the faU of the leaf, the threatener, the slayer, the vexer and the affUcter." Here we see how an evU and disreputable god, the patron of low castes and violent occupations, becomes associated with the uncanny forces of nature and is on the way to become an AU-God^ Rudra is frequently mentioned in the Atharva Veda. He is conceived much as in the Satarudriya, and is the lord of spirits and of animals. "For thee the beasts of the wood, the deer, swans and various winged birds are placed in the forest: thy Hving creatures exist in the waters : for thee the celestial waters flow. Thou shootest at the monsters of the ocean, and there is to thee nothing far or near^." These passages show that the main conceptions out of which the character of the later Siva is buUt existed in Vedic times. The Rudra of the Yajur and Atharva Vedas is not Brahmanic: he is not the god of priests and orderly ritual, but of wUd people and places. But he is not a petty provincial demon who afflicts rustics and their cattle. Though there is some hesitation between one Rudra and many Rudras, the destructive forces are unified in thought and the destroyer is not opposed to creation as a devU or as the principle of evU, but with profounder insight is recognized as the Lord and Law of aU Hving things. But though the outline of Siva is found in Vedic writings, later centuries added new features to his cult. Chief among these is the worship of a column known as the Linga, the emblem under which he is now most commonly adored. It is a phaUic ^ In the play Mricchakatika or The Clay Cart (probably of the sixth century a.d. ) a burglar invokes Kartikeya, the son of Siva, who is said to have taught different styles of house-breaking. ' A similarly strange coUocation of attributes is found in Daksha's hymn to Siva. Mahabharata, xn. Seo. 285. ' Atharva, v. xi. 2. 24. xxv] Siva and vishnu 143 symbol though usuaUy decent in appearance. The Vedas do not countenance this worship and it is not clear that it was even known to them'. It is first enjoined in the Mahabharata and there only in two passages^ which appear to be late additions. The inference seems to be that it was accepted as part of Hinduism just about the time that our edition of the Maha bharata was compiled^. The old theory that it was borrowed from aboriginal and especiaUy from Dravidian tribes* is now discredited. In the first place the instances cited of phalUc worship among aboriginal tribes are not particularly numerous or striking. Secondly, Unga worship, though prevalent in the south, is not confined to it, but flourishes in all parts of India, even in Assam and Nepal. Thirdly, it is not connected with low castes, with orgies, with obscene or bloodthirsty rites or with anything which can be caUed un-Aryan. It forms part of the private devotions of the strictest Brahmans, and despite the significance of the emblem, the worship offered to it is perfectly decorous*. The evidence thus suggests that this cultus grew up among Brahmanical Hindus in the early centuries of our era. The idea that there was something divine in virility and genera tion already existed. The choice of the symbol — the stone piUar — may have been influenced by two circumstances. Firstly, the Buddhist veneration of stupas, especiaUy miniature stupas, must have made famiUar the idea that a cone or column is a reUgious emblem ®, and secondly the Hnga may be compared to ^ It is not certain if the Sisnadevah whom Indra is asked to destroy in Rig. V. vn. 21. 5 and x. 99. 3 are priapio demons or worshippers of the phaUus. ' vn. sees. 202, 203, and xm. sec. 14. ' The inscriptions of Camboja and Champa seem to be the best proof of the antiquity of Linga worship. A Cambojan inscription of about 550 a.d. records the dedication of a Unga and the worship must have taken some time to reach Camboja from India. Some lingas discovered in India are said to be anterior to the Christian era. * See F. Kittel, Ueber den Ursprung der Linga Kultus, and Barth, Religions of India, p. 261. ' As is also its appearance, as a rule. But there are exceptions to this. Some Hindus deny that the Linga is a phallic emblem. It is hardly possible to maintain this thesis in view of suoh passages as Mahabh. xm. 14 and the innumerable figures in which there are both a Unga and a Yoni. But it is true that in its later forms the worship is purged of aU grossness and that in its earlier forms the symbol adored was often a stupa-like column or a pUlar with figures on it. • Such scenes as the reUef from Amaravati figured in Griinwedel, Buddhist art in India, p. 29, fig. 8, might easily be supposed to represent the worship of the linga, and some of Anoka's piUars have been worshipped as lingas in later times. 144 HINDUISM [ch. the carved pUlars or stone standards erected in honour of Vishnu. Some lingas are carved and bear one or four faces, thus entirely losing any phalUc appearance. The wide extension of this cult, though its origin seems late, is remarkable. Something similar may be seen in the worship of Ganesa : the first records of it are even later, but it is now universal in India. It may seem strange that a rehgion whose outward cere monies though unassuming and modest consist chiefly of the worship of the linga, should draw its adherents largely from the educated classes and be under no moral or social stigma. Yet as an idea, as a philosophy, Sivaism possesses truth and force. It gives the best picture which humanity has drawn of the Lord of this world, not indeed of the ideal to which the saint aspires, nor of the fancies with which hope and emotion people the spheres behind the veU, but of the force which rules the Universe as it is, which reproduces and destroys, and in performing one of these acts necessarily performs the other, seeing that both are but aspects of change. For aU animal and human existence' is the product of sexual desire: it is but the temporary and transitory form of a force having neither begiiming nor end but continuaUy manifesting itseU in indi viduals who must have a beginning and an end. This force, to which European taste bids us refer with such reticence, is the true creator of the world. Not only is it unceasingly performing the central miracle of producing new Hves but it accompanies it by unnumbered accessory miracles, which provide the new bom chUd with nourishment and make lowly organisms care for their young as if they were gifted with human InteUigence. But the Creator is also the Destroyer, not in anger but by the very nature of his activity. When the series of changes culmin ates in a crisis and an individual breaks up, we see death and destruction, but in reahty they occur throughout the process of growth. The egg is destroyed when the chicken is hatched: the embryo ceases to exist when the chUd is born; when the man comes into being, the child is no more. And for change, im provement and progress death is as necessary as birth. A world of immortals would be a static world. When once the figure of Siva has taken definite shape, 1 But not of course the soul which, according to the general Indian idea, exists before and continues after the lite of the body. xxv] Siva and vishnu 145 attributes and epithets are lavished on it in profusion. He is ^he great ascetic, for asceticism in India means power, and Siva is the personification of the powers of nature. He may alternate strangely between austerities and wUd debauch, but the senti- mentaUty of some Eo-ishnaite sects is alien to him. He is a magician, the lord of troops of spirits, and thus draws into his circle aU the old animistic worship. But he is also identified with Time (Mahakala) and Death (Mrityu) and as presiding over procreation he is Ardhanaresvara, half man, haU woman. Stories are invented or adapted to account for his various attributes, and he is provided with a divine famUy. He dwells on Mount Kaiiasa : he has three eyes : above the central one is the crescent of the moon and the stream of the Ganges descends from his braided hair: his throat is blue and encircled by a serpent and a necklace of skulls. In his hands he carries a three- pronged trident and a drum. But the effigy or description varies, for Siva is adored under many forms. He is Mahadeva, the Great God, Hara the Seizor, Bhairava the terrible one, Pasupati, the Lord of cattle, that is of human souls who are compared to beasts. Local gods and heroes are identified with him. Thus Gor Baba', said to be a deified ghost of the aboriginal races, reappears as Groresvara and is counted a form of Siva, as is also Khandoba or Khande Rao, a deity connected with dogs. Ganesa, ' ' the Lord of Hosts,' ' the God who removes obstacles and is represented with an elephant's head and accompanied by a rat, is recognized as Siva's son. Another son is Skanda or Kartikeya, the God of War, a great deity in Ceylon and southern India. But more important both for the absorption of aboriginal cults and for its influence on speculation and morahty is the part played by Siva's wife or female counterpart. The worship of goddesses, though found in many sects, is speciaUy connected with Sivaism. A figure analogous to the Madoima, the kind and compassionate goddess who helps and pities all, appears in later Buddhism but for some reason this train of thought has not been usual in India. Lakshmi, Sarasvati and Sita are benevolent, but they hold no great position in popular esteem^, and the being who attracts millions of wor- ' Crooke, Popular Religion and Folklore of Northern India, I. 84; n. 219. ' They are however of some importance in Vishnuite theology. For instance according to the school of Ramanuja it is the Sakti (Sri) who reveals the true doctrine to mankind. Vishnu is often said to have three consoi'ts, Sri, Bhu and LUa. E. U. 10 146 HINDUISM [on. shippers under such names as KaU, Durga, or Mahadevi, though she has many forms and aspects, is most commonly represented as a terrible goddess who demands offerings of blood. The worship of this goddess or goddesses, for it is hard to say if she is one or many, is treated of in a separate chapter. Though in shrines dedicated to Siva his female counterpart or energy (Sakti) also receives recognition, yet she is revered as the spouse of her lord to whom honour is primarily due. But in Saktist worship adoration is offered to the Sakti as being the form in which his power is made manifest or even as the essential God head. Let us now pass on to Vishnu. Though not one of the great gods of the Veda, he is mentioned fairly often and with respect. Indian commentators and comparative mythologists agree that he is a solar deity. His chief exploit is that he took (or perhaps in the earHer version habituaUy takes) three strides. This was originally a description of the sun's progress across the firma ment but grew into a myth which relates that when the earth was conquered by demons, Vishnu became incarnate as a dwarf and induced the demon king to promise him as much space as he could measure in three steps. Then, appearing in his true form, he strode across earth and heaven and recovered the world for mankind. His special character as the Preserver is already outlined in the Veda. He is always benevolent : he took his three steps for the good of men : he estabhshed and preserves the heavens and earth. But he is not the principal solar deity of the Rig Veda: Siirya, Savitri and Pushan receive more invocations. Though one hymn says that no one knows the Umits of his greatness, other passages show that he has no pre eminence, and even in the Mahabharata and the Vishnu-Purana itseff he is numbered among the Adityas or sons of Aditi. In the Brahmanas, he is somewhat more important than in the Rig Veda', though he has not yet attained to any position Hke that which he afterwards occupies. Just as for Siva, so for Vishnu we have no clear record of the steps by which he advanced from a modest rank to the ' E.g. Sat. Brah. i. 2. 5. See also the strange legend ib. xi. 1. 1 where Vishnu is described as the best of the gods but is eaten by Indra. He is frequently (e.g. in the Sata Brah) stated to be identical with the sacrifice, and this was probably one of the reasons lor his becoming prominent. xxv] Siva and vishnu ui position of having but one rival in the popular esteem. But the Hues on which the change took place are clear. Even in his own Church, Vishnu himself claims comparatively Uttle attention. He is not a force like Siva that makes and mars, but a benevolent and retiring personaUty who keeps things as they are. His worship, as distinguished from that of his incarnations, is not conspicuous in modern India, especiaUy in the north. In the south he is less overshadowed by Kjishna, and many great temples have been erected in his honour. In Travancore, which is formaUy dedicated to him as his special domain, he is adored under the name of Padmanabha. But his real claim to reverence, his appeal to the Indian heart, is due to the fact that certain deified human heroes, particularly Rama and Krishna, are identified with him. Deification is common in India'. It exists to the present day and even defunct Europeans do not escape its operation. In modem times, when the idea of reincarnation had become familiar, eminent men Uke Caitanya or Vallabhacarya were declared after their death to be embodiments of Krishna without more ado, but in earUer ages the process was probably double. First of all the departed hero became a powerful ghost or deity in his own right, and then this deity was identified with a Brahmanic god. Many examples prove that a remarkable man receives worship after death quite apart from any idea of incarnation. The incarnations of Vishnu are most commonly given as ten^ but are not all of the same character. The first five, namely, the Fish, Tortoise, Boar, Man-Lion and Dwarf, are mythical, and due to his identification with supernatural creatures playing a benevolent role in legends with which he had originaUy no con nection. The sixth, however, Paralu-rama or Rama with the axe, may contain historical elements. He is represented as a miUtant Brahman who in the second age of the world extermin- ^ See many modem examples in Crooke, Popular Religion and Folk Lore of Northern India, chap. rv. and Census of India, 1901, vol. vi. Bengal, pp. 196-8, where are described various deified heroes who are adored in Bengal, such as Goveiya (a bandit), SaUesh, Karikh, Ldrik, Amar Singh, and Gobind Raut (a slayer of tigers). Compare too the worship of Gopi Nath and Zinda Kaliana in the Panjab as described in Census of India, 1901, vol. xvn. pp. 118-9. 2 The Bhagavata Purana (i. ui.) and the Bhaktam41a (see J.B.A.S. 1909, pp. 621 ff.) give longer Usts ot 22 and 26, and the Pancaratra gives 39. See Ahirbudlmya Samhita, v. 50-55. 10—2 148 HINDUISM [ch. ated the Kshatriyas, and after reclaiming Malabar from the sea, settled it with Brahmans. This legend clearly refers to a struggle for supremacy between the two upper castes, though we may doubt if the triumphs attributed to the priestly champion have any foundation in fact. The Ramayana' contains a singular account of a contest between this Rama and the greater hero of the same name in which Para^u-rama admits the other's superiority. That is to say an epic edited under priestly super vision relates how the hero-god of the warriors vanquishes the hero-god of the priests, and this hero-god of the warriors is then worshipped by common consent as the greater divinity, but under priestly patronage. The tenacity and vitality of the Brahmans enabled them ultimately to lead the conqueror captive, and Rama-candra became a champion of Brahmanism as much as Para^u-rama. Very interesting too is the ninth avatara (to leave for a moment the strict numerical order) or Buddha^. The reason assigned in Brahmanic literature for Vishnu's appearance in this character is that he wished to mislead the enemies of the gods by false teaching, or that out of compassion for animals he preached the abohtion of Vedic sacrifices. Neither explanation is very plausible and it is pretty clear that in the period when degenerate Buddhism offered no objection to deification and mythology, the Brahmans sanctioned the worship of the Buddha under their auspices. But they did so only in a half-hearted way. The Buddha was so important a personage that he had to, be explained by the intervention, kindly or hostile, of a deity^. In his tenth incarnation or Kalki*, which has yet to take 1 Book I, cantos 74-76. 2 A paraUel phenomenon is the belief found in BaU, that Buddha is Siva's brother. " For Brahmanic ideas about Buddha see Vishnu Puraria, ni. 18. The Bhagavata Puraija, i. 3. 24 seems to make the Buddha incarnation future. It also counts Kapila and Rishabha, apparently identical with the founder of the Sankhya and the first Jain saint, as incarnations. The Padma Purana seems to ascribe not only Buddhism but the Maya doctrine of Sankara to delusions deUberately inspired by gods. I have not been able to find the passage in the printed edition of the Purana but it is quoted in Sanskrit by Aufrecht, Cat. Cod. Bib. Bodl. p. 14, and Muir, Original Sanskrit Texts, p. 198. * See Norman in Trama. TAiVd /«.<. Congress of Religions,!!, p. 8.5. lni\i.e Ind. Ant. 1918, p. 145 Jayaswal tries to prove that Kalki is a historical personage and identical with King Ya^odharman of Central India (about a.d. 500) and that the idea of xxv] Siva and vishnu 149 place, Vish^iu wiU appear as a Messiah, a conception possibly influenced by Persian ideas. Here, whore we are in the realm of pure imagination, we see clearly what the signs of his avataras are supposed to be. His mission is to sweep a,wa,j the wicked and to ensure the triumph of the pious, but he comes as a warrior and a horseman, not as a teacher, and if he protects the good he does so by destroying evil. He has thus all the attributes of a Kshatriya hero, and that is as a matter of fact the real character of the two most important avataras to which we now turn, Rama and Krishna. Rama, often distinguished as Rama-candra, is usually treated as the seventh incarnation and anterior to Krishna, for he was born in the second age of this rapidly deteriorating world, whereas Kiishna did not appear until the third. But his deification is later than that of Krishna and probably an imitation of it. He was the son of Dasaratha, King of Ayodhya or Oudh, but was driven into banishment by a j^alace intrigue. He married Sita, daughter of the King of Mithila. She ^^as carried off by Ravana, the demon tyrant of Ceylon, and Rama re-captured her with the aid of Hanuman, King of the Monkeys, and his hosts'. Is there any kernel of history in this story? An examination of Hindu legends suggests that they usually pre serve names and genealogies correctly but distort facts, and fantasticaUy combine independent narratives. Rama was a semi-divine hero in the tales of ancient Oudh, based on a real personaUty, and Ceylon was colonized bjr Indians of Aryan speech^. But can \^e assume that a king of Oudh really led an expedition to the far south, with the aid of ape-hke aborigines ? his being a, future saviour is late. This theory offers difficulties, for firstly thoiv is no proof thnt the passages of the Mfiliabharata which mention Kalki (iii. 190, 13101; m. 191, 13111; xn. 340, 12968) are additions later than Ya^odharman and secondly if Kalki was first a historical figure and tlicn projected into the futuix^ we shoiUd expect to hear that he wUl cotne again, but such language is not quoted. On the other hand it seems quite likely (1) that tJiere was on old tradition about a future saviour called Kalki, (2) that YajSodliarmnn after defeating the Huns assumed the r61e, (3) and that when it was found that the golden age had not i-eoommcnced he waa forgotten (as many pseudo-Messiahs have been) and Kalki again became a hope for the future. Vincent SmitJi (Hist, of India, ed. in. p. 320) intimates that YaSodharman performed considerable exploits but was inordinately boastful. ' AnotJier version of the story which omits the expedition to Laiika and maies Sitn the sister of Rfima is found iu tJie Dasaratha Jataka (1)41). ^ But this colonization is aftributtd by tradition to Viiaya, not Rama. 150 HINDUISM [CH. It is doubtful, and the narrative of the Ramayana reads like poetic invention rather than distorted history. And yet, what can have prompted the legend except the occurrence of some such expedition? In Rama's wife Sita, seem to be combined an agricultural goddess and a heroine of ancient romance, embodying the Hindu ideal of the true wife. We have no record of the steps by which Rama and Krishna were deified, although in different parts of the epic they are presented in very different aspects, sometimes as Httle more than human, sometimes as nothing less than the Supreme Deity. But it can hardly be doubted that this deification owes some thing to the example of Buddhism. It may be said that the development of both Buddhism and Hinduism in the centuries immediately preceding and following our era gives paraUel manifestations of the same popular tendency to deify great men. This is true, but the non-Buddhist forms of Indian rehgion while not objecting to deification did not particularly encourage it. But in this period, Buddhism and Jainism were powerful: both of them sanctioned the veneration of great teachers and, as they did not recognize sacrifice or adoration of gods, this veneration became the basis of their ceremonies and easily passed into worship. The Buddhists are not responsible for the introduction of deification, but the fact that it was to some extent the basis of their pubhc ceremonies must have gone far to make the worship of Rama and Krishna seem natural. It is commonly said that whereas the whole divine nature of Vishnu was embodied in KJrishna, Rama was only a partial incarnation. Half the god's essence took human form in him, the other half being distributed among his brothers. Krishna is a greater figure in popular esteem and receives the exclusive devotion of more worshippers. The name of Rama commands the reverence of most Hindus, and has a place in their prayers, but his figure has not been invested with the attributes (often of dubious moral value) which most attract sectarian devotion. His worship combines easily with the adoration of other deities. The great temple of Ramesvaram on Adam's Bridge is dedicated not to Rama himseff but to the Unga which he erected there, and Tulsi Das, the author of the liindi Ramayana, while in voking Rama as the Supreme Lord and redeemer of the world. xxv] Siva and vishnu 151 emphaticaUy states' that his worship is not antagonistic to that of Siva. No inscriptions nor ancient references testify to the worship of Rama before our era and in the subsequent centuries two phases can be distinguislied. First. RAma is a great hero, an mcamation of Vishnu for a particular pm'pose and analogous to the Vamana or any other avat&ra: deserving as such of aU respect but stUl not the object of any special cult. This is the view taken of Rama in the Mahabharata, the Puranas, the Raghuvamsa, and those parts of the Ramajj^ana which go beyond it are probably late additions-. But secondly Rama becomes for his worshippera the supreme deity. Ramfinuja (on the Vedanta sutras, n. 42) mentions him and Krishna as two great incarnations in which the supreme being became manifest, and since Krishna was certainly worshipped at this period as identical with the AU-God, it would appear that Rama held the same position. Yet it was not untU the fom-teenth or fifteenth century that he became for many sects the central and ultimate divine figure. In the more Uberal sects the worship of Rama passes easUy into theism and it is the direct parent of the Kabirpanth and Sikhism, but unlike Krishnaism it does not lead to erotic excess. RSma personifies the ideal of chivahy, Sita of chastity. Less edUyiiig forms of worship may attract more attention, but it must not be supposed that Rama is relegated to the penumbra of phUosophic thought. If anything so multiplex as Hinduism can be said to have a watchword, it is the cry. Ram, Ram. The story of his adventures has traveUed even further than the hero himself, and is known not only from Kashmir to Cape Comorin but from Bombay to Java and Indo-China where it is a common subject of art. In India the R;lmayana is a favourite recitation among aU classes, and dramatized versions of various episodes are performed as rehgious plays. Though two late Upanishads, the Rainapurvatapaniya and Rama- uttaratApaniya extol R&ma as the Supreme Being, there is no Ramapurana. The fact is significant, as showing that his worship did not possess precisely those features of priestly sectarianism which mark the PurSnas and perhaps that it is later than the ' See especiaUy book vi. p. 67. in Growse's T:\i7),slafion. • See Muir"s Sanslrit TexU, vol. n'. especiaUy pp. 441-491 152 HINDUISM [CH. Puranas, But it has inspired a large Uterature, more truly popular than anything that the Puranas contain. Thus we have the Sanskrit Ramayana itself, the Hindi Ramayana, the TamU Ramayana of Kamban, and works Uke the Adhyatma-Rama- yana and Yoga-Vasistha-Ramayana'- Of aU these, the Rama yana of Tulsi Das is speciaUy remarkable and I shaU speak of it later at some length. 4 Krishna, the other great incarnation of Vishnu, is one of the most conspicuous figures in the Indian pantheon, but his historical origin remains obscure. The word which means black or dark blue occurs in the Rig Veda as the name of an otherwise unknown person. In the Chandogya Upanishad^, Krishna, the son of Devaki, is mentioned as having been instructed by the sage Ghora of the Angirasa clan, and it is probably impHed that Krishna too belonged to that clan^. Later sectarian writers never quote this verse, but their silence may be due to the fact that the Upanishad does not refer to Krishna as if he were a deity, and merely says that he received from Ghora instruction after which he never thirsted again. The purport of it was that the sacrifice may be performed without rites, the various parts being typified by ordinary human actions, such as hunger, eating, laughter, HberaUty, righteousness, etc. This doctrine has some resemblance to Buddhist language* and if this Krishna is reaUy the ancient hero out of whom the later deity was evolved, there may be an aUusion to some simple form of worship which rejected ceremonial and was practised by the tribes to whom Krishna belonged. I shaU recur to the question of these tribes ^ Ekanatha, who Uved in the sixteenth century, caUs the Adhyatma R. a modem work. See Bhandarkar, FaisAn-oTOdSamsm, page 48. The Yoga-VasishthaR. purports to be instruction given by Vasishtha to Rama who wishes to abandon the world. Its date is uncertain but it is quoted by authors of the fourteenth century. It is very popular, especiaUy in south India, where an abridgment in TamU caUed Jfiiana-Vasishtha is much read. Its doctrine appears to be Vedantist with a good deal of Buddhist philosophy. Salvation is never to think that pleasures and pains are "mine." ^ ChancJ. Up. ni. 17. 6. ' The Kaush. Brahm. says that Krishna was an Angirasa xxx. g. The Ann. kramani says that the Krishna of Rig Veda, vm. 74 was an Angirasa. For Ghora Angirasa "the dread desoendent of the Angirases" see MacdoneU and Keith, Vedic Index, s.v. ¦¦ E.g. Dig. Nik. v. The Pancaratra expressly states that Yoga is worship of the heart and self-sacrifice, being thus a counterpart of the external sacrifice (bahya- yaga). xxv] Siva and vishnu 153 and the Bhagavata Sect below, but in this section I am con cerned with the personaUty of Krishna. vasudeva is a well-known name of Krishna and a siitra of Panini', especiaUy if taken in conjunction with the comment of PatanjaU, appears to assert that it is not a clan name but the name of a god. If so Vasudeva must have been recognized as a god in the fourth century b.c. He is mentioned in inscriptions which appear to date from about the second century b.c.^ and in the last book of the Taittiriya Aranyaka^, which however is a later addition of uncertain date. The name Krishna occurs in Buddhist writings in the form Kanha, phoneticaUy equivalent to Krishna. In the Digha Nikaya* we hear of the clan of the Kanhayanas (= Karshna- yanas) and of one Kanha who became a great sage. This person may be the Krishna of the Rig Veda, but there is no proof that he is the same as our Krishna. The Ghata-Jataka (No. 454) gives an account of Krishna's childhood and subsequent exploits which in many points corre sponds with the Brahmanic legends of his life and contains several familar incidents and names, such as Vasudeva, Baladeva, Kamsa. Yet it presents many pecuHarities and is either an independent version or a misrepresentation of a popular story that had wandered far from its home. Jain tradition also shows that these tales were popular and were worked up into different forms, for the Jains have an elaborate system of ancient patriarchs which includes Vasudevas and Baladevas. Krishna is the ninth of the Black Vasudevas* and is connected with Dvaravati or Dvaraka. He wiU become the twelfth tirthankara of the next world-period and a similar position will be attained by Devaki, Rohini, Baladeva and Javakumara, all members of his family. This is a striking proof of the popularity of the Krishna legend outside the Brahmanic reUgion. 1 Pan. IV. 3. 98, Vdsudevdrjundiihydm vun. See Bhandarkar, Vaishnavism and Saivism, p. 3 and J.R.A.S. 1910, p. 168. Siitra 95, just above, appears to point to bhakti, faith or devotion, felt for this Vasudeva. ^ EspeciaUy the Besnagar column. See Rapson, Ancient India, p. 156 and various articles in J.R.A.S. 1909-10. 3 x. 1. VI. ' m. i. 23, Ularo so Kanho isi ahosi. But this may refer to the Rishi mentioned in R. V. vm. 74 who has not necessarily anything to do with the god Krishna. * See Hemacandra Abhidhanacintamani, Ed. BoehtUngk and Rien, p. 128, and Bamett's translation of the Antagada Dasao, pp. 13-15 and 67-82. 154 HINDUISM [ch. No references to Krishna except the above have been found in the earHer Upanishads and Siitras. He is not mentioned in Manu but in one aspect or another he is the principal figure in the Mahabharata, yet not exactly the hero. The Ramayana would have no plot without Rama, but the story of the Maha bharata would not lose its unity if Krishna were omitted. He takes the side of the Pandavas, and is sometimes a chief some times a god but he is not essential to the action of the epic. The legend represents him as the son of Vasudeva, who be longed to the Sattvata sept' of the Yadava tribe, and of his wife Devaki. It had been predicted to Kamsa, king of Mathura (Muttra), that one of her sons would kill him. He therefore slew her first six children: the seventh, Balarama, who is often counted as an incarnation of Vishnu, was transferred by divine intervention to the womb of Rohini. Krishna, the eighth, escaped by more natural methods. His father was able to give him into the charge of Nanda, a herdsman, and his wife Yasoda who brought him up at Gokula and Vrindavana. Here his youth was passed in sporting with the Gopis or mUkmaids, of whom he is said to have married a thousand. He had time, however, to perform acts of heroism, and after kilUng Kamsa, he trans ported the inhabitants of Mathura to the city of Dvaraka which he had built on the coast of Gujarat. He became king of the Yadavas and continued his mission of clearing the earth of tyrants and monsters. In the struggle between the Pandavas and the sons of Dhritarashtra he championed the cause of the former, and after the conclusion of the war retired to Dvaraka. Internecine confiict broke out among the Yadavas and annihi lated the race. Krishna himself withdrew to the forest and was kiUed by a hunter called Jaras (old age) who shot him supposing him to be a deer. In the Mahabharata and several Puranas this bare outhne is distended with a plethora of miraculous incident remarkable even in Indian literature, and almost aU possible forms of divine and human activity are attributed to this many-sided figure. We may indeed suspect that his personahty is dual even in the simplest form of the legend for the scene changes from Mathura to Dvaraka, and his character is not quite the same in the two regions. It is probable that an ancient mihtary hero of the west ^ Apparently the same as the Vpshnig. xxv] Siva and vishnu iss has been combined with a deity or perhaps more than one deity. The pUe of story, sentiment and theology which ages have heaped up round Krishna's name, represents him in three principal aspects. Firstly, he is a warrior who destroys the powers of evU. Secondly, he is associated with love in all its forms, ranging from amorous sport to the love of God in the most spiritual and mystical sense. Thirdly, he is not only a deity, but he actuaUy becomes God in the European and also in the pantheistic acceptation of the word, and is the centre of a phUosophic theology. The first of these aspects is clearly the oldest and it is here, if anywhere, that we may hope to find some fragments of history. But the embeUishments of poets and story-teUers have been so many that we can only point to features which may indicate a substratum of fact. In the legend, Krishna assists the pandavas against the Kauravas. Now many think that the pandavas represent a second and later immigration of Aryans into India, composed of tribes who had halted in the Himalayas and perhaps acquired some of the customs of the inhabitants, including polyandry, for the five Pandavas had one wife in common between them. Also, the meaning of the name Krishna, black, suggests that he was a chief of some non-Aryan tribe. It is, therefore, possible that one source of the Krishna myth is that a body of invading Aryans, described in the legend as the pandavas, who had not exactly the same laws and beUefs as those already estabUshed in Hindustan, were aided by a powerful aboriginal chief, just as the Sisodias in Rajputana were aided by the BMls. It is possible too that Krishna's tribe may have come from Kabul or other mountainous districts of the north west, although one of the most definite pomts in the legend is his connection with the coast town of Dvaraka. The fortifica tions of this town and the fruitless efforts of the demon king, Saiva, to conquer it by seige are described in the Mahabharata', but the narrative is surrounded by an atmosphere of magic and miracle rather than of history^. ' m. XV. ' It would seem that the temple of Dvaraka was buUt between the composition of the narrative in the Mahabharata and of the Vishnu Purana, for whUe the former says the whole town was destroyed by the sea, the latter excepts the temple and says that whoever visits it is freed from aU his sins. See Wilson, Vishnu Purdtia, v. p. 155. 156 HINDUISM [CH. Though it would not be reasonable to pick out the less fantastic parts of the Kjishna legend and interpret them as history, yet we may fairly attach significance to the fact that many episodes represent him as in conflict with Brahmanic institutions and hardly maintaining the position of Vishnu incarnate'. Thus he plunders Indra's garden and defeats the gods who attempt to resist him. He fights with Siva and Skanda. He burns Benares and all its inhabitants. Yet he is caUed Upendra, which, whatever other explanations sectarian in genuity may invent, can hardly mean anything but the Lesser Indra, and he fills the humble post of Arjuna's charioteer. His kinsmen seem to have been of Uttle repute, for part of his mission was to destroy his own clan and after presiding over its annihilation in internecine strife, he was slain himseff. In all this we see dimly the figure of some aboriginal hero who, though ultimately canonized, represented a force not in complete harmony with Brahmanic civiHzation. The figure has also many solar attributes but these need not mean that its origin is to be sought in a sun myth, but rather that, as many early deities were forms of the sun, solar attributes came to be a natural part of divinity and were ascribed to the deified Krishna just as they were to the deified Buddha^. Some authors hold that the historical Krishna was a teacher, similar to Zarathustra, and that though of the mihtary class he was chiefly occupied in founding or supporting what was after wards known as the rehgion of the Bhagavatas, a theistic system inculcating the worship of one God, caUed Bhagavat, and perhaps identical with the Sun. It is probable that Krishna 1 A most curious chapter of the Vishnu Purana (iv. 13) contains a vindication of Krishna's character and a picture of old tribal Ufe. " Neither can I agree with some scholars that Krishna is mainly and primarily a deity of vegetation. All Indian ideas about the Universe and God emphasize the interaction of life and death, growth and decay, spring and winter. Krishna is undoubtedly associated with Ufe, growth and generation, but so is Siva the destroyer, or rather the transmuter. The account in the Mahabhashya (on Pan. m. 1. 26) of the masque representing the slaughter of Kamsa by Krishna is surely a slight foundation for the theory that Krishna was a nature god. It might be easUy argued that Christ is a vegetation spirit, for not only is Easter a spring festival but there are numerous aUusions to sowing and harvest in the Gospels and Paul Ulus trates the resurrection by the germination of com. It is a mistake to seek for uniformity in the history of religion. There were in ancient times different types of mind which invented different kinds of gods, just as now professors invent different theories about gods. xxv] Siva and vishnu 157 the hero was connected with the worship of a special deity, but I see no evidence that he was primarily a teacher'. In the earlier legends he is a man of arms: in the later he is not one who devotes his life to teaching but a forceful personage who explains the nature of God and the universe at the most un expected moments. Now the founders of rehgions such as Mahavlra and Buddha preserve their character as teachers even in legend and do not accumulate miscellaneous heroic exploits. SimUarly modern founders of sects, Hke Caitanya, though revered as incarnations, stiU retain their historical attributes. But on the other hand many men of action have been deified not because they taught anything but because they seemed to be more than human forces. Rama is a classical example of such deification and many local deities can be shown to be warriors, bandits and hunters whose powers inspired respect. It is said that there is a disposition in the Bombay Presidency to deify the Maratha leader Sivaji^. In his second aspect, Krishna is a pastoral deity, sporting among nymphs and cattle. It is possible that this Krishna is in his origin distinct from the violent and tragic hero of Dvaraka. The two characters have Httle in common, except their lawless ness, and the date and locaHty of the two cycles of legend are different. But the death of Kamsa which is one of the oldest incidents in the story (for it is mentioned in the Mahabhashya*) belongs to both and Kamsa is consistently connected with Muttra. The Mahabharata is mainly concerned with Krishna the warrior : the few aUusions in it to the freaks of the pastoral Krishna occur in passages suspected of being late interpolations and, even if they are genuine, show that Uttle attention was paid to his youth. But in later works, the relative importance is reversed and the figure of the amorous herdsman almost banishes the warrior. We can trace the growth of this figure in the sculptures of the sixth century, in the Vishnu and Bhagavata Puranas and the Gita-govinda (written about 1170). Even later is the worship of Radha, Kjishna's mistress, as a portion of the ' The Krishna of the Chandogya Upanishad receives instruction but it is not said that he was himself a teacher. " Hopkins, India Old and New, p. 105. ' Bhandarkar. Allusions to Krishna in Mahabhashya, Ind. Ant. 1874, p. 14. For the pastoral Kpsh^ia see Bhandarkar, Vaish'mvism and Saivism, chap. ix. 158 HINDUISM [ch. deity, who is supposed to have divided himseU into male and female halves'. The birth and adventures of the pastoral Krishna are located in the land of Braj, the district round Muttra and among the tribe of the Abhiras, but the warUke Krishna is connected with the west, although his exploits extend to the Ganges vaUey^- The Abhiras, now caUed Ahirs, were nomadic herdsmen who came from the west and their movements between Kathiawar and Muttra may have something to do with the double location of the Krishna legend. Both archaeology and historical notices teU us something of the history of Muttra. It was a great Buddhist and Jain centre, as the statues and viharas found there attest. Ptolemy caUs it thecity of thegods. Fa-Hsien (400 a.d.) describes it as Buddhist, but that faith was dechning at the time of Hsiian Chuang's visit (c. 630 A.D. ). The sculptural remains also indicate the presence of Grseco-Bactrian influence. We need not therefore feel surprise if we find in the reUgious thought of Muttra elements traceable to Greece, Persia or Central Asia. Some claim that Christianity should be reckoned among these elements and I shaU discuss the question elsewhere. Here I wiU only say that such ideas as were common to Christianity and to the rehgions of Greece and western Asia probably did penetrate to India by the northern route, but of specificaUy Christian ideas I see no proof. It is true that the pastoral Krishna is unUke aU earHer Indian deities, but then no close parallel to him can be adduced from elsewhere, and, take him as a whole, he is a decidedly un christian figure. The resemblance to Christianity consists in the worship of a divine child, together with his mother. But this feature is absent in the New Testament and seems to have been borrowed from paganism by Christianity. The legends of Muttra show even clearer traces than those already quoted of hostility between Krishna and Brahmanism. He forbids the worship of Indra*, and when Indra in anger sends down a deluge of rain, he protects the country by holding "¦ The divinity of Radha is taught speciaUy in the Brahmavaivarta Purana and the Narada paiicaratra, also called Jfianamritasara. She is also described in the Gopala-tapaniya Upanishad of unknown date. * ButKamsa appears in both series of legends, i.e., in the Ghata-Jataka whichoon- tains no hint of the pastoral legends but is a variant of the story of the warlike Krishna. ' Vishnu Purana, v. 10, 11 from which the quotations in the text are taken. Much of it is repeated in the Harivamsa. See for instance H. 3808. xxv] Siva and vishnu 159 up over it the hill of Goburdhan, which is still one of the great centres of pUgrimage'. The language which the Vishnu Purana attributes to him is extremely remarkable. He interrupts a sacrifice which his fosterfather is offering to Indra and says, "We have neither fields nor houses: we wander about happily wherever we Ust, traveUing in our waggons. What have we to do with Indra? Cattle and mountains are (our) gods. Brahmans offer worship with prayer: cultivators of the earth adore their landmarks but we who tend our herds in the forests and mountains should worship them and our kine." This passage suggests that Krishna represents a tribe of highland nomads who worshipped mountains and cattle and came to terms with the Brahmanic ritual only after a struggle. The worship of mountain spirits is common in Central Asia, but I do not know of any evidence for cattle-worship in those regions. Clemens of Alexandria^, writing at the end of the second century a.d., teUs us that the Indians worshipped Herakles and Pan. The pastoral Krishna has considerable resemblance to Pan or a Faun, but no representations of such beings are recorded from Graeco-Indian sculptures. Several Bacchic groups have however been discovered in Gandhara and also at Muttra* and Megasthenes recognized Dionysus in some Indian deity. Though the Bacchic revels and mysteries do not explain the pastoral element in the Kjishna legend, they offer a paraUel to some of its other features, such as the dancing and the crowd of women, and I am incHned to think that such Greek ideas may have germinated and proved fruitful in Muttra. The Greek king Menander is said to have occupied the city (c. 155 B.C.), and the sculptures found there indicate that Greek artistic forms were used to express Indian ideas. There may have been a similar fusion in reUgion. In any case. Buddhism was predominant in Muttra for several centuries. It no doubt forbade the animal sacrifices of 1 The Muttra cycle of legends cannot be very late for the inscription of Glai Lomor in Champa (811 a.d.) speaks of Narayana holding up Goburdhan and a Cambojan inscription of Prea Eynkosey (970 a.d.) speaks of the banks of the Yamuna where Krishna sported. These legends must have been prevalent in India some time before they traveUed so far. Some of them are depicted on a piUar found at Mandor and possibly referable to the fourth century a.d. See .Arch. Survey Ind. 1905-1906, p. 135. ^ Strom, m. 194. See M'Crindle, Ancient India, p. 183. ' Vincent Smith, Fine Art in India, pp. 134-138. 160 HINDUISM [CH. the Brahmans and favoured milder rites. It may even offer some explanation for the frivolous character of much in the Krishna legend'. Most Brahmanic deities, extraordinary as their conduct often is, are serious and imposing. But Buddhism claimed for itself the serious side of rehgion and while it tolerated local godUngs treated them as fairies or elves. It was perhaps while Krishna was a humble rustic deity of this sort, with no claim to represent the Almighty, that there first gathered round him the cycle of light love-stories which has clung to him ever since. In the hands of the Brahmans his worship has undergone the strangest variations which touch the highest and lowest planes of Hinduism, but the Muttra legend stiU retains its special note of pastoral romance, and exhibits ICrishna in two principal characters, as the divine child and as the divine lover. The mysteries of birth and of sexual union are congenial topics to Hindu theology, but in the cult of Muttra we are not concemed with reproduction as a world force, but simply with childhood and love as emotional manifestations of the deity. The same ideas occur in Christianity, and even in the Gospels Christ is compared to a bridegroom, but the Krishna legend is far more gross and naive. The infant Krishna is commonly adored in the form known as Makhan Chor or the Butter Thief^. This represents him as a crawUng child holding out one hand full of curds or butter which he has stolen. We speak of idoHzing a chUd, and when Hindu women worship this image they are unconsciously generalizing the process and worshipping chUdhood, its way ward pranks as weU as its loveable simpHcity, and though it is hard for a man to think of the freaks of the butter thief as a manifestation of divinity, yet clearly there is an analogy between these childish escapades and the caprices of mature deities, which are respectfully described as mysteries. If one admits the worship of the Bambino, it is not unreasonable to include in it admiration of his rogueries, and the tender playfulness which is permitted to enter into this cult appeals profoundly to ' In the Sutta-nipata Mara, the Evil One is called Kanha, the phonetic equivalent of Krishna in Prakrit. Can it be that Mara and his daughters have anything to do with Krish^ia and the Gopis ? ^ Compare the Greek stories of the infant Hermes who steals ApoUo's cattle and invents the lyre. Compare too, as having a general resemblance to fantastic Indian legends, the story of young Hephaestus. xxv] Siva and vishnu lei Indian women. Images of the Makhan Chor are sold by thou sands in the streets of Muttra. Even more popular is the image known as Kanhaya, which represents the god as a young man playing the flute as he stands in a careless attitude, which has something of Hellenic grace. Krishna in this form is the beloved of the Gopis, or milk-maids, of the land of Braj, and the spouse of Radha, though she had no monopoly of him. The stories of his froUcs with these damsels and the rites instituted in memory thereof have brought his wor ship into merited discredit. Krishnaism offers the most extensive manifestation to be found in the world of what W. James calls the theopathic condition as iUustrated by nuns Hke Marguerite Marie Alacoque, Saint Gertrude and the more distinguished Saint Theresa. "To be loved by God and loved by him to distraction (jusqu'a la foUe), Margaret melted away with love at the thought of such a thing.... She said to God, 'Hold back, 0 my God, these torrents which overwhelm me or else enlarge my capacity for their reception ' '." These are not the words of the Gita-govinda or the Prem Sagar, as might be supposed, but of a CathoUc Bishop describing the transports of Sister Mar guerite Marie, and they Ulustrate the temper of Krishna's worshippers. But the verses of the Marathi poet, Tukaram, who Hved about 1600 a.d. and sang the praises of Krishna, rise above this sentimentaHty though he uses the language of love. In a letter to Sivaji, who desired to see him, he wrote, "As a chaste wife longs only to see her lord, such am I to Vitthala^. AU the world is to me Vitthala and nothing else: thee also I behold in him." He also wrote elsewhere, "he that taketh the unprotected to his heart and doeth to a servant the same kind ness as to his own chUdren, is assuredly the image of God." More recently Ramakrishna, whose sayings breathe a wide inteUigence as weU as a wide charity, has given this rehgion of love an expression which, if somewhat too sexual to be perfectly in accordance with western taste, is nearly related to emotional Christianity. "A true lover sees his god as his nearest and dearest relative" he writes, "just as the shepherd women of ^ Mgr. Bongard, Hisloire de la Bienheureuse Marguirite Marie. Quoted by W. James, Varieties of Religious Experience, p. 343. 2 Vitthal or Vittoba is a local deity of Pandharpur in the Deccan (perhaps a deified Brahman of the place) now identified with Krishna. m. TT. 11 162 HINDUISM [oh. Vrindavana saw in Krishna not the Lord of the Universe but their own beloved.... The knowledge of God may be Hkened to a man, while the love of God is Hke a woman. Knowledge has entry only up to the outer rooms of God, and no one can enter into the inner mysteries of God save a lover.... Knowledge and love of God are ultimately one and the same. There is no difference between pure knowledge and pure love'." These extracts show how Krishna as the object of the soul's desire assumes the place of the Supreme Being or God. But this surprising transformation^ is not speciaUy connected with the pastoral and erotic Krishna: the best known and most thorough-going exposition of his divinity is found in the Bhagavad-gita, which represents him as being in his human aspect, a warrior and the charioteer of Arjuna. Probably some seventy-five milHons to-day worship Krishna, especially under the name of Hari, as God in the pantheistic sense and naturaUy the more his identity with the supreme spirit is emphasized, the dimmer grow the legendary features which mark the hero of Muttra and Dvaraka, and the human element in him is reduced to this very important point that the tie uniting him to his worshippers is one of sentiment and affection. In the foUowing chapters I shaU treat of this worship when describing the various sects which practise it. A question of some importance for the history of Kiishna's deification is the meaning of the name Vasudeva. One explanation makes it a patronymic, son of Vasudeva, and supposes that when this prince Vasudeva was deified his name, like Rama, was trans ferred to the deity. The other regards Vasudeva as a name for the deity used by the Sattvata clan and supposes that when Krishna was deified this already weU-known divine name was bestowed on him. There is much to be said for this latter theory. As we have seen the Jains give the title Vasudeva to a series of supermen, and a remarkable legend states* that a king caUed ' Life and Sayings of Rdmakrishna. Trans. P. Max Miiller, pp. 137-8. The EngUsh poet Crashaw makes free use of reUgious metaphors drawn from love and even Francis Thompson represents God as the lover of the Soul, e.g. in his poem Any Saint. ¦ * Though surprising, it can be paraUeled in modem times for Kabir (c. 1400) was identified by his later foUowers with the supreme spirit. ' Mahabhar. Sabhap. xiv. Vishnu Pur. v. xxxiv. The name also occurs in the Taittiriya Aranyaka (i. 31) a work of moderate if not great antiquity Nazaya^iaya vidmahe Vasudevaya dhimahi. xxv] Siva and vishnu i63 Paundraka who pretended to be a deity used the title Vasudeva and ordered Kjishna to cease using it, for which impertinence he was slain. This clearly implies that the title was something which could be detached from Krishna and not a mere patrony mic. Indian writings countenance both etymologies of the word. As the name of the deity they derive it from vas to dweU, he in whom aU things abide and who abides in all'. 5 Siva and Vishnu are not in their nature different from other Indian ideas, high or low. They are the offspring of philosophic and poetic minds playing with a luxuriant popular mythology. But even in the epics they have already become fixed points in a flux of changing fancies and serve as receptacles in which the most diverse notions are collected and stored. Nearly all phUosophy and superstition finds its place in Hinduism by being connected with one or both of them. The two worships are not characteristic of different periods : they coexist when they first become known to us as they do at the present day and in essential doctrines they are much ahke. We have no name for this curious double theism in which each party describes its own deity as the supreme god or AU-god, yet without denying the god of the other. Something similar might be produced in Christianity if different Churches were avowedly to worship different persons of the Trinity. Siva and Vishnu are sometimes contrasted and occasionaUy their worshippers quarreP. But the general inclination is rather ' See Vishnu Pur. vi. v. See also Wilson, Vishnu Purdtia, i. pp. 2 and 17. ' Thus the Saura Purana inveighs against the Madhva sect (xxxvin.-XL.) and caUs Vishnu the servant of Siva: a Puranic legal work called the Vriddha-Harita- Samhita is said to contain a polemic against Siva. Occasionally we hear of colUsions between the foUowers of Vishnu and Siva or the desecration of temples by hostile fanatics. But such conflicts take place most often not between widely different sects but between subdivisions of the same sect, e.g., Ten-galais and Vada-galais. It would seem too that at present most Hindus of the higher castes avoid ostentatious membership of the modem sects, and though they may practise special devotion to either Vishriu or Siva, yet they visit the temples of both deities when they go on pUgrimages. Jogendra Nath Bhattacharya in his Hindu Castes and Sects says (p. 364) that aristocratic Brahmans usuaUy keep in their private chapels both a salagram representing Vishnu and emblems representing Siva and his spouse. Hence different observers vary in their estimates of the importance of sectarian divisions, some holding that sect is the essence of modem Hinduism and others that most educated Hindus do not worship a sectarian deity. The Kurma Purana, Part i. chap. xxn. contains some curious rules as to what deities should be worshipped by the various classes of men and spirits. 11—2 164 HINDUISM [cH. to make the two figures approximate by bestowing the same attributes on both. A deity must be able to satisfy emotional devotion: hence the TamU Sivaite says of Siva the destroyer, "one should worship in supreme love him who does kindness to the soul." But then the feature in the world which most im presses the Hindu is the constant change and destruction, and this must find a place in the AU-god. Hence the sportive kindly Krishna comes to be declared the destroyer of the worlds'- It is as if in some vast Dravidian temple one wandered through two corridors differently ornamented and assigned to the priests of different rites but both leading to the same image. Hence it is not surprising to find that there is actuaUy a deity — if indeed the term is suitable, but European vocabularies hardly provide one which meets the case — caUed Harihara (or Sankara-Nara- yana), that is Siva and Vishnu combined. The Harivamsa contains a hymn addressed to him: fairly ancient sculptures attest the prevalence of his worship in the Deccan, especiaUy at Badami, he was once the chief deity of Camboja and he is stiU popular in south India. Here besides being worshipped under his own name he has undergone a singular transformation and has probably been amalgamated with some aboriginal deity. Under the designation of Ayenar (said to be a corruption of Harihara) he is extensively worshipped as a vUlage god and reputed to be the son of Siva and Vishnu, the latter having kindly assumed the form of a woman to effect his birth. Another form of this inchnation to combine and unite the various manifestations of the Divine is the tendency to worship groups of gods, a practice as old as the Vedas. Thus many temples are dedicated to a group of five, namely, Siva, Vishnu, Durga, Ganesa and the Sun and it is stated that every Hindu worships these five deities in his daily prayers^. The Trimiirti, or figure of Brahma, Siva and Vishnu, iUustrates the worship of groups. Its importance has sometimes been over-estimated by Europeans from an idea that it corresponded to the Christian Trinity, but in reahty this triad is late and has Uttle significance. No stress is laid on the idea of three in one and the number of persons can be increased. The Brahma-vaivarta Purana for instance adds KJrishna to Brahma, Siva and Vishnu. The union 1 Bhag.-gita, xi. 23-34. ^ See Srisa Chandra Vasu, Daily practice of the Hindus, p. 118. xxv] Siva and vishnu i65 of three personaUties is merely a way of summing up the chief attributes of the AU-God. Thus the Vishnu Purana' extols Vishnu as being " Hiranyagarbha, Hari and Sankara {i.e. Brahma, Vishnu and Siva), the creator, preserver and destroyer," but in another passage as him who is "Brahma, Isvara and spirit (Pums), who with the three Gunas (quahties of matter) is the cause of creation, preservation and destruction...." The origin of the triad, so far as it has any doctrinal or philosophical meaning, is probably to be sought in the personification of the three Gunas^. 1 n. 1 and i. 1. " See Maitrayana Up. v. 2. It is highly probable that the celebrated image at Elephanta is not a Trimurti at all but a Mahe^araiirti of Siva. See Gopinatha Rao, Hindu Iccmog. n. 382. CHAPTER XXVI FEATURES OF HINDUISM: RITUAL, CASTE, SECT, FAITH 1 In the last chapter I traced the growth of the great gods Siva and Vishnu. The prominence of these figures is one of the marks which distinguish the later phase of Indian rehgion from the earlier. But it is also distinguished by various practices, insti tutions and beliefs, which are more or less connected with the new deities. Such are a new ritual, the elaboration of the caste system, the growth of sects, and the tendency to make devotion to a particular deity the essence of reUgion. In the present chapter I shall say something of these phenomena. Hinduism has often and justly been compared to a jungle. As in the jungle every particle of soil seems to put forth its spirit in vegetable life and plants grow on plants, creepers and parasites on their more stalwart brethren, so in India art, commerce, warfare and crime, every human interest and aspiration seek for a manifestation in religion, and since men and women of all classes and occupations, aU stages of education and civilization, have contributed to Hinduism, much of it seems low, foohsh and even immoral. The jungle is not a park or garden. Whatever can grow in it, does grow. The Brahmans are not gardeners but forest officers. To attempt a history or description of Indian creeds seems an enterprise as vast, hope less and pathless as a general account of European poHtics. As for many centuries the Ufe of Europe has expressed itseU in politics, so for even longer ages the Hfe of India, which has more inhabitants than western Europe', has found expression in religion, speculation and philosophy, and has left of all this thought a voluminous record, mighty in bulk if wanting in dates and events. And why should it chronicle them? The truly religious mind does not care for the history of reUgion, • The population of India (about 315 milUons) is larger than that of Europe without Russia. CH. XXVI] RITUAL, CASTE, SECT, FAITH 167 just as among us the scientific mind does not dwell on the history of science. Yet in spite of their exuberance Hinduism and the jungle have considerable uniformity. Here and there in a tropical forest some weU-grown tree or brilHant flower attracts attention, but the general impression left on the traveller by the vegetation as he passes through it mile after mile is infinite repetition as weU as infinite luxuriance. And so in Hinduism. A monograph on one god or one teacher is an interesting study. But if we continue the experiment, different gods and different teachers are found to be much the same. We can write about Vishnuism and Sivaism as if they were different reUgions and this, though incomplete, is not incorrect. But in their higher phases both show much the same exceUences and when degraded both lead to much the same abuses, except that the worship of Vishnu does not aUow animal sacrifices. This is true even of externals. In the temples of Madura, Poona and Benares, the deities, the rites, the doctrines, the race of the worshippers and the archi tecture are aU different, yet the impression of uniformity is strong. In spite of divergences the reUgion is the same in all three places: is smacks of the soil and nothing Hke it can be found outside India. Hinduism is an unusual combination of animism and pan theism, which are commonly regarded as the extremes of savage and of phUosophic belieL In India both may be found separately but frequently they are combined in starthng juxtaposition. The same person who worships Vishnu as identical with the universe also worships him in the form of a pebble or plant'. The average Hindu, who cannot live permanently in the altitudes of pantheistic thought, regards his gods as great natural forces, akin to the mighty rivers which he also worships, irresistible and often beneficent but also capricious and destructive. Whereas Judaism, Christianity and Islam aU identify the moral law with the wUl and conduct of the deity, in Hinduism this is not com pletely admitted in practice, though a Hbrary might be fiUed 1 But compare the English poet "Flower in the crannied waU, I pluck you out of the crannies, but if I could understand What you are, root and all, and all in aU I should know what God and man ia." 168 HINDUISM [ch. with the beautiful things that have been said about man and God. The outward forms of Indian reUgion are pagan after the fashion of the ancient world, a fashion which has in most lands passed away. But whereas in the fourth century a.d. European paganism, despite the efforts of anti-christian eclectics, proved inelastic and incapable of satisfying new religious cravings, this did not happen in India. The bottles of Hinduism have always proved capable of holding all the wine poured into them. When a new sentiment takes possession of men's souls, such as love, repentance, or the sense of sin, some deity of many shapes and sympathies straightway adapts himself to the needs of his worshippers. And yet in so doing the deity, though he enlarges himself, does not change, and the result is that we often meet with strange anachronisms, as if Jephthah should listen appre ciatively to the Sermon on the Mount and then sacrifice his daughter to Christ. Many Hindu temples are served by dancing girls who are admittedly prostitutes', an institution which takes us back to the cultus of Corinth and Babylon and is without paraUel in any nation on approximately the same level of civiH zation. Only British law prevents widows from being burned with their dead husbands, though even in the Vedic age the custom had been discontinued as barbarous^. But for the same legislation, human sacrifice would probably be common. What the gods do and what their worshippers do in their service cannot according to Hindu opinion be judged by ordinary laws of right and wrong. The god is supra-moral: the worshipper when he enters the temple leaves conventionahty outside. Yet it is unfair to represent Hinduism as characterized by Hcence and cruelty. Such tendencies are counterbalanced by the strength and prevalence of ideas based on renunciation and seU-effacement. All desire, all attachment to the world is an evil; aU seU-assertion is wrong. Hinduism is constantly in extremes: sometimes it exults in the dances of Krishna or the destructive fury of Kali : more often it struggles for release from the transitory and for union with the permanent and real by 1 Efforts are now being made by Hindus to suppress this institution. " In the Vedic funeral ceremonies the wife lies down by her dead husband and is called back to the world of the living which points to an earlier form of the rite where she died with him. But even at this period, those who did not follow the Vedic customs may have kiUed widows with their husbands (see too Ath. Veda, XII. 3), and later, the invaders frora Central Asia probably reinforced the usage. The much-abused Tantras forbid it. XXVI] RITUAL, CASTE, SECT, FAITH 169 self-denial or rather seU-negation, which aims at the total suppression of both pleasure and pain. This is on the whole its dominant note. In the records accessible to us the transition from Brahman ism — that is, the reUgion of the Vedas and Brahmanas — to Hinduism does not appear as direct but as masked by Buddhism. We see Buddhism grow at the expense of Brahmanism. We are then conscious that it becomes profoundly modified under the influence of new ideas. We see it decay and the reUgion of the Brahmans emerge victorious. But that religion is not what it was when Buddhism first arose, and is henceforth generally known as Hinduism. The materials for studying the period in which the change occurred — say 400 B.C. to 400 a.d. — are not scanty, but they do not faciUtate chronological investigation. Art and architecture are mainly Buddhist until the Gupta period (c. 320 A.D.) and Hterature, though plentiful, is undated. The Mahabharata and Ramayana must have been edited in the course of these 800 years, but they consist of different strata and it is not easy to separate and arrange them without assuming what we want to prove. From 400 b.c (if not from an earUer date) onwards there grew up a great volume of epic poetry, founded on popular baUads, teUing the stories of Rama and the pandavas'. It was distinct from the canonical Uteratures of both Brahmans and Buddhists, but though it was not in its essential character religious, yet so general in India is the interest in rehgion that whole theological treatises were incorporated in these stories without loss, in Indian opinion, to the interest of the narrative. If at the present day a congregation is seen in * For the history of the Ramayana and Mahabharata and the dates assignable to the different periods of growth, see Wintemitz, Gesch. Ind. Lit. vol. i. p. 403 and p. 439. Also Hopkins' Great Epic of India, p. 397. The two poems had assumed soraething Uke their present form in the second and fourth centuries a.d. respectively. These are probably the latest dates for any substantial additions or alterations and there is considerable evidence that poems called Bharata and Ramayana were well known early in the Christian era. Thus in Asvaghosha's Siitralankara (story xxiv) they are mentioned as warUke poems inculcating unbuddhist views. The Ramayana is mentioned in the Mahavibhasha and was known to Vasubandhu (J.R.A.S. 1907, p. 99). A Cambojan inscription dating from the flrst years of the seventh century records arrangements made for the recitation of the Ramayana, Purana and com plete (afesha) Bharata, which implies that they were known in India considerably earUer. See Barth, Inscrip. Sanscrites de Cambodge, pp. 29-31. The Mahabharata itself admits that it is the result of gradual growth for in the opening section it says that the Bharata consists of 8800 verses, 24,000 verses and 100,000 verses. 170 HINDUISM [CH. a Hindu temple Ustening to a recitation, the text which is being chanted wiU often prove to be part of the Mahabharata. Such a ceremony is not due to forgetfulness of the Veda but is a repetition of what happened long before our era when rhapsodists strung together popular narratives and popular theology. Such theology cannot be rigidly separated from Brahmanism and Buddhism. It grew up under their infiuence and accepted their simpler ideas. But it brought with it popular behefs which did not strictly speaking belong to either system. By attacking the main Brahmanic doctrines the Buddhists gave the popular religion its opportunity. For instance, they condemned animal sacrifices and derided the idea that trained priests and compH cated rites are necessary. This did not destroy the influence of the Brahmans but it disposed them to admit that the Vedic sacrifices are not the only means of salvation and to authorize other rites and beliefs. It was about this time, too, that a series of invasions began to pour into India from the north-west. It may be hard to distinguish between the foreign behefs which they introduced and the Indian beliefs which they accepted and modified. But it is clear that their general effect was to upset traditional ideas associated with a ritual and learning which required lifelong study. 2 It has been well said' that Buddhism did not waste away in India until rival sects had appropriated from it everything they could make use of. Perhaps Hinduism had an even stronger doctrinal influence on Buddhism. The deiflcation of the Buddha, the invention of Bodhisattvas who are equivalent to gods and the extraordinary aUiance between late Buddhism and Sivaism, are all instances of the general Indian view overcoming the special Buddhist view. But Buddhism is closely connected with the theory of incarnations and the development of the Advaita philosophy, and in the externals of rehgion, in rites, ceremonies and institutions, its influence was great and lasting. We may take first the doctrine of Ahimsa, non-injury, or in other words the sanctity of animal Hfe. This beautiful doctrine, the glory of India, if not invented by the Buddha at least arose in schools which were not Brahmanic and were related to the Jain and ' Hardy, Indische Religionsgeschichie, p. 101. XXVI] RITUAL, CASTE, SECT, FAITH 171 Buddhist movements. It formed no part of the Vedic religion in which sacrifice often meant butchery. But in Hinduism, it meets with extensive though not universal acceptance. With the Vaishnavas it is an article of faith nor do the worshippers of Siva usuaUy propitiate him with animal sacrifices, though these are offered by the Saktas and also by the smaU class of Brahmans who stiU preserve the Vedic ritual'. Hardly any Hindus habituaUy eat meat and most abhor it, especiaUy beef. Yet beef-eating seems to have been permitted in Vedic times and even when parts of the Mahabharata were composed. Apart from animal sacrifices Buddhism was the main agent in effecting a mighty revolution in worship and ritual. One is tempted to regard the change as total and complete, but such wide assertions are rarely true in India : customs and institutions are not swept away by reformers but are cut down like the grass and Uke the grass grow up again. They sometimes die out but they are rarely destroyed. The Vedic sacrifices are still occasion aUy offered 2, but for many centuries have been almost entirely superseded by another form of worship associated with temples and the veneration of images. This must have become the dominant form of Hindu cultus in the first few centuries of our era and probably earHer. It is one of the ironies of fate that the Buddha and his foUowers should be responsible for the growth of image worship, but it seems to be true. He laughed at sacrifices and left to his disciples only two forms of rehgious exercise, sermons and meditation. For Indian monks, this was perhaps sufficient, but the laity craved for some outward form of worship. This was soon found in the respect shown to the memory of the Buddha and the reUcs of his body, although Hinduism never took kindly to rehc worship. We hear too of Cetiyas. In the Pitakas this word means a popular shrine uncon nected with either Buddhist or Brahmanic ceremonial, sometimes ' But some of these latter sacriflce images made of dough instead of living animals. * It is said that the Agnishtoma was performed in Benares in 1898, and in the last few years I am told that one or two Vedic sacrifices have been offered annuaUy in various parts of southern India. I have myself seen the sites where such sacrifices were offered in 1908-9 in Mysore city and in Chidambarara, and in 1912 at Wei near Poona. The most usual form of sacrifice now-a-days is said to be the Vajapeya. Much Vedic ritual is stUl preserved in the domestic life of the Nambathiri aud other Brahmans of southern India. See Cochin, Tribes and Castes, and Thurston, Castes and Tribes of southern India. 172 HINDUISM [ch. perhaps merely a sacred tree or stone, probably honoured by such simple rites as decorating it with paint or flowers. A Httle later, in Buddhist times, the Cetiya became a cenotaph or reUquary, generally located near a monastery and surrounded by a passage for reverential circumambulation. AUusions in the Pitakas also indicate that then as now there were fairs. The early Buddhists thought that though such gatherings were not edifjang they might be made so. They erected sacred buildings near a monastery, and held festivals so that people might coUect together, visit a holy place, and hear sermons. In the earUest known sanctuaries, the funeral monu ment (for we can scarcely doubt that this is the origin of the stupa)' has already assumed the conventional form known as Dagoba, consisting of a dome and chest of reUcs, with a spire at the top, the whole surrounded by raihngs or a colonnade, but though the carving is lavish, no figure of the Buddha himself is to be seen. He is represented by a symbol such as a footprint, wheel, or tree. But in the later school of sculpture known as Gandhara or Grseco-Buddhist he is frequently shown in a fuU length portrait. This difference is remarkable. It is easy to say that in the older school the Buddha was not depicted out of reverence, but less easy to see why such deUneation should have shocked an Indian. But at any rate there is no difficulty in understanding that Greeks or artists infiuenced by Greeks would think it obvious and proper to make an effigy of their principal hero. In these shrines we have if not the origin of the Hindu temple, at any rate a parallel development more nearly aUied to it than anything in the Vedic rehgion^. For the Buddhist shrine was a monument built over a receptacle containing reUcs and the essential feature of Hindu temples is a ceU containing an image or emblem and generally surmounted by a tower. The surrounding courts and corridors may assume gigantic propor tions, but the central shrine is never large. Images had no place ^ The outline of a stiipa may be due to imitation of houses constructed with curved bamboos as Vincent Smith contends (History of Fine Art, p. 17). But this is compatible with the view that stone buUdings with this curved outline had come to be used speciaUy as funeral monuments before Buddhism popularized in India and aU Eastern Asia the architectural form called stupa. 2 The temple of Aihole near Badami seems to be a connecting Unk between a Buddhist stiipa with a pradakshina path and a Hindu shrine. XXVI] RITUAL, CASTE, SECT, FAITH 173 in the Vedic sacrifices and those now worshipped in temples are generaUy smaU and rude, and sometimes (as at Bhuvaneshwar and Srirangam) the deity is represented by a block or carved stone which cannot be moved, and may have been honoured as a sacred rock long before the name of Vishnu or Siva was known in those regions'. The conspicuous statues often found outside the shrine are not generally worshipped and are merely ornaments. Buddhism did not create the type of ritual now used in Hindu temples, yet it contributed towards it, for it attacked the old Brahmanic sacrifices, it countenanced the idea that particular places and objects are holy, and it encouraged the use of images. It is strange that these wide-spread ideas should find no place in the Vedic rehgion, but even now-a-days whenever the old Vedic sacrifices are celebrated they are uncon- taminated by the temple ceremonial. More than this, the priests or Pujaris who officiate in temples are not always Brahmans and they rarely enjoy much consideration^. This curious and marked feature may be connected with the inveterate Indian feeHng that, though it is weU to multiply rites and rules for neophytes, no great respect is due to men occupied with mere ceremonial. But it also testifies to a dim consciousness that modern temples and their ceremonies have Uttle to do with the thoughts and mode of Ufe which made the Brahmans a force in India. In many ways the Brahmans dissociate themselves from popular reUgion. Those of good family will not perform reUgious rites for Sudras and treat the Brahmans who do so as inferiors*. The simplest ceremonial in use at the present day is that employed in some Sivaite temples. It consists in placing leaves on the linga and pouring holy water over it. These rites, which may be descended from prehistoric stone worship, are generaUy 1 In most temples (at least in southern India) there are two images: the milla- vigraha which is of stone and fixed in the sanctuary, and the utsava-vigraha which is smaUer, made of metal and carried in proce3.9ions. * Thus Bhattacharya (Hindu Castes and Sects, p. 127) enumerates eleven classes of Brahmans, who "have a very low status on account of their being connected with the great public shrines," and adds that mere residence in a place of pilgrimage for a few generations tends to lower the status of a Brahmanic famUy. ^ Thus in Bengal there is a special class, the Bama Brahmans, who perform reUgious rites for the lower castes, aud are divided into six classes according to the castes to whom they minister. Other Brahmans will not eat or intermarry with them or even take water from them. 174 HINDUISM [oh. accompanied by the reading of a Purana. But the commonest form of temple ritual consists in treating the image or symbol as an honoured human being'. It is awakened, bathed, dressed and put to bed at the close of day. Meals are served to it at the usual hours. The food thus offered is ca,\led prasdd (or favour) and is eaten by the devout. Once or twice a day the god holds a levee and on festivals he is carried in procession. These cere monies are specially characteristic of the worship of Krishna whose images receive all the endearments lavished on a pet child. But they are also used in the temples of Siva and Parvati, and no less than twenty-two of them are performed in the course of the day at the temple of Bhuvaneshwar in Orissa. It is clear that the spirit of these rites is very different from that which inspires pubhc worship in other civiUzed countries at the present day. They are not congregational or didactic, though if any of the faithful are in the temple at the time of the god's levee it is proper for them to enter and salute him. Neither do they recall the magical ceremonies of the Vedic sacrifices^. The waving of Hghts (arati) before the god and the burning of incense are almost the only acts suggestive of ecclesiastical ritual. The rest consists in treating a symbol or image as if it were a Hving thing capable of enjoying simple physical pleasures. Here there are two strata. We have reaUy ancient rites, such as the anointing or ornamenting of stones and offerings of food in sacred places. In this class too we may reckon the sacrifice of goats (and formerly of human beings) to KaU*. But on the other hand the growing idea of Bhakti, that is faith or devotion, imported a sentimental element and the worshipper endeavoured to pet, caress and amuse the deity. It is hard to see anything either healthy or artistic in this ' This is extraordinarUy like the temple ritual of the ancient Egyptians. For some account of the construction and ritual of south Indian temples see Richards in J. of Mythic Soe. 1919, pp. 158-167. ' But Vedic mantras are used in these ceremonies. The libations of water or other liquids are said to be accompanied by the mantras recited at the Soma sacrifice. ' At these sacrifices there is no elaborate ritual or suggestion of symboUsm. The animal is beheaded and the inference is that KaU likes it. SimUarly simple is the offering of coco-nuts to KaU. The worshipper gives a nut to the pujari who spUts it in two with an axe, spills the milk and hands back haU the nut to the worshipper. This is the sort of primitive offering that might be made to an African fetish. XXVI] RITUAL, CASTE, SECT, FAITH 175 emotional ritual. The low and foohsh character of many temple ceremonies disgusts even appreciative foreigners, but these services are not the whole of Hindu worship. All Hindus per form in the course of the day numerous acts of private devotion varying according to sect, and a pious man is not depen dent on the temple like a catholic on his church. Indian life is largely occupied with these private, intimate, individual observances, hardly noticeable as ceremonies and concerned with such things as dressing, ablution and the preparation of food. The monastic institutions of India seem due to Buddhism. There were wandering monks before the Buddha's time, but the practice of founding estabhshments where they could reside permanently, originated in his order. There appears to be no record of Hindu (as opposed to Buddhist) monasteries before the time of Sankara in the ninth century, though there must have been places where the leamed congregated or where wandering ascetics could lodge. Sankara perceived the advantage of the cenobitic Hfe for organizing reUgion and founded a number of maths or coUeges. Subsequent reUgious leaders imitated him. At the present day these institutions are common, yet it is clear that the wandering spirit is strong in Hindus and that they do not take to monastic discipline and fixed residence as readily as Tibetans and Burmese. A math is not so much a convent as the abode of a teacher. His pupils frequent it and may become semi-resident: aged pilgrims may make it their last home, but the inmates are not a permanent body foUowing a fixed rule Hke the monks of a Vihara. The Sattras of Assam, however, are true monasteries (though even there vows and monastic costume are unknown) and so are the estabhshments of the Swaminarayana sect at Ahmedabad and Wartai. The vast and compHcated organization of caste is mainly a post- Vedic growth and in the Buddha's time was only in the making'- His order was open to all classes aUke, but this does not imply that he was adverse to caste, so far as it then pre- ' See especiaUy the Ambattha Siitta (Dig. Nik. 3) and Rhys Davids's introduc tion. 176 HINDUISM [CH. vailed, or denied that men are divided into categories deter mined by their deeds in other births. But on the whole the influence of Buddhism was unfavourable to caste, especially to the pretensions of the Brahmans, and an extant polemic against caste is ascribed (though doubtfuUy) to Asvaghosha'. On the other hand, though caste is in its origin the expression of a social rather than of a reUgious tendency, the whole institution and mechanism have long been supported and exploited by the Brahmans. Few of them would dispute the proposition that a man cannot be a Hindu unless he belongs to a caste. The reason of this support is undisguised, namely, that they are the first and chief caste. They make their own position a matter of rehgion and claim the power of purifying and rehabihtating those who have lost caste but they do not usuaUy interfere with the rules of other castes or excommunicate those who break them^. That is the business of the Pancayat or caste council. Sometimes rehgion and caste are in opposition, for many modern religious leaders have begun by declaring that among believers there are no social distinctions. This is true not only of teachers whose orthodoxy is dubious, such as Nanak, the founder of the Sikhs, and Basava, the founder of the Lingayats*, but also of Vallabhacarya and Caitanya. But in nearly aU cases caste reasserts itseU. The reUgious teachers of the sect receive extravagant respect and form a body apart. This phenomenon, which recurs in nearly all communities, shows how the Brahmans established their position. At the same time social distinctions make themselves felt among the laity, and those who claim to be of good position dissociate themselves from those of lower birth. The sect ends by observing caste on ordinary occasions, and it is only in some temples (such as that of Jagannath at Puri)* that the worshippers mix and eat a sacred meal together. Sometimes, however, the sect which renounces caste becomes 1 See Weber, Die Vajrasuchi and Nanjio, Catal. No. 1303. In Ceylon at the present day only members of the higher castes can become Bhikkhus. ' But it is said that in Southern India serious questions of caste are reported to the abbot of the Sringeri monastery for his decision. ^ The modem Lingayats demur to the statement that their founder rejected caste. * So too in the cakras of the Saktists all castes are equal during the performance of the ceremony. XXVI] RITUAL, CASTE, SECT, FAITH 177 itseff a caste. Thus, the Sikhs have become almost a nation and other modern castes arising out of sects are the Atiths, who are Sivaites, the Saraks, who appear to have been originaUy Buddhists, and the Baishnabs (Vaishnavas), a name commonly given in Bengal to those foUowers of Caitanya who persist in the original rule of disregarding caste regulations within the sect, and hence now form a separate community. But as a rule sect and caste are not co-extensive and the caste is not a religious corporation. Thus the different subdivisions of the Baniyas belong to different sects and even in the same subdivision there is no rehgious uniformity'. Caste in its later developments is so complex and irregular, that it is impossible to summarize it in a formula or explain it as the development of one principle. In the earUest form known two principles are already in operation. We have first racial distinction. The three upper castes represent the invading Aryans, the fourth the races whom they found in India. In the modern system of caste, race is not a strong factor. Many who claim to be Brahmans and Kshatriyas have no Aryan blood, but stUl the .Aryan element is strongest in the highest castes and decreases as we descend the social scale and also decreases in the higher castes in proportion as we move from the north west to the east and south. But secondly in the three upper castes the dividing principle, as reported in the earhest accounts, is not race but occupation. We find in most Aryan countries a division into nobles and people, but in India these two classes become three, the priests having been able to assume a pro minence unknown elsewhere and to stamp on literature their claim to the highest rank. This claim was probably never admitted in practice so completely as the priests desired. It was certainly disputed in Buddhist times and I have myself heard a young Rajput say that the Brahmans falsified the Epics so as to give themselves the first place. It is not necessary for our purpose to describe the details of the modern caste system. Its effect on Indian rehgion has been considerable, for it created the social atmosphere in which the ' Some (Khandelwals, Dasa Srimalis and PaUiwals) include both Jains and Vaishnavas : the Agarwals are mostly Vaishnavas but some of them are Jains and some worship Siva and KaU. Jogendra Nath Bhattacharya, Hindu Castes and Sects, pp. 205 ff. B n. 12 178 HINDUISM [cH. various beUefs grew up and it has furnished the Brahmans with the means of estabhshing their authority. But many reUgious reformers preached that in rehgion caste does not exist — that there is neither Jew nor Gentile in the language of another creed — and though the apphcation of this theory is never complete, the imperfection is the result not of rehgious opposition but of social pressure. Hindu Hfe is permeated by the instinct that society must be divided into communities having some common interest and refusing to intermarry or eat with other com munities. The long Hst of modern castes hardly bears even a theoretical relation to the four classes of Vedic times'. Numerous subdivisions with exclusive rules as to intermarriage and eating have arisen among the Brahmans and the strength of this fissiparous instinct is seen among the Mohammedans who nominaUy have no caste but yet are divided into groups with much the same restrictions. This remarkable tendency to form exclusive corporations is perhaps correlated with the absence of poUtical Ufe in India. Such ideas as nationahty, citizenship, aUegiance to a certain prince, patriotic feelings for a certain territory are rarer and vaguer than elsewhere, and yet the Hindu is dependent on his feUows and does not Uke to stand alone. So finding Httle satis faction in the city or state he chngs the more tenaciously to smaUer corporations. These have no one character : they are not founded on any one logical principle but merely on the need felt by people who have something in common to associate together. Many are based on tribal divisions; some, such as the Marathas and Newars, may be said to be nationahties. In many the bond of union is occupation, in a few it is sectarian reUgion. We can stiU observe how members of a caste who migrate from their original residence tend to form an entirely new caste, and how intertribal marriages among the aborigines create new tribes. 1 The names used are not the same. The four Vedic castes are caUed Varna: the hundreds of modem castes are caUed Jdli. XXVI] RITUAL, CASTE, SECT, FAITH 179 Sect' must not be confounded with caste. Hindu sects are of many kinds; some, if not miUtant, are at least exceedingly self-confident. Others are so gentle in stating their views that they might be caUed schools rather than sects, were the word not too inteUectual. The notion that any creed or code can be quod semper, quod ubique, quod ab omnibus, is less prevalent than in Europe and even the Veda, though it is the eternal word, is admitted to exist in several recensions. Hinduism is possible as a creed only to those who select. In its Hteral sense it means simply aU the beUefs and rites recognized in India, too multifarious and inconsistent for the most hospitable and addled brain to hold. But the Hindus, who are as loth to aboHsh queer beUefs and practices as they are to take animal life, are also the most determined seekers after a satisfying form of rehgion. Brahmanic ritual and Buddhist monasticism de mand the dedication of a Hfe. Not every one can afford that, but the sect is open to aU. It attempts to sort out of the chaos of mythology and superstition something which aU can under stand and aU may find useful. It selects some aspect of Hinduism and makes the best of it. Sects usuaUy start by preaching theism and equahty in the sight of God, but in a few generations mythology and social distinctions creep in. Hence though the prevalence of sect is undoubtedly a feature of modem Hinduism it is also inteUigible that some observers should assert that most Hindus belong to the same general rehgion and that only the minority are definitely sectarian. The sectarian tendency is stronger in Vishnuism than in Sivaism. The latter has produced some definite sects, as, for instance, Lingayats, but is not Hke Vishnuism spUt up into a number of Churches each founded by a human teacher and provided by him with a special creed. Most Indian sects are in their origin theistic, that is to say, they take a particular deity and identify him with the Supreme Being. But the pantheistic tendency does not disappear. Popular religion naturaUy desires a personal deity. But it is significant that the personal deity frequently assumes pan theistic attributes and is declared to be both the world and the ^ Sampradaya seems to be the ordinary Sanskrit word for sectarian doctrine It means traditional teaching transmitted from one teacher to another. 12—2 180 HINDUISM [OH. human soul. The best known sects arose after Islam had entered India and some of them, such as the Sikhs, show a blending of Hindu and Moslem ideas. But if Mohammedan influence favoured the formation of corporations pledged to worship one particular deity, it acted less by introducing something new than by quickening a Hue of thought already existing. The Bhagavad-gita is as complete an exposition of sectarian pan theism as any utterances posterior to Mohammedanism. The characteristic doctrine of sectarian Hinduism is bhakti, faith or devotion. The older word Sraddhd, which is found in the Vedas, is less emotional for it means simply belief in the existence of a deity, whereas bhakti can often be rendered by love. It is passionate, self-obhvious devotion to a deity who in return (though many would say there is no bartering) bestows his grace {prasdda or anugraha). St Augustine in defining faith says: "Quid est credere in Deum? credendo amare, credendo diUgere, credendo in eum ire, et ejus membris incorporari'." This is an exceUent paraphrase of bhakti and the words have an oriental ring which is not quite that of the New Testament. Though the doctrine of bhakti marks the beginning of a new epoch in Hinduism it is not necessary to regard it as an importa tion or due to Christianity. About the time of the Christian era there was felt in many countries a craving for a gentler and more emotional worship and though the history of Bhaktism is obscure, Indian Hterature shows plainly how it may be a development of native ideas. Its first great textbook is the Bhagavad-gita, but it is also mentioned in the last verse of the Svetasvatara Upanishad and Panini appears to allude to bhakti felt for^ Vasudeva. The Katha Upanishad* contains the foUowing passage : "That Atman cannot be gained by the Veda, nor by under standing nor by much learning. He whom the Atman chooses, by him the Atman can be gained. The Atman chooses him as his own." Here we have not the idea of faith or love, but we have the negative statement that the Atman is not won by knowledge and the positive statement that this Atman chooses ^ I am discussing elsewhere the possible debt which Christianity and Hinduism may owe to one another. 2 Panini, iv. 3. 95-98. » Katha Up. 1. 1. 2, 23. XXVI] RITUAL, CASTE, SECT, FAITH 181 his own. In the Rig Veda' there is a poem put into the mouth of Vac or speech, containing such sentiments as " I give wealth to him who gives sacrifice.. ..I am that through which one eats, breathes, sees, and hears. ...Him that I love I make strong, to be a priest, a seer, a sage." This reads Hke an ancient preliminary study for the Bhagavad-gita. Like Krishna the deity claims to be in aU and, Hke him, to reward her votaries. It is true that the "Come unto me " is not distinctly expressed, but it is surely struggling for expression^. Again, in the Kaushitaki Upanishad (in. 1 and 2) Indra says to Pratardana, who had asked him for a boon, ' ' Know me only : that is, what I deem most beneficial to man, that he should know me. ...He who meditates on me as life and immortahty gains his full Ufe in this world and in heaven immortahty." Here the relation of the devotee to the deity is purely inteUectual not emotional, but the idea that inteUectual devotion directed to a particular deity wiU be rewarded is clearly present. In the Rig Veda this same Indra is caUed a deUverer and advocate; a friend, a brother and a father; even a father and mother in one. Here the worshipper does not talk of bhakti because he does not analyze his feehngs, but clearly these phrases are inspired by affectionate devotion. Nor is the spirit of bhakti absent from Buddhism. The severe doctrine of the older schools declares that the Buddha is simply a teacher and that every man must save himself. But since the teacher is the source of the knowledge which saves, it is natural to feel for him grateful and affectionate devotion. This sentiment permeates the two books of poems caUed Thera and Therigatha and sometimes finds clear expression*. In the commentary on the Dhammapada* the doctrine of salvation by devotion is affirmed in its extreme form, namely that a dying man who has faith in the Buddha wiU be reborn in heaven. But this commentary is not of early date and the doctrine quoted is probably an instance of the Hinayana borrowing the attractive features of the Mahayana. The sutras about Amitabha's paradise, which were composed about the time of the Christian era and owe something to Persian though not to Christian I R.V. X. 125. ' Compare too the hynms of the R.V. to Varuija as a rudimentary expression of Bhakti from the worshipper's point of view. > E.g. Theragatha, 818-841 and 1231-1245. ' i. 2. 182 HINDUISM [CH. influence, preach faith in Amitabha as the whole of rehgion. They who beheve in him and caU on his name wiU go to heaven. When bhakti was once accepted as a part of Indian religion, it was erected into a principle, analogous or superior to know ledge and was defined in Siitras' similar to those of the Saidihya and Vedanta. But its importance in phUosophy is smaU, whereas its power as an impulse in popular religion has been enormous. To estimate its moral and inteUectual value is difficult, for Hke so much in Hinduism it offers the sharpest contrasts. Its obvious manifestations may seem to be acts of devotion which cannot be commended ethically and beUef in puerile stories : yet we find that this offensive trash continuaUy turns into gems of reUgious thought unsurpassed in the annals of Buddhism and Christianity. The doctrine of bhakti is common to both Vishnuites and Sivaites. It is perhaps in general estimation associated with the former more than with the latter, but this is because the Bhagavad-gita and various forms of devotion to Krishna are weU known, whereas the Tamil Hterature of Dravidian Sivaism is ignored by many European scholars. One might be inclined to suppose that the emotional faith sprang up first in the worship of Vishnu, for the milder god seems a natural object for love, whereas Siva has to undergo a certain transformation before he can evoke such feelings. But there is no evidence that this is the historical development of the bhakti sentiment, and if the Bhagavad-gita is emphatic in enjoining the worship of Krishna only, the Svetasvatara and Maitrayaniya Upanishads favour Siva, and he is abundantly extolled in many parts of the Mahabharata. Here, as so often, exact chronology fails us in the early history of these sects, but it is clear that the practice of worshipping Siva and Vishnu, as being each by himseU aU- sufficient, cannot have begun much later than the Christian era and may have begun considerably earUer, even though people did not call themselves Saivas or Vaishnavas. ' They are called the Sandilya Sutras and appear to be not older than about the tweUth century a.d., but the tradition which connects them with the School of SandUya may be just, for the teaching of this sage (Chandog. Up. m. 14) lays stress on wUl and belief. Ramanuja (Sribhashya, ii. 2. 43) refers to Sandilya as the aUeged author of the Pancaratra. There are other Bhakti siitras caUed Naradiya and ascribed to Narada, pubUshed and translated in The Sacred Books of the Hindus, No. 23. They consist of 84 short aphorisms. Raj. Mitra in his notices of Sanskrit MSS. describes a great number of modem works deaUng with Bhakti. XXVI] RITUAL, CASTE, SECT, FAITH 183 Bhakti is often associated with the doctrine of the playful ness of God. This idea — so strange to Europe' — may have its roots partly in the odd non-moral attributes of some early deities. Thus the Rudra of the Satarudriya hymn is a queer character and a trickster. But it soon takes a philosophical tinge and is used to explain the creation and working of the universe which is regarded not as an example of capricious, ironical, inscrutable action, but rather as manifesting easy, joyous movement and the exuberant rhythm of a dance executed for its own sake. The European can hardly imagine a sensible person doing anything without an object: he thinks it almost profane to ascribe motiveless action to the Creator: he racks his brain to discover any purpose in creation which is moraUy worthy and moderately in accord with the facts of experience. But he can find none. The Hindu, on the contrary, argues that God being complete and perfect cannot be actuated by aims or motives, for aU such impulses imply a desire to obtain some thing, whereas a perfect and complete being is one which by its very definition needs neither change nor addition. Therefore, whatever activity is ascribed to the creator must not be thought of as calculating, purposeful endeavour, but as spontaneous, exultant movement, needing and admitting no explanation, and analogous to sport and play rather than to the proceedings of prudent people. This view of the divine activity is expounded by so serious a writer as Sankara in his commentary on the Vedanta Siitras, and it also finds mythological expression in numerous popular legends. The TamU Puranas describe the sixty-four miracles of Siva as his amusements : his laughter and joyous movements brighten all things, and the street minstrels sing "He sports in the world. He sports in the souP." He is supposed to dance in the Golden Hall of the temple at Chidam baram and something of the old legends of the Satarudriya 1 Yet it is found in Francis Thompson's poem caUed Any Saint So best God loves to jest With chUdren small, a freak Of heavenly hide and seek Fit For thy wayward wit. 2 Pope, The History of Manikka- Vafagar, p. 23. For the 64 sports of Siva see Siddhanta Dipika, vol. ix. 184 HINDUISM [CH. hangs about such popular titles as the Deceiver and the Maniac {Kalvar) and the stories of his going about disguised and visiting his worshippers in the form of a mendicant. The idea of sport and playfulness is also prominent in Vishnuism. It is a striking feature in the cultus of both the infant and the youthful Krishna, but I have not found it recorded in the severer worship of Rama. Another feature of Hindu sects is the extravagant respect paid to Gurus or teachers. The sanctity of the Guru is an old conviction in India. By common consent he is entitled to absolute obedience and offences against him are heinous crimes. But in sectarian Hterature there appears a new claim, namely, that the Guru in some way is or represents the god whose worship he teaches. If the deity is thought of primarily as a saviour, the Guru is said to deUver from suffering and heU: if he requires surrender and sacrifice, then person and possessions must be dedicated to the Guru. Membership of a sect can be attained only by initiation at the hands of a Guru who can teach a special mantra or formula of which each sect has its own. In some of the more modem sects the Guru need not be a Brahman, but if he cannot be venerated for his caste, the deficiency is compensated by the respect which he receives as a repository of oral teaching. The scriptural basis of many sects is dubious and even when it exists, many of the devout (especiaUy women) have not the inchnation or abihty to read and therefore take their rehgion from the Ups of the Guru, who thus becomes an oracle and source of truth. In Bengal, the famUy Guru is a regular institution in respectable castes. In many sects the founder or other prominent saint is described as an incarnation and receives veneration after death'- This veneration or deification of the Guru is found in most sects and assumes as extreme a form among the Saivas as among the Vaishnavas. The Saiva Siddhanta teaches that divine instruction can be received only from one who is both god and man, and that the true Guru is an incarnation of Siva. Thus the works of Manikka- Va9agar and Umapati speak of Siva coming to his devotees in the form of the Guru. In the sects that worship Krishna the Gurus are frequently caUed Gosain ' E.g. Ramanuja, Nammarvar, Basava. XXVI] RITUAL, CASTE, SECT, FAITH 185 (Goswami)'. Sometimes they are members of a particular family, as among the VaUabhacaryas. In other sects there is no hereditary principle and even a Sudra is eUgible as Guru. One other feature of Sectarian Hinduism must be mentioned. It may be described as Tantrism or, in one of its aspects, as the later Yoga and is a combination of practices and theories which have their roots in the old Hterature and began to form a connected doctrine at least as early as the eighth century a.d. Some of its principal ideas are as foUows : (i) Letters and syUables (and also their written forms and diagrams) have a potent influence both for the human organism and for the universe. This idea is found in the early Upanishads^ and is fuUy de veloped in the later Sectarian Upanishads. (U) The human organism is a miniature copy of the universe*. It contains many lines or channels (nadi) along which the nerve force moves and also nervous centres distributed from the hips to the head. (iii) In the lowest centre resides a force identical with the force which creates the universe*. When by processes which are partly physical it is roused and made to ascend to the highest centre, emancipation and bhss are obtained, (iv) There is a mysterious connection between the process of cosmic evolution and sound, especially the sacred sound Om. These ideas are developed most thoroughly in Saktist works, but are by no means pecuhar to them. They are found in the Pancaratra and the later Puranas and have influenced almost all modern sects, although those which are based on emotional devotion are naturaUy less inclined to favour physical and magical means of obtaining salvation. 1 Apparently meaning "possessor of cows," and originaUy a title of the youthful Krishna. It is also interpreted as meaning Lord of the Vedas or Lord of his own senses. * E.g. the beginning of the Chand. Up. about the syllable Om. See too the last section of the Aitareya Aran. The Yoga Upanishads analyse and explain Om and some Vishnuite Upanishads (Nrisimha- and Ramata-paniya) enlarge on the subject of letters and diagrams. ' The same idea pervades the old Uterature in a sUghtly different form. The parts of the sacrifice are constantly identified with parts of the universe or of the human body. * The cakras are mentioned in Act v of Malati and Madhava written early in the eighth century. The doctrine of the nadis occurs in the older Upanis'aads (e.g, Chand. and Maitrayana) in a rudimentary form. CHAPTER XXVII THE EVOLUTION OF HINDUISM. BHAGAVATAS AND PASUPATAS India is a Uterary country and naturaUy so great a change as the transformation of the old religion into theistic sects preaching salvation by devotion to a particular deity found expression in a long and copious Uterature. This Uterature supplements and supersedes the Vedic treatises but without impairing their theoretical authority, and, since it cannot compare with them in antiquity and has not the same historic interest, it has received little attention from Indianists until the present century. But in spite of its defects it is of the highest importance for an understanding of medieval and contemporary Hinduism. Much of it is avowedly based on the principle that in this degenerate age the Veda is difficult to understand', and that therefore God in His mercy has revealed other texts containing a clear compendium of doctrine. Thus the great Vishnuite doctor Ramanuja states authoritatively "The incontrovertible fact then is as follows : The Lord who is known from the Vedanta texts... recognising that the Vedas are difficult to fathom by aU beings other than himself... with a view to enable his devotees to grasp the true meaning of the Vedas, himseU composed the Pancaratra-Sastra^. ' ' This later sectarian Hterature faUs into several divisions. A. Certain episodes of the Mahabharata. The most celebrated of these is the Bhagavad-gita, which is probably anterior to the Christian era. Though it is incorporated in the Epic it is fre quently spoken of as an independent work. Later and less celebrated but greatly esteemed by Vishnuites is the latter part ^ An attempt was made to adapt the Veda to modem ideas by composing new Upanishads. The inspiration of such worlcs is not denied but they have not the same influence as the literature mentioned below. 2 Sri Bhashya, ii. 2. 43. So too the Vishnu Purana, i. 1 describes itseU as equal in sanctity to the Vedas. Sankara on Brah. Sutras, i. 3. 33 says that the Puranas are authoritative. OH. XXVII] BHAGAVATAS AND PASUPATAS 187 of book xn, commonly known as Narayaniya'. Both these episodes and others^ are closely analogous to metrical Upanishads. The Mahabharata even styles itself (i. 261) the Veda of Krishna (Karshna). The Ramayana does not contain religious episodes com parable to those mentioned but the story has more than once been re-written in a rehgious and phUosophic form. Of such versions the Adhyatmaramayana* and Yoga-va^ishtha-rama- yana are very popular. B. Though the Puranas* are not at all alike, most of them show clear affinity both as literature and as reUgious thought to the various strata of the Mahabharata, and to the Law Books, especiaUy the metrical code of Manu. These aU represent a form of orthodoxy which while admitting much that is not found in the Veda is stiU Brahmanic and traditibnaUst. The older Puranas {e.g. Matsya, Vayu, Markandeya, Vishnu), or at least the older parts of them, are the Hterary expression of that Hindu reaction which gained political power with the accession of the Gupta dynasty. They are less definitely sectarian than later works such as the Narada and Linga Puranas, yet aU are more or less sectarian. The most influential Purana is the Bhagavata, one of the great scriptures for all sects which worship Krishna. It is said to have been translated into every language of India and forty versions in BengaU alone are mentioned^. It was probably com- 1 See Grierson in Ind. Ant. 1908, p. 251 and p. 373. 2 E.g. the Sanatsujatiya andAnugita (both iu S.B.E. vm.). See Deussen, Vier philosophische Texte des Mahdbhdratam. ' Forming part of the Brahmanda Puraija. * See for a summary of them Wintemitz, Oesch. Ind. Lit. i. pp. 450^83. For the dates see Pargiter Dynasties of the Kali age. He holds that the historical portions of the older Puranas were compUed in Prakrit about 250 a.d. and re-edited in Sanskrit about 350. See also Vincent Smith, Early History, p. 21 and, against Pargiter, Keith in J.R.A.S. 1914, p. 1021. Alberuni (who wrote in 1030) mentions eighteen Puranas and gives two lists of them. Bana (c. 620 a.d.) mentions the recitation of the Vayu Purana. The commentary on the Svetasvatara Upan. ascribed to Saijkara quotes the Brahma P., Linga P. and Vishnu P. as authorities as weU as Puranic texts described as Vishnudharma and Sivadharmottara. But the authorship of this commentary is doubtful. The Puranic literature as we know it probably began with the Gupta dynasty or a century before it, but the word Purana in the sense of an ancient legend which ought to be learnt occurs as early as the Satapatha Brahmana (xi. 5. fr. 8) and even in A.V. xi. 7. 24. 5 See Dinesh Chandra Sen, Hist, Bengali Language and Lit. pp. 220-225. 188 HINDUISM [ch. posed in the eighth or ninth century'. A free translation of the tenth book into Hindi, caUed the Prem Sagar or Ocean of Love, is greatly revered in northern India 2. Other sectarian Puranas are frequently read at temple services. Besides the eighteen great Puranas there are many others, and in south India at any rate they were sometimes composed in the vernacular, as for instance the Periya Purana (c. 1100 a.d.). These vemacular Puranas seem to be collections of strangely fantastic fairy tales. C. The word Tantra originally meant a manual giving the essentials of a subject but later usage tends to restrict it to works, whether Hindu or Buddhist, inculcating the worship of Siva's spouse. But there are exceptions to this restriction: the Panca-tantra is a coUection of stories and the Lakshmi-tantra is a Vishnuite work*. The fact is that a whole class of Sanskrit rehgious Uterature is described by the titles Tantra, Agama and Samhita*, which taken in a wide sense are practicaUy synonymous, though usage is inclined to apply the first speciaUy to Saktist works, the second to Sivaite and the third to Vishnuite. The common character of aU these productions is that they do not attempt to combine Vedic rites and ideas with sectarian worship, but boldly state that, since the prescriptions of the Veda are too hard for this age, some generous deity has revealed an easier teaching. This teaching naturaUy varies in detail, but it usuaUy comprises devotion to some special form of the godhead and also a special ceremonial, which commences with initiation and includes the use of mystic formulae, letters and diagrams. * Pargiter, l,c, pp. xvu, xxviU. It does not belong to the latest class of Puranas for it seems to contemplate the performance of Smarta rites not temple ceremonial, but it is not quoted by Ramanuja (twelfth century) though he cites the Vishnu Purana. Probably he disapproved of it. * It was made as late as 1803 by Lallii Ji Lai, but is a rendering into Hindi of a version in the Braj dialect, probably made in the sixteenth centurj'. ' Another Vishnuite work is cited indifferently as Padma-tantra or Padma-sam- hita, and the Bhagavata Purana (i. 3. 8) speaks of the Sattvatam Tantram, which is apparently the Sattvata-sanahita. The work edited by Schrader is described as the Ahirbudhnya Samhitd of the Pdncardtra Agama, * See for sorae notices of these works A. Avalon's various pubUcations about Tantra. Siimvaaii, lyengSiT, Outlines of Indian Philosophy, 118-191. Govindacarya Svami on the Vaishnava Samhitas, J,R,A,S, 1911, pp. 935 ff. Schomerus, (laiva- Siddhdnta, pp. 7 ff. and Schrader's Introduction to the Pdncardtra. Whereas these works claim to be independent of the Veda, the Sectarian Upanishads (see vol, l. p. 76) are an attempt to connect post-Vedio sects with the Veda. xxvn] BHAGAVATAS AND PASUPATAS 189 Tantras, Agamas and Samhitas all treat of their subject-matter in four divisions' the first of which relates to the great problems of philosophy, the second to the discipline necessary for uniting the seU and God ; the third and fourth to ceremonial. These works have another feature in common, namely that they are Httle known except to those Hindus who use them for rehgious purposes and are probably not very anxious to see them published. Though they are numerous, few of them have been printed and those few have not been much studied by European scholars . I shaU say something more about them below in treating of the various sects. Some are of respectable antiquity but it is also clear that modem texts pass under ancient names. The Pancaratram and Pasupatam which are Vishnuite and Sivaite Samhitas are mentioned in the Mahabharata, and some extant Vishnuite Samhitas were perhaps composed in the fourth century a.d.^ Ramanuja as quoted above states that the Pancaratra- sastr a (apparently the same as the Pancaratra-tantra which he also mentions) was composed by Vasudeva himself and also cites as scripture the Sattvata, Paushkara and Parama Samhitas. In the same context he speaks of the Mahabharata as Bharata-Samhita and the whole passage is interesting as being a statement by a high authority of the reasons for accepting a non-Vedic work like the Pancaratra as revealed scripture. As already indicated European usage makes the words Tantra, Tantrism and tantric refer to the worship of goddesses. It would be better to describe this Hterature and worship as saktism and to use Tantrism for a tendency in doctrine and ceremonial which otherwise has no special name. I have been informed by TamU Pandits that at the present day the ritual in some temples is smarta or according to Smriti, but in the majority according to the Agamas or tantric. The former which is foUowed by many weU-known shrines (for instance in Benares and in the great temples of south India) conforms to the pre- 1 Jilana, Yoga, Carya, Kriya. The same names are used of Buddhist Tantras, except that Anuttara replaces Jnana. 2 See Schrader, Introd. to the Pdncardtra, p. 98. In the Raghuvamsa, x. 27. Agamas are not only mentioned but said to be extremely numerous. But in suoh passages it is hard to say whether Agama means the books now so-called or merely tradition. Alberuni seems not to have known of this Uterature and a Tantra for him is merely a minor treatise on astronomy. He evidently regards the Vedas, Puranas, philosophical Dar^anas and Epics as constituting the religious literature of India. 190 HINDUISM [CH. cepts of the Puranas, especiaUy on festival days. The officiants require no special initiation and bumt offerings are presented. But the Agamic ritual can be performed only by priests who have received initiation, burnt offerings rarely form part of the ceremony and vernacular hymns are freely used'. Such hymns however as weU as processions and other forms of worship which appeal directly to the religious emotions are certainly not tantric. Tantrism is a species of reUgious magic, differing from the Vedic sacrifices in method rather than principle^. For all that, it sets aside the old rites and announces itself as the new dispensation for this age. Among its principal features are the foUowing. The Tantras are a scripture for aU, and lay little stress on caste: the texts and the ritual which they teach can be understood only after initiation and with the aid of a teacher : the ritual consists largely in the correct use of speUs, magical or sacramental syUables and letters, diagrams and gestures: its object is less to beseech than to compel the god to come to the worshipper: another object is to unite the worshipper to the god and in fact transform him into the god : man is a microcosm corresponding to the macrocosm or universe : the spheres and currents of the universe are copied in miniature in the human body and the same powers rule the same parts in the greater and the lesser scheme. Such ideas are widely disseminated in almost aU modern sects*, though without ' Rajagopala Chariar ( Vaishnamte Reformers, p. 4) says that in Vishnu temples two rituals are used called Pancaratra and Vaikhanasa. The latter is apparently consistent with Smarta usage whereas the Pancaratra is not. From Gopinatha Rao's Elements of Hindu Iconography, pp. 56, 77, 78 it appears that there is a Vaikhanasagama parallel to the Paucaratragama. It is frequently quoted by this author, though as yet unpublished. It seems to be the ritual of those Bhagavatas who worship both Siva and Vishnu. It is said to exist in two recensions, prose and metrical, of which the former is perhaps the oldest of the Vaishnava Agamas. The Vaikhanasa ritual was once followed at Srirangam but Ramanuja substituted the Pancaratra for it. ^ Avalon, Principles of Tantra, p. xxvU describes it as "that development of the VaidUia Karmakanda which under the name of the Tantra Shastra is the scripture of the Kali age." This seems to me a correct statement of the tantric heory. ' Thus the Gautamiya Tantra which is held in high estimation by Vishnuite householders in Bengal, though not by ascetics, is a complete appUcation of Sakta worship to the cult of Kj-ishna. The Varahi Tantra is also Vishmute. See Raj. Mitra, Sanskrit MSS. of Bikaner, p. 583 and Notices of Sk. MSS. in. (1876), p. 99, and 1. colxxxvu. See too the usages of the Nambuthiri Brahmans as described in xxvn] BHAGAVATAS AND PASUPATAS 191 forming their essential doctrine, but I must repeat that to say aU sects are tantric does not mean that they are aU Saktist. But Saktist sects are fundamentaUy and thoroughly tantric in their theory and practice. D. Besides the Sanskrit books mentioned above numerous vernacular works, especiaUy coUections of hymns, are accepted as authoritative by various sects, and almost every language has scriptures of its own. In the south two Tamil hymnals, the Devaram of the Sivaites and Naiayira Prabandham of the Vishnuites, are recited in temples and are boldly stated to be revelations eqmvalent to the Veda. In northern India may be mentioned the Hindi Ramayana of Tulsi Das, which is almost universaUy venerated, the Bhakta-maia of Nabha Das', the Sur-sagar of Surdas and the Prem Sagar. In Assam the Nam Gosha of Madhab Deb is honoured with the same homage as a sacred image. The awkwardness of admitting direct inspuation in late times is avoided by the theory of spiritual descent, that is to say of doctrinal transmission from teacher to teacher, the divine revelation having been made to the original teacher at a discreetly remote epoch. In considering the evolution of modem Hinduism out of the old Vedic rehgion, three of the many factors responsible for this huge and compHcated result deserve special attention. The first is the unusual intensity and prevalence of the reUgious tempera ment. This has a double effect, both conservative and alterative : ancient customs receive an unreasonable respect: they are not abohshed for their immoraUty or absurdity; but since real interest impHes some measure of constructive power, there is a constant growth of new ideas and reinterpretations resulting in inconsistent combinations. The second is the absence of hierarchy and discipline. The guiding principle of the Brahmans has always been not so much that they have a particular creed to enforce, as that whatever is the creed of India they must be its ministers. NaturaUy every priest is the champion of his own god or rite, and such zeal may lead to occasional conflicts. But Cochin Tribes and Castes, n. pp. 229-233. In many ways the Nambuthiris preserve the ancient Vedic practices. * See Grierson's articles Gleanings frora the Bhaktamala in J.R.A.S. 1909-1910. 192 HINDUISM [ch. though the antithesis between the rituahsm of the older Brahmanism and the faith or philosophy of Sivaism and Vishnu ism may remind us of the differences between the Catholic Church and Protestant reformers, yet HistoricaUy there is no resemblance in the development of the antithesis. To some extent Hinduism showed a united front against Buddhism, but the older Brahmanism had no organization which enalsled it to stand as a separate Church in opposition to movements which it disHked. The third factor is the deeply rooted idea, which reappears at frequent intervals from the time of the Upanishads until to-day, that rules and rites and even creeds are somehow part of the lower and temporal order of things which the soul should transcend and leave behind. This idea tinges the whole of Indian philosophy and continually crops up in practice. The founder of a strange sect who declares that nothing is necessary but faith in a particular deity and that all ceremonies and caste observances are superfluous is not in the popular esteem a sub- verter of Hinduism. The history of both Sivaism and Vishnuism iUustrates these features. Siva begins as a wUd deity of non-moral attributes. As the rehgious sense develops he is not rejected Hke the less reputable deities of the Jews and Arabs but remains and coUects round himself other strange wild ideas which in time are made philosophical but not ethical. The rites of the new rehgion are, if not antagonistic, at least alternative to the ancient sacrifices, yet far from being forbidden they are performed by Brahmans and modern Indian writers describe Siva as pecuHarly the Brahman's god. FinaUy the Sivaite schools of the Tamil country reject in successive stages the grosser and more formal elements until there remains nothing but an ecstatic and mystical mono theism. Similarly among the Vishnuites Krishna is the centre of legends which have even less of conventional morality. Yet out of them arises a doctrine that the love of God is the one thing needful so similar to Christian teaching that many have supposed it must be borrowed. The first clear accounts of the worship of Siva and Vishnu are contained in the epics and indicate the existence of sectarian reUgion, that is to say of exclusive devotion to one or other deity. But there is also a tendency to find a place for both, a tendency which culminates in the composite deity Sankara- xxvn] BHAGAVATAS AND PASUPATAS 193 Narayana already mentioned. Many of the Puranas' reflect this view and praise the two deities impartiaUy. The Mahabharata not unfrequently does the same but the general impression left by this poem is that the various parts of which it consists have been composed or revised in a sectarian spirit. The body of the work is a narrative of exploits in which the hero Krishna plays a great part but revised so as to make him appear often as a deity and sometimes as the Supreme Spirit. But much of the didactic matter which has been added, particularly books xii and xni, breathes an equally distinct Sivaite spirit and in the parts where Krishna is treated as a mere hero, the principal god appears to be not Vishnu but Siva. The Mahabharata and Puranas contain legends which, though obscure, refer to conflicts of the worshippers of Siva with those who offered Vedic sacrifices as well as with the votaries of Vishnu, and to a subsequent reconcihation and blending of the various cults. Among these is the well-known story of Daksha's sacrifice to which Siva was not invited. Enraged at the omission he violently breaks up the sacrifice either in person or through a being whom he creates for the purpose, assaults the officiants and the gods who are present, and is pacified by receiving a share. Similarly we hear^ that he once seized a victim at a sacrifice and that the gods in fear aUotted to him the choicest portion of the offerings. These stories indicate that at one time Brahmans did not countenance his worship and he is even represented as saying to his wife that according to rule (dharmatah) he has no share in the sacrifice*. Possibly human victims were immolated in his honour, as they were in KaU's until recently, for in the Mahabharata* it is related how Krishna expostulated with Jarasandha who pro- ^ E.g, Markandeya, Vamana and Varaha. Also the Skanda Upanishad. ' Mahabh. Vanaparvan, 11001 ff. The Bhagavata Purana, Book iv. sec. 2-7 emphasizes more clearly the objections of the Rishis to Siva as an enemy of Vedic sacriflces and a patron of unhallowed rites. ' Mahabh. xu. sec. 283. In the same way the worship of Dionysus was once a novelty in Greece and not countenanced by the more conservative and respectable party. See Eur. Bacchae, 45. The Varaha-Purana relates that the Sivaite scriptures were revealed for the benefit of certain Brahmans whose sins had rendered thera incapable of performing Vedic rites. There is probably some trath in this legend iu 60 far as it means that Brahmans who were excommunicated for some fault were disposed to become the ministers of non-Vedic cults. * Mahabh. n. sees. 16, 22 ff. E. n. 13 194 HINDUISM [oh. posed to offer to Siva a sacrifice of captive kin^s. In the Vishnu- Purana, Krishna fights with Siva and bums Benares. But by the time that the Mahabharata was put together these quarrels were not in an acute stage. In several passages' Krishna is made to worship Siva as the Supreme Spirit and in others^ vice versa Siva celebrates the glory of Krishna. Vishnuites do not disbeheve in Siva but they regard him as a god of this world, whereas their own deity is cosmic and universal. Many Vishnuite works* are said to be revealed by Siva who acts as an intermediary between us and higher spheres. In the following sections I shaU endeavour to relate the beginnings of sectarianism. The sects which are now most important are relatively modern and arose in the tweUth century or later, but the sectarian spirit can be traced back several centuries before our era. By sectarians I mean wor shippers of Siva or Vishnu who were neither in complete sympathy with the ancient Brahmanism nor yet excommuni cated by it and who had new texts and rites to replace or at least supplement the Vedas and the Vedic sacrifices. It is probable that the different types of early Indian reUgion had originaUy different geographical spheres. Brahmanism flourished in what we caU the United Provinces : Buddhism arose in the regions to the east of this district and both Vishnuism and Sivaism are first heard of in the west. The earhest sect of which we have any record is that of the Bhagavatas, who were or became Vishnuite. At a date which it is impossible to fix but considerably before the epoch of panini, a tribe named the Yadavas occupied the country between Muttra and the shores of Gujarat. Septs of this tribe were caUed Vrishni and Sattvata. The latter name has passed into theology. Krishna belonged to this sept and it is probable that this name Vasudeva was not originaUy a patronymic but the name of a deity worshipped by it. The hero Krishna was identified with this god and subsequently when the Brahmans wished to bring this powerful sect within the pale of orthodoxy 1 Drona-p., 2862 ff. Anusasana-p., 590 ff. * E.g. Anueasana-p., 6806 ff. « E.g. the Ahirbudhnya Sanihita and Adhyatma Kamayaija. xxvn] BHAGAVATAS AND PASUPATAS 195 both were identified with Vishnu. In the Mahabharata' the rule ' or ritual (vidhi) of the Sattvatas is treated as equivalent to that of the Bhagavatas and a work called the Sattvata Sarnhita is stiU extant. Bhagavata appears to be the most general name of the sect or sects and means simply of the Lord (Bhagavat), that is worshippers of the one Lord^. Their religion is also called Ekantika dharma, or the rehgion with one object, that is ^monotheism*. A considerable Uterature grew up in this school and the principal treatise is often spoken of as Pancaratra because it was revealed by Narayana during five nights*. The name how ever appears to be strictly speaking appUcable to a system or body of doctrine and the usual term for the books in which this 'system is expounded is Samhita. All previous discussions and speculations about these works, of which little was known until recently, are superseded by Schrader's publication of the Ahirbudhnya Samhita, which appears to be representative of its class' . The names of over two hundred are cited and of these more than thirty are known to be extant in ms.^ The majority were composed in north-western India but the Pancaratra ^doctrine spread to the Dravidian countries and new Samhitas were produced there, the chief of which, the Isvara Samhita, can hardly be later than 800 a.d.' Of the older works Schrader ' ' Santipar. cccxxxvu, 12711 ff. In the Bhagavad-glta Krishna says that he is Vasudeva of the Vrishnis, xi. 37. I 2 Cf. the title Bhagavata Purana. . " Ekayana is mentioned several times in the Chandogya Up. (vii. 1, 2 and "afterwards) as a branch of reUgious or literary knowledge and in connection with Narada. But it is not represented as the highest or satisfying knowledge. * Even in the Satapatha Br. Narayapa is mentioned in connection with a sacrifice lasting five days, xm. 6. 1. ' The Samhitas hitherto best known to orientalists appear to be late and spurious. The Brihadbrahma Samhita published by the Anandasrama Press mentions Ramanu j a. The work printed in the Bibliotheca Indica as Narada pancaratra (although tits proper title apparently is Jfianamritasara) has been analyzed by Roussel in 'Milanges Harlez and is apparently a late Uturgical oompUation of little originality. Schrader's work was published by the Adyar Library in Madra3,1916. Apparently the two forms Pancaratra and Pancaratra are both found, but that with the long vowel is jthe more usual. Govindacarya's article in. J.R.A.S. 1911, p. 951 may also be consulted. ' The oldest are apparently the Paushkara, Varaha, Brahma, Sattvata, Jaya and Ahirbudhnya Saijihitas, all quoted as authoritative by either Ramanuja or Vedanta De^ika. t ' It is quoted as equal to the Vedas by Yamunacarya, so it must then have been in existence some centuries. 13—2 196 HINDUISM [oh. thinks that the Ahirbudhnya was written in Kashmir' between 300 and 800 a.d. and perhaps as early as the fourth century. It mentions the Sattvata and Jayakhya, which must therefore be older. The most remarkable feature of this Hterature is its elaborate doctrine of evolution and emanation from the Deity, the world process being conceived in the usual Hindu fashion as an alter nation of production and destruction. A distinction is drawn between pure and gross creation. What we commonly caU the Universe is bounded by the sheU of the cosmic egg and there are innumerable such eggs, each with its own heavens and its own tutelary deities such as Brahma and Siva who are sharply distinguished from Vishnu. But beyond this multitude of worlds are more mysterious and spiritual spheres, the highest heaven or Vaikuntha wherein dweUs God in his highest form (Para) with his Saktis^, certain archangels and Hberated souls. Evolution commences when at the end of the cosmic night the Sakti of Vishnu^ is differentiated from her Lord and assumes the two forms of Force and Matter*. He as differentiated from her is Vasudeva a personal deity with six attributes' and is the first emanation, or Vyuha, of the ineffable godhead. From him , proceeds Sankarshana, from Sankarshana Pradyumna, and from Pradyumna Aniruddha. These three Vyuhas take part in creation but also correspond to or preside over certain aspects of human personaUty, namely Sankarshana to the soul that animates aU beings, Pradyumna to inteUigence and Aniruddha to individuaUty. Strange to say these seem to be the names of ^ distinguished personages in the Sattvata or Vrishni clan®. Mere deification occurs in many countries but the transformation of heroes into metaphysical or psychological terms could hardly have happened outside India. Next to the Vyiihas come twelve 1 The story of Svetadvipa or White Island in the Santiparvan of the Mahabharata states definitely that Narada received the Pancaratra there. 4 ^ There is much diversity of statement as to whether there are one or many Saktis. ' Vishnu is the name of God in aU his aspects, but especiaUy God as the absolute. Vasudeva is used both of God as the absolute and also as the first emanation (Vyuha). i * KriyMakti and Bhuti^akti. ° Jnana, ai^varya, ^akti, bala, virya, tejas. These are caUed gunas but are not to be confounded with the three ordinary gunas. ' The words seem to have been originaUy proper names. See the articles in the J Petersburg Lexicon. , xxvn] BHAGAVATAS AND PASUPATAS 197 sub-Vyuhas, among whom is Narayana', and thirty-nine Avataras. All these beings are outside the cosmic eggs and our gross creation. As a prelude to this last there takes place the evolution of the aggregates or sources from which individual souls and matter are drawn, of space and of time, and finaUy of the elements, the process as described seeming to foUow an older form of the Sankhya philosophy than that known to us. The task of human souls is to attain liberation, but though the language of the Samhitas is not entirely consistent, the older view is that they become like to God, not that they are absorbed in him^. Thus it is not incorrect to say that the Bhagavata rehgion is monotheistic and recognizes a creator of souls. Indeed Sankara^ condemns it on the very ground that it makes indi vidual souls originate from Vasudeva, in which case since they have an origin they must also have an end. But Ramanuja in replying to this criticism seems to depart from the older view, for he says that the Supreme Being voluntarily abides in four forms which include the soul, mind and the principle of indi viduaUty. This, if not Pantheism, is very different from European monotheism*. The history of these Bhagavatas, Pancaratras or worshippers of Vishnu must have begun several centuries before our era, for there are aUusions to them in Panini and the Niddesa'. The names of Vasudeva and Sankarshana occur in old inscriptions® and the Greek Hehodoros caUs himseU a Bhagavata on the column found at Besnagar and supposed to date from the first , part of the second century b.c. The pancaratra was not Brahmanic in origin'' and the form ' Narayana like Vishiju is used to designate more than one aspect of God. Sometimes it denotes the Absolute. 2 The above brief sketch is based on Schrader's Int. to the Pdncardtra where the reader can find full detaUs. ' Coraraent on Vedanta sutras, ii. 2. 42. * And, as Sohrader observes, the evolutionary systera of the Pancaratra is practically concerned with only one force, the Sakti, which under the name Bhuti is manifested as the Universe and as Kriya vitalizes and governs it (p. 31). 5 On Sutta-nipata, 790, 792. The doctrine of the Vyiihas is expounded in the Mahabharata Santip. ccoXL. 36 ff., 70 ff. ; COOXLI. 26 ff. " Liider's List of Brahmi inscriptions. No. 6, supposed not to be later than 200 B.C. and No. 1112 supposed to be of the first century B.C. Sankarshana is abo mentioned in the KautiUya Artha^astra, xm. 3. ' Some Samhitas emphasize the distinction between the foUowers of the Veda and the enlightened ones who worship the Lord. See Schrader, Pdncardtra, p. 97. 198 HINDUISM [ch. of the sankhya phUosophy from which it borrowed was also un-Brahmanic. It seems to have grown up in north-western India in the centuries when Iranian influence was strong and may owe to Zoroastrianism the doctrine of the Vyuhas which finds a paraUel in the relation of Ahura Mazda to Spenta Mainyu, his Holy Spirit, and in the Fravashis. It is also remarkable that God is credited with six attributes comparable with the six Am esha Spentas. In other ways the Pancaratra seems to have some connection with late Buddhism. Though it lays Uttle stress on the worship of goddesses, yet aU the Vyiihas and Avataras are provided with Saktis, like the Buddhas and Bodhisattvas of tantric Buddhism, and in the period of quiescence which foUows on the dissolution of the Universe Vishnu is described under the name of Siinya or the void. It attaches great importance to the Cakra, the wheel or discus which denotes Vishnu's wiU to be', to evolve and maintain the universe, and it may have contri buted some ideas to the very late form of Buddhism caUed Kalacakra. This very word is used in the Ahirbudhnya Samhita as the name of one of the many wheels engaged in the work of evolution. Though the Pancaratra is connected with Krishna in its origin, it gives no prominence to devotion to him under that name as do modern sects and it knows nothing of the pastoral Krishna^. It recommends the worship of the four Vyiihas* presiding over the four quarters in much the same way that late Buddhism adores the four Jinas depicted in somewhat simUar forms. SimUarly the Sivaites say that Siva has five faces, namely Isana or Sadasiva (the highest, undifferentiated form of the deity) at the top and below Vamadeva, Aghora, Tatpurusha and Sadyojata, presiding respectively over the north, south, east and west. It is thus clear that in the early centuries of our era (or perhaps even before it) there was a tendency in Vishnuism, Sivaism and Mahayanist Buddhism alike to represent the ineffable godhead as manUested in four aspects somewhat more inteUigible to human minds and pro ducing in their turn many inferior marufestations. Possibly the ^ Syam iti Sankalpa, Ahirbudh. Sam. n. 7. In some late Upanishads (e.g. Naradaparivrajaka and Bjihatsannyasa) Cakri is used as a synonym for a PSicaratra. * The same is true of Ramanuja, who never quotes the Bhagavata Purana. ' See the quotations from the Sattvata Samhita in Sohrader, pp. 150-154. .'Vs in the Pancaratra there is the Para above the four V_vuhas, so some late forms of Buddhism regard Vairocana as the source of four Jinas. xxvn] BHAGAVATAS AND PASUPATAS 199 theory originated among the Vishnuites', but as often happened in India it was adopted by their opponents. None of these theories are of much importance as Hving behefs at the present day but their influence can be seen in iconography. As a sect the Pancaratras seem to have been a subdivision of the Bhagavatas and probably at the present day many Vishnuites would accept the second name but not the first. The Pancaratra is studied at only a few places in southern India but its doctrines permeate the popular work caUed Bhaktamaia and in view of the express approbation of Ramanuja and other authorities it can hardly be repudiated by the Sri- Vaishnavas. Bhagavata is sometimes used in the south as a name for Smartas who practise Vedic rites and worship both Siva and Vishnu^. In these early times there were strenuous theological struggles now forgotten, though they have left their traces in the legends which teU how the title of Krishna and others to divine honours was chaUenged. Amalgamation was the usual method of con- ciUation. Several gods grew sufficiently important to become in the eyes of their worshippers the supreme spirit and at least four were united in the deity of the Bhagavatas, namely, vasudeva, Krishna, Vishnu and Narayana. Of the first three I have spoken already. Narayana never became like Vishnu and Krishna a great mythological figure, but in the late Vedic period he is a personification of the primaeval waters from which aU things sprang or of the spirit which moved in them^. From this he easUy became the supreme spirit who animates aU the universe and the name was probably acceptable to those who desired a purer and simpler worship because it was con nected with comparatively few legends. But there is some confusion in its use, for it is appUed not only to the supreme being but to a double incarnation of him caUed Nara-Narayana, and images of the pair may stiU be seen in Vishnuite temples. ^ The Manicheans also had groups of five deities (see Chavannes and PeUiot in J.A. 1913, I. pp. 333-338) but they are just as likely to have borrowed from Bud dhism as vice versd. ^ See Bhattacharya, Hindu Castes and Sects, p. 565. ^ Manu, I. 10-11, identifies him with Brahma and says. "The waters are caUed Narah because they are produced from Nara, and he is caUed Narayana because they were his place of movement (ayana)." The same statement occurs in the Narayaniya. 200 HINDUISM [ch. They are said to have revealed the true doctrine to Narada and are invoked at the beginning of each book of the Mahabharata'. One of the main theses of the Narayaniya^ is the identity of Narayana and Vasudeva, the former being a Brahmanic, the latter a non-Brahmanic name for the Deity. The celebrated Bhagavad-gita* which is still held in such respect that, like the New Testament or Koran, it is used in law courts for the administration of oaths, is an early scripture of the Bhagavata sect. In it the doctrines of Krishna's divinity, the power of faith and the efficacy of grace are fuUy estabUshed. It is declared to be too hard for flesh and blood to find by meditation their way to the eternal imperceptible spirit, whereas Krishna comes straightway to those who make him their sole desire. "Set thy heart on me, become my devotee, sacrifice to me and worship thou me. Then shalt thou come to me. Truly I declare to thee thou art dear to me. Leave aU (other) rehgious duties and come to me as thy sole refuge. I wiU deUver thee from thy sins. Sorrow not." But the evolution of Sankarshana, etc., is not mentioned. The poem has perhaps been re-edited ' They are said to have been the sons of Dharma (religion or righteousness) and Ahimsa (not-injuring). This is obvious aUegory indicating that the Bhagavata religion rejected animal sacrifices. At the beginning of the Narayaniya (Santip. ccoxxxv.) it is said that Narayana the soul of the universe took birth in a quadruple form as the offspring of Dharma, viz. Nara, Narayana, Hari and Krishna. Nara and Narayana are often identified with Arjuna and Vasudeva. E.g. Udyogap. xxix. 19. ^ Mahabhar. xn. ^ It is an episode in Mahabhar. vi. and in its present form was doubtless elaborated apart from the rest. But we may surmise that the incident of Krishna's removing Arjuna's scruples by a discourse appeared iu the early versions of the story and also that the discourse was longer and profounder than would seem appropriate to the European reader of a tale of battles. But as the Vedanta phUosophy and the doctrine of Krishna's godhead developed, the discourse may have been amplified and made to include later theological views. Garbe in his German translation attempts to distinguish the different strata and his explanation of the inconsistencies as due to successive redactions and additions may contain some truth. But these inconsistencies in theology are comraon to aU sectarian writings and I think the main cause for them must be sought not so much in the alteration and combination of documents, as in a mixed and eclectic mode of thought. Even in European books of the first rank inconsistencies are not unknown and they need not cause surprise in works which were not written down but com mitted to memory. A poet composing a long reUgious poem in this way and feeling, as many Hindus feel, both that God is everything and also that he is a very present personal help, may very weU express himself differently in different parts. On the other hand the editors of suoh poems are undoubtedly tempted to insert in them later popular doctrines. xxvn] BHAGAVATAS AND pASUPATAS 201 and interpolated several times but the strata can hardly be distinguished, for the whole work, if not exactly paradoxical, is eclectic and continuaUy argues that what is apparently highest is not best for a particular person. The Hindus generaUy regard the contemplative Hfe as the highest, but the Bhagavad-gita is insistent in enjoining unselfish action : it admits that the supreme reaUty cannot be grasped by the mind or expressed in speech, but it recommends the worship of a personal deity. Even the older parts of the poem appear to be considerably later than Buddhism. But its mythology, if not Vedic, is also hardly Puranic and it knows nothing of the legends about the pastoral Krishna. It presupposes the Sankhya and Yoga, though in what stage of development it is hard to say, and in many respects its style resembles the later Upanishads. I should suppose that it assumed its present form about the time of the Christian era, rather before than after, and I do not think it owes anything to direct Christian influence. In its original form it may have been considerably older. The Bhagavad-gita identifies Krishna with Vasudeva and with Vishnu but does not mention Narayana and from its general style I should imagine the Narayaniya to be a later poem. If so, the evolution of Bhagavata theology wiU be that Krishna, a great hero in a tribe Ijong outside the sphere of Brahmanism, is fijst identified with Vasudeva, the god of that tribe, and then both of them with Vishnu. At this stage the Bhagavad-gita was composed. A later current of speculation added Narayana to the already complex figure, and a stiU later one, not accepted by aU sects, brought the pastoral and amorous legends of Krishna. Thus the history of the Bhagavatas illustrates the Indian disposition to combine gods and to see in each of them only an aspect of the one. But until a later period the types of divinity known as Vishnu and Siva resisted combination. The worshippers of Siva have in all periods shown less inchnation than the Vishnuites to form distinct and separate bodies and the earhest Sivaite sect of which we know anything, the Pasupatas', arose sUghtly later than the Bhagavatas. 1 The name appears not to be in common use now, but the Pasupata school ia reviewed in the Sarva-dar^ana-sangraha (c. 1330). 202 HINDUISM [ch. PatanjaU the grammarian (c. 150 B.C.) mentions devotees of Siva' and also images of Siva and Skanda. There is thus no reason to doubt that worshippers of Siva were recognized as a sect from at least 200 B.C. onwards. Further it seems probable that the founder or an early teacher of the sect was an ascetic caUed Lakulin or LakuHsa, the club-bearer. The Vayu Purana^ makes Siva say that he wiU enter an unowned corpse and become incarnate in this form at Kayarohana, which has been identified with Karvan in Baroda. Now the Vayu is beUeved to be the oldest of the Puranas, and it is probable that this Lakuhn whom it mentions Uved before rather than after our era and was especiaUy connected with the Pasupata sect. This word is derived from Pasupati, the Lord of cattle, an old title of Rudra afterwards explained to mean the Lord of human soiils. Tn the Santiparvan' five systems of knowledge are mentioned. Sankhya., Yoga, the Vedas, Pasupatam and Pan caratram, promulgated respectively by KapUa, Hiranyagarbha, Apantaratamas, Siva the Lord of spirits and son of Brahma, and "The Lord (Bhagavan) himseff." The author of these verses, who evidently supported the Pancaratra, considered that these five names represented the chief existing or permissible varieties of rehgious thought. The omission of the Vedanta is remarkable but perhaps it is included under Veda. Hence we may conclude that when this passage was written (that is probably before 400 a.d. and perhaps about the beginning of our era) there were two popular rehgions ranking in pubhc ^ Sivabhagavata, see his coinment on Panini, v. 3. 99 and v. 2. 76. The name is remarkable and suggests that the Sivaites may have imitated the Bhagavatas. ^ I. xxUi. 209. The Bihliotheca Ind. edition reads NakuU. Aufrecht (Bodl. MSS,) has LakuU. The same story is found in Linga P. chap. xxiv. LakuU is said to have had four pupUs who founded four branches. Lakulin does not play an important part in modem Sivaism but is mentioned in inscriptions from the tenth tUl the thirteenth centuries. The Sarva-dar.4ana-sangraha describes the NakuU^a-Pa^upata system and quotes NakuUsa who is clearly the same as LakuUn. The figures on Kushan coins representing Siva as holding a club may be meant for Lakulin but also may be influenced by Greek flgures of Herakles. See for Lakulin Fleet in J.R.A.S. 1907, pp. 419 ff. and Bhandarkar Vaishnavism and Saivism, pp. 115 ff. The coins of Wema Kadphises bear the title Mahi^vara, apparently meaning worshipper of the Great Lord. Temples in south India seem to have been named after Kayarohana in the seventh century a.d. See Gopinatha Rao, Hindu Icono graphy, II. p. 19. ° Mahabhar. xn. xxvn] BHAgAVATAS AND PASUPATAS 203 esteem with the philosophic and ritual doctrines of the Brahmans . The Mahabharata contains a hymn' which praises Siva under 1008 names and is not without resemblance to the Bhagavad- gita. It contains a larger number of strange epithets, but Siva is also extoUed as the AU-God, who asks for devotion and grants grace. At the close of the hymn Siva says that he has introduced the pasupata rehgion which partly contradicts and partly agrees with the institutions of caste and the Asramas, but is blamed by fools^. These last words hint that the Pasupatas laid themselves open to criticism bytheir extravagant practices, such as strange sounds and gestures'. But in such matters they were outdone by other sects caUed KapaUkas or Kaiamukhas. These carried skuUs and ate the flesh of corpses, and were the fore-runners of the filthy Aghoris, who were frequent in northern India especially near Mount Abu and Gimar a century ago and perhaps are not yet quite extinct. The biographers of Sankara* represent him as contending with these demoniac fanatics not merely with the weapons of controversy but as urging the princes who favoured him to extermina,te them. Hindu authorities treat the Pasupatas as distinct from the Saivas, or Sivaites, and the distinction was kept up in Camboja in the fourteenth century. The Saivas appear to be simply worshippers of Siva, who practice a sane ritual. In different parts of India they have pecuHarities of their own but whereas the Vaishnavas have spUt up into many sects each revering its own founder and his teaching, the Saivas, if not a united body, present few weU-marked divisions. Such as exist I shall notice below in their geographical or historical connection^. Most of them accept a system of theology or philosophy^ which starts 1 Mahabhar. xn. 13702 ff. It is recited by Daksha when he recognizes the might of Siva after the unfortunate incident of his sacrifice. 2 Santi-parvan, section oclxxxv especiaUy line 10, 470 ff. ' See Sarva-dar^ana-sangraha, chap. vi. and the comments of Ramanuja and Sankara on Vedanta Siitras, n. 2. 36. * E.g. Sankara-dig-vijaya. The first notice of these sects appears to be an inscrip tion at Igatpuri in the Nasik district of about 620 a.d. recording a grant for the worship of Kapale^vara and the maintenance of Mahavratins ( =:Kapalika3) in his temple. But doubtless the sects are much older. ' The principal are, the Pasupatas, the Saivasiddhantam of southern India and the Sivaism ot Kashmir. * The Sarva-dariana-sarigraha, chap. vn. gives a summary of it. 204 HINDUISM [on. with three principles, aU without beginning or end. These are Pati or the Lord, that is Siva: Pa^u, or the individual soul: Pa^a or the fetter, that is matter or Karma'. The task of the soul is to get free of its fetters and attain to the state of Siva. But this final deliverance is not quite the same as the identity with Brahman taught by the Vedanta: the soul becomes a Siva, equal to the deity in power and knowledge but still dependent on him rather than identical with him". Peculiar to Saiva theology is the doctrine of the five kaii- cukas' or envelopes which Hmit the soul. Spirit in itself is free: it is timeless and knows no restrictions of space, enjoyment, knowledge and power. But when spirit is contracted to indi vidual experience, it can apprehend the universe only as a series of changes in time and place: its enjoyment, knowledge and power are cramped and curtailed by the limits of personality. The terminology of the Saivas is original but the theory appears to be an elaboration of the Pancaratra thesis that the soul is surrounded by the sheath of Maya. The early literature of the worshippers of Siva (corresponding to the Samhitas of the Pancaratras) appears to have consisted of twenty-eight works composed in Sanskrit and called Agamas*. There is fairly good evidence for their antiquity. Tirumular, one of the earliest Tamil poets who is believed to have lived in the first centuries of our era, speaks of them with enthusiasm and the Buddhist Sanskrit works called Agamas (corresponding to ^ The Pfii5upatas seem to attach less importance to this triad, though as they speak of Pati, PaSu and the impurities of the soul there is not much difference. In their views of causation and free will they diffcicd slightly from the Saivas, since they held that Siva is the universal and absolute cause, the actions of indi viduals being effective only in so far as they are in conformity with the will ot Siva. The Saiva siddh&nta however holds that Siva's will is not irrespeotivo of mdividual Karma, although his independence is not tlim-eby diminished. Ho is like a man holding a magnet and directing the movements of neodlos. ^ Tlioro is some dilTorcuoe of language and pi-rhaps of docti-ine on this point in various Sivaite works. Both Sivaites and Paiicaratrins sometimes employ the language of Llio Advaita. But see Sohrader, Int. lo l' He married the daughter of a certain VaUabha who apparently was not the founder of the Sect, as is often stated. xxx] LATER VISHNUISM IN NORTH INDIA 255 same sacred food. In Caitanya's words "the mercy of God regards neither tribe nor family." His theology' shows Httle originahty. The deity is caUed Bhagavan or more frequently Hari. His majesty and omni potence are personified as Narayana, his beauty and ecstasy as Krishna. The material world is defined as bheddbhedaprakdAa, a manffestation of the deity as separate and yet not separate from him, and the soid is vibhinndmia or a detached portion of him. Some souls are in bondage to Prakriti or Maya, others through faith and love attain dehverance. Reason is useless in rehgious matters, but ruci or spiritual feeUng has a quick intuition of the divine. Salvation is obtained by Bhakti, faith or devotion, which embraces and supersedes aU other duties. This devotion means absolute seff-surrender to the deity and love for him which asks for no return but is its own reward. ' ' He who expects remunera tion for his love acts as a trader." In this devotion there are five degrees: (a) santi, calm meditation, (6) dasya, servitude, (c) sakhya, friendship, {d) vatsalya, love Uke that of a chUd for its parent, (e) madhurya, love like that of a woman for a lover. AU these sentiments are found in God and this combined ecstasy is an eternal principle identified with Hari himseff, just as in the language of the Gospels, God is love. Though Caitanya makes love the crown and culmination of rehgion, the worship of his foUowers is not Ucentious, and it is held that the right frame of mind is best attained by the recitation of Kjishna's names especiaUy Hari. The earHer centre of Caitanya's sect was his birthplace, Nadia, but both during his Ufe and afterwards his disciples frequented Brindaban and sought out the old sacred sites which were at that time neglected. At the beginning of the nineteenth century Lala Baba, a wealthy BengaU merchant, became a mendicant and visited Muttra. Though he had renounced the world, he stiU retained his business instincts and bought up the vUlages which contained the most celebrated shrines and were most frequented by pUgrims. The result was a most profitable ^ The theology of the sect may be studied in Baladeva's comraentary on the Vedanta siitras and his Praraeya RatnavaU, both contained in vol. v. of the Sacred Books of the Hindus. It would appear that the sect regards itself as a continuation of the Brahma-sampradaya but its tenets have more resemblance to those of VaUabha. 256 HINDUISM [ch. speculation and the estabUshment of Caitanya's Church in the district of Braj, which thus became the holy land of both the great Krishnaite sects. The foUowers of Caitanya at the present day are said to be divided into Gosains, or ecclesiastics, who are the descendants of the founder's original disciples, the Vrikats or cehbates, and the laity. Besides the cehbates there are several semi-monastic orders who adopt the dress of monks but marry. They have numerous maths at Nadia and elsewhere. Like the VaUabhis, this sect deifies its leaders. Caitanya, Nityananda and Advaita are caUed the three masters (Prabhii) and beUeved to be a joint incarnation of Krishna, though according to some only the first two shared the divine essence. Six of Caitanya's disciples known as the six Gosains are also greatly venerated and even ordinary rehgious teachers stiU receive an almost idolatrous respect. Though Caitanya was not a writer himseff he exercised a great infiuence on the literature of Bengal. In the opinion of so competent a judge as Dinesh Chandra Sen, BengaU was raised to the status of a Uterary language by the Vishnuite hynm- writers just as Pali was by the Buddhists. Such hymns were written before the time of Caitanya but after him they became extremely numerous' and their tone and style are said to change. The ecstasies and visions of which they tell are those described in his biographies and this emotional poetry has pro foundly influenced aU classes in Bengal. But there was and stiU is a considerable hostiUty between the Saktas and Vishnuites. A form of Vishnuism, possessing a special local flavour, is connected with the Maratha country and with the names of Namdev, Tukaram^ and Ramdas, the spiritual preceptor of Sivaji. The centre of this worship is the town of Pandharpur and I have not found it described as a branch of any of the four Vishnuite Churches : but the facts that Namdev wrote in Hindi as weU as in Marathi, that many of his hymns are included in the Granth, and that his sentiments show affinities to the ' No less than 159 padakartas or reUgious poets are enumerated by Dinesh Chandra Sen. Several coUections of these poems have been published of which the principal is caUed Padakalpataru. ^ ^ See Bhandarkar, Vaishri. and Saivism, pp. 87-99, and Nicol, Psalms of Maratha Saints which gives a bibUography. For Namdev see alsoMacauUffe, The Sikh Religion, vol. VI. pp. 17-76. For Ramdas see Rawlinson, Sivaji the Maratha, pp. 116 fi. xxx] LATER VISHNUISM IN NORTH INDIA 257 teaching of Nanak, suggest that he belonged to the school of Ramanand. There is however a difficulty about his date. Native tradition gives 1270 as the year of his birth but the language of his poems both in Marathi and Hindi is said to be too modem for this period and to indicate that he Uved about 1400', when he might easUy have felt the influence of Ramanand, for he traveUed in the north. Most of his poetry however has for its centre the temple of Pandharpur where was worshipped a deity called Vitthala, Vittoba or Pandurang. It is said that the first two names are dialectic variations of Vishnu, but that Pandurang is an epithet of Siva^. There is no doubt that the deity of Pandharpur has for many centuries been identified with Krishna, who, as in Bengal, is god the lover of the soul. But the hymns of the Marathas are less sensuous and Krishna is coupled not with his mistress Radha, but with his wife Rukmini. In fact Rukmini- pati or husband of Rukmini is one of his commonest titles. Namdev's opinions varied at different times and perhaps in different moods: Uke most rehgious poets he cannot be judged by logic or theology. Sometimes he inveigh's against idolatry — understood as an attempt to Hmit God to an image — but in other verses he sings the praises of Pandurang, the local deity, as the lord and creator of aU. His great message is that God — ¦ by whatever name he is caUed — ^is everywhere and accessible to aU, accessible without ceremonial or philosophy. "Vows, fasts and austerities are not needful, nor need you go on pilgrim age. Be watchful in your heart and always sing the name of Hari. Yoga, sacrifices and renunciation are not needful. Love the feet of Hari. Neither need you contemplate the absolute. Hold fast to the love of Hari's name. Says Nama, be steadfast in singing the name and then Hari will appear to you*." ' Bhandarkar, l.c. p. 92. An eaiUer poet of this country was Jnane^vara who wrote a paraphrase of the Bhagavad-glta in 1290. His writings are said to be the first great landmark in Marathi Uterature. ^ There is no necessary hostiUty between the worship of Siva and of Vishnu. At Pandharpur pUgrims -risit first a temple of Siva and then the principal shrine. This latter, Uke the temple of Jagannath at Puri, is suspected of having been a Buddhist shrine. It is caUed Vihara, the principal festival is in the Buddhist Lent and caste is not observed within its precincts. ' Quoted by Bhandarkar, p. 90. The subsequent quotations are from the same source but I have sometiraes sUghtly modified them and compared them with the original, though I have no pretension to be a Marathi scholar. B. n. • 17 258 HINDUISM [ch. Tukaram is better known than Namdev and his poetry which was part of the inteUectual awakening that accompanied the rise of the Maratha power is stiU a Uvmg force wherever Marathi is spoken. He Uved from 1607 to 1649 and was born in a famUy of merchants near Poona. But he was too generous to succeed in trade and a famine, in which one of his two wives died, brought him to poverty. Thenceforth he devoted himseff to prajdng and preaching. He developed a great aptitude for composing rhyming songs in irregular metre', and like Caitanya he held services consisting of discourses interspersed with such songs, prepared or extempore. In spite of persecution by the Brahmans, these meetings became very popular and were even attended by the great Sivaji. His creed is the same as that of Namdev and finds expression in verses such as these. "This thy nature is beyond the grasp of mind or words, and therefore I have made love a measure. I measure the Endless by the measure of love : he is not to be truly measured otherwise. Thou art not to be found by Yoga, sacrifice, fasting, bodily exertions or knowledge. 0 Kesava, accept the service which we render." But if he had no use for asceticism he also feared the passions. "The Endless is beyond; between him and me are the lofty mountains of desire and anger. I cannot ascend them and find no pass." In poems which are apparently later, his tone is more peaceful. He speaks much of the death of seff, of purity of heart, and of seff -dedication to God. "Dedicate aU you do to God and have done with it: Tuka says, do not ask me again and again: nothing else is to be taught but this." Maratha critics have discussed whether Tukaram foUowed the monistic philosophy of Sardiara or not and it must be con fessed that his utterances are contradictory. But the gist of the matter is that he disHked not so much monism as phUosophy. Hence he says "For me there is no use in the Advaita. Sweet to me is the service of thy feet. The relation between God and his devotee is a source of high joy. Make me feel this, keeping me distinct from thee." But he can also say almost in the language of the Upanishads. "When salt is dissolved in water, what remains distinct? I have thus become one in joy with thee and have lost myseff in thee. When fire and camphor are ^ CaUed Abhangs. xxx] LATER VISHNUISM IN NORTH INDIA 259 brought together, is there any black remnant? Tuka says, thou and I were one Ught." 5 There are interesting Vishnuite sects in Assam'. Until the sixteenth century Hinduism was represented in those regions by Saktism, which was strong among the upper classes, though the mass of the people stiU adhered to their old tribal worships. The first apostle of Vishnuism was Sankar Deb in the sixteenth century. He preached first in the Ahom kingdom but was driven out by the opposition of Saktist Brahmans, and found a refuge at Barpeta. He appears to have inculcated the worship of Krishna as the sole divine being and to have denounced idolatry, sacrffices and caste. These views were held even more strictly by his successor, Madhab Deb, a writer of repute whose works, such as the Namghosha and RatnavaU, are regarded as scripture by his foUowers. Though the Brahmans of Assam were opposed to the introduction of Vishnuism and a section of them con tinued to instigate persecutions for two centuries or more, yet when it became clear that the new teaching had a great popular foUowing another section were anxious that it should not pass out of sacerdotal control and organized it as a legitimate branch of Hinduism. WhUe fuUy recognizing the doctrine of justifica tion by faith, they also made provision for due respect to caste and Brahmanic authority. According to the last census of India^ the common view that Sankar Deb drew his inspiration from Caitanya meets with criticism in Assam. His biographies say that he Hved 120 years and died in 1569. It has been generaUy assumed that his age has been exaggerated but that the date of his death is correct. If it can be proved, as contended, that he was preaching in 1505, there would be no difficulty in admitting that he was independent of Caitanya and belonged to an earUer phase of the Vishnuite movement which produced the activity of VaUabha and the poetry of Vidyapati. It is a further argument for this independence that he taught the worship of Vishnu only and not of Radha and discountenanced the use of images. On the other hand it is stated that he sojourned in Bengal and it ' See EUot, Hinduism in Assam, J.R.A.S, 1910, pp. 1168-1186. 2 Census of India, 1911, Assam, p. 41, 17—2 260 HINDUISM [oh. appears that soon after his death his connection with the teaching of Caitanya was recognized in Assam. At present there are three sects m Assam. Firstly, the Mahapurushias, who foUow more or less faithfuUy the doctrines of Sankar and Madhab. They admit Sudras as rehgious teachers and abbots, and lay Httle stress on caste whUe not entirely rejecting it. They abstain almost entirely from the use of images in worship, the only exception being that a smaU figure of Krishna in the form of Vaikuntha Natha is found in their temples. It is not the principal object of veneration but stands to the left of a throne on which Ues a copy of the Namghosha'. This, together with the foot-prints of Sankar and Madhab, receives the homage of the faithful. The chief centre of the Mahapurushias is Barpeta, but they have also monasteries on the MajuU Island and elsewhere. Secondly, the Bamunia monasteries, with a large lay foUowing, represent a brahmamzed form of the Mahapurushia faith. This movement began in the Hfe-time of Madhab. Many of his Brahman disciples seceded from him and founded separate communities which insisted on the observance of caste (especiaUy on the necessity of rehgious teachers being Brahmans) but tolerated image-worship and the use of some kinds of flesh as food. Though this sect was perse cuted by the Ahom kings^, they were strong enough to maintain themselves. A compromise was effected in the reign of Rudra Singh (1696-1714), by which their abbots were shown all honour but were assigned the MajuH Island in the upper Brahmaputra as their chief, ff not only, residence. This island is stUl studded with numerous Sattras or monasteries, the largest of which contain three or four hundred monks, known as Bhakats (Bhaktas). They take no vows and wear no special costume but are obhged to be ceUbate while they remain in the sattra. The Mahapurushia and Bamunia monasteries are of simUar appearance, and in externals (though not in doctrine) seem to have been influenced by the Lamaism of the neighbouring regions of Sikhim and Tibet. The temples are long, low, wooden buildings, covered by roofs of corrugated iron or thatched, and 1 Some authorities state that the sacred book thus venerated is the Bhagavad- gita, but at Kamalabari I made careful enquiries and was assured it was the Namghosha. 2 EspeciaUy Gadadhar Singh, 1681-96. xxx] LATER VISHNUISM IN NORTH INDIA 261 containing inside a nave with two rows of wooden pillars which leads to a sanctuary divided from it by a screen. The third sect are the Moamarias, of poUtical rather than rehgious importance. They represent a democratic element, recruited from non- Hindu tribes, which seceded even in the life-time of Sankar Deb. They appear to reject nearly aU Hindu observances and to worship aboriginal deities as weU as Krishna. Little is known of their rehgious teaching, if indeed they have anything worthy of the name, but in the latter haU of the eighteenth century they distracted the kingdom of Assam with a series of rebeUions which were suppressed with atrocious cruelty. Caitanya is said to have admitted some Mohammedans as members of his sect. The precedent has not been followed among most branches of his later adherents but a curious haff-secret sect, found throughout Bengal in considerable numbers and caUed Kartabhajas', appears to represent an eccentric develop ment of his teaching in combination with Mohammedan ele ments. Both MosHms and Hindus belong to this sect. They observe the ordinary social customs of the class to which they belong, but it is said that those who are nominal MosHms neither cUcumcize themselves nor frequent mosques. The founder, caUed Ram Smaran Pal, was born in the Nadia district about 1700, and his chief doctrine is said to have been that there is only one God who is incarnate in the Head of the sect or Karta^. For the first few generations the headship was invested in the founder and his descendants but dissensions occurred and there is now no one head: the faithful can select any male member of the founder's famUy as the object of their devotion. The Karta claims to be the owner of every human body and is said to exact rent for the soul's tenancy thereof. No distinction of caste or creed is recognized and hardly any ceremonies are pre scribed but meat and wine are forbidden, the mantra of the sect is to be repeated five times a day and Friday is held sacred. These observances seem an imitation of Mohammedanism*. 1 See Census of India, 1901, Bengal, pp. 183^ and Bhattacharya, Hindu Castes and Sects, pp. 485-488. ' Karta, UteraUy doer, is the name given to the executive head of a joint famUy in Bengal. The sect prefer to caU themselves Bhabajanas or Bhagawanis. ' Another mixed sect is that of the Dhamis in the Panna state of BundeUjhand, founded by one Prannath in the reign of Aumngzeb. Their doctrine is a combination of Hinduism and Islam, tending towards Krishnaism. See RusseU, Tribes and Castes of Central Provinces, p. 217. CHAPTER XXXI AMALGAMATION OP HINDUISM AND ISLAM. KABIR AND THE SIKHS The Kartabhajas mentioned at the end of the last chapter show a mixture of Hinduism and Mohammedanism, and the mixture' is found in other sects some of which are of considerable im portance. A group of these sects, including the Sikhs and followers of Kabir, arose in the fifteenth and sixteenth centuries. Their origin can be traced to Ramanand but they cannot be caUed Vaishnavas and they are clearly distinguished from aU the rehgious bodies that we have hitherto passed in review. The tone of their writings is more restrained and severe: the worshipper approaches the deity as a servant rather than a lover: caste is rejected as useless: Hindu mythology is eschewed or used sparingly. Yet in spite of these differences the essential doctrines of Tulsi Das, Kabir and Nanak show a great resem blance. They aU beUeve m one deity whom they caU by various names, but this deity, though personal, remains of the Indian not of the Semitic type. He somehow brings the world of transmigration into being by his power of iUusion, and the business of the soul is to free itseff from the iUusion and return to him. Almost aU these teachers, whether orthodox or hetero dox, had a singular facihty for composing hymns, often of high Hterary merit, and it is in these emotional utterances, rather than in dogmatic treatises, that they addressed themselves to the peoples of northern India. The earhest of these mixed sects is that founded by Kabir*. He appears to have been a Mohammedan weaver by birth, 1 It is exempUfied by the curious word an-had limitless, being the Indian negative prefix added to the arable word had used in the Sikh Granth and by Caran Das as a name of God. 2 See especiaUy G. H. Westcott, Kabir and the Kabir Panth, and Maoauliffe, Sikh Religion, vol. VI. pp. 122-316. Also WUson, Essays on the religion ofthe Hindus, vol. I. pp. 68-98. Garciu de Tassy, Histoire de la Littirature Hindcme, ii. pp. 120-134. Bhandarkar, Vaishn, and Saivism, pp. 67-73. OH. XXXI] HINDUISM AND ISLAM 263 though tradition is not unanimous on this point'- It is admitted, however, that he was brought up among MosHms at Benares but became a disciple of Ramanand. This suggests that he Hved early in the fifteenth century*. Another tradition says that he was summoned before Sikander Lodi (1489-1517), but the detaUs of his Hfe are evidently legendary. We only know that he was married and had a son, that he taught in northern and perhaps central India and died at Maghar in the district of Gorakhpur. There is significance, however, in the legend which relates that after his decease Hindus and Mohammedans dis puted as to whether his body should be burned or buried. But when they raised the cloth which covered the corpse, they found underneath it only a heap of flowers. So the Hindus took part and burnt them at Benares and the MosHms buried the rest at Maghar. His grave there is stiU in Moslim keeping. In teaching Kabir stands midway between the two religions, but leaning to the side of Hinduism. It is clear that this Hindu bias became stronger in his foUowers, but it is not easy to separate his own teaching from subsequent embeUishments, for the numerous hymns and sayings attributed to him are coUected in compUations made after his death, such as the Bijak and the Adigranth of the Sikhs. In hymns which sound authentic he puts Hindus and MosHms on the same footing. "Kabir is a chUd of Ram and AUah," he says, "and accepteth aU Gurus and Pirs." "0 God, whether AUah or Ram, I Hve by thy name." "Make thy mind thy Kaaba, thy body its enclosing temple. Conscience its prime teacher. Then, O priest, call men to pray to that mosque Which hath five gates. The Hindus and Mussulmans have the same Lord." But the formahties of both creeds are impartiaUy condemned. "They are good riders who keep aloof from the Veda and Koran*." Caste, circumcision and idolatry are reprobated. The Hindu deities and their incarnations are aU dead : God was not * The name Kabir seems to me decisive. ^ Dadu who died about 1603 is said to have been fifth in spiritual descent from Kabir. ' Frora a hymn in which the spiritual Ufe is represented as a ride. MacauUffe, VI. p. 156. 264 HINDUISM [ch. in any of them'. Ram, it would seem, should be understood not as Ramacandra but as a name of God. Yet the general outlook is Hindu rather than Mohammedan. God is the magician who brings about this iUusory world in which the soul wanders*. "I was in immobile and mobile creatures, in worms and in moths; I passed through many various births. But when I assumed a human body, I was a Yogi, a Yati, a penitent, a Brahmacari : sometimes an Emperor and sometimes a beggar." Unhke the Sikhs, Kabir teaches the sanctity of Hfe, even of plants. "Thou cuttest leaves, 0 flower girl: in every leaf there is Hfe." Release, as for aU Hindus, consists in escaping from the round of births and deaths. Of this he speaks almost in the language of the Buddha*. "Though I have assumed many shapes, this is my last. The strings and wires of the musical instrument are all worn out: I am now in the power of God's name. I shall not again have to dance to the tune of birth and death. Nor shall my heart accompany on the drum." This deliverance is accompHshed by the union or identifica tion of the soul with God. "Remove the difference between thyself and God and thou shalt be united with him Him whom I sought without me, now I find vsdthin me Know God : by knovnng him thou shalt become as he. When the soul and God are blended no one can distinguish them*." But ff he sometimes writes Uke Sankara, he also has the note of the Psalms and Gospels. He has the sense of sin: he thinks of God in vivid personal metaphors, as a lord, a bride groom, a parent, both father and mother. "Save me, O God, though I have offended thee I forgot him who made me and did cleave unto strangers." '^ But Hari is sometiraes used by Kabir, especiaUy in the hymns incorporated in the Granth, as a name of God. ^ Though Kabir writes as a poet rather than as a phUosopher he evidently leaned to the doctrine of illusion (vivartavdda) rather than to the doctrine of mani festation or development (pariridmavdda). He regards Maya as soraething evU, a trick, a thief, a force which leads men captive, but which disappears with the knowledge of God. "The Ulusion vanished when I recognized him" (xxxix.). * He even uses the word nirvana. * Frora Kabir's acrostic. MacauUffe, vi. pp. 186 and 188. It is possible that this is a later ooraposition. XXXI] HINDUISM AND ISLAM 265 "Sing, sing, the marriage song. The sovereign God hath come to my house as my husband ¦ I obtained God as my bridegroom; so great has been my good fortune." "A mother beareth not in mind All the faults her son committeth. O, God, I am thy child: Why blottest thou not out my sins? " "My Father is the great Lord of the Earth ; To that Father how shall I go^? " The writings of Kabir's disciples such as the Sukh Nidhan attributed to Srut Gopal (and written according to Westcott about 1729) and the stUl later Amar Mul, which is said to be representative of the modern Kabirpanth, show a greater in clination to Pantheism, though caste and idolatry are stiU condemned. In these works, which relate the conversion of Dharm Das afterwards one of Kabir's principal foUowers, Kabir is identified with the Creator and then made a pantheistic deity much as Krishna in the Bhagavad-glta*. He is also the true Guru whose help is necessary for salvation. Stress is further laid on the doctrine of Sabda, or the divine word. Hindu theology was famUiar with this expression as signifying the etemal seU-existent revelation contained in the Vedas. Kabir appears to have held that articulate sound is an expression of the Deity and that every letter, as a constituent of such sound, has a meaning. But these letters are due to Maya: in reahty there is no pluraUty of sound. Ram seems to have been selected as the divine name, because its brevity is an approach to this unity, but true knowledge is to understand the Letterless One, that is the real name or essence of God from which aU differen tiation of letters has vanished. Apart from some special metaphors the whole doctrine set forth in the Sukh Nidhan ' MacauUffe, vi. pp. 230, 209, 202, 197. * Westcott, l.c. p. 144, "I am the creator of this world I am the seed and the tree aU are contained in rae — I Uve within all and aU live within me" and much to the same effect. Even in the hymns of the Adi Granth we find suoh phrases as "Now thou and I have become one." (MacauUffe, vi. p. 180.) This identification of Kabir with the deity is interesting as being a modem example of what probably happened in the case of Krishna. SimUarly those who coUected the hymns which form the sacred books of the Sikhs and Kabirpanthis repeated the process which in earUer ages produced the Rig Veda. 266 HINDUISM [ch. and Amar Mul is Httle more than a loose Vedantism, somewhat reminiscent of Sufiism'. The teachmg of Kabir is known as the Kabirpanth. At present there are both Hindus and Mohammedans among his foUowers and both have monasteries at Maghar where he is buried. The sect numbers in aU about a miUion*. It is said that the two divisions have Httle in common except veneration of Kabir and do not intermix, but they both observe the practice of partaking of sacred meals, holy water*, and consecrated betel nut. The Hindu section is again divided into two branches known as Father (Bap) and Mother (Mai). Though there is not much that is original in the doctrines of Kabir, he is a considerable figure in Hindi Uterature and may justly be caUed epoch-making as marking the first fusion of Hinduism and Islam which culminates and attains poHtical importance in the Sikhs. Other offshoots of his teaching are the Satnamis, Radha-swamis and Dadupanthis. The first were founded or reorganized in 1750 by a certain Jag-jivan-das. They do not observe caste and in theory adore only the True Name of God but in practice admit ordinary Hindu worship. The Radha-swamis, founded in 1861, profess a combination of the Kabirpanth with Christian ideas. The Dadupanthis show the influence of the military spirit of Islam. They were founded by Dadu, a cotton weaver of Ahmedabad who flourished in Akbar's reign and died about 1603. He insisted on the equaUty of man kind, vegetarianism, abstinence from alcohol and strict celibacy. Hence the sect is recruited by adopting boys, most of whom are trained as soldiers. In such conditions the Dadupanthis cannot increase greatly but they number about nine thousand and are found chiefly in the state of Jaipur, especiaUy in the town of Naraina*. 1 " The Atma mingles with Pararaatma, as the rivers fiow into the ocean. OiUy in this way can Pararaatma be found. The Atma without Sabda is blind and cannot find the path. He who sees Atma-Ram is present everywhere. AU he sees is Uke himself. There is nought except Brahma. I am he, I am the tme Kabir." Westcott, p. 168. ^ The Census of 1901 gives 843,171 but there is reason to think the real numbers are larger. ' Consecrated by washing in it wooden sandals supposed to represent the feet of Kabir. It is stated that they beUeve they eat the body of Kabir at their sacred meal which perhaps points to Christian influence. See RusseU, l.c. pp. 239-240. * See RusseU, Tribes and Castes of Central Provinces, p. 217, where it is said that some of them are householders. XXXI] HINDUISM AND ISLAM 267 The Sikh rehgion' is of special interest since it has created not only a political society but also customs so distinctive that those who profess it rank in common esteem as a separate race. The founder Nanak Hved from 1469 to 1538 and was born near Lahore. He was a Hindu by birth but came under Mohammedan influence and conceived the idea of reconciling the two faiths. He was attracted by the doctrines of Kabir and did not at first claim to teach a new rehgion. He wished to unite Hindus and MosHms and described himseU simply as Guru or teacher and his adherents as Sikhs or disciples. He spent the greater part of his Hfe wandering about India and is said to have reached Mecca. A beautiful story relates that he feU asleep with his feet turned towards the Kaaba. A moUah kicked him and asked how he dared to turn his feet and not his head towards God. But he answered, "Turn my feet in a direction where God is not." He was attended on his wanderings by Mardana, a lute-player, who accompanied the hymns which he never failed to compose when a thought or adventure occurred to him. These compositions are similar to those of Kabir, but seem to me of inferior merit. They are diffuse and inordinately long; the Japji for instance, which every Sikh ought to recite as his daily prayer, fiUs not less than twenty octavo pages. Yet beautUul and incisive passages are not wanting. When at the temple of Jagannath, he was asked to take part in the evening worship at which lights were waved before the god while fiowers and incense were presented on golden salvers studded with pearls. But he burst out into song*. "The sun and moon, 0 Lord, are thy lamps, the firmament thy salver and the orbs of the stars the pearls set therein. "The perfume of the sandal tree is thy incense; the wind is thy fan; aU the forests are thy flowers, 0 Lord of light." ¦ Though Nanak is fuU of Hindu aUusions he is more Moham medan m tone than KabU, and the ritual of Sikh temples is ^ See especiaUy MacauUffe, The Sikh Religion, six volumes. " MacauUffe, i. p. 82. 268 HINDUISM [OH. modeUed on the Mohammedan rather than on the Hmdu pattern. The opening words of the Japji are: "There is but one God, whose name is true, the Creator'" and he is regarded rather as the ruler of the world than as a spirit finding expression in it. "By his order" aU things happen. "By obeying him" man obtains happiness and salvation. "There is no Hmit to his mercy and his praises." In the presence of God "man has no power and no strength." Such sentiments have a smack of Mohammed and Nanak sometimes uses the very words of the Koran as when he says that God has no companion. And though the penetrating spirit of the Vedanta infects this regal monotheism, yet the doctrine of Maya is set forth in unusual phraseology : ' ' God himseU created the world and himseU gave names to things. He made Maya by his power : seated, he beheld his work with dehght." In other compositions attributed to Nanak greater promin ence is given to Maya and to the common Hindu idea that creation is a seU-expansion of the deity. Metempsychosis is taught and the divine name is Hari. This is characteristic of the age, for Nanak was nearly a contemporary of Caitanya and VaUabUacarya. For Kabir, the disciple of Ramananda, the name was Ram. Nanak was sufficiently conscious of his position as head of a sect to leave a successor as Guru*, but there is no indication that at this time the Sikhs differed materiaUy from many other religious bodies who reprobated caste and idolatry. Under the fourth Guru, Ram Das, the beginnings of a change appear. His strong personaUty collected many wealthy adherents and with their offerings he purchased the tank of Amritsar* and built in its midst the celebrated Golden Temple. He appointed his son Arjun as Guru in 1581, just before his death: the succession was made hereditary and henceforth the Gurus became chiefs rather than spiritual teachers. .Arjun assumed some of the insignia of royalty: a town grew up round the sacred 1 The original is Karta purukh (=purusha), the creative male. Thia phrase shows how Hindu habits of thought clung to Nanak. 2 The Guru of the SUdis are: (a) Nanak, 1469-1538, (6) Angada, 1538-1552, (c) Amardas, 1552-1575, (d) Ramdas, 1575-1581, (e) Arjun, 1581-1606, (/) Har- Govmd, 1606-1639, (g) Har-Rai, 1639-1663, (h) Har-Kisan, 1663-1666, (i) Teg- Bahadur, 1666-1675, (j) Govind Singh, 1675-1708. ' Amritasaras the lake of nectar. XXXI] HINDUISM AND ISLAM 269 tank and became the centre of a community ; a tax was coUected from aU Sikhs and they were subjected to special and often salutary legislation. Infanticide, for instance, was strictly for bidden. With a view of providing a code and standard Arjun compiled the Granth or Sikh scriptures, for though hymns and prayers composed by Nanak and others were in use there was as yet no authorized coUection of them. The example of Moham medanism no doubt stimulated the desire to possess a sacred book and the veneration of the scriptures increased with time. The Granth now receives the same kind of respect as the Koran and the first sight of a Sikh temple with a large open volume on a reading-desk cannot fail to recall a mosque. Arjun's compUation is caUed the Adi-granth, or original book, to distinguish it from the later additions made by Guru Govind. It comprises hymns and prayers by Nanak and the four Gurus who foUowed him (including Arjun himseU), Ramanand, Kabir and others, amounting to thirty-five writers in aU. The Hst is interesting as testifjdng to the existence of a great body of oral poetry by various authors ranging from Ramanand, who had not separated himseU from orthodox Vishnuism, to .Arjun, the chief of the Sikh national community. It was evidently felt that aU these men had one inspiration coming from one truth and even now unwritten poems of Nanak are current in Bihar. The Granth is written in a special alphabet known as Gurmukhi ' and contains both prose and poetical pieces in several languages: most are in old western Hindi* but some are in Panjabi and Marathi. But though in compiling a sacred book and in uniting the temporal and spiritual power Arjun was infiuenced by the spirit of Mohammedanism, this is not the sort of imitation which makes for peace. The combination of Hinduism and Islam resulted in the production of a special type of Hindu pecuHarly distasteful to MosHms and not much loved by other Hindus. Much of Arjun's activity took place in the later years of the Emperor Akbar. This most philosophic and tolerant of princes abandoned Mohammedanism after 1579, remitted the special ' It appears to be an arbitrary adaptation of the Deva-nagari characters. The shape of the letters is mostly the same but new values are assigned to them. ^ This is the description of the dialect given by Grierson, the highest authority in such matters. 270 HINDUISM [CH. taxes payable by non-MosUms and adopted many Hindu obser vances. Towards the end of his Hfe he promulgated a new creed known as the Din-i-ilahi or divine faith. This eclectic and composite rehgion bears testimony to his vanity as weU as to his large sympathies, for it recognized him as the viceregent or even an incarnation of God. It would appear that the singular Httle work caUed the AUopanishad or AUah Upanishad' was written in connection with this movement. It purports to be an Upanishad of the Atharva Veda and can hardly be described as other than a forgery. It declares that "the AUah of the prophet Muhammad Akbar* is the God of Gods " and identifies him with Mitra, Varuna, the sun, moon, water, Indra, etc. Akbar's reUgion did not long survive his death and never flourished far from the imperial court, but somewhat later (1656) Muhammad Dara Shukoh, the son of Shah Jehan, caused a Persian translation of about fUty Upanishads, known as the Oupnekhat*, to be prepared. The general temper of the period was propitious to the growth and immunity of mixed forms of behef, but the warlike and semi-poHtical character of the Sikh community brought trouble on it. Arjun attracted the unfavourable attention of Akbar's successor, Jehangir*, and was cast into prison where he died. The Sikhs took up arms and henceforth regarded themselves as the enemies of the government, but their strength was wasted by internal dissensions. The ninth Guru, Teg-Bahadur, was executed by Aurungzeb. Desire to avenge this martyrdom and the strenuous character of the tenth Guru, Govind Singh (1675- 1708), completed the transformation of the Sikhs into a church militant devoted to a holy war. Though the most aggressive and uncompromising features of Sikhism are due to the innovations of Govind, he was so far from being a theological bigot that he worshipped Durga and 1 See Rajendrala Mitra's article in J.A.S.B. XL. 1871, pp. 170-176, which gives the Sanskrit text of the Upanishad. Also Schrader, Catalogue of Adyar Library, 1908, pp. 136-7. Schrader states that in the north of India the AUopanishad is recited by Brahmans at the Vasantotsava and on other occasions: also that in southern India it is generaUy believed that Moslims are skiUed in the Atharva Veda. ^ I.e., not the Allah of the Koran. ' This Persian translation was rendered word for word into very strange Latin by AnquetU Duperron (1801-2) and this Latin version was used by Schopenhauer. * He is said to have prayed for the success of the Emperor's rebelUous son, XXXI] HINDUISM AND ISLAM 271 was even said to have offered human sacrifices. But the aim of aU his ordinances was to make his foUowers an independent body of fighting men. They were to return the salutation of no Hindu and to put to death every Mohammedan. The community was caUed Khalsa' : within it there was perfect equality : every -man was to carry a sword and wear long hair but short trousers. Converts, or recruits, came chiefly from the fighting tribes of the Jats, but in theory admission was free. The initiatory ceremony, which resembled baptism, was performed with sugar and water stirred with a sword, and the neophyte vowed not to worship idols, to bow to none except a Sikh Guru, and never to tum his back on the enemy. To give these institutions better religious sanction, Govind composed a supplement to the Granth, called Da^ama Padshah ka Granth or book of the tenth prince. It consists of four parts, aU in verse, and is said to inculcate war as persistently as Nanak had inculcated meekness and peace. To give his institutions greater permanence and prevent future alterations Govind refused to appoint any human successor and bade the Sikhs consider the Granth as their Guru. "Whatsoever ye shall ask of it, it wiU show you" he said, and in obedience to his command the book is stiU invested with a kind of personahty and known as Granth Sahib. Govind spent most of his time in wars with Aurungzeb marked by indomitable perseverance rather than success. Towards the end of his Hfe he retired into Malwa and resided at a place caUed Damdama. The accounts of his latter days are somewhat divergent. According to one story he made his peace with the Mughals and accepted a mihtary command under the successor of Aurungzeb but it is more commonly asserted that he was assassinated by a private enemy. Even more troublous were the days of his successor Banda. Since Govind had abohshed the Guruship, he could not claim to be more than a temporal chief, but what he lacked in spiritual authority he made amends for in fanaticism. The eight years of his leadership were spent in a war of mutual extermination waged with the MosHms of the Panjab and diversified only by internal dissen sions. At last he was captured and the sect was nearly annihUated by the Emperor Farukhsiyar. According to the ' This Arabic word is interpreted in this context as meaning the special portion (of God). 272 HINDUISM [oh. ordinary account this victory was foUowed by an orgy of torture and Banda was barbarously executed after witnessing during seven days the torments of his foUowers and kinsmen. We read with pleasure but increduHty that one division of the Sikhs beheve that he escaped and promulgated his pecuhar doctrines in Sind. Asiatics do not rehsh the idea that the chosen of God can suffer violent death. The further history of the Sikhs is poUtical rather than rehgious, and need not detain us here. Despite the efforts of the Mughals to exterminate them, they were favoured by the disturbed state of the country in the early decades of the eighteenth century, for the raids of Afghans and Persians con vulsed and paralyzed the empire of Delhi. The govemment of the Khalsa passed into the hands of a body of fanatics, caUed AkaUs, but the decision of grave matters rested with a council of the whole community which occasionaUy met at Amritsar. Every Sikh claimed to have joined the confederacy as an independent soldier, bound to fight under his mihtary leaders but otherwise exempt from control, and entitled to a share of land. This absolute independence, being unworkable in practice, was modified by the formation of Misals or voluntary associa tions, of which there were at one time twelve. From the middle of the eighteenth century onwards the Sikhs were masters of the Panjab and their great chief Ranjit Singh (1797-1839) succeeded in converting the confederacy into a despotic monarchy. Their power did not last long after his death and the Panjab was conquered by the British in the two wars of 1846 and 1849. With the loss of poHtical independence, the differences between the Sikhs and other Hindus tended to decrease. This was natural, for nearly aU their strictly religious tenets can be paralleled in Hinduism. Guru Govind waged no war against polytheism but wished to found a rehgious commonwealth equally independent of Hindu castes and Mohammedan sultans. For some time his ordinances were successful in creating a tribe, almost a nation. With the coUapse of the Sikh state, the old hatred of Mohammedanism remained, but the Sikhs differed from normal Hindus hardly more than such sects as the Linga yats, and, as happened with decadent Buddhism, the unobtrusive pressure of Hindu beUefs and observances tended to obUterate XXXI] HINDUISM AND ISLAM 273 those differences. The Census of India', 1901, enumerated three degrees of Sikhism. The first comprises a few zealots caUed Akahs who observe aU the precepts of Govind. The second class are the Guru Govind Sikhs, who observe the Guru's main commands, especiaUy the prohibition to smoke and cut the hair. Lastly, there are a considerable number who profess a respect for the Guru but foUow Hindu beUefs and usages wholly or in part. Sikhism indeed reproduces on a smaU scale the changeable- ness and complexity of Hinduism, and includes associations caUed Sabha, whose members aim at restoring or maintaining what they consider to be the true faith. In 1901 there was a tendency for Sikhs to give up their pecuHarities and describe themselves as ordinary Hindus, but in the next decade a change of sentiment among these waverers caused the Sikh community as registered to increase by thirty-seven per cent, and a period of rehgious zeal is reported*. ^ Census of India, 1901, Panjab report, p. 122. " Provincial Geographies of India, Panjab, Douie, 1916, p. 117. CHAPTER XXXII SAKTISM' Among the principal sub-divisions of Hinduism must be reckoned the remarkable religion known as Saktism, that is the worship of Sakti or Siva's spouse under various names, of which Devi, Durga and KaU are the best known. It differs from most sects in not being due to the creative or reforming energy of any one human founder. It claims to be a revelation from Siva himseU, but considered HistoricaUy it appears to be a compound of Hinduism with un-Aryan behefs. It acquired great influence both in the courts and among the people of north-eastern India but without producing personaUties of much eminence as teachers or writers. It would be convenient to distinguish Saktism and Tantrism, as I have aUeady suggested. The former means the worship of a goddess or goddesses, especially those who are regarded as forms of Siva's consort. Vishnuites sometimes worship female deities, but though the worship of Lakshmi, Radha and others may be coloured by imitation of Saktist practices, it is less conspicuous and seems to have a different origin. Tantrism is a system of magical or sacramental ritual, which professes to attain the highest aims of rehgion by such methods as spells, diagrams, gestures and other physical exercises. One of its bases is the assumption that man and the universe correspond as microcosm and macrocosm and that both are subject to the mysterious power of words and letters. These ideas are not modern nor pecuhar to any Indian sect. They are present in the Vedic ceremonial, in the practices of the Yoga and even in the teaching of the quasi-mussulman sect of Kabir, which attaches great importance to the letters of the divine name. They harmonize with the common Indian view that some form of disciphne or physical training is essential to ' See also chap. xxiv. as to Saktism and Tantrism iu Buddhism. Copious materials for the study of Saktism and Tantrism are being made available in the series of tantric texts edited in Sanskrit and Tibetan, and in some cases translated by the author who uses the pseudonym A. Avalon. ch. xxxn] Saktism 275 the rehgious Ufe. They are found in a highly developed form among the Nambutiris and other Brahmans of southern India who try to observe the Vedic rules and in the Far East among Buddhists of the Shingon or Chen-yen sect'. As a rule they receive the name of Tantrism only when they are elaborated into a system which claims to be a special dispensation for this age and to supersede more arduous methods which are pohtely set aside as practicable only for the hero-saints of happier times. Tantrism, Hke salvation by faith, is a simpHfication of reUgion but on mechanical rather than emotional Hues, though its deficiency in emotion often finds strange compensations. But Tantrism is analogous not so much to justification by faith as to sacramental ritual. The paraUel may seem shocking, but most tantric ceremonies are similar in idea to Christian sacraments and may be caUed sacramental as correctly as magical. Even in the AngHcan Church baptism includes sprinkling with water (abhisheka), the sign of the cross (nyasa) and a formula (mantra), and U any one supposes that a child so treated is sure of heaven whereas the future of the unbaptized is dubious, he holds Hke the Tantrists that spiritual ends can be attained by physical means. And in the Roman Church where the rite includes exorcism and the use of salt, oil and Hghts, the paraUel is stiU closer. Christian mysticism has had much to do with symboHsm and even with alchemy*, and Zoroastrianism, which is generaUy regarded as a reasonable rehgion, attaches extraordinary importance to holy spells*. So Indian rehgions are not singular in this respect, though the uncompromising thoroughness with which they work out this like other ideas leads to starthng results. The worship of female deities becomes prominent somewhat late in Indian Hterature and it does not represent — not to the same extent as the Chinese cult of Kwan-yin for example — the better ideals of the period when it appears. The goddesses of the Rig Veda are insignificant: they are Httle more than names, and grammaticaUy often the feminine forms of their consorts. But this Veda is evidently a special manual of prayer from which many departments of popular rehgion were excluded. In * See Annates du Musie Guimet, Tome vm. Si-Do-In-Dzon. Gestes de Pofficiant dans les c&emonies mystiques des sectes Tendai et Singon, 1899. ^ See UnderhUl, Mysticism, chaps, vi. and vu. ' See DhaUa, Zoroastrian Theology, p. 116. 18—2 276 HINDUISM [CH. the Atharva Veda many spUits with feminme names are invoked and there is an inchnation to personify bad quahties and disasters as goddesses. But we do not find any goddess who has attamed a position comparable with that held by Durga, Cybele or Astarte, though there are some remarkable hymns' addressed to the Earth. But there is no doubt that the worship of goddesses (especiaUy goddesses of fertiUty) as great powers is both ancient and widespread. We find it among the Egyptians and Semites, in Asia Mmor, m Greece, Italy, and among the Kelts. The goddess Anahit, who was worshipped with immoral rites in Bactria, is figured on the coins of the Kushans and must at one time have been known on the north-western borders of India. At the present day Sitaia and in south India Mariamman are goddesses of smaUpox who require propitiation, and one of the earhest deities known to have been worshipped by the Tamils is the goddess Kottavai*. Somewhat obscure but widely worshipped are the powers known as the Mothers, a title which also occurs in Keltic mythology. They are groups of goddesses varying in number and often malevolent. As many as a hundred and forty are said to be worshipped in Gujarat. The census of Bengal (1901) records the worship of the earth, sun and rivers as females, of the snake goddesses Manasa and Jagat Gauri and of numerous female demons who send disease, such as the seven sisters, Ola Bibi, Jogini and the Churels, or spirits of women who have died in childbirth. The rites celebrated in honour of these deities are often of a questionable character and include dances by naked women and offerings of spirituous hquors and blood. Similar features are found in other countries. Prostitution formed part of the worship of Astarte and Anahit : the Tauric Artemis was adored with human sacrifices and Cybele with seU-inflicted mutUations. Similarly offerings of blood drawn from the sacrificer's own body are enjoined in the Kahka Purana. Two stages can be dis tinguished in the relations between these cults and Hinduism. In the later stage which can be witnessed even at the present day an aboriginal goddess or demon is identified with one of the aspects (generaUy a "black" or fierce aspect) of Siva's ^ SpeciaUy Ath. Veda, xn. 1. " VUlage deities in south India at the present day are usuaUy female. See Whitehead, Village Gods, p. 21. xxxn] SAkTISM 277 spouse'. But such identification is facihtated by the fact that goddesses Hke KaU, Bhairavi, Chinnamastaka are not products of purely Hindu imagination but represent earlier stages of amalgamation in which Hindu and aboriginal ideas are already compounded. When the smaUpox goddess is identified with KaH, the procedure is correct, for some popular forms of KaU are Httle more than an aboriginal deity of pestilence draped with Hindu imagery and phUosophy. Some Hindu scholars demur to this derivation of Saktism from lower cults. They point to its refined and phUosophic aspects ; they see in it the worship of a goddess, who can be as mercfful as the Madonna, but yet, since she is the goddess of nature, combines in one shape Hfe and death. May not the grosser forms of Saktism be perversions and corruptions of an ancient and higher faith? In support of this it may be urged that the Buddhist goddess Tara is as a rule a beautiful and benevolent figure, though she can be terrible as the enemy of evU and has clear affinities to Durga. Yet the history of Indian thought does not support this view, but rather the view that Hinduism incorporated certain ancient ideas, true and striking as ancient ideas often are, but without purging them sufficiently to make them acceptable to the majority of educated Indians. The Yajur Veda* associates Rudra with a female deity caUed Ambika or mother, who is however his sister, not his spouse. The earhest forms of the latter seem to connect her with mountains. She is Uma Haimavati, the daughter of the Himalayas, and Parvati, she of the mountains, and was perhaps originaUy a sacred peak. In an interesting but brief passage of the Kena Upanishad (in. 12 and rv. 1) Uma Haimavati explains to the gods that a being whom they do not know is Brahman. In later times we hear of a similar goddess in the Vindhyas, Maharani Vindhyesvari, who was connected with human sacrifices and Thugs*. Siva's consort, like her Lord, has many forms classified as white or benignant and black or terrible. Uma belongs to the former class but the latter (such as KaU, ^ Thus Candi is considered as identical with the wood goddess BasuU, worshipped in the jungles of Bengal aud Orissa. See J.A. 1873, p. 187. 2 Vaj. Sanh. 3. 57 and Taittir. Br. i. 6. 10. 4. " Crooke, Popular Religion of Northern India, i. 63. Monier WUUams, Brahm, and Hinduism, p. 57 gives an interesting account of the shrine of KaU at Vindhyacal said to have been formerly frequented by Thugs. 278 HINDUISM [CH. Durga, Camunda, Cauda and Karaia) are more important'. Female deities bearing names Hke these are worshipped in most parts of India, HteraUy from the Himalaya to Cape Comorm, for the latter name is derived from Kumari, the Virgin goddess*. But the names Sakta and Saktism are usuaUy restricted to those sects in Bengal and Assam who worship the Consort of Siva with the rites prescribed in the Tantras. Saktism regards the goddess as the active manifestation of the godhead. As such she is styled Sakti, or energy (whence the name Sakta), and is also identified with Maya, the power which is associated with Brahman and brings the phenomenal world into being. Similar ideas appear in a philosophic form in the sankhya teaching. Here the soul is masculine and passive : its task is to extricate and isolate itseU. But Prakriti or Nature is feminine and active : to her is due the evolution of the universe : she involves the soul in actions which cause pain but she also helps the work of Hberation*. In its fully developed form the doctrine of the Tantras teaches that Sakti is not an emanation or aspect of the deity. There is no distinction between Brahman and Sakti. She is Parabrahman and pardtpard, Supreme of the Supreme. The birthplace of Saktism as a definite sect seems to have been north-eastern India* and though it is said to be extending in the United Provinces, its present sphere of influence is stiU ^ This idea that deities have different aspects in which they practicaUy become different persons is very prevalent in Tibetan mythology which is borrowed from medieval Bengal. ^ Though there are great temples erected to goddesses in S. India, there are also some signs of hostUity to Saktism. See the curious legends about an attendant of Siva caUed Bhriiigi who would not worship Parvati. Hultzsoh, South Indian Inscriptions, II. ii. p. 190. = There is a curious tendency in India to regard the male principle as quiescent, the female as active and stimulating. The Chinese, who are equaUy fond of using these two principles in their cosmologioal speculations, adopt the opposite view. The Yang (male) is positive and active. The Yin (female) is negative and passive. * The Mahanirvana Tantra seems to have been composed in Bengal since it recommends for sacrificial purposes (vi. 7) three kinds of fish said to be characteristic of that region. On the other hand Buddhist works caUed Tantras are said to have been composed in north-western India. Udyana had an old reputation for magic and even in modern times Saktism exists in western Tibet and Leh. It is highly probable that in all these districts the practice of magic and the worship of mountain goddesses were prevalent, but I find little evidence that a definite Sakta sect arose elsewhere than m Bengal and Assam or that the Saktist corruption of Buddhism prevailed elsewhere than in Magadha and Bengal. xxxn] Saktism 279 chiefly Bengal and Assam'. The population of these countries is not Aryan (though the Bengali language bears witness to the strong Aryan influence which has prevailed there) and is largely composed of immigrants from the north belonging to the Tibeto- Burman, Mon-Khmer and Shan famiUes. These tribes remain distinct in Assam but the BengaH represents the fusion of such invaders with a Munda or Dravidian race, leavened by a little Aryan blood in the higher castes. In aU this region we hear of no ancient Brahmanic settlements, no ancient centres of Vedic or even Puranic learning* and when Buddhism decayed no body of Brahmanic tradition such as existed in other parts of India imposed its authority on the writers of the Tantras. Even at the present day the worship of female spirits, only haU acknow ledged by the Brahmans, prevaUs among these people, and in the past the national deities of many tribes were goddesses who were propitiated with human sacrifices. Thus the Chutiyas of Sadiya used to adore a goddess, caUed Kesai Khati — the eater of raw flesh. The rites of these deities were originaUy performed by tribal priests, but as Hindu influence spread, the Brahmans graduaUy took charge of them without modifying their character in essentials. Popular Bengali poetry represents these goddesses as desiring worship and feeling that they are slighted: they persecute those who ignore them, but shower blessings on their worshippers, even on the obdurate who are at last compelled to do them homage. The language of mythology could not describe more clearly the endeavours of a plebeian cult to obtain recognition*. The Mahabharata contains hymns to Durga in which she is said to love offerings of flesh and wine*, but it is not Hkely that Saktism or Tantrism — that is a system with special scriptures ' But the Brahmans of isolated localities, like Satara in the Bombay Presidency, are said to be Saktas and the KaficuUyas of S. India are described as a Saktist sect. ^ The law-giver Baudhayana seems to have regarded Ariga and Vafiga with suspicion, i. 1. 13, 14. ' See especiaUy the story of Manasa Devi in Dinesh Chandra Sen (Beng. Lang. and Lit. 257), who says the earUest literary version dates from the twelfth century. But doubtless the story is much older. * Viratap. chap. vi. (not in aU mss.). Bhishmap. chap. xxni. Also in the Harivamsa, ¦vv. 3236 ff. Pargiter considers that the Devi-Mahatmya was probably composed in the fifth or sixth century. Chap. xxi. of the Lotus Siitra contains a speU invoking a goddess under many names. Though this chapter is an addition to the original work, it was translated into Chinese between 265 and 316. 280 HINDUISM [ch. and doctrines — was prevalent before the seventh century a.d. for the Tantras are not mentioned by the Chinese pilgrims and the lexicon Amara Kosha (perhaps c. 500 a.d.) does not recognize the word as a designation of rehgious books. Bana (c. 630) gives more than once in his romances Hsts of sectaries but though he mentions Bhagavatas and Pasupatas, he does not speak of Saktas'. On the other hand Tantrism infected Buddhism soon after this period. The earHer Tibetan translations of the Tantras are attributed to the ninth century. MSS. of the Kubjikamata and other Tantras are said to date from the ninth and even from the seventh century and tradition represents Sankaracarya as having contests with Saktas*. But many Tantras were written in the fifteenth century and even later, for the Yogini Tantra aUudes to the Koch king Bishwa Singh (1515-1540) and the Meru Tantra mentions London and the Enghsh. From the twelfth to the sixteenth century, when Buddhism, itself deeply infected with Tantrism, was disappearing, Saktism was probably the most powerful rehgion in Bengal, but Vishnu ism was gaining strength and after the time of Caitanya proved a formidable rival to it. At the beginning of the iUteenth century we hear that the king of the Ahoms summoned Brahmans to his Court and adopted many Hindu rites and behefs, and from this time onward Saktism was patronized by most of the Assamese Rajas although after 1550 Vishnuism became the religion of the mass of the people. Saktism never inspired any popular or missionary movement, but it was powerful among the aristocracy and instigated persecutions against the Vishnuites. The more respectable Tantras* show considerable resem blance to the later Upanishads such as the Nrisinhatapaniya and Ramatapaniya, which mention Sakti in the sense of creative energy*. Both classes of works treat of magical formulae (mantras) 1 But he does mention the worship of the Divine Mothers. Harshacar. vn. 250 and Kadamb. 134. ^ Hymns to the Devi are also attributed to him but I do not know what evidence there is for his authorship. ' As pointed out elsewhere, though this word is most commonly used of the Sakta scriptures it is not restricted to them and we hear of both Buddhist and Vaishnava Tantras. * The Adhyatma Ramayana is an instance of Saktist ideas in another theo logical setting. It is a Vishnuite work but Sita is made to say that she is Prakriti who does aU the deeds related in the poem, whereas Rama is Purusha, inactive and a witness of her deeds. xxxn] SAKTISM 281 and the construction of mystic diagrams or yantras. This re semblance does not give us much assistance in chronology, for the dates of the later Upanishads are very uncertain, but it shows how the Tantras are connected with other branches of Hindu thought. The distinction between Tantras and Puranas is not always weU-marked. The Bhagavata Purana countenances tantric rites' and the Agni Purana (from chapter xxi onwards) bears a strong resemblance to a Tantra. As a rule the Tantras contain less historical and legendary matter than the Puranas and more directions as to ritual. But whereas the Puranas approve of both Vedic rites and others, the Tantras insist that ceremonies other than those which they prescribe are now useless. They maintain that each age of the world has its own special revelation and that in this age the Tantra-sastra is the only scripture. Thus in the Mahanirvana Tantra Siva says*: "The fool who would foUow other doctrines heedless of mine is as great a sinner as a parricide or the murderer of a Brahman or of a woman.... The Vedic rites and mantras which were efficacious in the first age have ceased to have power in this. They are now as powerless as snakes whose fangs have been drawn and are like dead things." The Kuiarnava Tantra (l. 79 ff.) inveighs against those who think they wUl obtain salvation by Vedic sacrifices or asceticism or reading sacred books, whereas it can be won only by tantric rites. Various Hsts of Tantras are given and it is generaUy admitted that many have been lost. The most complete, but somewhat theoretical enumeration* divides India and the adjoining lands into three regions to each of which sixty-four Tantras are assigned. The best known names are perhaps Mahanirvana*, SaradatUaka^, Yogini, Kuiarnava* and Rudra-Yamala. A Tantra 1 XI. Ui. 47-8; xi. v. 28 and 31. Probably Vishnuite not Saktist Tantras are meant but the Purana distinguishes between Vedic revelation meant for previous sages and tantric revelation meant for the present day. So too KuUiika Bhatta the commentator on Manu who was a BengaU and probably Uved in the fifteenth century says (on Manu n. i.) that Sruti is twofold, Vedic and tantric. Srutisca dvividhd vaidiM tdntrikica, 2 jj_ i^ ' See for fuU list Avalon, Principles of Tantra, pp. Ixv-lxvu. A coUection of thirty-seven Tantras has been pubUshed at Calcutta by Babu RasikMohunChatterjee and a few have been published separately. ' Translated by Avalon, 1913, also by Manmatha Nath Dutt, 1900. ' Analysed in J.A.O.S. xxm. i. 1902. " Edited by Taranatha Vidyaratna, with introduction by A. Avalon, 1917. 282 HINDUISM [ch. is generaUy cast in the form of a dialogue in which Siva instructs his consort but sometimes vice versd. It is said that the former class are correctly described as Agamas and the works where the Sakti addresses Siva as Nigamas'. Some are also caUed Yamalas and Damaras but I have found no definition of the meaning of these words. The Prapaiicasara Tantra* professes to be a revelation from Narayana. Saktism and the Tantras which teach it are generaUy con demned by Hindus of other sects*. It is arguable that this condemnation is unjust, for Hke other forms of Hinduism the Tantras make the Hberation of the soul their object and prescribe a Hfe of religious observances including asceticism and medita tion, after which the adept becomes released even in this Hfe. But however much new tantric Hterature may be made accessible in future, I doubt if impartial criticism will come to any opinion except that Saktism and Tantrism coUect and emphasize what is superficial, trivial and even bad in Indian rehgion, omitting or neglecting its higher sides. If for instance the Mahanirvana Tantra which is a good specimen of these works be compared with Sankara's commentary on the Vedanta Sutras, or the poems of Tulsi Das, it will be seen that it is woefuUy deficient in the excellences of either. But many tantric treatises are chiefly concerned with charms, spells, amulets and other magical methods of obtaining wealth, causing or averting disease and destroying enemies, processes which even if efficacious have nothing to do with the better side of rehgion*. The rehgious Hfe prescribed in the Tantras^ commences with initiation and requires the supervision of the Guru. The object of it is Siddhi or success, the highest form of which is spiritual perfection. Siddhi is produced by Sddhana, or that method of 1 See Avalon, Principles of Tantra, p. Ixi. But these are probably special meanings attached to the words by tantric schools. Nigama is found pretty fre quently, e.g. Manu, iv. 19 and LaUta-vistara, xn. But it is not lUsely that it is used there in this special sense. 2 Edited by Avalon, 1914. 3 Satirical descriptions of Saktism are fairly ancient, e.g. Karpura Manjari, Harvard edition, pp. 25 and 233. * Tantrism has some analogy to the Feng-shui or geomancy of the Chinese. Both take ancient superstitions which seem incompatible with science and systema tize them into pseudo-sciences, remaining bhnd to the fact that the subject-matter is wholly imaginary. 5 For what foUows as for rauch else in this chapter, I am indebted to Avalon's translation of the Mahanirvana Tantra and introduction. xxxn] Saktism 283 training the physical and psychic faculties which reahzes their potentiahties. Tantric training assumes a certain constitution of the universe and the repetition in miniature of this constitu tion in the human body which contains various nervous centres and subtle channels for the passage of energy unknown to vulgar anatomy. Thus the Sakti who pervades the universe is also present in the body as Kundahni, a serpentine coil of energy, and it is part of Sadhana to arouse this energy and make it mount from the lower to the higher centres. Kundahni is also present in sounds and in letters. Hence if different parts of the body are touched to the accompaniment of appropriate mantras (which rite is called nyasa) the various Saktis are made to dweU in the human frame in suitable positions. The Tantras recognize that human beings are not equal and that codes and rituals must vary according to temperament and capacity. Three conditions of men, called the animal, heroic and divine', are often mentioned and are said to characterize three periods of Hfe — youth, manhood and age, or three classes of mankind, non-tantrists, ordinary tantrists, and adepts. These three conditions clearly correspond to the three Gunas. Also men, or rather Hindus, belong to one of seven groups, or stages, according to the rehgious practices which it is best for them to foUow. Saktists apparently demur* to the statement commonly made by Indians as weU as by Europeans that they are divided into two sects the Dakshinacarins, or right-hand worshippers, whose ritual is pubhc and decent, and the Vamacarins who meet to engage in secret but admittedly immoral orgies. But for practical purposes the division is just, although it must not be supposed that Dakshinacarins necessarily condemn the secret worship. They may consider it as good for others but not for themselves. Saktists apparently would prefer to state the matter thus. There are seven stages of rehgion. First come Vedic, Vishnuite and Sivaite worship, aU three inferior, and then Dakshinacara, interpreted as meaning favourable worship, that is favourable to the accompHshment of higher purposes, because the worshipper now begins to understand the nature of Devi, the great goddess. These four kinds of worship are all said to belong to pravritti or active Hfe. The other three, considered to be higher, require a special initiation and belong to nivritti, the ' Pa^u-, vira-, divya-bhava. ^ Avalon, Mahan. Tan. pp. Ixxix, Ixxx. 284 HINDUISM [ch. path of retum in which passion and activity are suppressed'. And here is propounded the doctrine that passion can be destroyed and exhausted by passion*, that is to say that the impulses of eating, drinking and sexual intercourse are best subjugated by indulging them. The fifth stage, in which this method is first adopted, is caUed Vamacara*. In the sixth, or Siddhantacara*, the adept becomes more and more free from passion and prejudice and is finaUy able to enter Kauiacara, the highest stage of aU. A Kaula is one who has passed beyond aU sects and belongs to none, since he has the knowledge of Brah man. "Possessing merely the form of man, he moves about this earth for the salvation of the world and the instruction of men^." These are aspirations common to aU Indian rehgion. The pecuHarity of the Tantras is to suppose that a ritual which is shocking to most Hindus is an indispensable preUminary to their attainment®. Its essential feature is known as pancatattva, the five elements, or pancamakdra the five m's, because they aU begin with that letter, namely, madya, mdmsa, matsya, mudrd, and maithuna, wine, meat, fish, parched grain and copulation. The celebration of this ritual takes place at midnight, and is caUed cakra or circle. The proceedings begin by the devotees seating themselves in a circle and are said to terminate in an indiscriminate orgy. It is only fair to say that some Tantras inveigh against drunkenness and authorize only moderate drinking'. In aU cases it is essential that the wine, fiesh, etc., 1 "The etemal rhythm of Divine Breath is outwards frora spirit to matter and inwards from matter to spirit. Devi as Maya evolves the world. As Mahamaya she recaUs it to herself Each of these movements is divine. Enjoyment and liberation are each her gifts." Avalon, Mahan. Tan. p. cxl. ' Yair eva patanam dravyaih siddhis tair eva codita — Kuiarnava Tantra, v. 48. There is probably something similar in Taoism. See Wieger, Histoire des Croyances religieuses en Chine, p. 409. The Indian Tantrists were aware of the dangers of their system and said it was as difficult as walking on the edge of a sword or holding a tiger. ' Vamacara is said not to mean left-hand worship but woman (vama) worship. This interpretation of Dakshina and Vamacara is probably fanciful. * Sometimes two extra stages Aghora and Yogacara are inserted here. ^ Mahan. Tan. x. 108. A Kaula raay pretend to be a Vaishnava or a Saiva. ' Although the Tantras occasionaUy say that mere ritual is not sufficient for the highest religions, yet indispensabh preliminary is often understood as meaning sure means. Thus the Mahanirvana Tantra (x. 202, Avalon's transl.) says "Those who worship the Kaulas with panca tattva and with heart uplUted, cause the salva tion of their ancestors and themselves attain the highest end." ' But on the other hand some Tantras or tantric treatises recommend crazy abominations. xxxn] SAKTISM 285 should be formally dedicated to the goddess : without this pre- Hminary indulgence in these pleasures is sinful. Indeed it may be said that apart from the ceremonial which they inculcate, the general principles of the Tantras breathe a Hberal and inteUigent spirit. Caste restrictions are minimized: travelUng is permitted. Women are honoured : they can act as teachers : the burning of widows is forbidden' : girl widows may remarry* and the murder of a woman is pecuHarly heinous. Prostitution is denounced. Whereas Christianity is sometimes accused of restricting its higher code to Church and Sundays, the opposite may be said of Tantrism. Outside the temple its morality is exceUent. A work Hke the Mahanirvana Tantra presents a refined form of Saktism modified, so far as may be, in conformity with ordinary Hindu usage*. But other features indubitably connect it with aboriginal cults. For instance there is a legend which relates how the body of the Sakti was cut into pieces and scattered over Assam and Bengal. This story has an uncouth and barbarous air and seems out of place even in Puranic mythology. It recaUs the tales told of Osiris, Orpheus and HaUdan the Black* and may be ultimately traceable to the idea that the dismemberment of a deity or a human representative ensures fertUity. UntU recently the Khonds of Bengal used to hack human victims in pieces as a sacrifice to the Earth Goddess and throw the shreds of flesh on the fields to secure a good harvest^- In Sanskrit Uterature I have not found any authority for the dismemberment of Sati earHer than the Tantras or Upapuranas {e.g. Kahka), but this late appearance does not mean that the legend is late in itseU but merely that it was not countenanced by Sanskrit writers until medieval times. Various reasons for the dismemberment are given and the incident is rather awkwardly tacked on to other stories. One common version relates that when Sati (one of the many forms of Sakti) died of vexation because her husband Siva was insulted by her ^ Mahanir. Tant. x. 79. Bhartra saha kule^ani na dahet kulakaminim. 2 Ib. XI. 67. ' E.g. It does not prescribe human sacrifices and counsels moderation in the use of wine and maithuna. * See Frazer's Adonis, Attis and Osiris, pp. 269-273 for these and other stories of dismemberment. * See Prazer, Golden Bough : Spirits of the Corn, vol. i. 245 and authorities quoted. 286 HINDUISM [ch. father Daksha, Siva took up her corpse and wandered dis tractedly carrying it on his shoulder'. In order to stop this penance Vishnu foUowed him and cut off pieces from the corpse with his quoit until the whole had faUen to earth in fifty-one pieces. The spots where these pieces touched the ground are held sacred and caUed piths. At most of them are shown a rock supposed to represent some portion of the goddess's body and some object caUed a bhairabi, left by Siva as a guardian to protect her and often taking the form of a Hngam. The most important of these piths are Kamakhya near Gauhati, Faljur in the Jaintia Parganas, and Kahghat in Calcutta*. Though the Sakti of Siva is theoreticaUy one, yet since she assumes many forms she becomes in practice many deities or rather she is many deities combined in one or sometimes a sovereign attended by a retinue of simUar female spirits. Among such forms we find the ten Mahavidyas, or personifications of her supernatural knowledge ; the Mahamatris, Matrikas or the Great Mothers, aUied to the aboriginal goddesses aUeady mentioned ; the Nayakas or mistresses ; the Yoginis or sorceresses, and fiends called Dakinis. But the most popular of her mani festations are Durga and KaU. The sects which revere these goddesses are the most important reUgious bodies in Bengal, where they number thirty -five milHon adherents. The Durgapiija is the greatest festival of the year in north-eastern India* and in the temple of Kahghat at Calcutta may be seen the singular spectacle of educated Hindus decapitating goats before the image of KaU. It is a black female figure with gaping mouth and protruded tongue dancing on a prostrate body*, and ' Images representing this are common in Assam. ^ Hsuan Chuang (Watters, vol. i. chap, vn) mentions several sacred places in N.W. India where the Buddha in a previous birth was disraembered or gave his flesh to feed mankind. Can these places have been simUar to the piths of Assam and were the original heroes of the legend deities who were dismembered like Sati and subsequently accommodated to Buddhist theology as Bodhisattvas ? ^ It is an autumnal festival. A special iraage of the goddess is made which is worshipped for nine days and then thrown into the river. For an account of the festival which makes its tantric character very clear see Durga Puja by Pratapa- chandra Ghosha, Calcutta, 1871. * One explanation given is that she was so elated with her victories over giants that she began to dance which shook the Universe. Siva in order to save the world placed himseU beneath her feet and when she saw she was trampUng on her husband, she stopped. But there are other explanations. Another of the strangely barbaric legends which cluster round the Sakti is xxxn] SAKTISM 287 adorned with skuUs and horrid emblems of destruction. Of her four hands two carry a sword and a severed head but the other two are extended to give blessing and protection to her wor shippers. So great is the crowd of enthusiastic suppHants that it is often hard to approach the shrine and the nationahst party in Bengal who clamour for parhamentary institutions are among the goddess's devotees. It is easy to criticize and condemn this worship. Its outward signs are repidsive to Europeans and its inner meaning strange, for even those who pray to the Madonna are startled by the idea that the divine nature is essentiaUy feminine'. Yet this idea has deep roots in the heart of Bengal and with it another idea: the terrors of death, plague and storm are haU but only haU revelations of the goddess-mother who can be smiling and tender as weU. Whatever may be the origin of KaU and of the strange images which represent her, she is now no she-devU who needs to be propitiated, but a reminder that birth and death are twins, that the horrors of the world come from the same source as its grace and beauty and that cheerful acceptance of the deity's terrible manifestations is an essential part of the higher spiritual Hfe*. These ideas are best expressed in the songs of Rama Prasada Sen (1718-1775) which "stiU reign supreme in the vUlages " of Bengal and show that this strange worship has ready a hold on milHons of Indian rustics* The directness and chUdHlce simpHcity of his poems have caused an Indian critic to compare him to Blake. "Though the mother beat the child," he sings, "the child cries mother, mother, and chngs stiU iUustrated by the figure caUed Chinnamastaka. It represents the goddess as carrying her own head which she has just cut off, while from the neck spout fountains of blood which are drunk by her attendants and by the severed head itseff. ' Yet the EngUsh mystic JuUan, the anchoress of Norwich (c. 1400), insists on the motherhood as well as the fatherhood of God. "God is our mother, brother and Saviour." "As verily God is our father, so verUy God is our mother." So too in an inscription fouud at Capua (C.I.N. 3580) Isis is addressed as una quae es omnia. The Power addressed in Swinburne's poems Mater Triumphalis, Hertha, The Pilgrims and Dolores is reaUy a conception very simUar to Sakti. ^ These ideas find frequent expression in the works of Bunkim Chandra Chatter- jee, Dinesh Chandra Sen and Sister Nivedita. ' See Dinesh Chandra Sen, Hist. Beng. Lang, and, Lit. pp. 712-721. Even the iconoclast Devendranath Tagore speaks of the Universal Mother. See Autobiog. p. 240. 288 HINDUISM [ch. tighter to her garment. True, I cannot see thee, yet I am not a lost child. I stiU cry mother, mother." "AU the miseries that I have suffered and am suffering, I know, 0 mother, to be your mercy alone." I must confess that I cannot fully sympathize with this worship, even when it is sung in the hymns of Rama Prasada, but it is clear that he makes it tolerable just because he throws aside all the magic and ritual of the Tantras and deals straight with what are for him elemental and emotional facts. He makes even sceptics feel that he has ready seen God in this strange guise. The chief sanctuary of Saktism is at Kamakhya (or Kamaksha) on a hiU which stands on the banks of the Brahmaputra, about two miles below Gauhati. It is mentioned in the Padma Purana. The temples have been rebuUt several times, and in the eighteenth century were munificently endowed by an Ahom king, and placed under the management of a Brahman from Nadia in Bengal, with reversion to his descendants who bear the title of Parbatiya Gosains. Considerable estates are stUl assigned to their upkeep. There are ten' shrines on the hUl dedicated to various forms of the Sakti. The situation is magnificent, com manding an extensive prospect over the Brahmaputra and the plains on either bank, but none of the buildings are of much architectural merit. The largest and best is the temple dedicated to Kamakhya herself, the goddess of sexual desire. It is of the style usual in northern India, an unUghted shrine surmounted by a dome, and approached by a rather ample vestibule, which is also imperfectly Ughted. An inscription has been preserved recording the restoration of the temple about 1550 but only the present basement dates from that time, most of the super structure being recent. Europeans may not enter but an image of the goddess can be seen from a side door. In the depths of the shrine is said to be a cleft in the rock, adored as the Yoni of Sakti. In front of the temple are two posts to which a goat is tied, and decapitated daily at noon. Below the principal shrine is the temple of Bhairavi. Human sacrifices were offered here in comparatively recent times, and it is not denied that they would be offered now U the law aUowed. Also it is not denied ^ So I was told, but I saw only six, when I visited the place in 1910. xxxn] SAKTISM 289 that the rites of the "five m's" already mentioned are frequently performed in these temples, and that Aghoris may be found in them. The spot attracts a considerable number of pilgrims from Bengal, and a wealthy devotee has buUt a vUla on the hiU and pays visits to it for the purpose of takmg part in the rites. I was informed that the most esteemed scriptures of the sect are the Yogini Tantra, the Mahanirvana Tantra, and the Kaiika Purana. This last work contains a section or chapter on blood', which gives rules for the performance of human sacrifices. It states however that they should not be performed by the first three castes, which is perhaps a way of saying that though they may be performed by non-Aryans under Brahmanic auspices they form no part of the Aryan rehgion. But they are recommended to princes and ministers and should not be performed without the consent of princes. The ritual bears Httle resemblance to the Vedic sacrifices and the essence of the ceremony is the presenta tion to the goddess of the victim's severed head in a vessel of gold, sUver, copper, brass or wood but not of iron. The axe with which the decapitation is to be performed is solemnly conse crated to KaU and the victim is worshipped before immolation. The sacrificer first thinks of Brahma and the other gods as being present in the victim's body, and then prays to him directly as being aU the gods in one. "When this has been done" says Siva, who is represented as himseU revealing these rules, "the victim is even as myseU." This identification of the human victim with the god has many analogies elsewhere, particularly among the Khonds*. It is remarkable that this barbarous and immoral worship, though looked at askance except in its own holy places, is by no means confined to the lower castes. A series of apologies composed in exceUent Enghsh (but sometimes anonymous) attest the sympathy of the educated. So far as theology and metaphysics are concemed, these defences are plausible. The Sakti is identified with Prakriti or with the Maya of the Advaita phUosophy and defined as the energy, coexistent with Brahman, which creates the world. But attempts to palHate the ceremonial, such as the argument that it is a consecration and Hmitation of the appetites because they may be gratified only in the service ^ Rudhkadhyaya. Translated in As. Researches, v. 1798, pp. 371-391. * See Frazer, op. cit. p. 246. B. n. 19 290 HINDUISM [ch. xxxn of the goddess, are not convincing. Nor do the Saktas, when able to profess their faith openly, deny the nature of their rites or the importance attached to them. An oft-quoted tantric verse represents Siva as saying Maithunena mahdyogi mama tulyo na samiayah. And for practical purposes that is the gist of Saktist teaching. The temples of Kamakhya leave a disagreeable impression— an impression of dark evU haunts of lust and bloodshed, akin to madness and unreUeved by any grace or vigour of art. For there is no attempt in them to represent the terrible or voluptu ous aspects of Hinduism, such as find expression in sculpture elsewhere. AU the buildings, and especiaUy the modem temple of KaU, which was in process of construction when I saw the place, testify to the atrophy and paralysis produced by erotic forms of rehgion in the artistic and inteUectual spheres, a phenomenon which finds another sad iUustration in quite different theological surroundings among the VaUabhacarya sect at Gokul near Muttra. It would be a poor service to India to paUiate the evUs and extravagances of Saktism, but stiU it must be made clear that it is not a mere survival of barbaric practices. The writers of the Tantras are good Hindus and declare that their object is to teach Hberation and union with the Supreme Spirit. The ecstasies induced by tantric rites produce this here in a pre- Hminary form to be made perfect in the Hberated soul. This is not the craze of a few hysterical devotees, but the faith of mUHons among whom many are weU educated. In some aspects Saktism is simUar to the erotic Vishnuite sects, but there is Httle real analogy in their ways of thinking. For the essence of Vishnuism is passionate devotion and seff-surrender to a deity and this idea is not prominent in the Tantras. The strange inconsistencies of Saktism are of the kind which are character istic of Hinduism as a whole, but the contrasts are more violent and the monstrosities more conspicuous than elsewhere; wUd legends and metaphysics are mixed together, and the peace that passes aU understanding is to be obtained by orgies and offerings of blood. CHAPTEE XXXIII HINDU PHILOSOPHY 1 PHiLOSOPHy is more closely connected with reUgion in India than in Europe. It is not a dispassionate scientific investigation but a practical religious quest. Even the Nyaya school, which is concemed chiefly with formal logic, promises that by the removal of false knowledge it can emancipate the soul and give the bhss of salvation. Nor are the expressions system or school of phUosophy, commonly used to render darsana, altogether happy. The word is derived from the root driS, to see, and means a way of looking at things. As such a way of looking is supposed to be both comprehensive and orderly, it is more or less what we caU phUosophical, but the points of view are so special and so various that the result is not always what we caU a phUosophical system. Madhava's' Hst of Darsanas includes Buddhism and Jainism, which are commonly regarded as separate reUgions, as weU as the Pasupata and Saiva, which are sects of Hinduism. The Darsana of Jaimini is merely a discussion of general questions relating to sacriflces : the Nyaya Darsana examines logic and rhetoric: the Paniniya Darsana treats of grammar and the nature of language, but claims that it ought to be studied ' ' as the means for attaining the chief end of man*." Six of the Darsanas have received special prominence and are often caUed the six Orthodox Schools. They are the Nyaya and Vaiseshika, Sankhya and Yoga, Piirva and Uttara Mmarnsa, In the Sarva-dar^ana-sangraha, the best known compendium of Indian phUosophy. ^ J. C. Chatterji's definition of Indian phUosophy (in his Indian Realism, p. 1 ) is interestmg. "By Hindu philosophy I mean that branch of the ancient learning of the Hindus which demonstrates by reasoning propositions with regard to "(a) what a man ought to do iu order to gain true happiness... or (6) what he "ought to reaUze by direct experience in order to be radicaUy and absolutely freed "from suffering and to be absolutely independent, such propositions being already "given and lines of reasoning in their support being established by duly qualified "authorities." 19—2 292 HINDUISM [CH. or Vedanta. The rest are either comparatively unimportant or are more conveniently treated of as rehgious sects. The six placed on the select Hst are sufficiently misceUaneous and one wonders what principle of classification can have brought them together. The first two have Httle connection with reUgion, though they put forward the emancipation of the soul as their object, and I have no space to discuss them. They are however important as showing that reahsm has a place in Indian thought in spite of its marked tendency to idealism'. They are concemed chiefiy with an examination of human faculties and the objects of knowledge, and are related to one another. The special doctrine of the Vaiseshika is the theory of atoms ascribed to Kanada. It teaches that matter consists of atoms (anu) which are etemal in themselves though all combinations of them are Hable to decompose. The Sankhya and Yoga are also related and represent two aspects of the same system which is of great antiquity and aUied to Buddhism and Jainism. The two Mimamsas are consecutive expositions of the teaching scattered throughout the Vedic texts respecting ceremonial and the knowledge of God respectively. The second Mimamsa, commonly caUed the Vedanta, is by far the more interesting and important. The common feature in these six systems which constitutes their orthodoxy is that they aU admit the authority of the Veda. This impUes more than our phrases revelation or inspiration of the Bible. Most of the Darsanas attach importance to the pramdnas, sources or standards of knowledge. They are variously enumerated, but one of the oldest definitions makes them three : perception (pratyaksha), inference (anumana) and scripture (sabda). The Veda is thus formaUy acknowledged to have the same authority as the evidence of the senses. With this is generaUy coupled the doctrine that it is etemal. It was not composed by human authors, but is a body of sound existing from eternity as part of Brahman and breathed out by him when he causes the whole creation to evolve at the beginning of a world period. The reputed authors are simply those who have, in Indian language, seen portions of this seU-existent teaching. This doctrine sounds more reasonable U restated in the form that words are the expression of thought, and that U thought is the etemal essence of both Brahman and the soul, * See Chatterji's work above cited. xxxm] HINDU PHILOSOPHY 293 a simUar eternity may attach to words. Some such idea is the origin of the Christian doctrine of the Logos, and in many rehgions we find such notions as that words have a creative efficacy', or that he who knows the name of a thing has power over it. Among Mohammedans the Koran is supposed to be not merely an inspired composition but a pre-existing book, revealed to Mohammed piecemeal. It is curious that both the sacred texts — the Veda and the Koran — to which this supernatural position is ascribed should be coUections of obviously human, incongruous, and often insignificant documents connected with particular occasions, and in no way suggesting or claiming that they are anterior to the ordinary Hfe of man on eartU. It is stiU more extraordinary that systems of philosophy should profess to base themselves on such works. But in reahty Hindu metaphysicians are not more bound by the past than their coUeagues in other lands. They do not take scripture and ask what it means, but evolve their own systems and state that they are in accordance with it. Sometimes scripture is ignored in the details of argument. More often the metaphysician writes a commentary on it and boldly proves that it supports his views, though its apparent meaning may be hostUe. It is clear that many philosophic commentaries have been written not because the authors reaUy drew their inspiration from the Upanishads or Bhagavad-gita but because they dared not neglect such important texts. AU the Vedantist schools labour to prove that they are in harmony not only with the Upanishads but with the Brahma-siitras. The philosophers of the Sankhya are more detached from Hterature but though they ignore the existence of the deity, they acknow ledge the Veda as a source of knowledge. Their recognition, however, has the air of a concession to Brahmanic sentiment. Isolated theories of the Sarikhya can be supported by isolated passages of the Upanishads, but no impartial critic can maintain that the general doctrines of the two are compatible. That the Brahmans should have been wiUing to admit the Sahkhya as a possible form of orthodoxy is a testimony both to its import ance and to their HberaUty. ^ It is this idea which disposes educated Hindus to beUeve in the magical or sacramental power of mystic syUables and letters, though the use of such speUs seems to Europeans incredible foUy. 294 HINDUISM [cH. It is remarkable that the test of orthodoxy should have been the acceptance of the authority of the Veda and not a confession of some sort of theism. But on this the Brahmans did not insist. The Vedanta is truly and intensely pantheistic or theistic, but in the other philosophies the Supreme Being is either ehminated or plays a smaU part. Thus while works which seem to be merely scientific treatises (Hke the Nyaya) set before themselves a rehgious object, other treatises, seemingly rehgious in scope, ignore the deity. There is a strong and ancient Hne of thought in India which, basing itseU on the doctrine of Karma, or the inevitable consequences of the deed once done, lays stress on the efficacy of ceremonies or of asceticism or of knowledge without reference to a Supreme Being because, if he exists, he does not interfere with the workings of Karma, or with the power of knowledge to release from them. Even the Vedanta, although in a way the quintessence of Indian orthodoxy, is not a scholastic philosophy designed to support recognized dogma and ritual. It is rather the orthodox method of soaring above these things. It contemplates from a higher level the Hfe of religious observances (which is the subject of the Purva Mimamsa) and recognizes its value as a prehminary, but yet rejects it as inadequate. The Sannyasi or adept foUows no caste observances, performs no sacrifices, reads no scriptures. His religion is to realize in meditation the true nature, and it may be the identity, of the soul and God. Good works are of no more importance for him than rites, though he does weU to employ his time in teaching. But Karma has ceased to exist for him: "the acts of a Yogi are neither black nor white," they have no moral quaUty nor consequences. This is dangerous language and the doctrine has sometimes been abused. But the point of the teaching is not that a Sannyasi may do what he Hkes but that he is perfectly emancipated from material bondage. Most men are bound by their deeds ; every new act brings consequences which attach the doer to the world of transmigration and create for him new existences. But the deeds of the man who is reaUy free have no such trammeUing effects, for they are not prompted by desire nor directed to an object. But since to become free he must have suppressed aU desire, it is hardly conceivable that he should do anything which could be caUed a sin. But this conviction that the task of the xxxm] HINDU PHILOSOPHY 295 sage is not to perfect any form of good conduct but to rise above both good and evil, imparts to the Darsanas and even to the Upanishads a singularly non-ethical and detached tone. The Yogi does no harm but he has less benevolence and active sympathy than the Buddhist monk. It was a feeUng that such an attitude has its dangers and is only for the few who have fought their way to the heights where it can safely be adopted, that led the Brahmans in aU ages to lay stress on the house holder's Hfe as the proper preparation for a philosophic old age. Despite utterances to the contrary, they never as a body approved the ideal of a Hfe entirely devoted to asceticism and not occupied with social duties during one period. The extra ordinary ease with which the higher phases of Indian thought shake off all formaUties, social, reUgious and ethical, was counterbalanced by the multitudinous regulations devised to keep the majority in a law-abiding Hfe. None of the six Darsanas concem themselves with ethics. The more important deal with the transcendental progress of sages who have avowedly abandoned the Hfe of works, and even those which treat of that lower Hfe are occupied with ritual and logic rather than with anything which can be termed moral science. We must not infer that Indian Hterature is altogether unmoral. The doctrine of Karma is intensely ethical and ethical discussions are more prominent in the Epics than in Homer, besides being the subject of much gnomic and didactic poetry. But there is no mistaking the fact that the Hindu seeks for salvation by knowledge. He feels the power of deeds, but it is only the lower happiness which Hes tn doing good works and enjoying their fruits. The higher bUss consists in being entirely free from the bondage of deeds and Karma. AU the Darsanas have as a common principle this idea of Karma with the attendant doctrines that rebirth is a consequence of action and that salvation is an escape from rebirth. They all treat more or less of the sources and standards of knowledge, and aU recognize the Veda as one of them. There is not much more that can be said of them aU in common, for the Vedanta ignores matter and the Sankhya ignores God, but they aU share a conviction which presents difficulties to Europeans. It is that the state in which the mind ceases to think discursively and is concentrated on itseU is not only desirable but the summum 296 HINDUISM [ch. bonum. The European is incHned to say that such a state is distinguished from non-existence only by not being permanent. But the Hindu will have none of this. He holds that mind and thought are material though composed of the subtlest matter, and that when thought ceases, the immaterial soul (purusha or atman) far from being practicaUy non-existent is more truly existent than before and enjoys untroubled its own existence and its own nature. Of the three most important systems, the Sankhya, Yoga and Vedanta, the first and last are on most points opposed: both are ancient, but perhaps the products of different inteUectual centres. In one sense the Yoga may be described as a theistic modification of the Sankhya: from another and perhaps juster point of view it appears rather as a very ancient science of asceticism and contemplation, susceptible of combination with various metaphysical theories. We may consider first of aU the Sahkhya'. Tradition ascribes its invention to Kapila, but he is a mere name unconnected with any date or other circumstance. It is probable that the principal ideas of the Saiikhya germinated several centuries before our era but we have no evidence whatever as to when they were first formulated in Siitras. The name was current as the designa tion of a philosophical system fairly early* but the accepted text-books are aU late. The most respected is the Sankhya- pravacana*, attributed to KapUa but generaUy assigned by European critics to the fourteenth century a.d. Considerably more ancient, but stUl clearly a metrical epitome of a system aUeady existing, is the Sankhya-Karika, a poem of seventy verses which was translated into Chinese about 560 a.d. and may be a few centuries older. Max MiiUer regarded the Tattva- samasa, a short tract consisting chiefly of an enumeration of ' See especially Garbe, Die Saiikhya Philosophic, 1894; and Keith, The Sdmkhya System, 1919, which however reached me too late for me to make any use of it. ^ E.g. in the Bhagavad-glta and Svetasvatara Upanishads. According to tra dition KapUa taught Asuri and he, Pancasikha, who made the system celebrated. Garbe thinks PanoaSikha may be assigned to the first century a.d. ' This appears to be the real title of the Siitras edited and translated by BaUantyne as "The Sankhya Aphorisms of KapUa." xxxm] HINDU PHILOSOPHY 297 topics, as the most ancient Sankhya formulary, but the opinion of scholars as to its age is not unanimous. The name Sankhya is best interpreted as signifying enumeration in aUusion to the predUection of the school for numbered lists, a predilection equaUy noticeable in early Buddhism. The object of the system set forth in these works is strictly practical. In the flrst words of the Sankhya-pravacana, the complete cessation of suffering is the end of man, and the Sankhya is devised to enable him to attain it. Another formula divides the contents of the Sankhya into four topics — {a) that from which man must liberate himseU, or suffering, (6) Hbera tion, or the cessation of suffering, (c) the cause of suffering, or the failure to discriminate between the soul and matter, {d) the means of Hberation, or discriminating knowledge. This division obviously resembles the four Truths of Buddhism. The object proposed is the same and the method analogous, though not identical, for Buddhism speaks as a rehgion and lays greater stress on conduct. The theory of the Saiddiya, briefly stated, is this. There exist, uncreated and from aU eternity, on the one side matter and on the other individual souls. The world, as we know it, is due entirely to the evolution of matter. Suffering is the result of souls being in bondage to matter, but this bondage does not affect the nature of the soul and in one sense is not real, for when souls acquire discriminating knowledge and see that they are not matter, then the bondage ceases and they attain to etemal peace. The system is thus founded on duaUsm, the eternal antithesis between matter and soul. Many of its detaUs are comprised in the simple enumeration of the twenty-five Tattvas or principles' as given in the Tattva-samasa and other works. Of these, one is Purusha, the soul or seU, which is neither produced nor pro ductive, and the other twenty-four are aU modifications of Prakriti or matter, which is unproduced but productive. Prakriti means the original ground form of external existence (as distinguished from Vikriti, modified form). It is uncreated and indestructible, but it has a tendency to variation or evolu- ^ Or topics. It is difficult to find any one English word which covers the twenty- five tattvas, for they include both general aud special ideas, mind and raatter on the one hand; special organs on the other. 298 HINDUISM [ch. tion. The Sankhya holds in the strictest sense that ex nihil fit. Substance can only be produced from substance and properly speaking there is no such thing as origination but only manUestation. Causahty is regarded solely from the point of view of material causes, that is to say the cause of a pot is clay and not the action of the potter. Thus the effect or product is nothing else than the cause in another shape : production is only manUestation and destruction is the resolution of a product into its cause. Instead of holding like the Buddhists that there is no such thing as existence but only becoming, the Sankhya rather affirms that there is nothing but successive manifestations of real existence. If clay is made into a pot and the pot is then broken and ground into clay again, the essential fact is not that a pot has come into existence and disappeared but that the clay continuously existing has undergone certain changes. The tendency to evolution inherent in matter is due to the three gunas. They are sattva, explained as goodness and happi ness ; rajas, as passion and movement ; and tamas, as darkness, heaviness and ignorance. The word Guna is not easy to translate, for it seems to mean more than quahty or mode and to signify the constituents of matter. Hence one cannot help feeHng that the whole theory is an attempt to explain the unity and diversity of matter by a phrase, but aU Hinduism is permeated by this phrase and theory. When the three gunas are in equdibrium then matter — Prakriti — is quiescent, undifferentiated and un- mardfested. But as soon as the equilibrium is disturbed and one of the gunas becomes preponderant, then the process of differentiation and manifestation begins. The disturbance of equihbrium is due to the action of the individual Purushas or soids on Prakriti, but this action is mechanical and due to proximity not to the voUtion of the souls and may be compared to the attraction of a magnet for iron'. Thus at the beginning of the evolutionary process we have quiescent matter in equih brium : over against this are souls innumerable, equaUy quiescent but exerting on matter a mechanical force. This upsets the equihbrium and creates a movement which takes at first the form of development and later of decay and coUapse. Then matter returns to its quiescent state to be again excited by the Purushas and commence its world-making evolution anew. The 1 Sankh. Pravao. i. 96. xxxm] HINDU PHILOSOPHY 299 doctrine that evolution, dissolution and quiescence succeed one another periodicaUy is an integral part of the Saukhya'. The unmodified Prakriti stands first on the Hst of twenty -five principles. When evolution begins it produces first Buddhi or inteUect, secondly Ahamkara, which is perhaps best rendered by individuahty, and next the five Tanmatras or subtle elements. Buddhi, though meaning inteUect, is used rather in the sense of ascertaining or perception. It is the faculty by which we dis tinguish objects and perceive what they are. It differs also from our conception of inteUect in being, like Ahamkara and aU the subsequent developments of Prakriti, material, and must not be confused with the immaterial Purusha or soul. It is in fact the organ of thought, not in the sense of the brain or anything tangible, but a subtle substratum of aU mental processes. But in what sense is it possible to say that this Buddhi exists apart from individuals, who have not come into being at this stage of cosmic evolution? This difficulty is not met by talking, as some commentators do, of cosmic as well as individual Buddhi, for even ff aU Prakriti is illuminated by Buddhi at this stage it is difficult to see what result can occur. To make the process of development coherent we must think of it not as a series of ChronologicaUy successive stages but rather as a logicaUy con nected series and an analysis of completely evolved beings, just as we might say that bones are covered with flesh and flesh with skin, without affirming that the bones have a separate and prior existence. Ahamkara, which is, Uke Buddhi, strictly speaking a physical organ, means Ego-maker and denotes the sense of personahty and individuahty, almost the will. In the language of Indian philosophy it is the delusion or misconception which makes the soul imagine itseU a personal agent and think, I see, / hear, I slay, / am slain, whereas the soul is ready incapable of action and the acts are those of Prakriti. The five subtle elements are the essences of sound, touch, colour, savour and odour conceived as physical principles, imperceptible to ordinary beings, though gods and Yogis can perceive them. The name Tanmatra which signifies that only indicates that they are concemed exclusively with one sense. ^ Garbe, Die Sdhkhya Philosophie, p. 222. He considers that it spread thence to other schools. This involves the assumption that the Sankhya is prior to Buddhism and Jainism. 300 HINDUISM [ch. Thus whereas the gross elements, such as earth, appeal to more than one sense and can be seen, felt and smelt, the subtle element of sound is restricted to the sense of hearing. It exists in aU things audible but has nothing to do with their tangibUity or visibiUty. There remain sixteen further modifications to make up the fuU Hst of twenty-four. They are the five organs of sense', the five organs of action*, Manas or mind, regarded as a sixth and central sense, and also as the seat of wiU, and the five gross elements — earth, water, Hght, air and ether. The Sankhya dis tinguishes between the gross and the subtle body. The latter, caUed Hngasarira, is defined in more than one way, but it is expressly stated in the Karikas* that it is composed of "Buddhi and the rest, down to the subtle elements." It practicaUy corresponds to what we caU the soul, though totaUy distinct from Purusha or soul in the Sankhya sense. It constitutes the character and essential being of a person. It is the part which transmigrates from one gross body to another, and is responsible for the acts committed in each existence. Its union with a gross body constitutes birth, its departure death. Except in the case of those who attain emancipation, its existence and transmigra tion last for a whole world-period at the end of which come quiescence and equihbrium. In it are imprinted the Samskaras*, the predispositions which pass on from one existence to another and are latent in the new-bom mind Hke seeds in a field. By foUowing the evolution of matter we have now accounted for inteUect, individuaUty, the senses, the moral character, wiU, and a principle which survives death and transmigrates. It might therefore be supposed that we have exhaustively analysed the constitution of a human being. But that is not the view of the sankhya. The evolution of Buddhi, Aharnkara, the subtle body and the gross- body is a physical process and the result is also physical, though parts of it are of so fine a substance that ordinary senses cannot perceive them. This physical organism becomes a Uving being (which term includes gods and animals) when it is connected with a soul (purusha) and consciousness depends on this connection, for neither is matter when isolated conscious, nor is the soul, at least not in our sense of the word. ^ Ears, skin, eyes, tongue and nose. " Voice, hands, feet, organs of excretion and generation. » Verse 40. » Cf. the Buddhist Sankharas. xxxm] HINDU PHILOSOPHY 301 Though the soul is neither the Hfe which ends at death (for that is the gross body) nor yet the Hfe which passes from existence to existence (for that is the subtle body) yet it is the vitahzing element which renders Ufe possible. The sankhya Hke Jainism regards souls as innumerable and distinct from one another. The word Purusha must have originaUy referred to the manikin supposed to inhabit the body, and there is some reason to think that the earhest teachers of the sahkhya held that it was infinitely small. But in the existing text-books it is described as infinitely large. It is immaterial and without beginning, end, parts, dimensions, or quahties, incapable of change, motion, or action. These definitions may be partly due to the influence of the Vedanta and, though we know Httle about the historical development of the Saiikhya, there are traces of a compromise between the old teaching of a soul held in bondage and struggling for release and later conceptions of a soul which, being infinite and passionless, hardly seems capable of submitting to bondage. Though the soul cannot be said to transmigrate, to act, or to suffer, stiU through consciousness it makes the suffering of the world felt and though in its essence it remains etemaUy unchanged and unaffected, yet it experiences the reflection of the suffering which goes on. Just as a crystal (to use the Indian simUe) aUows a red flower to be seen through it and remains unchanged, although it seems to become red, so does the soul remain unchanged by sorrow or joy, although the Ulusion that it suffers or rejoices may be present in the consciousness. The task of the soul is to free itseU .from iUusion, and thus from bondage. For strictly speaking the bondage does not exist : it is caused by want of discrimination. Like the Vedanta, the Sahkhya regards aU this troubled Ufe as bemg, so far as the soul is concemed, mere illusion. But while the Vedanta bids the soul know its identity with Brahman, the Sankhya bids it isolate itseU and know that the acts and feelings which seem to be its own have ready nothing to do with it. They are for the soul nothing but a spectacle or play originating in its connection with Prakriti, and it is actuaUy said', "Wherefore no soul is bound, or is Hberated or transmigrates. It is Prakriti, which has many bodily forms, which is bound, Hberated and trans- 1 Sankh. Kar. 62. 302 HINDUISM [oh. migrates." It is in Buddhi or inteUect, which is a manifestation of Prakriti, that the knowledge of the difference between the soul and Prakriti must arise. Thus though the Sankhya reposes on a fundamental duahsm, it is not the dualism of good and evU. Soul and matter differ not because the first is good and the second bad, but because the first is unchangeable and the second constantly changing. Matter is often personified as a woman. Her motives are unselfish and she works for the Hberation of the soul. "As a dancer after showing herseU on the stage ceases to dance, so does Prakriti cease when she has made herseU manifest to the soul." That is to say, when a soul once understands that it is distinct from the material world, that world ceases to exist for that particular soul, though of course the play continues for others. "Generous Prakriti, endowed with Gunas, causes by manifold means without benefit to herseU, the benefit of the soul, which is devoid of Gunas and makes no retum'." The condition of the Hberated soul, corre sponding to the moksha and nirvdi^a of other systems, is described as Kaivalya, that is, complete separation from the material world, but, as among Buddhists and Vedantists, he who has learnt the truth is Hberated even before death, and can teach others. He goes on Hving, just as the wheel continues to revolve for some time after the potter has ceased to tum it. After death, complete Hberation without the possibihty of re-birth is attained. The sankhya manuals do not dweU further on the character of this Uberation: we only know that the etemal soul is then completely isolated and aloof from aU suffering and material things. Liberation is compared to profound sleep, the difference being that in dreamless sleep there is a seed, that is, the possi bihty of retum to ordinary Ufe, whereas when Hberation is once attained there is no such return. Both in its account of the world process and in its scheme of salvation the Sarikhya ignores theism in the same way as did the Buddha. Indeed the text-books go beyond this and practicaUy deny the existence of a personal supreme deity. We are told* that the existence of God cannot be proved, for whatever exists must be either bound or free and God can be neither. We cannot think of him as bound and yet he cannot be free Hke an emanci pated soul, for freedom impHes the absence of desire and hence 1 Sankh. Kar. 59-61. " Sankh. Pravac. i. 92-95. xxxm] HINDU PHILOSOPHY 303 of the impulse to create. Similarly' the consequences of good and evil deeds are due to Karma and not to the government of God. Such a ruler is inconceivable, for if he governs the world according to the action of Karma his existence is superfluous, and ff he is affected by selfish motives or desire, then he cannot be free. It is true that these passages speak of there being no proof of God's existence and hence commentators both Indian and European who shrink from atheism represent the Saiikhya as suspending judgment. But U a republican constitution duly describes the President and other authorities in whom the powers of govemment are vested, can we argue that it is not un- monarchical because it does not expressly say there is no king ? In the sahkhya there is no more place for a deity than for a king in a repubhcan constitution. Moreover, the Siitras en deavour to prove that the idea of God is inconceivable and seU-contradictory and some commentaries speak plainly on this subject*. Thus the Sankhya-tattva-kaumudi commenting on Karika 57 argues that the world cannot have been created by God, whether we suppose him to have been impeded by selfish ness or kindness. For if God is perfect he can have no need to create a world. And U his motive is kindness, is it reasonable to caU into existence beings who while non-existent had no suffering, simply in order to show kindness in reheving them from suffering ? A benevolent deity ought to create only happy creatures, not a mixed world like the one we see*. Arguments like this were not condemned by the Brahmans so strongly as we should expect, but they did not Hke them and though they did not excommunicate the Sahkhya in the same way as Buddhism, they greatly preferred a theistic variety of it caUed Yoga. The Yoga and Sankhya are mentioned together in the Svetasvatara Upanishad*, and the Bhagavad-gita^ says that he sees truly who sees them as one. The difference Hes in treatment ' Sankh. Pravac. v. 2-12. ' Thus Sankh. Pravac. v. 46, says Tatkartufc purushasyabhavat ana the com mentary explains fevara-pratishedhad iti feshah "supply the words, because we deny that there is a supreme God." ' Nevertheless the commentator Vijnana-Bhikshu (c. 1500) tries to explain away this atheisra and to reconcUe the Sankhya with the Vedanta. See Garbe's preface to his edition of the Sankhya-pravacana-bhashya. ' VI. 13. 6 y. 5. 304 HINDUISM [CH. rather than in substance. Whereas the Sankhya is mainly theoretical, the principal topic of the Yoga is the cultivation of that frame of mind which leads to emancipation and the methods and exercises proper to this end. Further, the Yoga recognizes a deity. This distinction may seem of capital importance but the god of the Yoga (caUed Isvara or the Lord) is not its founda tion and essence as Brahman is of the Vedanta'. Devotion to God is recognized as one among other methods for attaining emancipation and if this particular procedure, which is men tioned in relatively few passages, were omitted, the rest of the system woidd be unaffected. It is therefore probable that the theistic portions of the Yoga are an addition made under Brahmanic influence. But taking the existing Siitras of the two phUosophies, together with their commentaries, it may be said that the Yoga impHes most of the Sankhya theory and the Sankhya most of the Yoga practice, for though it does not go into details it prescribes meditation which is to be perfected by regulating the breathing and by adopting certain postures. I have aUeady spoken of the methods and disciphne prescribed by the Yoga and need not dweU further on the topic now. That Buddhism has some connection with the Sankhya and Yoga has often been noticed*. Some of the ideas found in the sankhya and some of the practices prescribed by the Yoga are clearly anterior to Gotama and may have contributed to his mental development, but circumspection is necessary in the use of words Hke Yoga, Sankhya and Vedanta. If we take them to mean the doctrinal systems contained in certain siitras, they are clearly aU later than Buddhism. But if we assume, as we may safely do, that the doctrine is much older than the manuals in which we now study it, we must also remember that when we leave the texts we are not justified in thinking of a system but merely of a Hne of thought. In this sense it is clear that many ideas of the Sankhya appear among the Jains, but the Jains know nothing of the evolution of matter described by the sankhya manuals and think of the relation of the soul to matter ^ Isvara is apparently a purusha like others but greater in glory and untouched by human infirmities. Yoga sutras, i. 24-26. ' It is a singular fact that both the Sankhya-karika-bhashya and a treatise on the Vaifeshika phUosophy are included in the Chinese Tripitaka (Nanjio, Cat. Nos. 1300 and 1295). A warning is however added that they are not "the law of the Buddha." xxxm] HINDU PHILOSOPHY 305 in a more materialistic way. The notion of the separate etemal soul was the object of the Buddha's persistent polemics and was apparently a popular doctrine when he began preaching. The ascetic and meditative exercises prescribed by the Yoga were also known before his time and the Pitakas do not hide the fact that he received instruction from two Yogis. But though he was acquainted with the theories and practices which grew into the Yoga and Sahkhya, he did not found his religion on them for he rejected the idea of a soul which has to be dehvered and did not make salvation dependent on the attainment of trances. If there was in his time a systematic Sankhya philosophy explaining the nature of suffering and the way of release, it is strange that the Pitakas contain no criticism of it, for though to us who see these ancient sects in perspective the resemblance of Buddhism to the Saiikhya is clear, there can be Httle doubt that the Buddha would have regarded it as a most erroneous heresy, because it proposes to attain the same objects as his own teaching but by different methods. Sankhya ideas are not found in the oldest Upanishads, but they appear (though not in a connected form) in those of the second stratum, such as the Svetasvatara and Katha. It there fore seems probable, though not proven, that the origin of these ideas is to be sought not in the early Brahmanic schools but in the inteUectual atmosphere non-theistic, non-sacerdotal, but audaciously speculative which prevailed in the central and eastern part of northern India in the sixth century b.c. The sahkhya recognizes no merit in sacrifices or indeed in good works of any kind, even as a prehminary disciphne, and in many details is un-Brahmanic. Unlike the Vedanta Sutras, it does not exclude Sudras from higher studies, but states that there are eight classes of gods and five of animals but only one of men. A teacher must have himseU attained emancipation, but there is no provision that he must be a Brahman. Perhaps the fables and parables which form the basis of the fourth book of the sankhya Siitras point to some more popular form of instruc tion similar to the discourses of the Buddha. We may suppose that this ancient un-Brahmanic school took shape in several sects, especiaUy Jainism and Buddhism, and used the Yoga discipline. But the value and efficacy of that discipUne were admitted almost universaUy and several centuries later it was E. n. 20 306 HINDUISM [ch. formulated in the Sutras which bear the name of Pataiijah in a shape acceptable to Brahmans, not to Buddhists. If, as some scholars think, the Yoga sutras are not earlier than 450 a.d.' it seems probable that it was Buddhism which stimulated the Brahmans to codUy the principles and practice of Yoga, for the Yogacara school of Buddhism arose before the fifth century. The sankhya is perhaps a somewhat similar brahmanization of the purely speculative ideas which may have prevailed in Magadha and Kosala*. Though these districts were not strong holds of Brahmanism, yet it is clear from the Pitakas that they contained a considerable Brahman population who must have been influenced by the ideas current around them but also must have wished to keep in touch with other Brahmans. The sahkhya of our manuals represents such an attempt at conciha- tion. It is an elaboration in a different shape of some of the ideas out of which Buddhism sprung but in its later history it is connected with Brahmanism rather than Buddhism. When it is set forth in Sutras in a succinct and isolated form, its divergence from ordinary Brahmanic thought is striking and in this form it does not seem to have ever been influential and now is professed by only a few Pandits, but, when combined in a Hterary and eclectic spirit with other ideas which may be incompatible with it in strict logic, it has been a mighty influence in Indian rehgion, orthodox as weU as unorthodox. Such con ceptions as Prakriti and the Gunas colour most of the post- Vedic rehgious Hterature. Their working may be plainly traced in the Mahabharata, Manu and the Puranas*, and the Tantras identify with Prakriti the goddesses whose worship they teach. The unethical character of the Sankhya enabled it to form the strangest aUiances with aboriginal beUefs. 1 See Jacobi, J.A.O.S. Dec. 1910, p. 24. But if Vasubandhu lived about 280-360, as is now generaUy beUeved, aUusions to the Yogacara school in the Yoga siitras do not oblige us to place the siitras much later than 300 a.d. since the Yogacara was founded by Asanga, the brother of Vasubandhu. 2 I find it hard to accept Deussen's view (Philosophy of the Upanishads, chap, x) that the Sankhya has grown out of the Vedanta. = See e.g. Vishnu Purana, i. chaps. 2, 4, 5. The Bhagavad-gita, though almost the New Testaraent of Vedantists, uses the words Sankhya and Yoga in several passages as meaning speculative truth and the reUgious Ufe and is concemed to show that they are the same. See n. 39; m. 3; v. 4, 5. xxxm] HINDU PHILOSOPHY 307 Unhke the Sankhya, the Vedanta is seen in its most influential and perhaps most advantageous aspect when stated in its most abstract form. We need not enquire into its place of origin for it is clearly the final inteUectual product of the schools which produced the Upanishads and the Hterature which preceded them, and though it may be difficult to say at what point we are justified in applying the name Vedanta to growing Brah manic thought, the growth is continuous. The name means simply End of the Veda. In its ideas the Vedanta shows great breadth and freedom, yet it respects the prejudices and pro prieties of Brahmanism. It teaches that God is all things, but interdicts this knowledge to the lower castes: it treats rites as a merely prehminary discipline, but it does not deny their value for certain states of Hfe. The Vedanta is the boldest and the most characteristic form of Indian thought. For Asia, and perhaps for the world at large. Buddhism is more important but on Indian soil it has been vanquished by the Vedanta, especiaUy that form of it known as the Advaita. In aU ages the main idea of this philo sophy has been the same and may be summed up in the formula that the soul is God and that God is everything. If this formula is not completely accurate' — and a sentence which both trans lates and epitomizes ahen metaphysics can hardly aspire to complete accuracy — the error Hes in the fact to which I have caUed attention elsewhere that our words, God and soul, do not cover quite the same ground as the Indian words which they are used to translate. Many scholars, both Indian and European, wiU demur to the high place here assigned to the Advaita philosophy. I am far from claiming that the doctrine of Sankara is either primitive or unchaUenged. Other forms of the Vedanta existed before him and became very strong after him. But so far as a synthesis of opinions which are divergent in details can be just, he gives a just synthesis and elaboration of the Upanishads. It is true that his teaching as to the higher and lower Brahman and as to Maya has affinities to Mahayanist Buddhism, and that later sects were '•¦ It is perhaps hardly necessary to add that there has been endless discussion as to the sense aud manner in which the soul is God. 20—2 308 HINDUISM [ch. repeUed by the severe and impersonal character of his philosophy, but the doctrine of which he is the most thorough and eminent exponent, namely that God or spirit is the only reahty and one with the human soul, asserts itseU in almost aU Hindu sects, even though their other doctrines may seem to contradict it. This line of thought is so persistent and has so many ramifications, that it is hard to say what is and what is not Vedanta. If we take Uterature as our best guide we may distinguish four points of importance marked by the Upanishads, the Brahma-Siitras, Sankara and Ramanuja. I have said something elsewhere of the Upanishads. These works do not profess to form a systematic whole (though later Hinduism regards them as such) and when European scholars speak of them coUectively, they generaUy mean the older members of the collection. These may justly be regarded as the ancestors of the Vedanta, inasmuch as the tone of thought prevalent in them is incipient Vedantism. It rejects duahsm and regards the universe as a unity not as pluraUty, as something which has issued from Brahman or is pervaded by Brahman and in any case depends on Brahman for its significance and existence. Brahman is God in the pantheistic sense, totaUy disconnected with mythology and in most passages impersonal. The knowledge of Brahman is salvation : he who has it, goes to Brahman or becomes Brahman. More rarely we find statements of absolute identity such as "Being Brahman, he goes to Brahman'." But though the Upanishads say that the soul goes to or is Brahman, that the world comes from or is Brahman, that the soul is the whole universe and that a knowledge of these truths is the one thing of importance, these ideas are not combined into a system. They are simply the thoughts of the wise, not always agreeing in detaU, and presented as independent utterances, eacU with its own value. One of the most important of these wise men is Yajnavalkya*, the hero of the Brihad Aranyaka Upanishad and a great name, to whom are ascribed doctrines of which he probably never heard. The Upanishad represents him as developing and com pleting the views of Sandilya and Uddaiaka Aruni. The former taught* that the Atman or SeU within the heart, smaUer than 1 Brihad Aran. rv. 4. 6; ib. i. iv. 10. "I am Brahman." ' See above Book n. chaps, v and vi. » Chand. Up. m. 14. xxxm] HINDU PHILOSOPHY 309 a grain of mustard seed, is also greater than aU worlds. The brief exposition of his doctrine which we possess starts from and emphasizes the human seU. This seU is Brahman. The doctrine of Uddaiaka' takes the other side of the equation : he starts with Brahman and then asserts that Brahman is the soul. But though he teaches that in the beginning there was one only without a second, yet he seems to regard the subsequent pro ducts of this Being as external to it and permeated by it. But to Yajnavalkya is ascribed an important modification of these doctrines, namely, that the Atman is unknowable and trans cendental*. It is unknowable because since it is essentiaUy the knowing subject it can be known only by itseU: it can never become the object of knowledge and language is inadequate to describe it. All that can be said of it is neti, neti, that is no, no : it is not anything which we try to predicate of it. But he who knows that the individual soul is the Atman, becomes Atman ; being it, he knows it and knows aU the world : he perceives that in aU the world there is no plurahty. Here the later doctrine of Maya is adumbrated, though not formulated. .Any system which holds that in reahty there is no plurahty or, Hke some forms of Mahayanist Buddhism, that nothing ready exists impHes the operation of this Maya or iUusion which makes us see the world as it appears to us. It may be thought of as mere ignorance, as a faUure to see the universe as it reaUy is: but no doubt the later view of Maya as a creative energy which fashions the world of phenomena is closely connected with the haU-mytho- logical conceptions found in the Pancaratra and Saiva phUosophy which regard this creative iUusion as a female force — a goddess in fact — ^inseparably associated witU the deity. The phUosophy of the Upanishads, Hke aU reUgious thought in India, is avowedly a quest of happiness and this happiness is found in some form of union with Brahman. He is perfect bhss, and whatever is distinct from him is fuU of suffering*. But this sense of the suffering inherent in existence is less marked in the older Upanishads and in the Vedanta than in Buddhism and the Sankhya. Those systems make it their basis and first principle : in the Vedanta the temperament is the same 1 Chand. Up. VI. 2 See Deussen, Philosophy of the Upanishads. ' Ato'nyad artam. Brihad Ar. m. several times. 310 HINDUISM [ch. but the emphasis and direction of the thought are different. The sahkhya looks at the world and says that salvation Ues m escape into something which has nothing Ui common with it. But the Vedantist looks towards Brahman, and his pessimism is merely the feeUng that everything which is not whoUy and really Brahman is unsatisfactory. In the later developments of the system, pessimism almost disappears, for the existence of suffering is not the first Truth but an iUusion: the soul, did it but know it, is Brahman and Brahman is bliss. So far as the Vedanta has any definite practical teaching, it does not whoUy despise action. Action is indeed inferior to knowledge and when knowledge is once obtained works are useless accessories, but the four stages of 'a Brahman's career, including household Hfe, are approved in the Vedanta Sutras, though there is a disposition to say that he who has the necessary rehgious aptitudes can adopt the ascetic Hfe at any time. The occupations of this ascetic Hfe are meditation and absorption or samadhi, the state in which the meditating soul becomes so completely blended with God on whom it meditates, that it has no consciousness of its separate existence'. As indicated above the so-caUed books of Sruti or Vedic literature are not consecutive treatises, but rather responsa prudentium, utterances respecting ritual and theology ascribed to poets, sacrificers and philosophers who were accepted as authorities. When these works came to be regarded as an orderly revelation, even orthodoxy could not shut its eyes to their divergences, and a comprehensive exegesis became necessary to give a conspectus of the whole body of truth. This investiga tion of the meaning of the Veda as a connected whole is caUed Mimamsa, and is divided into two brancbes, the earHer (piirva) and the later (uttara). The first is represented by the Piirva- mimamsa-siitras of Jaimini* which are caUed earUer (piirva) not in the chronological sense but because they deal with rites which come before knowledge, as a preparatory stage. It is interesting to find that Jaimini was accused of atheism and defended by Kumarila Bhatta. The defence is probably just, for Jaimini does ^ Maitrayana. Brah. Upanishad, vi. 20. "Having seen his own seU as The Self he becomes selfiess, and because he is selfless he is without Umit, without cause, absorbed iu thought." '^ There ia nothing to fix the date of this work except that Kumarila in com- mentmg on it in the eighth century treats it as old and authoritative. It was perhaps composed in the early Gupta period. xxxm] HINDU PHILOSOPHY 311 not so much deny God as ignore him. But what is truly extraordinary, though characteristic of much Indian Hterature about ritual, is that a work deaUng with the general theory of rehgious worship should treat the deity as an irrelevant topic. The Piirva-mimarnsa discusses ceremonies prescribed by an etemal seU-existing Veda. The reward of sacrifice is not given by God. When the result of an act does not appear at once, Jaimini teaches that there is all the same produced a super- sensuous principle caUed apurva, which bears fruit at a later time, and thus a sacrifice leads the offerer to heaven. This theory is ready tantamount to placing magic on a phUosophic basis. Badarayana's siitras, which represent the other branch of the Mimarnsa, show a type of thought more advanced and pro found than Jaimini's. They consist of 555 aphorisms — less than a fifth of Jaimini's voluminous work — and represent the out come of considerable discussion posterior to the Upanishads, for they cite the opinions of seven other teachers and also refer to Badarayana himseU by name. Hence they may be a compendium of his teaching made by his pupds. Their date is unknown but Sankara evidently regards them as ancient and there were several commentators before him'- Like most sutras these aphorisms are often obscure and are hardly intended to be more than a mnemotechnic summary of the doctrine, to be supple mented by oral instruction or a commentary. Hence it is difficult to define the teaching of Badarayana as distinguished from that of the Upanishads on the one hand, and that of his commentators on the other, or to say exactly what stage he marks in the development of thought, except that it is the stage of attempted synthesis*. He teaches that Brahman is the origin of the world and that with him should aU knowledge, reUgion and effort be concemed. By meditation on him, the soul is released and somehow associated with him. But it is not clear that we have any warrant for finding in the sutras (as does Sankara) the distinction between the higher and lower Brahman, or the doctrine of the unreahty of the world (Maya) or the absolute identity of the individual soul with Brahman. We are ' Keith in J.R.A.S. 1907, p. 492 says it is becoming more and more probable that Badarayana cannot be dated after the Christian era. Jacobi in J.A.O.S. 1911, p. 29 concludes that the Brahma-siitras were composed between 200 and 450 a.d. ^ Such attempts must have begun early. The Maitrayana Upanishad (n. 3) talks of Sarvopanishadvidya, the science of aU the Upanishads. 312 HINDUISM [ch. told that the state of the released soul is non-separation (avibhaga) from Brahman, but this is variously explained by the commentators according to their views. Though the siitras are the acknowledged text-book of Vedantism, their utterances are in practice less important than subsequent explanations of them. As often happens in India, the comment has overgrown and superseded the text. The most important of these commentators is Sankaracarya'. Had he been a European phUosopher anxious that his ideas should bear his name, or a reformer Uke tUe BuddUa with Httle respect for antiquity, he would doubtless have taken his place in history as one of the most original teachers of Asia. But since his whole object was to revive the traditions of the past and suppress his originahty by attempting to prove that his ideas are those of Badarayana and the Upanishads, the magnitude of his contribution to Indian thought is often under-rated. We need not suppose that he was the inventor of aU the ideals in his works of which we find no previous expression. He doubtless (like the Buddha) summarized and stereotyped an existing mode of thought but his summary bears the unmistakeable mark of his own personality. Sankara's teaching is known as Advaita or absolute monism. Nothing exists except the one existence called BraUman or Paramatman, the Highest SeU. Brahman is pure being and thought (the two being regarded as identical), without quahties. Brahman is not inteUigent but is intelUgence itseU. The human soul (jiva) is identical with the Highest SeU, not merely as a part of it, but as being itseU the whole universal indivisible Brahman. This must not be misunderstood as a blasphemous assertion that man is equal to God. The soul is identical with Brahman only in so far as it forgets its separate human exist ence, and aU that we caU seU and individuaUty. A man who has any pride in himself is ipso facto differentiated from Brahman as much as is possible. Yet in the world in which we move we see not only differentiation and multipHcity but also a plurahty of individual souls apparently distinct from one another and from Brahman. This appearance is due to the principle of Maya which is associated with Brahman and is the cause of the phenomenal world. If Maya is translated by iUusion it must 1 See above, p. 207 ff. xxxm] HINDU PHILOSOPHY 313 be remembered that its meaning is not so much that the world and individual existences are iUusory in the strict sense of the word, as phenomenal. The only true reality is seU-conscious thought without an object. When the mind attains to that, it ceases to be human and individual : it is Brahman. But when ever it thinks of particular objects neither the thoughts nor the objects of the thoughts are real in the same sense. They are appearances, phenomena. This universe of phenomena includes not only aU our emotions and aU our perceptions of the external world, but also what might be supposed to be the deepest truths of rehgion, such as the personahty of the Creator and the wanderings of the soul in the maze of transmigration. In the same sense that we suffer pain and pleasure, it is true that there is a personal God (Isvara) who emits and reabsorbs the world at regular intervals, and that the soul is a Hmited existence passing from body to body. In this sense the soul, as in the SaidLhya phUosophy, is surrounded by the upddhis, certain hmiting conditions or disguises, which form a permanent psychical equipment with which it remains invested in aU its innumerable bodies. But though these doctrines may be true for those who are in the world, for those souls who are agents, enjoyers and sufferers, they cease to be true for the soul which takes the path of knowledge and sees its own identity with Brahman. It is by this means only that emancipation is attained, for good works bring a reward in kind, and hence inevitably lead to new embodiments, new creations of Maya. And even in knowledge we must distinguish between the knowledge of the lower Brahman or personal Deity (Isvara) and of the higher Uidescribable Brahman'. For the orthodox Hindu this distinc- 1 The same distinction occurs in the works of Meister Eckhart (f 1327 a.d.) who in many ways approximates to Indian thought, both Buddhist and Vedantist. He makes a distinction between the Godhead and God. The Godhead is the revealer but unrevealed: it is described as "wordless" (Yajnavalkya's neti, neti), "the name less nothing," " the immoveable rest." But God is the manifestation of the Godhead, the uttered word. "All that is in the Godhead is one. Therefore we can say nothing. "He is above aU names, above aU nature. God works, so doeth not the Godhead. "Therein are they distinguished, in working and in not working. The end of all "things is the hidden darkness of the etemal Godhead, unknown and never to be "known." (Quoted by Rufus Jones, Studies in Mystical Religion, p. 225.) It may be doubted if Sankara's distinction between the Higher and Lower Brahman is to be found in the Upanishads but it is probably the best means of harmonizing the discrepancies in those works which Indian theologians feel bound to explain away. 20—5 314 HINDUISM [CH. tion is of great importance, for it enables him to reconcile passages Ui the scriptures which otherwise are contradictory. Worship and meditation which make Isvara their object do not lead directly to emancipation. They lead to the heavenly world of Isvara, in which the soul, though glorified, is stiU a separate individual existence. But for him who meditates on the Highest Brahman and knows that his true seU is that Brahman, Maya and its works cease to exist. When he dies nothing differentiates him from that Brahman who alone is bliss and no new individual existence arises. The crux of this doctrine is in the theory of Maya. If Maya appertains to Brahman, if it exists by his wUl, then why is it an evil, why is release to be desired? Ought not the individual souls to serve Brahman's purpose, and would not it be better served by Hving gladly in the phenomenal world than by passing beyond it ? But such an idea has rarely satisfied Indian thinkers. If, on the other hand, Maya is an evd or at least an imperfection, if it is Hke rust on a blade or dimness in a mirror, U, so to speak, the edges of Brahman are weak and break into fragments which are prevented by their own feebleness from reaUzing the unity of the whole, then the mind wonders uneasUy U, in spite of aU assurances to the contrary, this does not imply that Brahman is subject to some external law, to some even more mysterious Beyond. But Sankara and the Brahma-sutras wiU not tolerate such doubts. According to them. Brahman in making the world is not actuated by a motive in the ordinary sense, for that woidd imply human action and passion, but by a sportive impulse': "We see in every-day Hfe," says Sankara, "that certain doings of princes, who have no desires left unfulfiUed, have no reference to any extraneous purpose but proceed from mere sportfulness. We further see that the process of inhalation and exhalation is going on without reference to any extraneous purpose, merely foUowing the law of its own nature. Analogously, the activity of the Lord also may be supposed to be mere sport, proceeding from his own nature without reference to any purpose*." This ' Vedanta siitras, ii. 1. 32-3, and Sahkaras's commentary, S.B.E. vol. xxxiv. pp. 356-7. Ramanuja holds a simUar view and it is yery common in India, e.g. Vishnu Pur. i. chap. 2. ' See too a remarkable passage in his comment on Brahma-sutras, n. 1. 23. "As soon as the consciousness of non-difference arises in us, the transmigratory "state of the individual soul and the creative quaUty of Brahman vanish at once. xxxm] HINDU PHILOSOPHY 315 is no worse than many other explanations of the scheme of things and the origui of evd but it is not ready an explanation. It means that the Advaita is so engrossed in ecstatic contempla tion of the omnipresent Brahman that it pays no attention to a mere by-product Hke the physical universe. How or why that universe with aU its imperfections comes to exist, it does not explain. Yet the boldness and ample sweep of Sankara's thought have in them something greater than logic', something recalling the grandeur of plains and seas limited only by the horizon, nay rather those abysses of space wherein on clear nights worlds and suns innumerable are scattered like sparks by what he would caU God's playfulness. European thought attains to these altitudes but cannot Uve in them for long: it demands and fancies for itseU just what Sankara wdl not grant, the motive of Brahman, the idea that he is working for some consummation, not that he was, is and wiU be etemaUy complete, unaffected by the drama of the universe and yet identical with souls that know him. Even in India the austere and impersonal character of Sankara's system provoked dissent: He was accused of being a Buddhist in disguise and the accusation raises an interesting question* in the history of Indian philosophy to which I have referred in a previous chapter. The affinity existing between the Madhyamika form of Buddhist metaphysics and the earHer Vedanta can hardly be disputed and the only question is which borrowed from the other. Such questions are exceedingly difficult to decide, for from time to time new ideas arose in India, permeated the common inteUectual atmosphere, and were worked up by aU sects into the forms that suited each best. In the present instance aU that can be said is that certain ideas about the unreaUty of the world and about absolute and relative "the whole phenomenon of pluraUty which springs from wrong knowledge being "snblated by perfect knowledge and what becomes then of the creation and the "faults of not doing what is beneficial and the like? " 1 Although Sankara's comraentary is a piece of severe ratiocination, especiaUy in its controversial parts, yet he holds that the knowledge of Brahman depends not on reasoning but on scripture and intuition. "The presentation before the mind of the Highest Self is effected by meditation and devotion." Brah. Sut. in. 2. 24. See too his comments ou I. 1. 2 and u. 1. 11. ^ See Sukhtankar, Teachings of Veddnta according to Rdmdnuja, pp. 17-19. WaUeser, Der aeltere Veddnta, and De la VaUde Poussin m J.B.A.S. 1910, p. 129. 316 HINDUISM [ch. truth appear in several treatises both Brahmanic and Buddhist, such as the works of Sardiara and Nagarjuna and the Gauda- padakarikas, and of these the works attributed to Nagarjuna seem to be the oldest. It must also be remembered that according to Chinese accounts Bodhidharma preached at Nanking in 520 a doctrine very similar to the advaita of Sardsara though ex pressed in Buddhist phraseology. Of other forms of Vedantism, the best known is the system of Ramanuja generaUy caUed Visishtadvaita'. It is an evidence of the position held by the Vedanta phUosophy that rehgious leaders made a commentary on the Sutras of Badarayana the vehicle of their most important views. Unlike Sankara, Rama nuja is sectarian and identifies his supreme deity with Vishnu or Narayana, but this is Httle more than a matter of nomen clature. His interpretation is modern in the sense that it pursues the Une of thought which leads up to the modem sects. But that Hne of thought has ancient roots. Ramanuja foUowed a commentator named Bodhayana who was anterior to Sankara, and in the opinion of so competent a judge as Thibaut he gives the meaning of Badarayana in many points more exactly than his great rival. On the other hand his interpretation often strains the most important utterances of the Upanishads. Ramanuja admits no distinction between Brahman and Isvara, but the distinction is abohshed at the expense of aboHshing the idea of the Higher Brahman, for his Brahman is practicaUy the Isvara of Sankara. Brahman is not without attributes but possessed of aU imaginable good attributes, and though nothing exists apart from him, Uke the antithesis of purusha and prakriti in the Saiikhya, yet the world is not as in Sankara's system merely Maya. Matter and souls {cit and cuiit) form the body of Brahman who both comprises and pervades 1 This term is generally rendered by quaUfied, that is not absolute. Monism. But South Indian scholars give a sUghtly different explanation and maintain that it is equivalent to Visishtayor advaitam or the identity of the two quaUfied (viHshta) conditions of Brahman. Brahman is qualified by cit and acit, souls and matter, which stand to him in the relation of attributes. The two conditions are Kdrydvasthd or period of cosmic manifestation in which cit and cuiit are manifest and Kararid- vasthd or period of cosmic dissolution, when they exist only in a subtle state within Brahman. These two conditions are not different (advaitam). See Srinivas Iyengar, J.R.A.S. 1912, p. 1073 aud also Sri Bdmdnujdcdrya: His PhUosophy by Rajagopa- lacharyar. xxxm] HINDU PHILOSOPHY 317 aU things, which are merely modes of his existence'. He is the inner mler (antaryamin) who is in aU elements and aU human souls*. The texts which speak of Brahman as being one only without a second are explained as referring to the state of pralaya or absorption which occurs at the end of each Kalpa. At the conclusion of the period of pralaya he re-emits the world and individual souls by an act of vohtion and the souls begin the round of transmigration. Salvation or release from this round is obtained not by good works but by knowledge and meditation on the Lord assisted by his grace. The released soul is not identified with the Lord but enjoys near him a personal existence of etemal bhss and peace. This is more like European theism than the other doctrines which we have been considering. The difference is that God is not regarded as the creator of matter and souls. Matter and souls consist of his substance. But for aU that he is a personal deity who can be loved and worshipped and whereas Sankara was a rehgious phdosopher, Ramanuja was rather a phUosophic theologian and founder of a church. I have aUeady spoken of his activity in this sphere. 4 The epics and Puranas contain philosophical discussions of considerable length which make Httle attempt at consistency. Yet the line of thought in them aU is the same. The chief tenets of the theistic Sankhya- Yoga are assumed: matter, soul and God are separate existences: the soul wishes to move towards God and away from matter. Yet when Indian writers glorify the deity they rarely abstain from identUying him with the universe. In the Bhagavad-gita and other philosophical cantos of the Mahabharata the contradiction is usuaUy left without an attempt at solution. Thus it is stated categoricaUy* that the world consists of the perishable and imperishable, i.e., matter and soul, but that the supreme spirit is distinct from both. 1 Compare the phrase of Keats in a letter quoted by Bosanquet, Gifford Lectures for 1912, p. 66. "As various as the Uves of men are, so various become their souls and thus does God make individual beings, souls, identical souls of the sparks of his own essence." 2 Ttus tenet is justified by Brihad Aran. Up. m. 3 ff. which is a great text for Ramanuja's school. "He who dweUs in the earth (water, etc.) and within the earth (or, is different from the earth) whom the earth knows not, whose body the earth is, who rales the earth within, he is thyseff, the ruler within, the immortal." !.-^ta, XV. 16, 17. 318 HINDUISM [cH. Yet in the same poem we pass from this antithesis to the monism which declares that the deity is aU things and "the seU seated in the heart of man." We have then attained the Vedantist point of view. Nearly aU the modem sects, whether Sivaite or Vishnuite, admit the same contradiction into their teaching, for they reject both the atheism of the Sankhya and the immaterial- ism of the Advaita (since it is impossible for a practical rehgion to deny the existence of either God or the world), while the irresistible tendency of Indian thought makes them describe their deity in pantheistic language. AU strive to find some metaphysical or theological formula which wiU reconcUe these discrepant ideas, and nearly aU Vishnuites profess some special variety of the Vedanta caUed by such names as Visishtadvaita, Dvaitadvaita, Suddhadvaita and so on. They differ chiefly in their definition of the relation existing between the soul and God. Only the Madhvas entirely discard monism and profess duahty (Dvaita) and even Madhva thought it necessary to write a commentary on the Brahma-siitras to prove that they support his doctrine and the Sivaites too have a commentator, NUakantha, who interprets them in harmony with the Saiva Siddhanta. There is also a modem commentary by Somanaradittyar which expounds this much twisted text agreeably to the doctrines of the Lingayat sect. In most fundamental principles the Sivaite and Saktist schools agree with the Visishtadvaita but their nomenclature is different and their scope is theological rather than phdosophical. In aU of them are felt the two tendencies, one wishing to dis tinguish God, soul and matter and to adjust their relations for the purposes of practical rehgion, the other holding more or less that God is aU or at least that aU things come from God and return to him. But there is one difference between the schools of sectarian philosophy and the Advaita of Saidiara which goes to the root of the matter. Sankara holds that the world and individual existences are due to iUusion, ignorance and mis conception: they vanish in the Hght of true knowledge. Other schools, while agreeing that in some sense God is aU, yet hold that the universe is not an Ulusion or false presentment of him but a process of manifestation or of evolution starting from him'. It is not precisely evolution Ui the European sense, but rather 1 The two doctrines are caUed Vivartavdda and Parindmavdda, xxxm] HINDU PHILOSOPHY 319 a rhythmic movement, of duration and extent inexpressible in flgures, in which the Supreme Spirit alternately emits and re absorbs the universe. As a rule the higher rehgious life aims at some form of union or close association with the deity, beyond the sphere of this process. In the evolutionary process the Vaishnavas interpolate between the Supreme Spirit and the phenomenal world the phases of conditioned spirit known as Sankarsha,na, etcf; in the same way the Sivaite schools increase the twenty-four tattvas of the Sankhya to thirty-six'. The first of these tattvas or principles is Siva, corresponding to the highest Brahman. The next phase is Sadasiva in which differentiation commences owing to the movement of Sakti, the active or female principle. Siva in this phase is thought of as having a body composed of mantras. Sakti, also known as Bindu or Suddhamaya, is sometimes regarded as a separate tattva but more generaUy as inseparably united witU Siva. The third tattva is Isvara, or Siva in the form of a lord or personal deity, and the fourth is Suddhavidya or true knowledge, explained as the principle of correlation between the experiencer and that which is experienced. It is only after these that we come to Maya, meaning not so much iUusion as the substratum in which Karma inheres or the protoplasm from which aU things grow. Between Maya and Purusha come five more tattvas, caUed envelopes. TheU effect is to enclose and Hmit, thus tuming the divine spirit into a human soul. Saktist accounts of the evolutionary process give greater prominence to the part played by Sakti and are usuaUy meta- physiological, U the word may be pardoned, inasmuch as they regard the cosmic process as the growth of an embryo, an idea which is as old as the Vedas*- It is impossible to describe even in outline these manifold cosmologies but they generaUy speak of Sakti, who in one sense is identical with Siva and merely his active form but in another sense is identified with Prakriti, coming into contact with the form of Siva caUed Prakasa or hght and then sohdifjdng into a drop (Bindu) or germ which divides. At some point in this process arise Nada or sound, and 1 These are only the more subtle tattvas. There are also 60 gross ones. See for the whole subject Schomerus Der Qaiva-Siddhanta, p. 129. ^ It also finds expression in myths about the division of the deity into male and female halves, the cosmic egg, etc., which are found in all strata of Indian literature. 320 HINDUISM [ch. Sabda-brahman, the sound-Brahman, which manifests itself in various energies and assumes in the human body the form of the mysterious coiled force caUed KundaUni'- Some of the older Vishnuite writmgs use simUar language of Sakti, under the name of Lakshmi, but m the Visishtadvaita of Ramanuja and subse quent teachers there is Httle disposition to dweU on any feminme energy in discussing the process of evolution. Of aU the Darsanas the most extraordmary is that caUed Rasesvara or the mercurial system*. According to it quicksilver, if eaten or otherwise appHed, not only preserves the body from decay but deUvers from transmigration the soul which inhabits this glorified body. Qmcksilver is even asserted to be identical with the supreme seU. This curious Darsana is represented as revealed by Siva to Sakti and it is only an extreme example of the tantric doctrine that spiritual results can be obtained by physical means. The practice of taking mercury to secure health and long Hfe must have been prevalent in medieval India for it is mentioned by both Marco Polo and Bemier*. A people among whom the Vedanta could obtain a large foUowing must have been prone to think Httle of the things which we see compared with the unseen of which they are the manifestation. It is, therefore, not surprising U materiaUsm met with smaU sympathy or success among them. In India the extravagances of asceticism and of mystic sensuahsm aUke find devotees, but the simple phdosophy of Let us eat and drink for to-morrow we die, does not commend itseU. Nevertheless it is not wholly absent and was known as the doctrine of Brihaspati. Those who professed it were also caUed Carvakas and Loka yatikas*. Brihaspati was the preceptor of the gods and his ^ An account of tantric cosmology can be found in Avalon, Mahdn. Tantra, pp xix-xxxi. See also Avalon, Prapancasdra Tantra, pp. 5 ff. ; Srinivasa Iyengar, Indian Philosophy, pp. 143 and 295 ff. ; Bhandarkar, Vaishn. and Saiiiism, pp. 145 ff. ' Sarva-dar^ana-sangraha, chap. ix. Por this doctrine in China see Wieger Histoire des Croyances religieuses en Chine, p. 411. 3 See Yule's Marco Polo, n. pp. 365, 369. * See Rhys Davids' note in his Dialogues of the Buddha on Digha Nikdya, Sutta V. pp. 166 ff. He seems to show that Lokayata meant originaUy natural philosophy as a part of a Brahman's education and only graduaUy acquired a bad meaning. The Arthasastra also recommends the Sankhya, Yoga and Lokayata systems. xxxm] HINDU PHILOSOPHY 321 connection with this sensuahstic philosophy goes back to a legend found in the Upanishads' that he taught the demons false knowledge whose "reward lasts only as long as the pleasure lasts " in order to compass their destruction. This is similar to the legend found in the Puranas that Vishnu became incarnate as Buddha in order to lead astray the Daityas. But though such words as Oarvaka and Nastika are used in later Hterature as terms of leamed abuse, the former seems to denote a definite school, although we cannot connect its history with dates, places or personahties. The Carvakas are the first system examined in the Sarva-darsana-sangraha, which is written from the Vedantist standpoint, and beginning from the worst systems of phUosophy ascends to those which are relatively correct. This account contains most of what we know about their doctrines*, but is obviously biassed : it represents them as cynical volupt uaries holding that the only end of man is sensual enjoyment. We are told that they admitted only one source of knowledge, namely perception, and four elements, earth, water, fire and air, and that they held the soul to be identical with the body. Such a phrase as my body they considered to be metaphorical, as apart from the body there was no ego who owned it. The soul was supposed to be a physical product of the four elements, just as sugar combined with a ferment and other ingredients produces an intoxicating Hquor. Among verses described as "said by Brihaspati" occur the foUowing remarkable Unes: "There is no heaven, no hberation, nor any soul in another world, Nor do the acts of the asramas or castes produce any reward. If the animal slain in the Jyotishtoma sacrifice will go to heaven, Why does not the sacrificer immolate his own father ? While life remains let a man live happily: let him feed on butter even if he runs into debt. When once the body becomes ashes, how can it ever return? " The author of the Dabistan, who Hved in the seventeenth century, also mentions the Carvakas in somewhat simUar terms^. Brahmanical authors often couple the Carvakas and Bud dhists. This lumping together of offensively heretical sects may 1 Maitr. Up. vn. 8. ^ See also SuaU in Musion, 1908, pp. 277 ff. aud the article MateriaUsm (Indian) in E.R.E. Por another instance of ancient materiaUsm see the views of Piyasi set forth in Dig. Nik. xxni. The Bphad Ar. Up. m. 2. 13 implies that the idea of body and spirit being disintegrated at death was known though perhaps not relished. ' Translation by Shea and Troyer, vol. n. pp. 201-2. 322 HINDUISM [ch. xxxm be merely theological animus, but stiU it is possible that there may be a connection between the Carvakas and the extreme forms of Mahayanist nihihsm. Schrader' in analysing a singular work, called the Svasamvedyopanishad, says it is "inspired by the Mahayanist doctrine of vacuity {Mnya-vdda) and proclaims a most radical agnosticism by asserting in four chapters (a) that there is no reincarnation (existence being bubble-Hke), no God, no world : that all traditional Hterature {Sruti and Smriti) is the work of conceited fools; (6) that Time the destroyer and Nature the originator are the rulers of aU existence and not good and bad deeds, and that there is neither bed nor heaven; (c) that people deluded by flowery speech cHng to gods, sacred places, teachers, though there is in reahty no difference at aU between Vishnu and a dog ; {d) that though aU words are untrue and aU ideas mere iUusions, yet Uberation is possible by a thorough reahzation of Bhdvddvaita." But for this rather sudden con cession to Hindu sentiment, namely that dehverance is possible, this doctrine resembles the tenets attributed to the Carvakas. ^ Sanskrit Manuscripts in the Adyar Library, 1908, pp. 300-1. OAMBBIDOB: PEINTBD by j. B. peace, M.A., AT THE UUrFBESITY PEBSS 9002 00560 2207