ON being mindless: The debate on the reemergence of

consciousness from the attainment of cessation in the
Abhidharmako`sabhaa.syam and its commentaries
By Paul Griffiths
Philosophy East and West
Volume 33, no.4 (October 1983)
P.379-394
(C) by the University of Hawaii Press


P.379 I. THE PROBLEM The Indian Buddhist meditative traditions bear strong witness to the existence and desirability of a trance state or states which are entirely without consciousness. This fact has been the source of many problems, both theoretical and practical, for Buddhist philosophers and meditation teachers, most of which originate in the uneasy tension visible between this type of mindless trance state, together with the techniques which produce it, and the more central and "orthodx" techniques of Buddhist meditation which aim at insight and are concerned to develop the practitioner's analytic capacities rather than to bring all mental operations to a complete halt.(1) This paper is not intended as a full discussion of the entire range of problems associated with the higher trance states(2) but rather as a study of one particular issue occasioned by the existence of mindless trance states in the tradition. This issue is: if the attainment of cessation (nirodhasamaapatti.h.(3)) here considered as a paradigm example of a mindless trance state) really is a state wherein all mental events have come to a complete halt, how is it that the practitioner's mental life can begin again after having completely ceased? How can mind arise from mindlessness? By what system of causality can something which has become completely nonexistent come into being once again? Is it necessary to postulate a continuing mental "something" in order to account for the ability of the practitioner of such trance states to enter and leave them at will? The intent of this paper is first to show that the Abhidharmako`sabhaa.syam--an influential summa of Buddhist philosophical theory from fifth-century India-and two of its important commentaries have to say to these issues, and then to make some attempt at clarifying and interpreting the issues at stake here in non-Buddhist language. A preliminary attempt will be made at relating these problems to similar ones which have arisen during the course of Western philosophical discussion of the relationship between mind and body. The analogies will, we hope, prove illuminating, and will show that Buddhists--or at least Indian Buddhist Abhidharmikas--are a good deal closer to being dualists than is often supposed. II. THE SOURCES Our discussion will be based on one short section of the Abhidharmako`sabhaa.syam (hereafter cited as AKBh):(4) that part of the second chapter which discusses the attainment of cessation as an example of a cittaviprayuktasa^mskaara.h. While we are fortunate enough to have access to this text in its original Sanskrit, a consultation of its Tibetan translation(5) often clarifies obscurities and corrects some of the many errors in Pralhad Pradhan's edition of the Sanskrit text. Therefore, although the translations in this paper are based primarily upon P.380 the Sanskrit, we have consulted the Tibetan version throughout and will refer to it where appropriate. The two commentaries to Vasubandhu's text we have consulted are: Ya`somitra's Abhidharmako`sasphu.taarthavyaakhyaa (hereafter cited as AKV),(6) written from the Sautraantika viewpoint and also extant in the original Sanskrit. In this case also we have profited from a consultation of the Tibetan translation.(7) Also, Sthiramati's Abhidharmako`sabhaa.sya.tiikaa (hereafter cited as AK.T) , extant only in Tibetan,(8) has been used and will be quoted and referred to where appropriate. This latter is a remarkably full and interesting commentary which as yet has hardly been utilized by Western scholarship. It is written from the Yogaacaara viewpoint and thus provides an interesting contrast to the AKV. Occasional reference will also be made to other texts when this becomes necessary, notably to Vasubandhu's Karmasiddhiprakar.ana (hereafter cited as KSP) (9) and its commentary by Sumati`siila (Karmasiddhi.tiikaa, hereafter cited as KS.T),(10) and to the Vaibhaa.sika Abhidharmadiipa.(11) Finally, anyone who works on the AKBh must gratefully acknowledge the debt owed to Louis de La Vallee Poussin for his translation of the entire work into French from Hsuan-tsang's Chinese version. His work was a landmark in Western Abhidharma studies and its existence has meant that almost no Western scholars since Poussin's time have done serious work on the AKBh. This is unfortunate because Poussin's work has at least two serious drawback: The first is that it was done before the recovery of the original Sanskrit and thus is not based upon a study of the text in the language in which it was composed. The second is that Poussin did not translate many of the technical terms which are so important for a full understanding of the AKBh. For the most part he was content to leave them in (reconstructed!) Sanskrit, and this makes his work of dubious value for the purpose of actually coming to an understanding of the philosophical debates which fill the AKBh. There is a pressing need for serious scholarly work on the AKBh which both takes the text seriously in its original language and is concerned to communicate its meaning. This paper is meant as a contribution to such study. III. THE DEBATE In Ya`somitra's AKV we find the following programmatic statement of positions of the schools on the issue which concerns us: On this matter the Vaibhaa.sikas and others say that the attainments of unconsciousness and cessation, together with the state of unconsciousness, are mindless. The Elder Vasumitra and others say that [these attainments] possess mind an unmanifest mental consciousness. The Yogaacaaras say that they possess mind--the store-consciousness. This is the division of the schools.(12) Vasubandhu's exposition of the first-Vaibhaa.sika-position reads as follows: How is it that mind can arise once again from a mind that has been brought to a halt for a long period of time? The Vaibhaa.sikas claim that because past [dhar- P.381 mas] continue to exist there is an immediately antecedent and homogenous condition [for the occurrence of such an event].(13) To understand what this Vaibhaa.sika view means, we need to know something of the Buddhist theory of causation, since the Vaibhaa.sikas are here claiming that a particular kind of cause--the samanantarapratyaya.h--accounts for the reemergence of mind from the mindless trances. We find a discussion of the nature of samanantarapratyaya.h in the commentary to AK2.62a-b, where the compound is glossed as a karmadhaaraya.h: sama`scaayamanantra`sca pratyaya iti sammanantarapratyaya.h(14) Hence the meaning is that samanantarapratyaya.h is a condition (pratyaya.h) which is of the same kind as (samam) and immediately antecedent to (anantaram--without interval. either temporal or spatial) its effect.(15) Normally, every mental event both has and is a samanantarapratyaya.h: that is, it is caused by the immediately preceding event in the relevant mental continuum, in conjunction with other significant nonmental causes, and itself causes the immediately subsequent event. The exceptions to this rule are: the last moment in the mental continuum of an Arhat before he attains final nirvaa.na(16) --which possesses an immediately antecedent and homegenous condition but is not itself such since consciousness ceases at that point--and, depending upon which school's viewpoint is followed, the last moment, of consciousness before entering the attainment of cessation (which, like the last conscious moment of an Arhat, may be considered to have, but not be, an immediately antecedent and homogenous condition) and the first moment of consciousness upon emerging from that trance (which may be considered to be, but not have, an immediately antecedent and homogenous condition) . Thus we find the following view attributed to the Sa.mtaanasabhaagikas in the AKBh: for this school, the first moment of consciousness which occurs upon leaving the attainment of cessation (vyutthaanacittam) has as its immediately antecedent and homogenous condition the last moment of consciousness before entering that trance (samaapatticittam). For this school, the fact that the two moments of consciousness are separated by a more or less extensive period of time--during which there is no consciousness at all--does not affect the fact that one is the immediately antecedent and homogenous cause of the other. On this view, separation of an effect from its cause by a dissimilar dharma does not constitute real separation; thus there is no real separation of the samapaatticittam from the vyutthaanacittam. This school further claims that any given immediately antecedent and homogenous cause can only give rise to an effect of the same type as itself; thus a moment of sensation gives rise only to another moment of sensation, a moment of conceptualization only to another moment of conceptualization, and so forth.(17) We do not know much else about the Sa.mtaanasabhaagikas. Ya`somitra merely tells us that they are "certain Abhidharmikas" (kecidabhidharmikaa.h) .(18) and Sthiramati adds nothing of significance.(19) To return to our main text: we now understand what a samanantarapratyaya.h is and can see that the Vaibhaa.sikas hold a view of it which is not dissimilar to that P.382 attributed to the Sa.mtaanasabhaagikas in the passage discussed above. Further information on the latter school is lacking for the present, but it certainly seems right to attribute their views on the immediately antecedent and homogenous condition to the Vaibhaa.sikas; it may even be historically correct to say that the Sa.mtaanasabhaagikas were simply a subgroup of the Vaibhaa.sika school. In any case, the Vaibhaa.sika view amounts to this: every mental event must have an immediately antecedent and homogenous condition, including the first moment of consciousness after the nirodhasamaapatti.h. Further, such a condition must be of essentially the same type as its effect, and so great stress is placed on the 'homogenous' (samam) element of the samanantarapratyaya.h compound, and rather less on the "immediately antecedent" (anantaram) element. For the Vaibhaa.sikas it is acceptable that the cause of a given mental event, X, may be another mental event, Y, which occurred at a considerable distance in time from X. This also explains the stress placed by this school on the continued existence of past dharmas or events; for something to be causally efficacious at a distance in time it must in some sense still be in existence.(20) Sthiramati's comment on the Vaibhaa.sika assertion of the existence of past dharmas in this context betrays his Yogaacaara affiliation without adding a great deal to what has been said here; he merely asserts that the Vaibhaa.sikas do not allow the existence of the aalayavij~naanam or store-consciousness.(21) We may now move to the second view discussed by Vasubandhu in the AKBh: Others say "How is is it that physical form can arise once again for those who have been reborn in places like the formless realms and whose physical form has been brought to a halt for a long period of time? It arises precisely from mind, not from physical form [since mind continues to exist in the formless realms]. Such is also the case with the mind [of one in the attainment of cessation]. It arises from the body with its senses, not from mind." Previous teachers have said: "Both of these mutually seed one nother: namely, the mind and the body with its senses."(22) Ya`somitra identifies the "others" with the Saurtaantikas,(23) and by the use of bruuma.h, "we say, " shows that he agrees with their view. According to Ya`somitra, the Sautraantikas deny that a past dharma can have causal efficacy; when a particular state of mind has ceased to exist its causal efficacy ceases with it. Therefore the Vaibhaa.sika view, that the immediately antecedent and homogenous condition for the first moments of consciousness after the attainment of cessation is the (long ceased) last moment of consciousness before entry into the said trance, cannot, for the Sautraantikas, be admitted. In contrast to this view, they offer the idea that mind is capable of sowing seeds which can remain dormant within the physical continuum even when all other mental events have ceased, and only much later mature and give rise to other mental events. If we unpack this seed metaphor in a little more detail. we see that it strongly implies that the attainment of cessation is mindless. What happens in this view is that the last moments of consciousness before entry into the nirodhasamaapatti.h plant seeds in the continuing stream of physical events. In time these seeds ripen and P.383 produce their fruit--the reemergence of consciousness. If this is a correct understanding of the Sautraantika view,(24) then it would seem that Lamotte and Poussin are wrong in suggesting that all Sautraantikas held the view that there is some subtle form of consciousness extant in the attainment of cessation. The Sautraantika view under discussion here certainly does not support such an interpretation. We shall have to return to this issue when we come to consider the views of the Bhadanta Vasumitra. We have seen, then, that Ya`somitra identifies himself with the Sautraantika view discussed in this section of the AKBh. Sthiramati makes an interesting contrast because he offers some criticism of the seed image, and suggests some undesirable logical consequences of the Sautraantika view. After offering the usual grammatical comments(25) and agreeing with Ya`somitra that these "former teachers" are in fact Sautraantikas, Sthiramati says this: If [as the Sautraantikas claim] consciousness can arise from the body with its senses without reference to the cause which assures homogeneity of species, then when there exist both basis and object consciousness would occur simultaneously everywhere. But if mind arises subsequently by way of connection to that state of mind which existed prior [to it], then since there is no immediately antecedent and homogenous condition for the second [i.e., subsequent state of mind], the conclusion is that, even when basis and object exist there would be no simultaneous arising [of the relevant consciousness]. And if it is asked how, in the absence of mind, [mind] can arise from a seed by means of the mindless body with its senses, then [the answer is] that this is not possible because there is no distinct cause [for such arising to occur].(26) If we have understood Sthiramati correctly, he here makes three separate points: first, that the Sautraantika view must be wrong because it ignores the necessity to take into account the nikaayasabhaagahetu.h(27) (ris mthun pa'i rgyu) , that type of causality which ensures that an effect is in some significant way like its cause. This causal principle, for example, is the one which ensures that the sexual intercourse of two human beings always produces other human beings as its result, and not, say, elephants. To assert that mind can arise from the body with its senses is to assert that it can arise from something quite other than itself; the result, thinks Sthiramati, is that the kind of consciousness which occurs at any given moment need not be causally related to the combination of sense-organ ("basis," aa`sraya.h, rten) and sense-object ("object," aalambanam, dmigs pal. For example, when the eye and a visual sense-object come into contact, the nikaayasabhaagahetu.h (among other things) ensures that visual consciousness results, and not, say, auditory or olfactory consciousness. But if the nikaayasabhaagahetu.h is ignored, as the Sautraantikas appear to do with their theory of mind originating from body (or, more precisely, from mental "seeds" in a physical continuum), then any kind of consciousness whatever could result--or many types at once. Sthiramati's second point is that on the Sautraantika view, if we disallow the body as the immediately antecedent and homogenous condition for the reemer- P.384 gence of consciousness, there can be no such condition. And if there is no such condition, according to the Yogaacaara position which Sthiramati is here explaining, consciousness can never reemerge from the attainment of cessation. Sthiramati's third point is to criticize the Sautraantika image of seed. In order that consciousness may reemerge after the attainment of cessation, some distinct or specific cause is necessary; the general fact of the existence of the "seeded body" will not suffice to meet this requirement. What mechanism accounts for the ripening of seeds of consciousness at a given moment? The Sautraantikas offer no account. We shall return to Sthiramati's criticisms of the Sautraantika position at a later stage in this paper. For the moment, we simply need to note that Sthiramati's own position has not yet been stated; his criticisms are a propadeutic for the later introduction of the aalayavij~naanam. To summarize what we have learned about the Sautraaintika position thus far: first, the Sautraantikas seem to agree with the Vaibhaa.sikas that the attainment of cessation is mindless. They suggest that the reemergence of consciousness is caused by mental seeds preserved in the continuum of physical events which is all that exists in the nirodhasamaapatti.h. To return to our main text, Vasubandhu next summarizes the position of the Bhadanta Vasumitra: The Bhadanta Vasumitra, on the other hand, says in the Parip.rcchaa: "This problem [i.e., of how consciousness can reemerge from the attainment of cessation] occurs for one who thinks that the attainment of cessation is mindless. For me, though, the attainment possesses mind."(28) There are some problems with the identification of the Vasumitra mentioned here, and with his scholastic affiliation. Ya`somitra tells us that Vasumitra was the author of a number of works, among them the Pa~ncavastuka, and that the Parip.rccha is named in order to give a specific reference for this quotation.(29) Sthiramati merely expounds Vasumitra's position in a little more detail, showing that he thinks the arising of consciousness from a state that is entirely without it would be impossible. Sthiramati offers no criticism, except to note that Vasumitra's concept of the mind that does exist in the attainment of cessation is not the same as the aalayavij~naanam, but instead a kind of unmanifest mental consciousness.(30) In this attribution of the theory of aparisphu.tamanovij~naanam to Vasumitra both Sthiramati and Ya`somitra agree,(31) and this theory alone is enough to make it likely that the Vasumitra referred to here is not the Vasumitra frequently quoted in the Mahaavibhaa.saa. In that text we find attributed to a Vasumitra a theory on the nirodhasamaapatti.h which seems precisely the same as the one attributed to the Vaibhaa.sikas in the AKBh and is therefore in direct contradiction to that attributed to Vasumitra here.(32) This problem has led both Poussin and Lamotte to suggest that the Vasumitra referred to here must have been a Sautraantika.(33) There seems to be no hard evidence to support this view of our Vasumitra's scholastic affiliation. We have already seen that neither Ya`somitra nor Sthiramati identifies Vasumitra as a Ssutraantika, and indeed both P.385 attribute to the Sautraantikas a view which is quite different from that of Vasumitra. Noriaki Hakamaya has pointed out that the quotation from Vasumitra's Parip.rcchaa occurs not only in the AKBh (as translated above) but also in Vasubandhu's KSP.(34) In the KSP itself Vasumitra--or rather the quotation from the Parip.rcchaa--is not said to belong to a particular school, but in Sumati`siila's KS.T it is attributed to a school which upholds the existence of external objects, (35) not a characteristic usually taken to be definitive of the Sautraantikas. It seems clear, then, that there is no unambiguous attribution of Vasumitra's view to the Sautraantika school in any of the relevant texts, and there is plenty of evidence which suggests that the Sautraantikas in fact held a different view. It is likely that Poussin and Lamotte were wrong to call Vasumitra a Sautraantika, although the exact nature of his scholastic affiliation must await further research.(36) Vasubandhu next turns to the objections offered by one Gho.saka to Vasumitra's view: The Bhadanta Gho.saka says that this does not follow since the Lord has said: "When consciousness exists there is also contact--which is the conjunction of the three [i.e., sense-organ, sense-object, consciousness].(37) Further. sensation, conceptualization and volition have contact as their cause." Hence [if consciousness does exist in the attainment of cessation as Vasumitra suggests] the cessation of sensation and conceptualization therein could not occur.(38) Ya`somitra's comment offers nothing of note, merely pointing out that the appellation 'cessation of sensation and conceptualization' is commonly applied to the attainment of cessation.(39) Sthiramati's explanation is more illuminating: A detailed explanation of the words: "Sensation and so forth have contact as their cause..." has been given [in the sutra]. Since the result follows automatically when an unobstructed cause exists, then in accordance with contact [which is the unobstructed cause of sensation and conceptualization] the cessation of sensation and conceptualization cannot occur in the attainment of cessation.(40) Ghosaka's view is now clear: he thinks that the existence of consciousness necessarily implies also the existence of indriyam and artha.h, sense-organ and sense-object, and that the existence of these three together is a sufficient condition for the existence of contact (spar`sa.h/reg pa). Contact, in turn, in Ghosaka's view, is a sufficient condition for the existence of sensation, conceptualization, and so forth. The conclusion is that if we allow any kind of consciousness to exist in the attainment of cessation--as we have seen that Vasumitra wishes to do--there can be no possibility of sensation and conceptualization ceasing as they are supposed to do in this trance state. Gho.saka is here expounding the standard Vaibhaa.sika view on the nature of consciousness and the mechanisms by which it arises,(41) and, if this view is followed in the way in which Gho.saka presents it, we have no choice but to conclude that the attainment of cessation really is as mindless (acittakam) as the Vaibhaa.sikas suggest. P.386 The next stage in the debate preserved in the AKBh is a reply from Vasumitra: [Vasumitra replies] But it [i.e., the existence of consciousness in the attainment of cessation] could occur. Just as in the case of thirst, which is said to have sensation as its cause: when sensation exists for an Arhat thirst does not arise. Just so, when contact exists sensation and so forth need not necessarily arise because this [case] is not different from that [case].(42) Ya`somitra does not comment upon this section and Sthiramati merely unpacks the position presented in the AKBh at some length.(43) Drawing partially upon what he says, we may explain Vasumitra's reply in the following manner: Gho.saka claimed that contact is a sufficient condition for the existence of sensation and conceptualization. Vasumitra admits that contact is a condition or cause (pratyaya.h) for the existence of sensation and conceptualization, but denies that it is a sufficient condition. He illustrates this by taking the analogous case of "thrist" (t.r.s.na.h), which is conditioned by sensation; here also, claims Vasumitra, there is no relation of necessary concomitance since in the case of an Arhat it is possible for the cause (in this case sensation) to exist without its effect (in this case thirst). The analogy holds, he claims, in the case of contact and sensation. Sthiramati summarises his position thus: Just as with sensation, not all instances of contact are causes of sensation.(44) If Vasumitra's position is correct, then his contention that it is possible for the attainment of cessation to contain some sort of consciousness without this necessarily also implying the existence of sensation and conceptualization may stand. But the section of the AKBh we are considering in this paper ends with some final criticisms of Vasumitra's view: But this [case] is not the same, since it is said that thirst arises in dependence upon sensation which is born from connection with ignorance, but with regard to the arising of sensation it is not said that contact is differentiated [i.e., differentiated in the sense that some types of contact give rise to sensation and some do not]. Therefore the Vaibhaa.sikas say that the attainment of cessation is mindless.(45) Ya`somitra's comment attributes this reply to Gho.saka. He adds: "With regard to the arising of sensation it is not...." means that with regard to the arising of sensation contact is not differentiated by saying that a particular kind of contact is the cause of sensation.(46) This is very clear. Vasumitra claims that contact can occur without sensation necessarily following, using the model of sensation and thirst which have no necessary connection. Gho.saka refutes this by pointing out that in the case of thirst it is not said that all types of sensation give rise to it, but only that sensation linked to ignorance has this power. That is to say, a distinction or differentiation is made among different types of sensation with regard to their ability to give rise to thirst, but no such differentiation is made with regard to contact as the cause of sensation; in the latter case we have a straightforward relationship of sufficient causality. P.387 Sthiramati's comment reinforces this understanding of the somewhat elliptical material in AKBh. He says: [According to Gho.saka] it is not possible to establish that sensation which is not an example of ignorance can be a cause of thirst.... Therefore, since this absence of differentiation [in contact as the cause of sensation] has been approved, and because all types of contact are causes of sensation, [Vasumitra's counter-example] of the causal relationship between sensation and thirst is not appropriate.(47) Here again we have the point that only in some cases is sensation a condition for the arising of thirst but that in all instances of contact sensation necessarily arises. Sthiramati's comment on this section concludes with a brief discussion of the sense in which the Vaibhaa.sikas assert that the attainment of cessation is mindless, and then passes to other matters. Here we conclude our examination of the sources, and we may pass to a restatement and discussion of the positions outlined here. IV. THE POSITIONS RESTATED We have now presented an outline account of the debate as it occurs in our sources. What has been said up to this point should be reasonably comprehensible anyone with some background in Buddhism, but is likely to remain almost completely unintelligible to, let us say, the Western philosopher specializing in philosophy of mind. The terms of reference are very different in each sphere of discourse. This is unfortunate, as the debate under discussion in this paper has many clear similarities to related debates in Western philosophy, and it is probable that both Buddhists and Western philosophers could profit from a fuller awareness of each other's achievements in the philosophy of mind. This concluding section of our paper is therefore intended as a restatement of the issues underlying this debate and the positions taken within it in terms that might be comprehensible to Western philosophers. We have seen that the debate centred around a particular trance state called the nirodhasamaapatti.h--the attainment of cessation. It is important to be clear about what this is. Its definitions in the earliest texts available to us make clear that it is a condition entirely without consciousness. differing from death only in that certain processes of the autonomic nervous system continue. At first sight it seems as though we are talking about a cataleptic trance in which the individual exhibits no reactions to external stimuli and is incapable of initiating action. Someone in a long-term coma would be another example. Such individuals, like the Buddhist who has attained cessation, exhibit no mental activity. But there is an important difference: for Western psychology, on the whole, catalepsy and coma are not thought of as though they are entirely without consciousness. It may be said that the psychotic in catalepsy has retreated into the depths of his psyche and erected powerful defence mechanisms against the external world, but it is not usually said that his consciousness has ceased altogether. The presup- P.388 position is always that within or beneath the apparent mindlessness mental processes continue; the subconscious mind still functions. Indeed: all therapy for cataleptic psychoses and long-term coma patients is based on such presup-positions. For some Buddhists, as we have seen, the attainment of cessation is thought of as a complete bringing to a halt of all mental events without remainder. The nearest analogy in Western thought would probably be the idea of brain death, that condition in which a patient is declared clinically dead but whose nervous system may be kept in operation--for a time at least-by artificial means. An extreme example of such a condition would be the case of a person who underwent a neat, surgical decapitation, and whose heartbeat and so forth were thereafter kept in operation artificially. For Western psychology and philosophy such a condition is by definition irreversible. After such a complete cessation of consciousness there is no way that it can reemerge. So, on the whole, Western thought would equate the complete cessation of consciousness with death. For the Buddhist philosopher, on the other hand, the problem is still more pressing, since he conceives death on a different theoretical model. In Buddhist thought,(48) death does not involve the complete cessation of consciousness but is simply one more moment in a causally linked stream of mental and physical events. Consciousness--in various more or less attenuated forms--continues through it. The problem for Buddhists then becomes not whether the attainment of cessation is the same as death, but whether such a condition is possible at all. For the fundamental presupposition behind all of Buddhist soteriology is that the continuum continues; there is no escape from continuing consciousness unless and until one enters nirvaa.na.(49) Therefore all Buddhist schools presuppose tha it must be possible for consciousness to reemerge after the meditator has entered the attainment of cessation, for if consciousness does not reemerge, the causal chain has been broken and the meditator has attained his goal of final liberation. To make the problem clearer let us formalize it: P1 Every event has a cause. P2 There are only two kinds of event: mental and physical. P3 Any given chain of caused events is beginningless and endless unless brought to a final halt in nirvaa.na. P4 It is possible that, in a given continuum X, at a given time T, there be a complete absence of mental events while physical events continue. All the Buddhist schools we are considering in this paper hold P1-P3; these are fundamental presuppositions. Let us see what happens if P4 also is assented to, as we have seen it to be by the Vaibhaa.sikas and by at least some Sautraantikas. There seem to be only three possible alternatives: P5 All of P1-P4 are true, and the cessation of mental events described in P4 is permanent and to be identified with nirvaa.na. P.389 No buddhist would wish to take this step. To identify nirvana with a specific trance state is heresy (though see note 49). Therefore we do not find an instance of this option in the texts we are considering. P6 All of P1-P4 are true and the cessation of mental events described in P4 is temporary; their reemergence is caused by continuing physical events. We find no clear instance of this position being taken either, and that this is so reveals something interesting about the basic Buddhist conception of the relationship between physical events and mental events. In terms of Western philosophy of mind it appears that Buddhists are, on the whole, modified dualists. They are strongly aware of the essential phenomenological difference between mental and physical events, and thus have many of the problems which have faced Western dualists in accounting for the causal relationship between the two. If it is the case that only things which are in some sense alike (samam) can be causally related, then it is difficult to see how consciousness can arise in a situation where there exists nothing but the physical. This is the main reason why P6 is not held as a possibility by the schools referred to in the sources discussed in the third section of this paper. This might seem to suggest that Buddhists must of necessity be adherents of parallelism--that view which states that physical and mental events run along in parallel streams, contemporaneous with each other but without causal connection, since it is not possible for events as radically different as the mental and the physical to causally influence one another. There are clearly many problems with such a view, problems which have been rehearsed ad nauseam in the history of Western philosophy, and it can be no part of this paper to review them yet again. It must suffice to point out that the Buddhist view is not quite so simplistic. One example should be enough to show this. Let us take the case of a person looking at a splash of red paint on a white wall. The Buddhist would call this an instance of visual consciousness (cak.survij~naanam) and say that it has as necessary precondition for its occurrence the contact between three things: mind (the general continuum of mental events) , sense-organ (in this case the eye, part of the continuum of physical events), and sense-object (the splash of paint) . This shows that interaction between the mental and the physical is possible. but it also makes very clear that for any kind of consciousness to occur, a preceding moment of consciousness is necessary. The mental and the physical can enter into various combinations with one another, but the requirement that each moment of consciousness have an immediately antecedent and homogenous cause shows why none of the schools are willing to assert that a mental event can arise solely from a physical event. Therefore, P6 is not a genuine option. P7 Ail of P1-P4 are true, and the cessation of mental events described in P4 is temporary. The reemergence of mental events is caused by other mental events. P.390 We have seen that both the Vaibhaa.sikas and Sautraantikas adhere to this view and that neither of their accounts is altogether satisfactory. The Vaibhaa.sika account postulates causation at a substantial temporal distance without explaining exactly how this can work. The Sautraantika image of "seeds" of consciousness preserved in a purely physical medium is just that--an image, unsystematized and probably incapable of being so. It explains nothing. In a sense, although both these schools wish to affirm P4, they cannot wholeheartedly do so. They have to preserve some element of existence of the mental--whether as seed or as past, but still in some sense extant and efficacious--cause, in order to account for the reemergence of consciousness without slipping back into the unacceptable P6. The final possibility open in this matter is to affirm P1-P3 and to deny P4. This means that the attainment of cessation as classically conceived in Buddhism becomes an impossibility because of the difficulties we have examined, For both Vasumitra, with his "unmanifest mental consciousness," and the Yogaacaarins, with their "store-consciousness," the complete cessation of consciousness is not possible short of nirvaa.na,(50) and with this understanding of the attainment of cessation we are back to the Western model of a short-lived cataleptic trance. There are clearly substantially fewer philosophical problems with this last position. The reemergence of consciousness is not a problem if it never fully ceased in the first place. The interesting point about this position is the issue of what kind of consciousness can remain in the attainment of cessation, and it is here that there are significant differences between Vasumitra and the Yogaacaarins. For Vasumitra the model of consciousness used is still that of the Vaibhaa.sikas and indeed that of the Abhidharmikas per se. It is, as we have seen in our brief discussion of visual consciousness, a type of interactional model in which consciousness is occasioned by the interaction of mental and physical events. The corollary of this is that for the Vaibhaa.sikas and Vasumitra, consciousness is intentional in Brentano's sense of that term--to be conscious is to be conscious of something. This explains the basic point of the long debate between Gho.saka and Vasumitra; Vasumitra says that consciousness is still possible in the attainment of cessation. Gho.saka replies that since all consciousness has an object, and since being conscious of something implies having sensation, if we allow consciousness in the attainment of cessation it seems that none of the other mental functions can cease. Vasumitra, hampered by his intentional model of consciousness, has no effective reply. It is only the Yogaacaarins, with a radically different idea of what consciousness is, who can escape this dilemma. The aalayavij~naanam is not necessarily a dualistic consciousness; it can exist without sensation or conceptualization; and it is fully capable of providing the necessary causal impetus for the reemergence of ordinary everyday consciousness from the attainment of cessation. To summarize: this debate on the reemergence of consciousness from the attainment of cessation is important from a number of different angles. For the P.391 Buddhologist it sheds new light on the history of scholastic Buddhism; for the philosopher it lays bare some key Buddhist assumptions about the nature of consciousness and the causal interactions between mind and body; and for the historian of religions it shows the vitally important role that the meditative experience of the religious virtuoso can play in shaping new theories of consciousness. NOTES 1. A full discussion of this issue is beyond the scope of the present article. For a partial discussion of the problem as it appears in Theravaada sources, see Paul Griffiths, "Concentration or Insight: The Problematic of Theravaada Buddhist Meditation-Theory," Journal of the American Academy of Religion 49, no. 4(1981): 605-624. 2. A fair amount of scholarly work has been done on the problems associated with the attainment of cessation. The following are important: Louis de La Vallee Poussin, "Musiila et Naarada, " Melanges Chinois et Bouddhique 5 (1936-1937), especially p. 210ff, L'Abhidharmako`sa de Vasubandhu, nouvelle ed. (Bruxelles: Institut Belge des Hautes Etudes Chinois, 1971), Tome 1, pp. 203-214; Tome 4, pp. 223-227; Tome 5, pp. 207-208; Etienne Lamotte, Lee Somme du Grand Vehicule D'Asa^nga (Louvain: Bureaux de Museon, 1938) (Tome 2, pp. 15-16 of the Notes et References section contains an especially useful set of references) ; Lambert Schmithausen, Der Nirvaa.na-Abschnitt in der Vini`scayasamgraha.nii der Yogaacaarabhuumi.h (Wien: Hermann Bohlaus, 1969) pp. 122-124; "Zur Struktur der erlosenden Erfahrung im Indischen Buddhismus," in Transzendenzerfahrung, Vollzugshorizont des Heils, hrsg. Gerhard Oberhammer (Wien: Publications of the De Nobili Research Library, 1978), pp. 97-119, "On Some Aspects of Descriptions of Theories of 'Liberating Insight' or 'Enlightenment' in Early Buddhism," in Studien Zum Jainismus und Buddhismus, hrsg. Klaus Bruhn und Albrecht Wezler (Wiesbaden: Franz Steiner, 1981), pp. 199-250 (these last two papers--the latter develops ideas outlined in the former--are especially suggestive. It is to be hoped that Professor Schmithausen will continue his investigations in this field) ; Stefan Anacker. "Vasubandhu's Karmasiddhiprakara.na and the Problem of the Highest Meditations," Philosophy East & West 22, no. 3 (October 1972): 247-258; Winston King, "The Structure and Dynamics of the Attainment of Cessation in Theravaada Meditation," Journal of the American Academy of Religion 45, no. 2, suppl. ( 1977): 707- 725. (much of the material in this paper is repeated in the relevant chapter of Professor King's later book, Theravaada Meditation: the Buddhist Transformation of Yoga (Pennsylvania: Pennsylvania State University Press, 1980); Noriaki Hakamaya, "Nirodhasamaapatti--its Historical Meaning in the Vij~naptimaatrataa system." Journal of Indian and Buddhist Studies 23. no. 2 (1975): 33-43 (this paper of Professor Hakamaya's, together with many personal discussions I have been fortunate enough to have with him, has had a great deal of influence on the present work. I wish to record my gratitude to him without, of course, suggesting that he should be held responsible for any of the ideas expressed herein). 3. The nirodhasamaapatti.h is not the only mindless trance state discussed in the AKBh. Among the cittaviprayuktasa.mskaaraa.h we also find the asa.mj~nisamaapatti.h and its result, aasa.mj~nikam, fully discussed at AKBh 68.12ff. The reasons for the distinctions made between these states and the attainment of cessation lie beyond the scope of this paper; it is sufficient to note that they are phenomenologically identical. 4. All references to this text in what follows will be to page and line of the text edited by Pralhad Pradhan, Abhidharmako`sabhaa.syam of Vasubandhu, 2nd ed. (Patna: Kashi Prasad Jayaswal Research Institute. 1975). We shall be considering 72.19-73.4. 5. Chos mngon pa'i mdzod kyi bshad pa, Peking No. 5591, Mngon pa'i bstan bcos GU 27b6-NGU 109a8. We shall be considering GU 88b1-89al. 6. All references to this text in what follows will be to page and line of the text edited by Swami P.392 Dwarikadas Shastri, Abhidharmako`sa and Bhaa.sya of Acharya Vasubandhu with Sphutaarthaa Commentary of AAcaarya Ya`somitra, 4 vols (Varanasi: Bauddha Bharati, 1970-1973) . We shall be considering 245.4-246.29. 7. Chos mngon pa'i mdzod kyi `grel bshad, Peking No. 5593, Mngon pa'i bstan bcos CU la1-CHU 408a8. We shall be considering CU 174b2-175a7. 8. Chos mngon pa'i mdzod kyi bshad pa'i rgya cher `grel pa don gyi de kho na nyid zhes bya ba, Peking No. 5875, Ngo mtshar bstan bcos TO la1-THO 565a8. We shall be considering TO 265b6-267a6. The full Sanskrit title of this commentary appears to have been Abhidharmako`sabhaa.sya.tiikaatattvaarthanaama. 9. Edited and translated into French by Etienne Lamotte, "Le Traite de L'Acte de Vasubandhu, " Melanges Chinois et Bouddhique 4 (1935-1936) : 151-263. References will be to page and line of Lamotte's edition of the Tibetan text. 10. Las grub pa'i bshad pa, Peking No. 5572. Sems tsam KU 69a6-117b1. 11. References will be to page and line of the text edited by Padmanabh S. Jaini, Abhidharmadiipa with Vibhaa.saprabhaav.rtti, 2nd ed. (Patna: Kashi Prasad Jayaswal Research Institute, 1977). 12. tatraacittakaanyeva nirodhaasa.mj~nisamaapattyaasa.mj~nikaaniiti vaibhaa.sikaadaya.h aparisphu.tamanovij~naanasacittakaaniiti sthaviravasumitraadaya.h aalayavij~naanasacittakaaniiti yogaacaaraa.h--iti siddhaantabheda.h/ (AKV 245.21-23//CU 174b2-4) 13. katham idaanii.m bahukaala.m niruddhaaccittaat punarapi citta.m jaayate/atiitasyaapyastitvaat i.syate vaibhaa.sikai.h samanantarapratyayatvam/(AKBh 72.19-21//GU 88b1-2) 14. AKBh 98.10-11//GU 114a4-5. 15. We may note that Vasubandhu's gloss here reveals a folk etymology rather than a true one. If samanantara were really derived from sama plus anantara we would expect samaanantara rather than samanantara. Ya`somitra recognizes the difficulty and says samaanaarthe sam`sabda.h--"The prefix samhas the meaning of homogenous" (AKV 342. 16). The Tibetans also translate the sam- prefix in this case as if it were sama- or samaana- using mtshungs pa. As always, therefore, the meaning of a given term to its users is more important than its "true" etymology. 16. AKBh 98.9-10//GU 114a3-4. 17. AKBh 98.23-28//GU 114b5-7. 18. AKV 344.11-14. 19. TO 358a1-2. 20. AKBh 295.20ff. makes this very clear. For an illuminating discussion of the rationale behind the Vaibhaa.sika assertion of sarvam asti see Paul M. Williams, "On the Abhidharma Ontology," Journal of Indian Philosophy 9(1981): 227-257. 21. 'das pu yang yod pa'i phyir zhes bya ba 'byung ste/gal te yod na ji ltar 'das pa ying zhe na/byed pa gas pa'i 'das pa zhes bya'i/rang gi ngo be yongs su btang ha'i phyir ni ma yin no/de'i phyir gags ma thag pa lar `gags nas yun ring du lon pa yang bye brag tu smra ha rnams mtshungs pa de ma thag pa'i rkyen nyid du `dod do-kun gzhi rnam par shes pa yod pa ma yin pa sems la bzhag nas bye brag tu smra ba smos so/(AK.T TO 265b6-8) 22. apare punaraahu.h/katha.m taavadaarupyopapannaanaa.m ciraniruddhe'pi ruupe punarapi ruupa.m jaayate/cittaadeva hi tajjaayate na ruupaat/eva.m cittam-apyasmaadeva sendriyaatkaayaajjaayate na cittaat/anyonyabiijaka.m hyetadubhaya.m yad uta citta.m ca sendriya`sca kaaya iti puurvaacaaryaa.h/(AKBH 72.21-4//GU 88b2.4) 23. AKV 246.15//CU 174b7-8. 24. For further information we may look to KSP, which quotes a similar opinion: kha cig na re de'i sa hen dhang po gzugs can la gnas pa las te (KSP 193.14) . The KS.T glosses kha cig with "some particular Sautraantikas," thus indicating that at least some sautraantikas held this view. Clearly, as the rest of the KSP shows, not all did so. 25. AK.T TO 265b8-266a1. 26. gal te ris mthun pa'i rgya la ma bltos par dbang po bcas pa i lus las rnam par shes pa skye ba nyid yin na/rten dang dmigs pa gcig car gnas pa na yul thams cad la rnam par shes skye bar `gyur ro/sems snga ma gang yin pa de la rag las pas sems phyi ma skye bas na/gnyis pa la mtshungs pa de ma thag per rkyen med pas rten dang dmigs pa yod kyang cig car mi skye bas tha/bar `gyur ro/gal te sems yod pa ma yin yang sems med pa'i dbang po dang bcas pa'i lus kyi sa bon las so zhe na/'di yang mi rigs te/khyad par gyi rgyu med pa'i phyir ro/so zhe na/`di yang mi rigs te/khyad par gyi rgyu med pa'i phyir ro/(AK.T TO 266a2-4) P.393 27. On sabhaagataa see AKBh 67.14ff//GU 83b6ff. 28. bhadantavasumitrastvaaha parip.rcchaayaa.m yasyaacittakaa nirodha-samaapattistasyai.sa do.so mama tu sacittakaa samaapattiriti/(AKBh 72.24-26//GU 88b4-5) 29. AKV 246.20-3//GU 175a2-4. 30. de la btsun pa dbyig bshes ni yid kyi rnam par shes pa'i sems dang bcas pa `dod kyi/kun gzhi rnam par shes pa'i sems dang bcas pa ni ma yin no/(AK.T TO 266a8) 31. See AKV 245.22//CU 175a2-4. 32. See the quotation from the Mahaavibhaa.saa given by Poussin in L'Abhidharmako`sa, Tome 1, p. 212n2, and more fully by Hakamaya in "Nirodhasamaapatti--its Historical Meaning in the Vij~naptimaatrataa System,'' pp. 36-37. 33. For Poussin's view see L'Abhidarmako`sa, Tome 1, p. xlv; for Lamotte's view see "Le Traite," p. 237n77. 34. Noriaki Hakamaya, "On a Verse Quoted in the Tibetan Translation of the mahaayaanasa.mgrahopanibandhana." Journal of Indian and Buddhist Studies 22, no. 2 (1974): 61-21, and p. 19n6. The two Tibetan versions of this quotation (KSP 193.33-1 94.3//AKBh GU 88b4-5) are substantially the same and also certainly rest upon an identical Sanskrit original. 35. phyi rol gyi don yod par smra ba (KS.T KU 94b8). 36. We may note finally that in the Abhidharmadiipa, a Vaibhaa.sika rejoinder to the AKBh, the view that the nirodhasamaapatti.h possesses mind is attributed to the Ko`sakaara.h-atra puna.h ko`sakaara.h pratijaaniite sacittakeya.m samaapatti.h iti (AD 93.14). It seems to us that this is the view attributed to Vasumitra in the section of the AKBh under discussion in this paper. 37. See AKBh 143.2ff. 38. bhadanta gho.saka aaha tadida.m nopapadyate/sati hi vij~naane trayaa.naam sa.mnipaata.h spar`sa.h/ spar`sapratyayaa ca vedanaa sa.mj~naa cetanetyukta.m bhagavataa/ata.h sa.mj~naavedanayorapyatra nirodho na syaat(AKBh 72.26-8//GU 88b5-7) There are two illegibilities in the Tibetan text at this point, easily reconstructed from the Sanskrit given above. There is also one minor difference between the Sanskrit and the Tibetan. For the Sanskrit trayaa.naam sa.mnipaata.h spar`sa.h (which places sa.mnipaata.h in apposition to spar`sa.h) the Tibetan translation reads gsum 'dus pa las reg pa, constructing an ablative case which, if anything, makes the sense clearer. 39. AKV 246.23-5//CU 175a4. There is a difference between Sanskrit and Tibetan here. The Sanskrit reads tasmaad bhadantagho.saka eva.m prasa^nga.m karoti. The Tibetan reads de'i phyir de ltar thal bar byed par yin no, omitting the reference to Gho.saka. 40. reg pa'i rkyen gyis tshor ba dang zhes bya ba rgyas par 'byung ba la/gegs byed pa med pa'i rgyu yod no 'bras bu gdon mi za bar `byung bas `gog pa'i snyoms par `jug pa'i rdzas `di la `du shes dang tshor ba dag kyang `gog par mi `gyur/(AK.t TO 266b2-3) 41. See AKBh 143.2ff. and Th. Stcherbatsky, Buddhist Logic, vol. 2 (New York: Dover Publications, 1962). pp. 311-312. 42. athaapisyaat/yathaa vedanaapratyayaa t.r.s.netyuktam/satyaamapi tu vedanaayaam arhato na t.r.s.notpattireva.m satyapi spar`se vedanaadayo na syur iti/tasyaavi`se.sitatvaat/(AKBh 72.28-73.2/GU 88b7-8) The Tibetan translation differs slightly here; for tasyaavi`se.sitatvaat, we read khyad par du byas pa'i phyir de ni ma yin te. 43. AK.T TO 266b3-7. 44. tshor ba bzhin du reg pa thams cad tshor ba'i rkyen ma yin ni/(AK.T TO 266b6) 45. avidyaasa.mspar`saja.m hi vedita.m pratiityotpannaa t.r.s.netyukta.m na tu vedanotpattau spar`so vi`se.sita ityasamaanametat/tasmaadacittakaa nirodhasamaapattiriti vaibhaa.sikaa.h/(AKBh 73.2-4//GU 88b8-89a1) 46. na tu vedanotpattaaviti/na tu vedanotpattau spar`so vi`se.sita.h iid.r`sa.h spar`so vedanaapratyaya iti/(AKV 246.27-8//CU 175a6-7) The Tibetan translation here omits na tu vedanotpattau spar`so vi`se.sita.h, the sense is not affected. 47. ma rig pa mi dpe'i tshor ba ni srid pa'i rkye yin no... de'i phyir bye brag med par nye bar blangs pa'i phyir reg pa thams cad tshor ba'i rkyen yin pas dpe `di mtshungs pa ma yin no/(AK.T TO 267a1 -2) 48. Almost any sentence beginning "In Buddhist thought..." is very easily falsifiable, and this one is no exception. The whole of the concluding section of this paper moves on a high level of generality and should be taken and discussed on that level. P.394 49. This raises the fascinating possibility that the nirodhasamaapatti.h may be assimilated to nirvaa.na. There is in fact some evidence that by some Buddhists at some periods the two were thought to be identical (see Paul Griffiths, "Buddhist Jhaana: A Form-Critical Study, " forthcoming in Religion). 50. And maybe not even there for a Yogaacaarin.