Articles by alphabetic order
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0


Cosmology and Meditation From the Aggañña-Sutta to the Mahāyāna

From Chinese Buddhist Encyclopedia
Jump to navigation Jump to search



Rupert Gethin


Cosmology and Meditation: From the Aggañña-Sutta to the Mahāyāna


THE SIGNIFICANCE OF BUDDHIST COSMOLOGY


Now there comes a time, Vasettha, when after a long period of time this world contracts. When the world contracts beings are for the most part born in the realm of Radiance There they exist made of mind, feeding on joy, self-luminous, moving through the air, constantly beautiful; thus they remain for a long, long time. Now there comes a time, Vasettha, when after a long period of time this world expands. When the world expands beings for the most part fall from the realm of Radiance and come here [to this realm]; and they exist made of mind, feeding on joy, self-luminous, moving through the air, constantly beautiful; thus they remain for a long, long time.1 An earlier version of this article was read at the Tenth Congress of the International Association of Buddhist Studies in Paris in 1991. I am grateful to Steven Collins, Lance Cousins, Nobumi Iyanaga, Rita Langer, Oran Rotem, Paul Williams, and Nobuyoshi Yamabe for comments, criticism, or help with tracing references in the course of writing this article. 1 D 3:84-85: "hoti kho so Vasettha samayo yam kadaci karahaci dighassa addhuno accayena ayam loko samvattati. samvattamane loke yebhuyyena satta abhassara-samvattanika honti. te tattha honti manomaya piti-bhakkha sayam-pabha antalikkha-cara subhatthayino ciram digham addhanam titthanti. hoti kho so Vasettha samayo yam kadaci karahaci dighassa addhuno accayena ayam loko vivattati. vivattamane loke yebhuyyena satta abhassara-kaya cavitva itthattam agacchanti. te ca honti manomaya piti-bhakkha sayyampabha antalikkha-cara subhatthayino ciram digham addhanam titthanti." All references to Pali and Sanskrit texts use the abbreviations listed in app. A of this article. For full citations, see app. A. References are to volume and page of the cited edition, except in the case of the Abhidharmakosa and Visuddhimagga; references to the former are to chapter and verse, and to the latter, to chapter and section of the Warren-Kosambei dition and Nanamoli translation.


This striking and evocative passage introduces the well-known account of the evolution of the world and human society found in the Aggafiiiasutta of the Pali Digha Nikaya.2 It marks the beginning of a particular line of thought within Buddhist tradition concerning the world and its cycles of expansion and contraction. It is this line of thought that I wish to investigate in the present article. It can sometimes seem that, as "literate, demythologized and Aristotelianized academics"-to borrow a characterization from G. S. Kirk3 we become peculiarly insensitive to the kind of poetic and imaginative world which, for perhaps most human beings for most of human history, has constituted "reality." It is perhaps not an accident then that, despite the fact that certain studies of contemporary Buddhism in Sri Lanka, Burma, and Thailand have drawn attention to the importance of the traditional cosmology to the worldview of present-day Theravada Buddhists, 4 the subject of Buddhist cosmology has received relatively little attention from textual scholars.5 Significantly, one of the few works de- 2 This initial formula must be regarded as constituting a significant piece of floating tradition that forms part of the common heritage of ancient Buddhism. Apart from its occurrence in all four surviving recensions of the Aggafifia-sutta-see K. Meisig, Das Sitra von den vier Standen: Das Aggainia-sutta im Licht seiner chinesischen Parallelen (Wiesbaden: Harrassowitz, 1988)-we find the same formula (though with a slightly different account of the process of world expansion) used in two other suttas of the Digha Nikaya: the Brahmajala and Pdtika (D 1:17 and D 3:28-29; the expansion formula here reads: "vivattamane loke suinnam brahma-vimanam patubhavati. ath' afifataro satto ayukkhaya va puiinakkhaya va abhassara-kaya cavitva sufiiam brahma-vimanam upapajjati. so tattha hoti manomayo piti-bhakkho sayam-pabho antalikkha-caro subhatthayi ciram digham addhanam titthati"). Two Aiguttara passages (A 4:89; 5:60) also make use of parts of the formula, while Vibh 415 (cf. D 3:88), which states that human beings at the beginning of an aeon are born lacking the male or female faculty, also alludes to it. Outside the Nikfyas and Agamas, looking beyond the Pali tradition we find the formula used in the Mahavastu (see Le Mahdvastu, ed. E. S6nart, 3 vols. [[[Wikipedia:Paris|Paris]], 1882-97], 1:52, 338-39) and referred to and commented on by Vasubandhu in the Abhidharmakosa (Abhidh-k 3:97c-d-98a-b; see Louis de La Vall6e Poussin, trans., L'Abhidharmakosa de Vasubandhu: Traduction et Annotations, 6 vols. [Brussels: Institut belge des hautes etudes chinoises, 1971], 2:203-4, and Abhidharmakosa and Bhasya of Acdrya Vasubandhu with Sphutartha Commentary of Acdrya Yasomitra, ed. D. Shastri, 3 vols. [[[Varanasi]]: Bauddha Bharati, 1970-72], 2:554). 3 G. S. Kirk, Myth: Its Meaning and Functions in Ancient and Other Cultures (London, Berkeley, and Los Angeles: Cambridge University Press and University of California Press, 1970), p. 281. 4 See R. F Gombrich, Precept and Practice: Traditional Buddhism in the Rural Highlands of Ceylon (Oxford: Oxford University Press, 1971), pp. 153-91; S. J. Tambiah, Buddhism and the Spirit Cults in North-East Thailand (Cambridge: Cambridge University Press, 1970), pp. 32-52.

5 The pioneering work is W. Kirfel, Kosmographie der Inder (Bonn: K. Schroeder, 1920), but this devotes rather little space to Buddhist sources in comparison to Brahmanical and Jain materials and is now rather dated. It is Louis de La Vallee Poussin's work on the Abhidharmakosa that has given us the most substantial material on Vaibhasika cosmology; see his L'Abhidharmakosa de Vasubandhu, Vasubandhu et YaSomitra: Troisieme chapitre de l'Abhidharmakosa: Kdrikd, Bhdsya et Vydkhyd (Brussels: Acad6mie royale de Belgique, 1919), and "Cosmology and Cosmogony (Buddhist)," in Encyclopaedia of Religion


voted to Buddhist cosmology to be published in more recent years is not a study of ancient Pali or Sanskrit sources but a translation of a fourteenth-century Thai classic, Phya Lithai's Traibhimikatha or Thrai Phum Phra Ruang ("Three Worlds according to King Ruang").6 The overall paucity of scholarly materials dealing with Buddhist cosmology would seem to reflect a reluctance on the part of modern scholarship to treat this dimension of Buddhist thought as having any serious bearing on those fundamental Buddhist teachings with which we are so familiar: the four noble truths, the eightfold path, no-self, dependent arising, and so on. The effect of this is to divorce the bare doctrinal formulations of Buddhist thought from a traditional mythic context. This can result in serious distortions: the picture that has sometimes been painted of especially early Buddhism and Theravada Buddhism is somewhat one-dimensional and flat. However, the principle that the study of the imagery employed in early Buddhist texts is a useful way of deepening our understanding of the more overtly conceptual teachings of the Nikayas has already been used to good purpose by Steven Collins in his discussion of house imagery, vegetation imagery, and water imagery in and Ethics, ed. J. Hastings, 13 vols. (Edinburgh: T. & T. Clark, 1908-27), 2:129-38. The relevant portions of Nlanamoli's translation of Buddhaghosa's Visuddhimagga constitute the only readily available and accessible sources for the developed Theravadin system; see The Path of Purification (Colombo: Semage, 1964), 7:40-44, 13:29-65, p. 214, n. 14. The two more comprehensive studies of the details of the Nikayas' cosmological outlook, Joseph Masson's La religion populaire dans le canon bouddhique pali (Louvain: Bureaux du Mus6on, 1942); and M. M. J. Marasinghe's Gods in Early Buddhism: A Study in Their Social and Mythological Milieu as Depicted in the Nikayas of the Pali Canon (Vidyalankara: University of Sri Lanka, 1974) tend to approach their subject from the standpoint that talk of gods and the like in the Nikayas is something of a concession to "popular" Buddhism rather than an integral part of Buddhist thought-this is explicitly revealed in the title of Masson's book and is perhaps less true of Marasinghe's work; both these books, however, represent useful collections of material on cosmological ideas as presented in the Nikayas. The figure of Mara has received some additional attention: T. O. Ling, Buddhism and the Mythology of Evil: A Study in Theravada Buddhism (London: George Allen & Unwin, 1962); J. W. Boyd, Satan and Mara: Christian and Buddhist Symbols of Evil (Leiden: Brill, 1975). R. Kloetzli's more recent Buddhist Cosmology: From Single World System to Pure Land (Delhi: Motilal Banarsidass, 1983), while providing a useful summary and overview of Buddhist cosmological ideas from the Nikayas through to the developed Mahayana, from my perspective passes rather quickly over the early materials and the Abhidharma. One of the most interesting treatments of cosmology in the Nikayas to have been published in recent years is Peter Masefield's "Mind/ Cosmos Maps in the Pali Nikayas," in Buddhist and Western Psychology, ed. N. Katz (Boulder, Colo.: Prajiia Press, 1983), pp. 69-93. See also R. F. Gombrich, "Ancient Indian Cosmology," in Ancient Cosmologies, ed. C. Blacker and M. Loewe (London, 1975), pp. 110-42. 6 F. E. Reynolds and M. B. Reynolds, Three Worlds according to King Ruang: A Thai Buddhist Cosmology (Berkeley: Asian Humanities Press, 1982). One of the sources employed by Phya Lithai was the earlier Pali Lokapanniatti; see E. Denis, trans. and ed., La Lokapaiinati et les iddes cosmologiques du bouddhisme ancien, 2 vols. (Lille, 1977).


the context of the Nikayas' presentation of the teaching of "no-self."7 Advocating an approach not dissimilar to Collins's, Stanley Tambiah has commented that the traditional Buddhist cosmological scheme "says figuratively and in terms of metaphorical images the same kind of thing which is stated in abstract terms in the doctrine. The basic doctrinal concepts of Buddhism ... which are alleged to explain man's predicament and to direct his religious action, are also embedded in the cosmology (and its associated pantheon)."8 It seems to me that in this he can only be right, and one of the things I will do in this article is to explore further the relationship in Buddhist thought between the realms of abstract theory, on the one hand, and cosmological myth, on the other. To ignore the mythic portions of ancient Buddhist texts is to fail in a significant way to enter into their thought-world. My particular focus will be certain cosmological ideas concerning the expansion and contraction of the universe and their implications for our understanding of the nature and significance of the fourth "meditation" (jhdna/dhydna) in the account of the stages of the Buddhist path as presented in the Nikayas and Abhidharma. What also emerges, I will argue, is a clearer perspective on the development of certain ideas usually considered characteristic of certain strands of Mahayana Buddhist thought: the tathdgatagarbha and an idealist ontology.


COSMOLOGY IN THE NIKAYAS AND ABHIDHARMA The Nikayas and Agamas contain very many cosmological details, but it is not until the period of the Abhidharma that we get attempts to organize these details into a systematic whole. Yet what Masson's and Marasinghe's studies of "gods" in the Nikayas reveal is that, notwithstanding the fact that the Nikayas nowhere give a systematic exposition of their cosmology,9 all the basic principles and not a few of the details of the developed cosmology of the Abhidharma are to be found scattered throughout the Nikayas.l? I reckon the basic principles to be three. First, there are a number of different realms of existence that constitute a hierarchy; there are lower realms-the realms of animals (tiracchdnayoni) and of hungry ghosts (pettivisaya) and various hells (niraya); there is the realm of men (manussa) and, above, the various heaven realms of the 7 S. Collins, Selfless Persons: Imagery and Thought in Theravada Buddhism (Cambridge: Cambridge University Press, 1982), pp. 165-76, 218-24, 247-61. 8 Tambiah, Buddhism and the Spirit Cults in North-East Thailand, pp. 34-35. 9 Although we do find the beginnings of systematization in the Ainguttara-Nikayas, ee Marasinghe, pp. 244-81. 10 For a summary of the cosmological details as found in the Nikayas, see ibid., pp.


devas and brahmds.11 Second, beings are continually reborn in these various realms in accordance with their actions-the ten unskillful (akusala) courses of action (kammapatha) lead to rebirth in one of the lower realms, and the ten skillful (kusala) courses of action lead to rebirth as a human being or in the lower heavens, while meditation attainments (jhana) lead to rebirth in the higher heavens as a brahma.12 The third principle is that which is inherent in the formula from the Aggafina-sutta that I quoted above. The various levels of existence arrange themselves in "world-systems" (loka-dhatu); there are innumerable world-systems which all expand and contract across vast expanses of time.13 This basic cosmological scheme is not confined to one isolated Nikaya context; it is something alluded to and assumed by very many of the Nikaya formulas. It is perhaps most conveniently summed up in the well-known formula which states that the Buddha, "having himself fully understood and directly experienced this world with its devas, its Mara and Brahma, this generation with its samanas and brahmanas, with its princes and peoples, makes it known."14

What I want to argue below does not hinge on establishing that the Buddha himself or the earliest phase of Buddhist thought subscribed to this specific cosmological view; I am concerned with how the tradition read the texts as a coherent whole rather than with their relative chronology and evolution. But I would add that I can see no particular reason for thinking that this basic conception of the universe does not belong to the earlier strata of the Nikayas. There are no a priori historical grounds for regarding the principles of this cosmology as improbable in the mouth of the Buddha; as Marasinghe has commented, "From a study of the Jain, Ajivika, and the Buddhist ideas of cosmological thinking, it may be surmised that, by the time of the Buddha, there was a rich floating mass of cosmological ideas in the Gangetic regions from which most religious teachers drew quite freely."15 l See esp. Masson, pp. 18-38, and the chart facing p. 144 for details of the various hierarchical lists found in the Nikayas. 12 See in particular A 2:126, 230; 4:39, 241; cf. Marasinghe, pp. 244-68, and chart facing p. 62.

13 See Marasinghe, p. 44; D 2:139, 253; M 3:101-2; A 1:227, 5:59. 14 For example, D 1:62: "so imam lokam sadevakam samarakam sabrahmakam sassamana-brahamanimp ajam sadeva-manussam sayam abhiniiias acchikatva pavedeti." I follow the commentary (Sv 1:174) in taking sadeva in the sense of sammuti-deva. It is possible to take samaraka and sabrahmaka as indicating a plurality of maras and brahmas, respectively (on the grounds that the Nikayas clearly do recognize a plurality of brahmds and maras); on the other hand, it seems to me probable that in the present context we should take imam lokam as implying simply "this [one] world-system" that we occupy; see Boyd, pp. 100-111; cf. the discussion of the terms loka, loka-dhatu, and cakkavala below, n. 34. 15 Marasinghe (n. 5 above), p. 260; cf. pp. 59, 259-61.


On the evidence of the Rg-Veda, Upanisads, and Jain sources such cosmological ideas might easily have been borrowed and adapted from the cultural milieu in which we understand the Buddha to have formulated his teachings. But this is perhaps to put it too negatively. In many respects the kind of cosmology that I have indicated above seems actually fundamental to Buddhist thought. On the evidence of the Nikayas (and apparently the Chinese Agamas) we know of no Buddhism or Buddha that did not teach a belief in rebirth, or conceive of rebirth as fluid among different realms, whether animal, hellish, human, or heavenly.16 While certain of the details of the Aggainia-sutta's account of the evolution of human society may be, as Gombrich has persuasively argued, satirical in intent, there is nothing in the Nikayas to suggest that these basic cosmological principles that I have identified should be so understood; there is nothing to suggest that the Aggafinia-sutta'sin troductory formula describing the expansion and contraction of the world is merely a joke.17 We should surely expect early Buddhism and indeed the Buddha to have some specific ideas about the nature of the round of rebirth, and essentially this is what the cosmological details presented in the Aggafifia-sutta and elsewhere in Nikayas constitute. They represent a concretized and mythic counterpart to the more abstract formulation of, say, dependent arising (paticcasamuppada). What functions do the various levels of existence and the gods play in the Nikayas? There is no one simple answer to this question, but I shall answer initially by stating more fully what I identified above as the second principle of Buddhist cosmology, namely, that particular kinds of action of body, speech, and mind lead to certain kinds of rebirth. The passages I referred to in this connection effectively draw up a hierarchy of kamma that corresponds very closely to the hierarchy of levels of existence. At the bottom of this hierarchy we have unskillful kammas leading to rebirth in the realms of hell, hungry ghosts, and animals; next we have the skillful kammas of generosity (dana) and the precepts (slla) practiced to various degrees and leading to rebirth as a human being or as a deva in one of six realms of heaven; finally the practice of meditation (bhavana) and the development of the various jhdnas leads to rebirth among "the gods of Brahma's retinue" (brahmakdyika deva) and beyond. At this point we should remind ourselves that kamma is for the Nikayas-as for Buddhist thought generally-at root a mental act or intention; acts of body and speech are performed in response to and conditioned by the quality of the underlying intention or will (cetana); they are unskillful or skillful because they are motivated by unskillful or 16 See, e.g., the Mahasihanada-sutta (M 1:68-83). 17 See app. B, "How Old Is Buddhist Cosmology? A Note on the Aggafifia-Sutta."


skillful intentions.18 Acts of body and speech are, as it were, the epiphenomena of particular kinds of mentality; they are driven by specific psychological states. In a very real sense acts of body and speech are acts of will. Thus the hierarchy is essentially one of certain kinds of mentality (understood as kamma) being related to certain levels of existence; this is most explicit in the case of the various jhanas and Brahma realms. This way of thinking demonstrates the general principle of an equivalence or parallel in Buddhist thought between psychology on the one hand and cosmology on the other. Many of the stories about devas from different heavens in the Nikayas lend themselves very readily to a kind of "psychological" interpretation, that is, to interpretation in terms of certain mental states; in certain contexts this interpretation is explicit in the texts themselves. In the vana-samyutta of the Samyutta Nikaya there is a whole series of accounts of devas visiting bhikkhus dwelling in the forest in order to admonish the bhikkhus for their laziness.19 Here the devas serve to arouse skillful states of mind in the bhikkhu that spur him on in his practice. Similarly in the Mara- and Bhikkhuni-samyuttasM ara is represented as appearing on the scene and tempting bhikkhus, bhikkhunis, and the Buddha, with the world of the five senses.20 Here then Mara appears to act as the five hindrances (nivarana) which are precisely the mental states that one must overcome in order to attain jhina, and it is precisely jhdna that-at least according to a later understanding-takes one temporarily beyond the world of the five senses and out of Mara's reach.21 To read these texts in loosely psychological terms is not, I think, to engage in acts of gratuitous "demythologizing"; the Buddhist tradition itself at an early date was quite capable of demythologizing-so much so that one hesitates to use such a term in this context. It is rather, I think, that this kind of psychological interpretation was for the Nikayas inherent in the material itself. When questioned as to the nature of Mara, the Buddha responds in abstract terms that have to do with general psychological experience: "One says, 'Mara! Mara!' lord. Now to what extent, lord, might Mara or the manifestation of Mara exist?' 'Where the eye exists, 18 A 3:415: "cetanfham bhikkhave kammam vadami. cetayitva kammam karoti kayena vac1a9y a manasa" (cf. Abhidh-k 4:1). S 1:197-205; Marasinghe, pp. 207-13. 20 S 1:111-13, 116-18, 130-31; 132-33; Marasinghe, pp. 185-98. 21 According to the stock Nikaya formula (e.g., D 1:73), by abandoning the five hindrances one attains the first jhana thereby passing from the kamdvacara to the rupivacara; the developed cosmological tradition states that Mara dwells as a rebellious prince among the paranimittavasavattin gods (S 1:133, 1:33-34); see Boyd (n. 5 above), pp. 81-84, 111-19; G. P. Malalasekera, Dictionary of Pali Proper Names, 2 vols. (London: Pali Text Society, 1974), 2:613.


Samiddhi, where visible forms, eye consciousness and dhammas cognizable by the eye exist, there Mara or the manifestation of Mara exist.'"22 Again the Suttanipata defines the armies of Mara that assault the Bodhisatta in what are essentially psychological terms: 435. Dwellingt hush avinga ttainedt he higheste xperience,m y mindh as no regard for sensual desires. See the purity of a being. 436. Sensuald esirei s called your [Mara'sf]i rsta rmy,d iscontenty our second; your third is called hunger and thirst, your fourth craving. 437. Your fifth is called tiredness and sleepiness, your sixth fear. Your seventh is doubt, deceit and obstinacy your eighth . . . 439. Namuci, this is your army-the attacking (force) of the Dark One [[[Mara]]].N ot being a hero one does not conqueri t, but havingc onqueredit one gains happiness.23

In the Samyutta Nikaya, the daughters of Mara too are presented as having a similar psychological reality: "Then Craving, Discontent, and Lust, the daughters of Mara, approached the Blessed One [the Buddha]. Having approached they spoke thus to the Blessed One: 'Ascetic, we would serve at your feet.' Now the Blessed One paid no attention, since he was freed in the unsurpassable, complete destruction of attachments."24I t is surely to read nothing into these texts to say that the descriptions of the Bodhisatta's/Buddha'se ncounter with Mara's armies and daughters represent vivid descriptions of the psychology of the Buddha before and after his awakening. The Bodhisatta/Buddha has wrestled with certain mental states-Mara, his armies, his daughters-and defeated them. That is to say, particular psychological states are described in terms of an encounter with beings with cosmological significance-or vice versa.25 22 S 4:38-39: "maro maro ti vuccati. kittavata nu kho bhante maro va assa marapafifiatti va ti. yattha kho Samiddhi atthi cakkhum atthi ripa atthi cakkhu-vifiianam atthi cakkhu-vififiana-vififiatabbdah amma, atthi maro va mara-pafifiattvi a." 23 Sn 435-39: "tassa m' evam viharato pattass' uttama-vedanam / kamesu napekhate cittam passa sattassa suddhatam / / kama te pathama sena dutiya arati vuccati / tatiya khuppipasa te catutthi tanha pavuccati / / paficami thina-middham te chatthabhiru pavuccati / sattami vicikiccha te makkho thambho te atthamo //... / / esa namuci sena kanhassabhippaharani / na nam asuro jinati jetva ca labhate sukham" (trans. adapted from K. R. Norman, trans., The Group of Discourses: Revised Translation with Introduction and Notes [[[Oxford]]: Pali Text Society, 1992]). 24 S 1:124.26-30: "atha kho tanha ca arati ca raga ca mara-dhitaro yena bhagava ten' upasamkamimsu. upasamkamitva bhagavantam etad avocum: pade te samana paricarema ti. atha kho bhagava na manasakasi yatha tar anuttare upadhi-samkhaye vimutto." See also Sn 835; Nd 1:181. 25 The fact that the armies of Mara here in part overlap with the five hindrances of sensual desire (kdma-cchanda), aversion (vydpada), tiredness and sleepiness (thina-middha), excitement and depression (uddhacca-kukkucca), and doubt (vicikiccha) underlines the point made already about the particular psychological interpretation of Mara in terms of the five hindrances.


I do not wish, however, to suggest that a psychological interpretation of such figures as Mara is the whole story. I am not claiming that all ancient readers or hearers of these "texts" would have conceived of Mara's daughters and armies simply as mystic symbols of particular mental states. No doubt for many, Mara, his daughters, and his armies would have had a reality as autonomous beings apart from their own mental states. I do want to claim, however, that a psychological interpretation would have made sense to the authors and readers of these texts. Yet in making such a claim I do not wish to imply that a psychological reading somehow reveals the "true" and "real" significance of the various cosmological beings-the significance intended by the Buddha but which the Buddhist tradition had to compromise in the face of popular belief, and which we in the late twentieth century are at last privileged to access. The Buddhism of the Nikayas embraces the notion of rebirth, and the account of different realms of existence occupied by a variety of beings is integral to that. The categories of "mythic symbol" and "literally true" are modern and are bound up with a complex ontology that has been shaped by a particular intellectual and cultural tradition. Thus to approach what, for the want of a better term, we call the mythic portions of the Nikayas with the attitude that such categories as "mythic symbol" and "literally true" are absolutely opposed is to adopt an attitude that is out of time and place. It seems to me that in some measure we must allow both a literal and a psychological interpretation. Both are there in the texts. The equivocation between cosmology and psychology is particularly clear in a passage of the Kevaddha-sutta.26T he Buddha tells of a certain bhikkhu who wished to discover where the four great elements (mahabhita) ceased without remainder (aparisesa nirujjhanti). It seems that we must understand this as wishing to know the full extent of the conditioned world-both physical and mental. The bhikkhu appears to have been a master of meditation, for we are told that he attained a state of concentration in which the path leading to the gods appeared to his concentrated mind ("tatharupam samadhim samapajji yatha samahite citte deva-yaniyo maggo paturahosi"). He then proceeds to approach the gods of ever higher levels to pose his question until eventually he finds himself in the presence of Mahabrahma himself, who confesses that he cannot answer the question and suggests that he return to the Buddha to put this question to him. The Buddha answers that the four elements cease, not "out there" in some remote outpost of the universe, but in "consciousness" (viiinna).27 This account states very clearly how specific psychological states-in this instance, the mind concentrated in the 26 D 1:215-23. 27 "viiiiiinam anidassanam anantam sabbato paham" (D 1:23), interpreted by Buddhaghosa (Sv 2:393) as referring to nibbana.


various levels of meditation-give access to particular cosmological realms. Thus the bhikkhu is explicitly described as at once making a journey through various levels of the cosmos and making an inner, spiritual journey-a journey of the mind. In the light of an extremely suggestive article by Peter Masefield, it seems that instead of being misled into searching for meaning in terms of the categories of literal truth and mythic symbol, we should understand the Nikayas' reference both to a cosmic hierarchy of beings (humans, devas and brahmis) and to a psychological hierarchy of mental states (levels of jhana) as paralleling the Upanisadic categories of "with reference to the gods" (adhidaivatam) and "with reference to the self" (adhyatmam): that is, "reality" may be viewed either from the perspective of an exterior world (brahman) or from the perspective of an interior world (atman) that are in some sense-though, in the case of Buddhist thought, not an absolutist metaphysical one-the same.28 Thus Masefield suggests that to talk or conceive of Mara as a cosmic entity on the one hand and as psychological forces on the other is essentially to shift from the adhidaivatam to the adhyatmam perspective.29 I am persuaded that Masefield has indeed identified here a way of thinking that runs very deep in the Indian philosophical tradition, and while the importance of this way of thinking may be acknowledged in the context of the Vedas and Hindu and Buddhist tantra, it is insufficiently understood in the context of early Buddhism. Turning from the Nikayas to the Abhidharma, two full systematic accounts of Buddhist cosmology survive: that of the Theravadin Abhidhamma and that of the Sarvastivada-Vaibhasika Abhidharma. These two accounts are remarkably similar in broad outline and in fact also agree on many points of detail. This again suggests that the basic cosmology should be regarded as having been formulated relatively early since it forms part of the common heritage of ancient Buddhism. In what follows, I shall be drawing on both the Pali Theravadin traditions and also, at points, Vasubandhu's Abhidharmakosa for the traditions of the Sarvastivadins.

One of the general concerns of the Abhidharma is to provide a detailed and complex hierarchy of consciousness. The classic Theravada scheme of eighty-nine or 121 "consciousnesses" (citta) begins with unskillful consciousnesses at the bottom, followed by consciousnesses that concern the mechanics of bare awareness of the objects of the five senses, and then by skillful sense-sphere consciousnesses; next come the various formsphere and formless-sphere consciousnesses that constitute the jhinas, 28 Masefield (n. 5 above). The Upanisadic locus classicus for the terms is Brhadaranyaka 2.3. 29 Masefield, p. 93, n. 32.


or meditation attainments; finally, we have the world-transcending (lokuttara) consciousnesses that constitute the mind at the moment of awakening itself.30 The basic structure of this hierarchy of consciousness parallels quite explicitly the basic structure of the cosmos: consciousness belongs to the sense sphere (kamavacara), the form sphere (ripavacara), or the formless sphere (arupavacara); beings exist in the sense world (kama-dhatu, kama-loka), the form world (ripa-dhdtu, riipa-loka), or the formless world (aripa-dhatu, aripa-loka). As well as laying down a more precise hierarchy of consciousness, the Abhidhamma also finalized the structure of the cosmos: both Theravadin and northern sources detail thirty-one basic realms.31 The basic structure of this cosmos, along with its psychological parallel, is set out in figure 1. In detailing the types of consciousness that beings reborn in particular realms are able to experience, the Abhidhamma provides a further indication of the parallel between the psychological order and the cosmological order.32 Beings in the lowest realms (hell beings, animals, hungry ghosts, Asuras) can only experience sense-sphere consciousness; beings in the human realm and the heavens of the sense sphere characteristically experience sense-sphere consciousness but can in special circumstances (i.e., when attaining jhana) experience form-sphere and formless-sphere consciousness. Beings in the form and formless worlds characteristically experience form and sense-sphere consciousness respectively; both may experience certain forms of both skillful and unskillful sense-sphere consciousness, but not those associated with hatred and unpleasant feeling.33 The logic governing this arrangement is as follows: A being in one of the lower realms must experience at least a modicum of skillful consciousness or else, never being able to generate the skillful kamma necessary to condition rebirth in a higher realm, he or she is stuck there forever. Similarly, beings in the Brahma worlds must experience some unskillful consciousness, otherwise their kamma would be exclusively skillful, and they would be able to remain forever in these blissful realms where no 30 Abhidh-s 1-5 (citta-samgaha-vibhdga); cf. Vism 14:83-110. 31 Vibh 422-26; Vism 7:40-44, 13:29-65; Abhidh-s 22-24; Abhidh-k 3:1-3. Theravadin sources enumerate eleven realms in the kamadhdtu (four descents, the human realm and six heavens), sixteen in the rupadhitu (three each for the first three jhana realms and seven-including unconscious beings and five Pure Abodes-for the fourth), and four in the arapadhatu; Abhidh-k enumerates ten in the kamaloka (missing is the realm of asuras from the descents), seventeen in the ripaloka (exchanging unconscious beings for two further basic fourth dhydna realms), and four in the arupaloka; bhasya to Abhidh-k 3:2b-d records that the Kasmiris accepted only sixteen realms in the fourth dhydna while La Vallee Poussin, trans. (n. 2 above), 2:3, n. 1, records a number of other slight variations in the northern sources.

32 Abhidh-av 182-289 ("bhumi-puggala-vasena cittuppatti-niddeso"). 33 The kind of consciousness that is characteristic of a being is essentially a function of a being's bhavatiga-citta;s ee R. Gethin, "Bhavatigaa nd Rebirthi n the Abhidhamma,"in The Buddhist Forum, vol. 3 (London: School of Oriental and African Studies, 1995), pp. 11-35.


COSMOLOGY WORLD (dhatu) |REALM (bhumi) LIFE SPAN Neither Consciousness nor Unconsciousness (nevasan~ndntsainniyatana) 84,000 aeons WWORLDS LD Nothingness (akincannayatana) 60,000 aeons O(arupadhtu) Infinite Consciousness (vinnidnaincyatana) 40,000 aeons (arcpeadhatu) IIn finite Space ((aakka(asrcainpaahvcagcyaartaa)n a) 20,000 aeons

The Supreme (akanittha)

The Clear-sighted (sudassin)

The Lovely (sudassa)

The Serene (atappa)

The Durable (aviha)

PURE ABODES

(suddhdvasa)

16,000 aeons 8,000 aeons

4,000 aeons

2,000 aeons

1,000 aeons

Unconscious Beings (asanna-satta) 500 aeons I 500 aeons I ft Complete Beauty (subha-kinha) 64 aeons
destruction Boundless Beauty (appamana-subha) 32 aeons by wind Limited Beauty (paritta-subha) 16 aeons

1 Streaming Radiance (dbhassara) 8 aeons
destruction Boundless Radiance (appamanabha) 4 aeons by water Limited Radiance (parittdbha) 2 aeons
The masters of the creations of others (paranimmita-vasavattin)
Those who delight in creation (nimmana-ratin)
The Delighted (tusita) l
The Yama Gods (yama) DE
The Thirty-Three Gods (tavatimsa) (
The Gods of the Four Kings (catummaharajika)
Human Beings (manussa)
Jealous Gods (asura)
Hungry Ghosts (petti-visaya)
Animals (tiracchanayoni)
Hell Beings (niraya)
1 aeon
1/2 aeon
1/3 aeon
HAPPY
ISTINIES
sugati)
DESCENTS
(apaya)
128,000 divine years
64,000 divine years
16,000 divine years
8,000 divine years
2,000 divine years
500 divine years
variable
unspecified
unspecified
unspecified
unspecified

- a
FIG. 1.-The thirty-one realms of existence, according to the following Pali sources: Abhidhammatthasangaha Visuddhimagga 7:40-44, 8:29-65; Dighanikayatthakathatika 1:217.

WORLD OF PURE FORM

(rupadhatu)

Great Reward (vehapphala)

ft
destruction

by fire

Great Brahma (mahabrahma)

Brahma's Ministers (brahma-purohita)

Brahma's Retinue (brahma-parisajja)

WORLD OF THE FIVE SENSES


(kamadhatu)



unpleasant bodily or mental feeling ever occurs, escaping dukkha permanently rather than only temporarily (albeit for an aeon or two). Finally, beings such as humans who are in the middle of the hierarchy are evenly poised; they may experience the most unskillful kinds of consciousness or they may experience the most skillful-they may go right to the bottom or right to the top. A point of particular significance that emerges from this is that, from the perspective of Abhidharma, to shift from talk about levels of existence to talk about levels of the mind is to continue to talk about the same thing but on a different scale. What is involved in moving from the psychological order (the hierarchy of consciousness) to the cosmological order (the hierarchy of beings) is essentially a shift in time scales. The mind (of certain beings) might range through the possible levels of consciousness in a relatively short period-possibly in moments. A being, in contrast, exists at a particular level in the cosmos for rather longer- 84,000 aeons in the case of a being in the realm of "neither consciousness nor unconsciousness"-and to range through all the possible levels of being is going to take a very long time indeed.34 The fact that what we are talking about here is a change of scale is exactly brought out by the Abhidharma treatment of "dependent arising" (pratityasamutpada). This law that governs the process of things, whether the workings of the mind or the process of rebirth, is always the same. Thus the Abhidharma illustrates the operation of the twelve links of dependent arising either by reference to the way in which beings progress from life to life or by reference to the progress of consciousness from moment to moment: from one perspective we are born, live, and die over a period of, say, eighty years; from another we are born, live, and die in every moment.35 In chapter 3 of the Abhidharmakosa, Vasubandhu in fact discusses these different scales for the interpretation of pratityasamutpdda precisely in the context of his exposition of cosmology (vv. 20-38). In general, traditional Buddhist cosmology as expounded in the Nikayas and Abhidhamma must be understood as at once a map of all realms of existence and an account of all possible experiences.


THE EXPANSION AND CONTRACTION OF WORLD-SYSTEMS


According to Buddhist cosmological systems the universe is constituted by innumerable "world-systems" or "world-spheres" (loka-dhatu, cakkavala) comprising just thirty-one levels of existence.36 Much as the mind 34 Vibh 426; Abhidh-s 24. 35 See Vibh 135-92; Vibh-a 199-200; bhasya to Abhidh-k 3:24 (La Vallee Poussin, trans., 2:65-66); cf. R. Gethin, The Buddhist Path of Awakening: A Study of the Bodhi- Pakkhiyd Dhammd (Leiden: Brill, 1992), p. 351. 36 Quite what constitutes a "world-system" is not clear. The term cakkavala does not appear to occur in the four primary Nikayas. Strictly a cakkavala (cf. Skt cakravala and Buddhist Sanskrit cakravada) refers to the range of mountains surrounding the world; the


is not static or stable, neither, on a grander scale, are world-systems; they themselves go through vast cycles of expansion and contraction. According to the exegetical traditions of both the Theravadins and Sarvastivadins, the formula I quoted from the Aggainnia-suttar,e ferring as it does to the rebirth of beings in the realm of Radiance (abhassara/abhdsvara)37 at the time of world contraction, describes this contraction as the result of destruction by fire. Both Buddhaghosa and Vasubandhu provide some further details about how the destruction proceeds.38 According to Buddhaghosa, world-systems contract in great clusters-he speaks of a billion (koti-sata-sahassa) world-systems contracting at a time.39 Both writers describe how, when they contract, world-systems contract from the bottom upward. Thus in the case of destruction by fire, the fire starts in the lower realms of the sense sphere and having burned up these, it invades the form realms; but having burned up the realms corresponding to the first jhdna/dhydna, it stops. The realms corresponding to the second, third, and fourth jhdnas, and the four formless realms, are thus spared the destruction. But destruction by fire is not the only kind of destruction, merely the most frequent-water and wind also wreak their havoc. When the destruction is by water, the three realms corresponding to the second jhdna are also included in the general destruction, while the destruction by wind invades and destroys even the realms corresponding to the third jhdna. Overall, only the seven realms corresponding to the fourth term is then used to refer to a single "world-system" as constituted by the various realms that make up the world of sense-desire; Buddhaghosa says that there are an infinite number of such world-systems (Vism 7:40-44). The term used as a gloss for cakkavala by Buddhaghosa here is loka-dhatu, which seems to be the preferred term in the Nikayas. Thus the Afiguttara Nikaya (5:59-60) talks of a Mahabrahma ruling over a thousand such world-systems, while the Majjhima Nikaya (3:101-2) talks of Brahmas ruling over as many as a hundred thousand world-systems. It thus seems that world-systems that are distinct and self-contained at the lower realms of existence are not necessarily so at higher levels of existence. However, Buddhist tradition does not conclude that one should therefore talk of there being only one all-embracing Brahma-world. In fact, A 5:59 already talks in terms of thousands of Brahma-worlds, and the ancient conception of the thousandfold world-system, the twice-thousandfold world-system (embracing 1 million worldsystems), and the thrice-thousandfold world-system (embracing 1 trillion world-systems according to Pali sources and 1 billion according to northern) (see A 1:227-28, Mp 2:340- 41, Abhidh-k 3:73-74) seems to imply a kind of pyramidal structure of world-systems: units of thousands of world systems (i.e., sense-sphere world-systems) are governed by a Mahabrahma, and units of a thousand such Brahma realms are in turn governed by Brahmas of yet higher realms, and so on. Whatever, as the Atthasalini says (pp. 160-61), there is no end to the hundreds and thousands of world-systems: if four Mahabrahmas in Akanittha were to set off at a speed which allowed them to traverse a hundred thousand world-systems in the time it takes a swift arrow to pass over the shadow of a palm tree, they would reach nibbdna without ever seeing the limit of world-systems. 37 Sv 1:110; Vism 13:30; bhasya to Abhidh-k-bh 3:90c-d. 38 Vism 13:32-55; Abhidh-k-bh 3:89-90, 100-102; cf. Reynolds and Reynolds (n. 6 above), pp. 305-27. 39 Vism 13:31, 40-41, 55.


jhana and the four formless realms are never subject to this universal destruction.40 So what becomes of the beings that occupy the lower realms when fire, water, and wind wreak their destruction? They cannot just disappear from samsdra; they must go somewhere. Here we touch upon a question which posed something of a problem in the Buddhist tradition and to which its answers are not entirely consistent. The simple answer that Buddhaghosa gives in the Visuddhimagga is that at the time of the destruction of a world-system by fire, all the beings that occupy the lower realmsincluding hell beings (nerayika)-are reborn in the Abhassara Brahma realm (corresponding to the second jhdna) or above it. But since rebirth in a Brahma realm can only occur as a result of the practice of thejhanas, Buddhaghosa has a problem. The chaos and hardships that are a prelude to the destruction of the world are hardly conducive to the practice of jhdna. Moreover, certain beings simply do not have the capacity to attain jhdna even if they try.

There is no rebirth in the Brahma world without jhdna, and some beings are oppressed by the scarcity of food, and some are incapable of attaining jhdna. How are they reborn there? By virtue of jhana acquired in the Deva world. For at that time, knowing that in a hundred thousand years the aeon will come to an end, the sense-sphere gods, called "Marshals of the World," loosen their headdresses and, with disheveled hair and pitiful faces, wiping their tears with their hands, clothed in red and wearing their garments in great disarray, come and frequent the haunts of men saying, "Good sirs, a hundred thousand years from now the aeon will come to an end: this world will be destroyed, the great ocean will dry up, and Sineru, king of mountains, will be burnt up and destroyed. The destruction of the world will reach the Brahma world. Develop loving kindness, good sirs. Develop compassion, sympathetic joy, and equanimity. Take care of your mothers and fathers; honor the elders of the family." Hearing their words, both men and the deities of the earth are for the most part moved; they become kind to one another, and making merit by loving kindness and so on, they are reborn in the Deva world. There they enjoy the food of the gods and having completed the initial work on the air kasina, they attain jhdna. However there are others who are reborn in the Deva world by virtue of their kamma "that is to be experienced at an unspecified time," for there is certainly no being wandering in samsdra devoid of kamma that is to be experienced at an unspecified time. They also similarly acquire jhdna there [in the Deva world]. 40 Vism 13:55-62 describesd estructionb y fire, water, and wind; Vism 13:65 and Abhidh-k-bh3 :102 detail the sequencea nd frequencyo f destructionb y these threee lementsa nda rei n completea greements:e ven cycles of seven destructionbs y firef ollowed by one by water( fifty-sixd estructions)f;o llowedb y one cycle of seven destructionsb y fire followedb y one by wind (sixty-fourd estructions)t;h ust he Brahmasw ho live in the Subhakinha/Subhakrtsnar ealms-the highest of the third jhana/dhyana realms-have a life spano f sixty-foura eons.


So all beings are reborni n the Brahmaw orldb y virtueo f the attainmenot f jhana.41 For Buddhaghosa, at the time of the contraction of a world-system, all the beings occupying the lower realms should be understood as being reborn in those higher Brahma worlds that escape the destruction-this is true even of the beings in the lower realms of hell. When all else fails, this comes about by virtue of the fact that there is no being in samsdra that has not at some time or other performed the kamma necessary for rebirth in the happy realms of the sense sphere. Thus even beings born in hell realms as the result of unwholesome kamma will always have a latent good kamma that can come to fruition at the time of the pending contraction of the world-system; this is their "kamma to be experienced at an unspecified time" (aparapariya-vedaniya-kamma).4S2u ch beings are first reborn in a sense-sphere heaven, where they subsequently cultivate jhdna leading to rebirth in the Brahma worlds. What follows from this view of the matter is that all beings in samsdra are regarded as having dwelt at some time in the Brahma realms corresponding to the second, third, and fourth jhanas; moreover, periodically-though the periods may be of inconceivable duration-all beings are regarded as returning to these realms. It seems, however, that some in the Buddhist tradition were not entirely happy with the understanding of the matter presented by Buddhaghosa. Commenting on the phrase, "when the world contracts beings are for the most part born in the realm of Radiance," as it occurs in the Brahmajala Sutta, Buddhaghosa states that "'for the most part' [yebhuyyena] is said because there are other beings who are born either in higher Brahma realms or in the formless realms."43D hammapala,h owever, in his subcommentary on the text by Buddhaghosa, adds:

41 Vism 13:33-35: "jhanam vina natthi brahma-loke nibbatti; etesafi ca keci dubbhikkha- pilita keci abhabba jhanfdhigamaya. te katham tattha nibbattanti ti. devaloke patiladdha- jhana-vasena. tada hi vassa-sata-sahassass' accayena kapputthanam bhavissati ti. loka-byuha nama kamavacara-deva mutta-sira vikinna-kesa ruda-mukha assuni hatthehi punchamanar atta-vattha-nivatthaa tiviya virupa-vesa-dharinoh utva manussa-pathev icaranta evam arocenti: marisa marisa ito vassa-sata-sahassassa accayena kappa-vutthanam bhavissati; ayam loko vinassissati, maha-samuddo pi ussussissati, ayaii ca maha-pathavi sineru ca pabbata-raja uddayhissanti vinassissanti, yava brahma-loka loka-vinaso bhavissati. mettam marisa bhavetha, karunam muditam upekkham marisa bhavetha, mataram upatthahatha pitarum upatthahatha, kule jetthapacayino hotha ti. tesam vacanam sutva yebhuyyena manussa ca bhumma-devata ca samvega-jata afiamaniiiam mudu-citta hutva mettadini puiinani karitva deva-loke nibbattanti. tattha dibba-sudha-bhojanarmbh uiijitva vayo-kasine parikammam katvajhanam patilabhanti. tad-afife pana aparapariya-vedaniyena kammena deva-loke nibbattanti. aparapariya-vedaniya-kamma-rahithoi samsare samsaranto satto nama natthi. te pi tattha tath' eva jhanam patilabhanti. evam deva-loke patiladdha- jjhana-vasena sabbe pi brahma-loke nibbattanti ti." 42 On aparapariya-vedaniya-kammas, ee Vism 19:14, Abhidh-s 5:52, Abhidh-s-t 131- 32. 43 Sv 1:110: "yebhuyyena ti ye upari brahma-lokesu va aruppesu va nibbanti, tad-avasese sandhaya vuttam."


"or in world-systemso thert hant hose in the processo f contracting"is the alternativet o be understoodb y the wordo r. Fori t is not possiblet o considert hat all beings in the descents at that time are born in form or formless existence, since it is impossible for those beings in the descents with the longest life spans to be reborn in the human realm.44 Dhammapala's problem with Buddhaghosa's account seems to be that it fails to take account of the case of beings who, for example, commit one of the five great anantariya-kammas (killing one's mother, father, an arhat, wounding a Buddha, splitting the Samgha) toward the end of an aeon. Such beings must as a result surely be born in the hell realms, and yet the aeon might end before they had lived out the result of that kamma. Dhammapala therefore concludes that such beings must be reborn in the hells of other world systems.45 Looking further afield in Buddhist sources we find other instances of both Buddhaghosa's position and Dhammapala's position on what happens to beings in the lower realms when a world-system contracts. For example, in chapter 3 of the Kosa, Vasubandhu writes: When not a single being remainsi n the hells, the worldh as contractedt o this extent:n amelyb y the contractiono f the hells. At thatt ime any being who still has karmat hatm ustb e experiencedin a hell is throwni nto the hells of another world-system[ thati s not contracting].46 In chapter 8, however, Vasubandhu comments that at the time of the contraction of a world-system, "all beings of the lower realms produce dhydna of the form-realm because of the special occurrence of skillful dharmas."47 Yasomitra comments that in these circumstances dhydna 44 DAT 1:201: "aruppesuv a ti va-saddena samvattamna-lokadhi aa-aloak-laodkhahdih tesu va ti vikappanam veditabbam. na hi sabbe apaya-sattat ada rupaprupa-bhavesuup pajjantit i sakka vifiniatum, apayesu dighatamfyukanam manussalokuppattiya asambhavato." The fact that the DAT comments here in this way when Vism-t fails to make any comment on Buddhaghosa's account of the contraction of the world is perhaps further evidence that the authors of the Nikaya tikas and Vism-t are not the same; see L. S. Cousins, "Dhammapala and the Tika Literature," Religion 2 (1972): 159-65; P. Jackson, "A Note on Dhammapala( s)," Journal of the Pali Text Society 15 (1990): 209-11. 45 Compare Kv 476.

46 Abhidh-k-bh 3:89: "yada narakesv eka-sattvo navasisto bhavati iyatayam lokah samvrtto bhavati / yaduta naraka-samvarttanya/ yasya tadanim niyatam naraka-vedaniyam karma dhriyate sa lokadhatv-antara-narakesuk sipyate." 47 This is stated by way of explanation of the last of three ways in which dhyana belonging to the ripadhatu may be produced: by the force of conditions (hetu), defined as repeated practice (abhlksndbhyasa), by the force of karma leading to rebirth in a higher realm coming to fruition, and also by the nature of things (dharmata) (Abhidh-k-bh 8: 38c-d: "rupadhataud hyanotpadaname tabhyam ca hetu-karma-balabhyamd harmatayapic a samvartani-kale. tadanim hi sarva-sattva evadhara-bhumikas tad dhyanam utpadyanti kusalanam dharmanam udbhuta-vrttitvat.")


arises without any instruction because of the existence of the trace (vasana) of previous dhydna attainment.48 Another cosmological treatise current in Southeast Asia is the eleventh- or twelfth-century Lokapaiiatti. Like the Visuddhimagga of Buddhaghosa, the Lokapainnatti states that at the time of the contraction of a world-system, beings in the lower realms are reborn first in the kamadhatu and then in the Abhassara realm after practicing the second jhdna; there is no mention of being reborn in the hells of other world systems.49 The much later Theravadin source, "Three Worlds according to King Ruang," on the other hand, takes the line of Dhammapala and chapter 3 of the Kosa, stating that hell beings may be reborn in the hells of world-systems that are not contracting.

What are relative merits of these two perspectives regarding what happens to beings in lower realms at the time of world contraction? The position represented by Dhammapala, Kosa chapter 3 and the Triphum of Phya Lithai-namely, that they are reborn in the lower realms of worldsystems that are not in the process of contracting-appears to be more in keeping with the laws of karma and, for this reason, the more carefully considered: beings who murder their mothers, fathers, arhats, wound a Buddha, or split the Samgha must surely experience the results of their actions whether or not a world-system contracts.51 Yet this makes the 48 Vyakhya to Abhidh-k-bh 8:38c-d: "upadesam antarenayatah purva-dhyana-vasanayam satyam dhyanotpattir iti."

49 Although composed in Pali, the Lokapainnattia ppears to be based directly on Sanskrit traditions rather than the traditions of the Sri Lankan Theravfda; it corresponds closely to the Lokaprajnapti translated into Chinese by Paramartha in 558 C.E. (Denis, trans. and ed. [n. 6 above], 2:ii). The position recorded here on what happens to hell beings at the time of the contraction of a world-system appears to reflect exactly the position of Paramartha'st ranslationo f the Lokaprajiapti (Denis, trans. and ed., 1:194, 2:225-26). 50 Reynolds and Reynolds (n. 6 above), p. 308. 51 The Nikayas and Agamas for their part prefer to speak of the length of time beings will suffer in hell realms by way of simile rather than specific numbers of years or aeons (see Kokaliya-sutta, S 1:149-53; A 5:170-74; Sn 123-31; cf. bhasya to Abhidh-k 3:84). Vibh 422:26, which deals with age limits in the various realms of existence, says nothing about the hell realms, and begins with the human realm; the commentary (Vibh-a 521) states that kamma is what determines the life span of beings in the descents-as long as kamma is not exhausted beings do not pass from those realms; the Anutika apparently adds (see Nanamoli, trans., The Dispeller of Delusion, 2 vols. (London: Pali Text Society, 1987-90), 2:299, n. 7) that the life span in Avici is an antarakappa (a sixty-fourth of a mahakappa). Abhidh-s 23 (chap. 5, verse 21) states that there is no definite age limit for beings in the four descents and for humans; the length of time spent in these realms is dependent on the specific kamma that brought about the rebirth. As far as human beings are concerned this comment seems to be made with reference to the tradition-found in the Cakkavattisihanada-sutta(D 3:58-79) and Mahapadana-sutta( D 2:1-54)-that the life span of humans varies from ten years to 80,000 years at different periods within an aeon, and thus does not mean that humans can outlive the aeon. Vasubandhu too states (Abhidh-k 3:83) that the life span of beings in Avici is one antarakalpa (an eightieth of a mahakalpa according to northern tradition). Malalasekera (n. 21 above) comments (s.v. Avici, Devadatta) that Devadatta is destined to suffer in Avici for 100,000 aeons, but the source he cites (Dhp-a 1:148) strictly says only that at the end of 100,000 aeons Devadatta


alternative tradition-that all beings are reborn in the Brahma realmsall the more interesting and, I think, significant. It is, as it were, the lectio difficilior. Why should Buddhaghosa, Vasubandhu, and the Lokapainatti preserve and hand down a tradition that is so obviously problematic? In order to answer this question I would like to turn first to consider the theoretical account of the stages of the Buddhist path, since it seems to me that, viewed in the light of each other, the accounts of the stages of the path and the process of the expansion and contraction of the universe reveal clues about the unspoken assumptions that lie at the heart of Indian Buddhist thought.


COSMOLOGY AND THE BUDDHIST PATH


What should perhaps be regarded as the classic Nikaya account of the stages of the Buddhist path is found repeated in various suttas of the slakkhandha-vagga of the Digha Nikaya, and also, with slight variations, in several suttas of the Majjhima Nikaya.52 This account can be summarized in simple terms as follows: on the basis of the practice of good conduct (sila), the bhikkhu practices meditation; by this means, he abandons the five hindrances and attains the first jhdna. Attaining, successively, the second and third jhanas, the bhikkhu is described as further refining his concentrated mind until he eventually attains and abides in the fourth jhana. This is described as a state of "purity of equanimity and mindfulness" (upekkha-sati-pdrisuddhi); "he suffuses his body with his mind that has been thoroughly purified and cleansed."53 We then have a description of a series of eight (in the Digha) or three (in the Majjhima)54 different attainments, each one of which is introduced by precisely the same formula: "When his mind has become concentrated thus, when it is thoroughly purified and cleansed, stainless, the defilements absent, when it has become sensitive, workable, steady, having attained imperturbability, he inclines and applies his mind to... "55 In other words, having stilled the mind to the level of the fourth jhana, the bhikkhu has brought his mind to an extremely refined state that is suitable and fit for various tasks: the development of knowledge of the interdependence of will become a paccekabuddha, and not that he will spend that period continuously in Avici.

52 D 1, passim; M 1:178-84, 344-48, 3:33-36, 134-37; cf. M 1:267-71. See Gethin, The Buddhist Path to Awakening (n. 35 above), pp. 207-8. 53 D 1:75-76: "so imam eva kayam parisuddhena cetasa pariyodatena pharitva nisinno hoti."

54 At M 3:36 there is just one attainment. The attainments are the eight vijjas (Vism 7:20), the last six of which are often referred to as abhiniii (e.g., D 3:281) and the last three as vijjd (e.g., M 55 1:482). D 1:76-83 (passim): "evar samahite citte parisuddhe pariyodate anafigane vigatupakkilese mudubhute kammaniye thite anejjappatte."


consciousness and the body; the creation of a mind-made body; the acquiring of certain extraordinary powers (the iddhis and other abilities, elsewhere termed higher knowledges or abhifnns). Lastly he may apply this mind to the gaining of the knowledge of the destruction of the asavas, the knowledge of suffering, the arising of suffering, the cessation of suffering, and the way leading to the cessation of suffering; he then knows that for him birth is destroyed and that there is no future rebirth after the present one.

The story of the bhikkhu in the Kevaddha-sutta to which I referred earlier is in fact a rather precise parable of this understanding of the progress of the Buddhist path. The bhikkhu of the Kevaddha-sutta resorts to increasingly subtler states of consciousness and/or levels of the cosmos in order to seek an answer to the question of the ultimate nature of the universe; and yet, having come to the furthest reaches of the universe, he does not find his question satisfactorily answered but must return to the Buddha and be instructed to reorient his quest. Similarly, the bhikkhu who attains jhana does not come to the end of the path but must turn his attention elsewhere in order finally to understand the nature of suffering, its cause, it cessation, and the path leading to its cessation. It is in the light of this close correspondence that exists in Buddhist literature between journeys through the realms of the cosmos and inner journeys of the mind that the significance of the accounts of the expansion and contraction of the universe begins to be revealed. Stanley Tambiah has already drawn attention to this in some comments made in his study of the Thai forest monastic tradition-comments which are, however, brief and do not articulate the nature of the parallels entirely accurately. 57 Buddhist cosmology-in general, but especially in the account of the contraction and expansion of world-systems-provides us with a poetic, imaginative, and mythic counterpart to accounts of the stages of jhana attainment. Reading accounts of the Buddhist path alongside tales of the universe's end and beginning is the way to enter more fully into the thought-world of ancient Indian Buddhism. In particular, what is revealed in the cosmological accounts is the understanding of the nature of the fourth jhdna: both the theoretical accounts of the stages of the path and the mythic descriptions of the contraction of the world-system converge on the fourth jhana. That the mythic account of the contraction of a world-system can be read as paralleling a meditator's progress through the successive dhyanas 56 D 1:84: "khina jati vusitam brahmacariyam katam karaniyam naparam itthattaya ti." 57 S. J. Tambiah, The Buddhist Saints of the Forest and the Cult of Amulets (Cambridge: Cambridge University Press, 1984), pp. 49-52. Tambiah confusingly describes the Abhassara realm as arupa at one point and creates, to my mind, a rather misleading "dyadic opposition between material states and formless states."


is brought out explicitly in the following passage from the Abhidharmakosa which comments on how, at the time of contraction, fire, water, and wind destroy the successively higher levels of the world-system: In the first dhydna thinking and reflection are imperfections; these are similar to fire since they burn through the mind. In the second dhydna joy is the imperfection; this is like water since, by association with tranquility, it makes the senses soft.... In the third dhydna out-breaths and in-breaths [are imperfections]; these are actually winds. In this way the subjective [adhydtmika] imperfection in a dhydna attainment is of the same nature as the objective [bdhya] imperfection in the corresponding dhyana rebirth.

A mediator's entering the fourth jhana thus marks the temporary attainment of a state of consciousness that is secure in its freedom from disturbances and defilements. For just as the realms of existence corresponding to the fourth jhdna can never be reached by the ravages of fire, water, or wind, so the mind in the fourth jhdna is undisturbed either by the gross objects of the five senses or the subtler movements of the mind still remaining in the first, second, and third jhdnas. What is more, viewed from the cosmological perspective of the expansion and contraction of the world-system and the periodic return of beings to the Brahma realms, in stilling the mind to the level of the fourth jhdna, the bhikkhu is returning to a state experienced long ago. The cultivation of thejhdnas becomes almost a kind of Platonic recollection of something long forgotten, of something one does not remember one knows. The recovery of the fourth jhdna is a return to a basic or fundamental state-a stable and imperturbable state of the universe and also of the mind.

In saying, however, that the realms of existence corresponding to the fourth jhana are always there, it is, of course, necessary to keep firmly in mind Buddhist principles of impermanence. The realms of the fourth jhana do not have some kind of mysterious existence of their own; these realms always exist in the sense that there are always beings "in" these realms, although the particular beings occupying these realms continually 58 Bhasya to Abhidh-k 3:100c-d: "prathame hi dhyane vitarka-vicara apaksalah / te ca manasah paridahakatvad agni-kalpah / dvitiye pritir apaksala / sa praSrabdhi-yogenasraya- mrdu-karanada p-kalpa /.. / trtiye dhyane asvasa-prasvasah/ te ca vayava eva / iti yasyam dhyana-samapattau yathabhuta adhy8tmiko 'paksalah tasyam dhyanopapattau tathabhuto bahya iti" (cf. Abhidh-di 115-16). 59 Incidentally, this way of looking at the progress of the practice of meditation as a return to a kind of primordial state is not without parallels elsewhere in Indian tradition. The practice of yoga as presented in the Yoga-sitras of Patafijali is also essentially a species of return: a reversal of the stages of the evolution of the tattvas from prakrti. Thus the full manifestation of prakrti with the appearance of the five senses and their respective objects is what characterizes ordinary human consciousness; by the practice of samddhi the yogin gradually, stage by stage, regains the primordial equilibrium of the three gunas in unmanifest prakrti. The knowledge that discriminates between purusa and prakrti can then be achieved.


change and no individual being can permanently exist in such a realm. The fourth jhdna realms thus do not constitute some kind of permanent substrate of the universe; it is simply that there are always beings "there," or rather beings that exist in the manner of the fourth jhdna. For the Abhassara or Vehapphala, realms are not so much places as modes or ways of being.60 So, to say that periodically the world contracts back as far as the Vehapphala realm is exactly to say that periodically beings return to this manner of being. It is in this sense that the levels associated with the fourth jhdna are basic, fundamental, almost, one might say, primordial. This, it seems, is precisely why they can serve as the stepping-off point for gaining the four formless attainments,61f or developing various extraordinary meditational powers,62 for realizing the liberating knowledge of the path. This, it seems, is precisely why, at the time of his parinibbana, the fourth jhana is the final active state of mind to be experienced by a living Buddha.63

I am now in a position to return to the question I posed above concerning Buddhaghosa's (and others') account of the process of the contraction of world-systems: Why does he preserve an apparently problematic account? The view handed down by Buddhaghosa, which he has no doubt received from the Sinhala atthakathd sources he had before him, seems concerned to emphasize that no being in samsara is without the necessary kamma to enable a skillful rebirth in the kamadhatu as a basis for subsequent rebirth in the realms corresponding to the fourth jhdna; and that there is no being in samsdra without experience of the realms of the fourth jhana-of the states which give close access to the liberating insight of bodhi. In other words, all beings have the capacity to become awakened and indeed all have somewhere in them an experience of a state of mind that is in certain important respects "close" to the awakening state of mind. THE MAHAYANA To anyone familiar with the Mahayana, the suggestion that beings always have within them a capacity to become awakened sounds strangely familiar, and at this point I would like to consider certain parallels that can, 60 Vasubandhu does, however, designate the realms of the rupadhatu as "places" or "locations" (sthana); the drupyadhatu, on the other hand, is without location (asthana). This would seem to be because to the extent that beings of the rupadhatu possess rupa-skandha (they possess the senses of sight and hearing) they must have location. Compare Abhidh-k 2:2-3, 7:3; Y. Karunadasa, The Buddhist Analysis of Matter (Colombo: Department of Cultural Affairs, 1967), pp. 161-62. 61 For example, Vism 10; one should note here that in certain contexts (e.g., Abhidh-s 5; Narada, trans., A Manual of Abhidhamma [[[Kandy]]: Buddhist Publication Society, 1980], p. 64) the four formless attainments are treated simply as modifications of the fourth (or, according to the Abhidhamma reckoning, fifth) jhdna. 62 Vism 12:2, 12-13, 58; Gethin, The Buddhist Path to Awakening (n. 35 above), p. 102. 63 D 2:156.


I think, be found between the cosmological ideas I have been discussing and certain ideas that find expression in Mahayana sutras. Buddhaghosa's account of what happens to beings when a world-system contracts bears a certain resemblance to aspects of an idea we are accustomed to associate with the Mahayana, namely, the tradition of tathdgatagarbha-"that within each being which enables enlightenment to take place."64 Although formulated rather differently, something of the tathagatagarbha way of thinking is, I suggest, present in the cosmological traditions of the Abhidharma. In the context of the Nikaya and Abhidharma understanding of the development of the stages of the Buddhist path, the function of a "trace" left by previous dhydna practice experienced long ago, or of a skillful karma "to be experienced at an unspecified time" which makes for the attainment of the fourth dhydna state, is in significant respects similar to that of the tathagatagarbha in Mahayana thought: both may facilitate and effect enlightenment for deluded beings. This is not to suggest that Buddhaghosa here espouses a doctrine of tathagatagarbha or that tathagatagarbha views have influenced him or that he has influenced the development of tathagatagarbha theory. Rather there appears to be a common Buddhist theme here that finds expression in one way in Buddhaghosa's account of the contraction of a world-system and in another way in the theory of tathagatagarbha.65 While we cannot say that Buddhaghosa's account of the expansion and contraction of a world-system is in all respects equivalent to the theory of tathagatagarbha, we can say that in certain respects it is; there is a certain overlap here. A second area of interest centers on the understanding of the "pure abodes" (suddhavdsa/suddhavdsa) in the Nikayas, Abhidharma, and Mahayana. The Buddhist yogin who has mastered the fourth jhdna has withdrawn the mind from the world of the senses, from the world of ordinary ideas and thoughts, and returned it, as it were, to a refined and fundamental state. From this state of mind he now has the possibility of seeing the world more clearly, seeing it as it truly is, and even, to a limited extent, by the practice of the various meditational powers (such as creating mind-made bodies, etc.), of constructing a different world. This way of thinking is continued and taken further in Mahayana Buddhist thought. For it is in the realm of the fourth dhydna that Bodhisattvas become Buddhas and create their "Buddha fields" and "pure lands." In non-Mahayana texts the five "pure abodes" are regarded as the abodes of "never-returners"( anagdmin), beings who are all but awakened, beings who are in their last life and who will certainly attain arhatship 64 P. M. Williams, Mahayana Buddhism: The Doctrinal Foundations (London: Routledge, 1989), p. 98. 65 Both these expressions are connected with another expression of this theme, namely, tile SautrHntikath eory of "seeds"; cf. P. S. Jaini, "The SautrantikaT heory of Bija," Bulletin of the School of Oriental Studies 22 (1959): 237-49.


before they pass away.66 Rather interestingly, then, according to certain traditions of the developed Mahayana, the Akanistha realm-the highest of the "pure abodes" and of the realms of the fourth dhyana-is occupied not by never-returners about to become arhats but by tenth-stage Bodhisattvas about to become samyaksambuddhas. Having attained Buddhahood in the Akanistha realm, they send out their "creation bodies" (nirmana-kaya) to the lower realms for the benefit of sentient beings. Santaraksita in the Tattvasamgraha explains as follows: 3549. Since their existence is outside samsdra, which consists of the five destinies, the death of Buddhas is not admitted by us; therefore it is their creations that are perceived.

3550. In the lovely city of Akanistha, free from all impure abodes-there Buddhas awaken; but here [in this world] creations awaken.


Kamalasila goes on to comment: Samsara consists of the five destinies comprising hells, hungry-ghosts, animals, gods and men; and since Buddhas exist outside this their mortality is not accepted. How then does one learn of their birth in the family of Suddhodana and others? Accordingly he says that it is their creations that are perceived. Supporting this from scripture he utters the words beginning, "In the Akanistha..." There are gods called the Akanisthas; in a certain place among them the gods are called "those belonging to the pure abodes," for here only the pure noble ones dwell. Among them the highest place is called the Palace of the Great Lord, and there only Bodhisattvas in their last existence who are established in the tenth bhimi are born, while here [in this world] by reason of their sovereignty in that place their creations gain knowledge. Such is the tradition.68 66 Vism 22:56-57; Malalasekera (n. 21 above), s.v. "suddhavasa"; Marasinghe (n. 5 above), p. 262; Abhidh-k-bh 6:42-44 (La Vallee Poussin, trans. [n. 2 above], 4:221-28). 67 Tattvasafigraha 2:1107 (vv. 3549-50): "paficagaty-atma-samsara-bahir-bhavanna martyata / buddhanam isyate 'smabhir nirmanam tu tatha matam / / akanisthe pure ramye 'Suddhavasa-vivarjite/ budhyante tatra sambuddhan irmitas tv iha budhyate."I read ramye 'suddhavdsavivarjitef or Shastri'sr amyes uddhavasavivarjite, although a Tibetant ranslation of apparently the same verse does not recognize the sandhi: "Rejecting the pure abodes, he rightly and completely awakened in the ecstatic abode of Akanistha." (See mKhas grub rje's Fundamentals of the Buddhist Tantras, trans. F. D. Lessing and A. Wayman [The Hague: Mouton, 1968], pp. 22-23.) The implication that the Akanistha realm is somehow apart from the pure abodes is surely problematic, while the phrase "akanistha-bhavane divye sarva-papa-vivarjite"( LafikavataraS utra 269.4) would seem to confirm my emendation. 68 Tattvasafgraha2 :1107: "naraka-preta-tiryag-deva-manusya-bhedenpaa ficagaty-atmakah samsarah tad-bahir-bhutas ca buddha bhagavata ity asiddham martyatvam esam / katham tarhi Suddhodanadi-kulotpattire sam grfuyate/ ity aha nirmanam tu tatha matam iti / etad evfgamena samspandyann aha akanistha ity adi / akanistha nama devah tesam ekadese Suddhfvasa-kayika nama devah / atra hy arya eva guddha avasanti / tesam upari mahesvara-bhavanam nama sthanam / tatra carama-bhavika eva daSabhtmi-pratisthita bodhisattva utpadyante / iha tu tad-adhipatyena tatha nirmanam upalabhyata ity agamah" (cf. G. Jha, trans., The Tattvasahgraha of Shantaraksita with the Commentary of Kamalashila, 2 vols. [[[Delhi]]: Motilal Banarsidass, 1986], 2:1547; Williams, pp. 180-81).


Significantly, a level associated with the fourth dhydna is once more conceived of as in some sense fundamental and primordial-the level upon which the creative activity of Buddhas is based. The extent and precise interpretation of the tradition that Buddhas become enlightened in Akanistha is not, however, entirely clear; the ancient accounts of the career of the Bodhisattva are varied and not always consistent. The exact source of Santaraksita'sq uotation from "scripture" (dgama) is not traced, although the Lafikavatara Sitra (Sagathaka vv. 38-40, 772-74) similarly states that beings become Buddhas in the Pure Abodes "among the Akanisthas of the form-realm" while their creations awaken in this world: 772. I am of Katyayana'fsa mily;i ssuing from the PureA bodeI teach beings dharmat hatl eads to the city of nirvana. 773. This is the ancientp ath;t he Tathagataasn dI have taughtn irvanai n three thousands utras. 774. Thus not in the realm of the senses nor in the formless does a Buddha awaken,b ut amongt he Akanisthaso f the formr ealmw ho aref ree of passionh e awakens.

Taking this tradition at face value, what seems to be being said is that full Buddhahood is attained by a tenth-stage Bodhisattva in the Akanistha realm; after this the "created" or "emanated" body (nirmanakaya) performs the acts of a Buddha beginning with the descent to this world from Tusita, the Heaven of Delight. In other words, Siddhartha Gautama from the time of his conception and birth is a nirmana-kaya of an already fully awakened Buddha. However, such an understanding is not entirely consistent with what is said in the Prajnfapramita literature or in the Dasabhumika about the final stages of the career of the Bodhisattva. The Pancavims'atasdhasrikd-Prajinpdramita ppearst o make no mention of the Pure Abodes or Akanistha in this connection, and it is the ninth-stage Bodhisattva that descends into the womb, takes birth, and sits beneath the tree of awakening, reaching the tenth stage when he becomes a Tathagata.70

69 Laiikvatara Suitra 269.4-9, 361.1-6: "akanista-bhavane divye sarva-papa-vivarjit / nirvikalpah sada yuktag citta-caitta-vivarjitah / balabhijin-vasi-praptah tat-samadhi-gatimgatah / tatra budhyanti sambuddha nirmitas tv iha budhyate / / nirmana-kotyo hy amita buddhanam niscaranti ca / sarvatra balah smvanti dharmam tebhyah pratisrutvd / ... katyayanasya gotro 'ham suddhavasad vinissrtah/ desemi dharmam sattvanam nirvana-puragaminam / / pauranikam idam vartma aham te ca tathagathah / tribhih sahasraih sutranam nirvanam atyadegayan / / kama-dhatau tatharupye na vai buddho vibudhyate / rupa-dhatvakanisthesu vita-ragesu budhyate." 70 6. Lamotte, trans. Le traite de la grande vertu de sagesse de Ndgarjuna (Mahaprajnaparamitasastra), 5 vols. (Louvain: Bibliotheque du Museon and Publications de 1'Institut Orientaliste, 1944-80), 5:2431-32, 2438, 2442-43; E. Conze, trans., The Large Sutra


Although the Dasabhumika once again does not mention the Pure Abodes or Akanistha in connection with the bhumis, it does talk of Bodhisattvas established in the tenth stage as being "mostly the Great Lord [[[mahesvara]]], king of the gods [[[deva-raja]]]."71V arious passages (which must be the source of the Tattvasamgraha tradition quoted above) consistently identify these terms as epithets of the chief of the gods of the Pure Abodes.72 But for the Dasabhumika it is the Bodhisattva of the tenth stage (and not the ninth stage as in the Prajnaparamitaw) ho manifests in a single world-system all the acts of Tathagatas from abiding in the Tusita realm to Parinirvana (the final attainment of nirvana at death), but he appears to do this as Bodhisattva, remaining such and not becoming a full Buddha in the process.73 Moreover, At will he displays the array of the realms of all the Buddhas at the end of a single hair; at will he displays untold arrays of the realms of the Buddhas of all on Perfect Wisdom with the Divisions of the Abhisamayalamkdra (Delhi: Motilal Banarsidass, 1979), p. 165. Eighth-stage Bodhisattvas are here described as enjoying the play of the higher knowledges (abhijinkridanata), seeing Buddha fields (buddha-ksetradarsanatd), and producing their own Buddha fields in accordance with what they have seen ("tesam buddha-ksetranamy atha-drs.tnam sva-ksetra-parinispadanata")T. he commentarial

  • Mahaprajnaparamitdasstra( see Lamotte, trans., 5:2433-35, 2439, 2444) fills

this out and explains that at the eighth stage the Bodhisattva sees the bodies of the Buddhas as "creations" (nirmdna), and that he accomplishes the concentration that fills the universe with his own magical creations, like a magician producing apparitional armies, palaces, and cities; from now on he knows the precise circumstances of any new birth he will assume. During the ninth stage he is a Bodhisattva in his last existence (caramabhavika); finally, seated beneath the tree of enlightenment, he at last enters into the tenth stage, the stage of the Cloud of Dharma (dharma-megha bhumi). The Mahdprajidparamitadgstra here appears to impose the standard nomenclature of the Dasabhumika Sutra on the ten bhumis of the Prajnaparamita, despite the fact that the details of the Dagabhumika scheme are manifestly different. 71 Dasabhumikasutra 94.20-95.6: "yasyam pratisthito bodhisattvo bhuyastvena mahesvaro bhavati deva-rajah."C ompareD aSabhiimisvaro1 99.2-5; T. Cleary,t rans., The Flower Ornament Scripture: A Translation of the Avatamsaka Sutra, 3 vols. (Boston: Shambala, 1984-87), 2:111. 72 Lalitavistara 79.6-7: "jata-matrasya bodhisattvasya mahesvaro deva-putrah suddhavfsa- kfyikin deva-putran amantryaivam aha." G. Bays, trans., The Lalitavistara Sutra: The Voice of the Buddha: The Beauty of Compassion, 2 vols. (Berkeley: Dharma, 1983), 1:164. See also F Edgerton, Buddhist Hybrid Sanskrit Dictionary (New Haven, Conn.: Yale University Press, 1953), s.v. "mahesvara." The Lalitavistara's account of the Pure Abodes is interesting in itself. The Lalitavistara begins with the Buddha attaining a samadhi called "the manifestationo f the ornamentso f a Buddha"( buddhdlamkdravyuha(B) ays, trans.,p . 2); the lights that subsequently issue from his body attract the attention of various gods of the Pure Abodes who come to him and request the Buddha to teach the Lalitavistara, a teaching that "cultivates the skillful roots of the Bodhisattva" (bodhisattva-kusala-mula-samudbhavana) (p. 3). The gods of the Pure Abodes lead the way in coming to honor the newly born Bodhisattva (p. 79) while later they create the four omens that prompt the Bodhisattva to go forth (p. 136). Nobuyoshi Yamabe has drawn my attention to Lamotte, trans., 1:519, which associates tenth-stage Bodhisattvas called MaheSvaradevarajaws ith the Pure Abodes. 73 Dasabhumikasuitra 90.11-15: "dharma-meghayam bodhisattva ekasyam api lokadhatau tusita-vara-bhavana-vasam upadaya cyavanacankramana-garbhasthiti-janmabhiniskramanabhisambodhy- adhyesana-mahadharmacakra-pravartana-mahaparinirvana-bhumir


kinds;a t will in the twinklingo f an eye he createsa s manyi ndividualsa s there arep articlesi n untoldw orld-systems.... In the arisingo f a thoughth e embraces the ten directions;i n a momento f thoughth e controlst he manifestationo f innumerablep rocesseso f completea wakeninga nd final nirvana.... In his own body he controlsc ountlessm anifestationos f the qualitieso f the Buddhaf ields of innumerablBe lessed Buddhas.74 If this is what tenth-stage Bodhisattvas do, then what do Buddhas do? Ignoring the poetic imagination of the Dasabhimika, the short answer seems to be much, much more of the same-so much so that one cannot properly begin to conceive of what Buddhas truly do. Nevertheless, it appears that we are to understand that at some point in the process-the repeated process of manifesting the acts of Buddhas and carrying out their work-these tenth-stage Bodhisattvas do actually become Buddhas. At this point it is useful, I think, to consider the witness of the later Indo-Tibetan tradition. mKhas grub rje's "Fundamentals of the Buddhist Tantras" (rGuyd sde spyi'i ram par gzhag pa rgyas par brjod) is an early fifteenth-century dGe lugs work which devotes its first chapter to the question of how the Sravakas and then the Mahayana (considered by way of the "Paramita" and "Mantra" schools) understand the final stages of the process of the Blessed teacher's becoming a fully awakened one (abhisambuddha).75 Let me go straight to mKhas grub rje's account of the understanding of this process according to the Mantra school. mKhas grub rje takes it as axiomatic for the Mahayana that full awakening is gained in Akanistha. But how precisely does it come about there? mKhas grub rje details the position of the Yoga and Anuttarayoga Tantras according to a number of Indian commentators (eighth to tenth century). For present concerns some indication of his account of Sakyamitra's and Buddhaguhya's understanding of the Yoga Tantras will suffice. According to them, Siddhartha Gautama, a tenth-stage Bodhisattva from the time of his birth, having practiced austerities for six years, then established himself in the imperturbable concentration (aninijyo-ndma-samadhi) of the fourth dhydna.

At thatt ime, the Buddhaso f all the ten directionsa ssembleda, rousedh im from thats amidhi by snappingt heirf ingers,a nd said to him, "Youc annotb ecomea iti sarva-tathagata-karyam adhitisthati" (cf. Dasabhumisvaro 191.6-8; Cleary, trans., 2:107). 74 Dasabhumikasitra 91.4-7, 14-18: "akaiiksann ekavalapatha ekasarvabuddhavisayavyuham adarSayati/ akaiksan yavad anabhilapyan sarvakarabuddhavisayavyuhana darsayati / akaiiksan yavanty abhilapyasu lokadhatusu paramanurjamsi tavata atmabhavan ekaksanalavamuhurtenan irmite / ... / cittotpade ca dagadikspharanamg acchati / cittaksane capramanaa bhisambodhiry avan mahaparinirvanavyuhana dhitisthati/ ... / svakaye capramananam buddhanam bhagavatam aprameyan buddhaksetragunavyuhana dhitisthati" (cf. Dasabhumisvaro 192.11-13, 193.3-6; Cleary, trans., 2:108). 75 Lessing and Wayman, trans. (n. 67 above), pp. 16-39.


ManifestC ompleteB uddhab y this samadhia lone."" Thenh ow shallI proceed?" he imploredt hem.T hey guidedh im to the Akanisthah eaven.M oreoverw, hile his maturationb ody (vipaka-kdyas)t ayedo n the banko f the same Nairafijana River, the mental body (manomaya-kaya) of the Bodhisattva Sarvarthasiddha proceededt o the Akanisthah eaven. Aftert he Buddhaso f the ten directionsh ad given garmenitn itiation( vastraabhiseka) a nd diademi nitiation( mukuta-abhisekath),e y badeh im entert he intensec ontemplatioinn sequenceo f the five AbhisambodhAi. fterc ompletingth e five Abhisambodhhi,e becamea ManifestC ompleteB uddhaa s Mahavairocana, the Sambhoga-kaya.76 Insofar as this account sees Gautama as a Bodhisattva who has taken a human birth in his last existence and the enlightenment as straightforwardly founded on the actual attainment of the fourth dhydna, it is closer (than, say, the Pancavimsatika or Dasabhumika accounts) to the Nikaya description; the Bhayabherava-sutta of the Majjhima Nikaya describes the Bodhisatta as gaining the fourth jhana and then, on the basis of that attainment, the three knowledges which culminate in the knowledge of the destruction of the dsavas.77 If we now consider the above range of material on the process of the Bodhisattva's final attainment of Buddhahood, it seems that it embraces two basic views. According to one view the Bodhisattva in his "final existence" (i.e., before finally transcending existence) is reborn in the Akanistha heaven where he finally becomes a Buddha; he subsequently manifests various creations which appear to be born, go forth, practice, meditation, and become Buddhas. According to the other the Bodhisattva in his last existence is actually born as a human being; seated beneath the tree of awakening he ascends in meditation with a mind-made body to the Akanistha heaven where he finally becomes a Buddha, while his "real" human body remains seated beneath the tree. Yet to state the positions thus baldly actually infringes a deeply rooted ambiguity and equivocation that runs through the cosmological material I have been considering in the course of this article. For where is the true Buddha? In Akanistha? Or seated beneath the tree of awakening? How does one come to Akanistha? By traveling through space? Or by journeying in the mind? Let me emphasize here that I am asking these questions of the ancient texts and not raising the problem of how the modern Buddhist tradition should set about finding an understanding of its ancient cosmology that is compatible with the "findings" or modern science, whatever precisely those are. And my point is that to ask such questions in such terms betrays a particular metaphysics and ontology which is precisely not the metaphysics and ontology of the Indian Buddhist tradition. 76 Ibid., p. 27. See also T. Skorupski, "Sakyamuni's Enlightenment according to the Yoga Tantra," Sambhasa (Nagoya University, Indian Buddhist Studies) 6 (1985): 87-94. 77 See M 1:21-24.


In the course of this article I have been trying to explore the way in which psychology and cosmology parallel each other in Buddhist thought-something that Peter Masefield has already tried to elucidate in the Nikayas by reference to the Upanisadic terms adhydtmam and adhidaivatam. I have suggested that in the Abhidharma the shift from psychology (levels of citta) to cosmology (levels of the lokadhdtu) can be viewed as a shift of time scale. The effect of my discussion is not to reveal something new but to bring into sharper focus something that lies at the heart of Indian Buddhist thought, namely, a basic ambiguity about matters of cosmology and psychology, about the objective outer world and the subjective inner world. This is true to the extent that the key to understanding both is to recognize that there is a fundamental and profound equivalence between cosmology and psychology. In conclusion I should like to risk a few general comments about the metaphysics and ontology of Indian Buddhism. I do not want to imply here that all Indian Buddhism shares an explicit and definite metaphysics and ontology, but I am suggesting that there is a general, underlying orientation, which tends to locate reality in the mind and its processes rather than in something "out there" which is other than the mind. We may want to persist in asking questions in the latter terms, yet it is significant that the tradition itself never quite does. On the contrary, it seems to take for granted and as natural an ambiguity between cosmology and psychology, for what is the difference between really being in Akanistha and experiencing one is really in Akanistha?

To put it another way, there is a loosely "idealist" tendency to all Indian Buddhist thought. It is no accident that one of the most important and influential philosophical schools of Indian Buddhism, the Yogacara, expounded an idealist ontology. For the Yogacara the only reality anything ultimately has is psychological. Yogacara thought is essentially a product of and a continuation of an Abhidharmic way of thinking; it gives explicit expression in systematic and philosophical form to a tendency that runs through the whole of Buddhist thought. The Theravadin Abhidhamma tends to sidestep the issue of the ultimate ontological status of the external world and the world of matter; the question is never explicitly raised. Yet for the Theravadin Abhidhamma-and as I understand it this would also be true of the Vaibhasika Abhidharma-the physical world each being lives in and experiences is one that is the result of his or her past kamma performed by deed, word, and thought; regardless of the ultimate ontological status of the external world and the world of matter, the particular physical sensations that beings experience are constructed mentally insofar as each one is the result of past kamma. In technical Abhidhamma terms our basic experience of the physical world is encompassed by just ten classes of sense-sphere consciousness that are the results (vipaka) of twelve unskillful and eight skillful classes of


sense-sphere consciousnesses: what we thought in the past has created the world we live in and experience in the present; what we think in the present will create the world we shall live in in the future.78O r, as Dhammapada (vv. 1-2) famously put it, "dhammas have mind as their forerunner, mind as preeminent, mind as their maker" ("manopubbamgama dhamma manosettha manomaya"). That is, Indian Buddhist thought is in unanimous agreement that ultimately the particular world each of us experiences is something that we individually and collectively have created by our thoughts. The parallel that exists in Buddhist thought between cosmology and psychology is simply a reflection of this basic fact of the Abhidharma understanding of the nature of existence. Indologists are familiar with the Upanisadic interiorization of the Vedic sacrificial ritual; students of Hindu and Buddhist Tantra take for granted the correspondences that are made between the body of the yogin and the universe as microcosm and macrocosm respectively.79 Yet the similarities between this and certain ways and patterns of thinking found in early and Abhidharmic Buddhist thought are rarely recognized in the existing scholarly literature. These similarities consist in the general tendency to assimilate some kind of internal world to an external world, and in the principle that places mind and psychology-the way the world is experienced-first. The assimilation of cosmology and psychology found in early Buddhist thought and developed in the Abhidharma must be seen in this context to be fully understood and appreciated. I can do no better than to finish with the words of the Buddha: That the end of the world... is to be known, seen or reached by travellingthat I do not say.... And yet I do not say that one makes an end of suffering withoutr eachingt he end of the world.R ather,i n this fathom-longb ody, with its consciousness and mind, I declare the world, the arising of the world, the ceasing of the world and the way leading to the ceasing of the world.

University of Bristol 78 See, e.g., Abhidh-s 2 on "motivationless consciousness" (ahetuka-citta) and Abhidh-s chap. 4, on the "consciousness process" (citta-vithi); cf. L. S. Cousins, "The Patthana and the Development of the Theravadin Abhidhamma," Journal of the Pali Text Society (1981), pp. 22-46.

79 See, e.g., J. Varenne, Yoga and the Hindu Tradition (Chicago: University of Chicago Press, 1976), pp. 127-63; J. Brereton, "The Upanisads," in Approaches to the Asian Classics, ed. W. T. de Bary and I. Bloom (New York: Columbia University Press, 1990), pp. 115-35. 80 S 1:62 = A 2:48: "naham tam gamanena lokassa antam inateyyam dattheyyam patteyyan ti vadami ti. na kho panaham avuso appatva lokassa antam dukkhass' antakiriyam vadami. api khvaham avuso imasmim yeva vyamamatte kalevare sanfiimhi samanake lokam ca pafnnapemlio ka-samudayamc a loka-nirodhamc a loka-nirodha-gaminimc a patipadam."


APPENDIX A

BIBLIOGRAPHY

ABBREVIATIONS

A

Abhidh-av

Abhidh-di

Abhidh-k-(bh)

Abhid-s-(t)

As
D
DAT
Dhp-a
Ky
M
Mp
Pp
S
Sn
Sv
Vibh
Vibh-a
Vism
Vism-t
Aiiguttara-Nikaya
Abhidhammavatara
Adhidharmadipa
Abhidharmakos`a-(bhdsya)
Abhidhammatthasafigaha-Q(ikd)
Atthasalini
Digha-Nikaya
DIghanikdyatthakathittka
Dhammapadatthakathd
Kathavatthu
Majjhima-Nikdya
Manorathapiirani
Papaficastdani
Samyutta-Nikdya
Suttanipdta
Sumafigalavilasini
Vibhaiiga
Vibhanigatthakath(a= Sammohavinodani)
Visuddhimagga
Visuddhimagga-tika (= ParamatthamafijUisdtika)
PALI AND SANSKRIT TEXTS
Abhidhammatthasatigaha of Bhadant&cariya Anuruddha and the Abhidhammatthavibh&
vant-tika of Bhadantiicariya Sumaiigalascmi. Edited by H. Saddhatissa.
Oxford: Pali Text Society, 1989.
Abhidhammivatelra: Buddhadatta's Manuals. Edited by A. P Buddhadatta. 2 vols.
London: Pali Text Society, 1915-28. Vol. 1.
Abhidharmadtpa with Vibhdisdprabhdvrtti. Edited by P S. Jaini. Patna: Kashi Prasad Jayaswal Research Institute, 1959.
Abhidharmakosaa nd BhdiFyao f Acifrya Vasubandhuw ith Sphutartha Commentary
of Acarya Ya?omitra. Edited by D. Shastri. 3 vols. Varanasi: Bauddha Bharati, 1970-72.
Atthasilini: Buddhaghosa's Commentary on the Dhammasahzgani. Edited by
E. Miiller. London: Pdli Text Society, 1979.
Afiguttara-nikdya. Edited by R. Morris and E. Hardy. 5 vols. London: Pali Text Society, 1885-1900.
Dasabhfimikasiitra. Edited by J. Rahder. Leuven, 1926.
DaabhFimitvaro nCima mahiaydnastitram. Edited by RyUiko Kondo. Tokyo:
Daijyo Bukkyo Kenyo-Kai, 1936.
Dhammapadatthakathti. The Commentary on the Dhammapada. Edited by H. C.
Norman. 4 vols. London: Pali Text Society, 1906.




Digha-nikaya. Edited by T W. Rhys Davids et al. 3 vols. London: Pali Text Society, 1890-1911.
Kathavatthu. Edited by A. C. Taylor. 2 vols. London: Pali Text Society, 1894-97.
Lalitavistara. Edited by P. L. Vaidya. Darbhanga: Mithila Institute of Postgraduate
Studies and Research in Sanskrit Learning, 1958.
Majjhima-nikaya. Edited by V. Trenckner and R. Chalmers. 3 vols. London: Pali Text Society, 1888-1902.
Sammohavinodani Abhidhamma-Pitake Vibhahgatthakatha. Edited by A. P.
Buddhadatta. London: Pali Text Society, 1923.
Samyutta-nikaya. Edited by L. Feer. 5 vols. London: Pali Text Society, 1884-98.
Sutta-Nipata. Edited by D. Anderson and H. Smith. London: Pali Text Society,
1913.

Tattvasahgraha of Acarya Santaraksita with the Commentary 'Panjika' of Sri Kamalaslla. Edited by D. Shastri. 2 vols. Varanasi: Bauddha Bharati, 1968. Vibhahga. Edited by C. A. F Rhys Davids. London: Pali Text Society, 1904. Visuddhimagga of Buddhaghosdcariya. Edited by H. C. Warren and D. Kosambi. Cambridge, Mass.: Harvard University Press, 1950. Visuddhimagga-ktik. Buddhaghosacariya's Visuddhimaggo with Paramatthamanjusdtlka of Bhadantacariya Dhammapala. Edited by Rewatadhamma. 3 vols. Varanasi: Varanaseya Sanskrit Vigvavidyalaya, 1969-72.


APPENDIX B

HOW OLD IS BUDDHIST COSMOLOGY? A NOTE ON THE AGGANNA-SUTTA

The writings of a number of scholars seem to imply that the Nikaya cosmology should not be attributed to the Buddha himself. Konrad Meisig, continuing the work of Ulrich Schneider, argues that the account of the evolution of the world and human society introduced by the formula I quoted at the start of this article should not be regarded as forming part of the "original" Aggainna-sutta.81 Schneider's and Meisig's arguments are complex and involved but appear to me to be neither individually nor collectively conclusive. The fact remains that the cosmogonic myth forms a significant part of all four versions of the text that Meisig examines; in other words, we have no hard evidence of an Aggafinasutta- or whatever its "original" title-without the cosmogonic myth. On the other hand, we do have some hard evidence for the cosmogonic myth apart from the Agganfia-sutta.82E ven when it is not accepted as forming parto f an "original" Aggaifia-sutta, it must be acknowledged that the tradition it represents is well attested. The whole notion of an original version of a sutta raises interesting questions. The kind of model with which Meisig would seem to be working regards the original Aggaiina-sutta as a discourse delivered by the Buddha himself on one particular occasion (at Savatthi since all versions are agreed in locating it there?), which was remembered by his followers and for a while handed down


faithfully by them, until someone or some group still in the pre-Asokan period appended to it a cosmogonic myth.83 But this kind of model is perhaps inappropriate to the composition and transmission of oral literature and may also be historically naive. A more appropriate general model for an original sutta might be of a "text" representing the substance of a discourse or teaching that the Buddha himself may have given on a number of different occasions and which in part at least draws on a stock of images and formulas which the Buddha himself employed in a variety of contexts as he considered appropriate. Whether or not the Buddha himself composed his teachings in this way, it is clear that someone started doing so at some point, since many of the discourses of the Pali Nikayas and Chinese Agamas are manifestly put together in this way. This, however, is a matter that needs more systematic research. It may well be that Schneider's and Meisig's analysis goes some way to revealing the blocks of tradition which have been put together to form the Aggafina-sutta; but to expose these blocks of tradition does not of itself tell us anything about who put them together and when. In the end, Schneider's and Meisig's understanding of the original Aggafifia-sutta amounts to a judgment about how well the blocks of tradition have been put together; their view is that they have been put together badly and that the two basic parts of the discourse are ill-fitting. Yet even if we agree with this judgment, the bare fact that a sutta is badly put together does not of itself preclude the possibility that it is the original work of the Buddha; a claim that the Buddha cannot possibly have made such a mess of it is an appeal to the transcendent notion of Buddhahood rather than a conclusive historical argument. To say that the Aggafiia-sutta is composed of two parts must surely be largely uncontroversial. Clearly paragraphs 1-9 and 27-32 do form something of a unity and could intelligibly stand on their own; again, the cosmogonic myth of paragraphs 10-26 is an intelligible unit such that the Buddhist tradition itself abstracted portions of it to be used outside this context. But it seems to me purely arbitrary to pick on the first as original and relegate the second to the status of later interpolation. One might just as well argue the Buddha originally gave a discourse consisting of a cosmogonic myth that was later wrapped up in an ethical disquisition on the four classes (vanna) by certain of his followers who did not appreciate myth. This reveals what one suspects might be the true basis for the conclusion that it is the section of the Aggafifia-sutta concerned with the four classes that constitute the original sutta: the "ethical" portion of the discourse is to be preferred to the "mythic" precisely because it is ethical, and, as we all know, the earliest Buddhist teachings were simple, ethical teachings, unadulterated by myth and superstition; we know that early Buddhist teaching was like this because of the evidence of the rest of the canon. Here the argument becomes one of classic circularity: we arrive at a particular view about the nature of early Buddhism by ignoring portions of the canon and then use that view to argue for the lateness of the portions of the canon we have ignored. Richard Gombrich has countered the Schneider/Meisig view of the Aggafifiasutta by arguing that the two parts of the discourse have been skillfully put together and that the cosmogonic myth works as an integral part of the discourse


taken as a whole.84 According to Gombrich the first half of the discourse introduces the problem of the relative status of brahmanas and suddas; this question is then dealt with in a tongue-in-cheek satirical manner by the Aggafnia myth. Gombrich regards the overall form of the Aggafinia-suttaa s we have it as attributable to the Buddha himself and thus original. But for Gombrich the text is "primarily satirical and parodistic in intent," although in time the jokes were lost on its readers and the myth came to be misunderstood by Buddhist tradition "as being a more or less straight-faced account of how the universe, and in particular society, originated."85F ollowing Gombrich, Steven Collins has discussed the Aggafifia Sutta in some detail as a "humorous parable," finding in certain of its phrases echoes of Vinaya formulas.86 Gombrich's arguments for the essential unity of the Aggainia text as we have it are extremely persuasive, yet I would disagree with the implication that we should regard the mythic portions of the Aggafina-sutta as solely satirical.

Certainly it seems to me that Gombrich must be right in arguing that there is a good deal of intended humor in the Aggaifnia-suttaa, nd certainly I would not want to argue that the cosmogonic myth was never intended to be understood as literal history in the moder sense. How could it have been? Yet it still seems to me unlikely that, for the original compiler(s) of and listeners to the discourse, the mythic portion of the sutta could have been intended to be understood or actually understood in its entirety as a joke at the expense of the poor old brahanas. As Gombrich so rightly says, if we want to discover the original meaning of the Buddha's discourses we need to understand the intellectual and cultural presuppositions shared by the Buddha and his audience. While in absolute terms this is an impossible task, since we can never entirely escape our own intellectual and cultural presuppositions and be reborn in the world of the Buddha-at least in the short term-we can still surely make some progress in trying to rediscover that world.

The question I would therefore ask is, Do we have any particular historical reasons for supposing that it is unlikely that the Buddha should have recounted a more or less straight-faced cosmogonic myth? My answer is that we do not. Indeed, I want to argue the opposite: what we can know of the cultural milieu in which the Buddha operated and in which the first Buddhist texts were composed suggests that someone such as the Buddha might very well have presented the kind of myth contained in the Aggafinia-sutta as something more than merely a piece of satire. Far from being out of key with what we can understand of early Buddhist thought from the rest of the Nikayas, the cosmogonic views offered by the Aggafifia-sutta in fact harmonize extremely well with it. I would go further and say that something along the lines of what is contained in the Aggafina myth is actually required by the logic of what is generally accepted as Nikaya Buddhism. 84 Richard Gombrich, "The Buddha's Book of Genesis?" Indo-Iranian Journal 35 (1992): 159-78; see also his Theravada Buddhism: A Social History from Ancient Benares to Modern Colombo (London: Routledge & Kegan Paul, 1988), p. 85. 85 Gombrich, "The Budda's Book of Genesis?" pp. 163, 161. 86 Steven Collins, "The Discourse on What Is Primary (Aggafiiia-Sutta): An Annotated Translation," Journal of Indian Philosophy 21 (1993): 301-93.


It might be countered that the Buddha's refusal to answer categorically certain questions-including questions about whether or not the world was eternal and infinite-indicates that the Buddha was not interested in metaphysical questions and instructed his monks not to waste their energy on them. The account of the world on a cosmic scale found in the Aggafifia-sutta is then to be seen as not in keeping with the spirit of the Buddha's instructions and therefore as the creation of curious bhikkhus who, unable to restrain their imaginations, ignored the express instructions of their teacher. Such an outlook both misunderstands the nature of the, usually, ten "undetermined questions" and misrepresents the Aggafifia-sutta. This sutta does not expressly answer the question of whether or not the world is eternal and infinite, and as Steven Collins has argued, the real reason for the refusal to give a categorical answer to the questions is that they are, from the standpoint of Buddhist thought, linguistically ill-formed.87 Thus it is not because the Buddha does not know the answer to these questions that he refuses to answer them but because the terms employed in the questions have in the Buddhist view of things no ultimate referent: it simply does not make sense to ask whether the world is eternal or not because there is no one "thing" to which the word world refers. The notion "world" is just like the notion "self": it is not of itself an ultimately real thing but merely a concept, a mental construct. The ten undetermined questions thus, it seems to me, have no direct bearing on the cosmological ideas expounded in the Aggafifia-sutta.




Source