To view the diacritics on this page, you must install the Indic Times font on your machine and have a browser capable of displaying the Unicode (utf-8) character set.
paul.williams@bristol.ak.uk
John Pettit has written an eminently intelligent, scholarly, and detailed review of my book Altruism and Reality. For this I am extremely grateful. I am not sure, however, that I agree with some of his more critical points, and it may be of some value to try and formulate a tentative response to what seems to me to be his two main contentions: (a) that Mi pham, contrary to what I imply in my book, does not hold in any obvious way to a gzhan stong position, at least if the gzhan stong position is held to involve some sort of absolutism; and (b) that my attack on "SĀntideva's inference from no self to altruism at BCA 8:101-3 is misplaced because I have ignored the relevance of the innate misapprehension of self, to be undermined for the Madhyamaka in particular through meditation in addition to philosophical analysis.
Pettit is a very considerable authority on Mi pham -- indeed one might say a "devotee" of Mi pham. He recently completed a doctorate on Mi pham's interpretation of Madhyamaka, and the published version of his thesis is expected soon (I believe with Wisdom Publications). It will without doubt make an important contribution to the understanding and appreciation of this significant Tibetan thinker, arguably philosophically the most interesting Tibetan Madhyamaka in recent centuries. I am quite prepared to consider that my understanding of Mi pham in relationship to his view of the ontological status of the ultimate truth is simplistic. Indeed Pettit particularly takes me to task for comments made in the first and earliest written of my papers published in Altruism and Reality. A modification of a number of views and my approach can be detected through the papers published in this book, which were written over six years, and I had already begun to consider that the issues of gzhan stong, absolutism, and, inter alia, Mi pham were more complicated than they had initially appeared to me by the time I wrote my book on Mi pham's treatment of rang rig, The Reflexive Nature of Awareness (Curzon, 1998). This book was written some four years after the paper to which Pettit takes particular objection as regards my views on Mi pham and gzhan stong. Pettit also plans to review The Reflexive Nature of Awareness in a forthcoming edition of the Journal of Buddhist Ethics. In that book I set out in a long footnote (pp. 199-206) to try and defend my association of Mi pham with the gzhan stong position. I hope I can be forgiven for repeating most of my note here:
Perhaps I should say a little more here about my association (if cautious and tentative) of Mi pham with a teaching of Ultimate Reality and even a version of the gzhan stong position. This latter in particular will be thought controversial. Geoffrey Samuel is not alone, for example, when he states that "…Mi pham, the most important Nyingmapa scholar of the late nineteenth century and a prominent member of the Rimed movement . . . argued for a Madhyamaka interpretation of the Nyingmapa teachings, in this respect opposing the tendency of other Rimed teachers such as the Kagyüdpa lama Jamgön Kongtrul to employ more positive (shentong) phraseology" (Williams, 1993, p. 465). There is no doubt that Mi pham argued for a Madhyamaka interpretation of rNying ma pa philosophy; the question is what his understanding of Madhyamaka was, taken as a whole with its place beside Tantric and rDzogs chen theory and praxis. For example, take Mi pham's dBu ma'i lta khrid zab mo: Having explained the process of Madhyamaka analysis to discover emptiness of things, absence of inherent existence, Mi pham comments that "while empty, appearing; while appearing, empty; one should contemplate appearance and emptiness in the manner of an illusion -- this is the metaphoric (rnam grangs; verbal; conceptual) ultimate" (stong bzhin snang la snang bzhin stong // snang stong sgyu ma'i tshul du bsgom // 'di ni rnam grangs pa'i don dam). This ultimate emptiness, absence of inherent existence, the ultimate of a PrĀsa"ngika like Chandrakiirti or Tsong kha pa, the status of things which exist due to dependent origination -- is the same ultimate that Mi pham speaks about in his commentary to BodhicaryĀvatĀra 9:140 as a nyi tshe ba'i stong pa [nyid], a limited or relative, i.e. not truly ultimate, ultimate. This expression is used disparagingly of false claims to real ultimacy in the La"nkĀvatĀrasuutra and SamĀdhirĀjasuutra, and in those suutras the Buddha says that the nyi tshe ba'i stong pa nyid is to be abandoned [see Williams 1983, pp. 133ff, which treats Mi bskyod rdo rje's similar use of nyi tshe ba'i rnam grangs pa'i stong nyid in attacking Tsong kha pa's understanding of the ultimate as emptiness. For Mi pham's use, see also Williams 1998a]. Use of such an expression -- employed critically for the wrong view in the suutras -- clearly indicates an ultimate which is not at all the final true ultimate. In his comments on BCA 9:140 (p. 86) Mi pham explicitly links rnam grangs pa such as absence of a pot, or empty [emptiness] of a pot with the nyi tshe ba'i stong pa [nyid] as both finally delusory since the entities which they negate are delusory, although true to his use of PrĀsa"ngika thought, Mi pham does hold that these negations are true relatively/conventionally, and thus emptiness is true as far as things which come within its range are concerned. Thus it is clear that for Mi pham the PrĀsa"ngika emptiness put forward by, say, Tsong kha pa -- a nonaffirming negation (med dgag) -- is not a true ultimate, nor is it the true ultimate, since Mi pham is prepared to speak of an actual ultimate (a rnam grangs min pa'i don dam) which must therefore be different from this emptiness. In his dBu ma'i lta khrid zab mo, for example, Mi pham continues with the caution that with the previous emptiness, (the rnam grangs pa'i don dam [stong pa nyid]), since one has not passed beyond conceptual construction (kun tu rtog) there is no seeing the true nature (chos nyid) which is free from verbal differentiations (spros bral) (kun tu rtog las ma brgal phyir // spros bral chos nyid mthong ba min). Note that Mi pham does not use the expression chos nyid here for the emptiness which is a mere negation, mere absence of inherent existence. The emptiness which is a mere negation is not the highest, final emptiness. This is not because of the way it is seen, or the fact that in speaking about it has been conceptualised. It is something to do with what it actually is. Absence of inherent existence is, in its very nature, by definition, a pure negation. And as a pure negation it is, as Mi pham knew from BodhicaryĀvatĀra 9:140, in its very nature, by definition, relative to the negandum -- in this case, fictitious inherent existence. Thus absence of inherent existence, emptiness, is in its very nature, by definition, according to "SĀntideva, ultimately delusory and, for Mi pham, is in its very nature, by definition a nyi tshe ba'i stong pa nyid. As a negation, absence of inherent existence is relative to its counterpositive. For Mi pham this makes it a conceptual emptiness, a conceptual (rnam grangs) ultimate. Thus it is clear, I think, that for Mi pham if there is a rnam grangs min pa'i don dam it cannot be the same as mere absence of inherent existence, and it cannot therefore be a mere nonaffirming negation (med dgag). In the dBu ma'i lta khrid zab mo Mi pham refers to this rnam grangs min pa'i don dam as the true nature (chos nyid) of all dharmas (chos). And although it is free from affirmations of existence, nonexistence and so on (yod med la sogs khas len bral) and has a meaning (or referent) which is strictly unutterable (brjod med don) Mi pham is also unequivocal in referring to it as "gnosis (ye shes/j~nĀna) which is nonconceptual absorption, known as reflexive awareness (cf. the "mere gnosis which is reflexive awareness" (so so rang rig pa'i ye shes tsam) of rDzogs chen: de ni chos kun chos nyid [9a] de // rnam grangs min pa'i don dam ba // so sor rang gi rig bya ba // mnyam gzhag mi rtog ye shes yin). Thus a ye shes -- an expression which must have primary mentalistic usage -- is said to be the rnam grangs min pa'i don dam. We have seen that absence of inherent existence is defective because it is a relative ultimate and comes within the range of conceptualisation. Nondual ye shes is either truly ultimate or it is not. Since it is a consciousness of some sort it is clearly not a pure negation. This, apart from anything else, is why Mi pham elsewhere (in 1976, for example) can deny that the rnam grangs min pa'i don dam is a mere nonaffirming negation, a pure negative bound up with duality. If absence of inherent existence is (with Chandrakiirti and Tsong kha pa) the only ultimate then a ye shes no matter how exalted should be a conventional truth and there could be no true ultimate at all. The only ultimate would be an ultimate said by Mi pham and others to be a relative nyi tshe ba'i stong pa nyid. But clearly for Mi pham this nondual ye shes is not only an ultimate, but the finally true nonconceptual ultimate, beyond the nyi tshe ba'i stong pa nyid, and an affirming negative (ma yin dgag). But the ultimate of, say, Tsong kha pa, the emptiness which is a mere negation of inherent existence, is a thoroughly uncompromising ultimate. Anything both existing and other than absence of inherent existence itself is, for Tsong kha pa, a conventional truth, and anything which is putatively both said to exist and not to come within the range of absence of inherent existence would be a true, inherently existent ultimate truth. According to "SĀntideva in BodhicaryĀvatĀra 9:140, followed by Mi pham, the relative (= rnam grangs) emptiness is delusory, as is that which is empty (its basis). Thus if the nonconceptual ye shes were empty of inherent existence, that relative emptiness alone would be the only ultimate, while nonconceptual ye shes (= the rnam grangs min pa'i don dam) would become delusory. I do not think this is what Mi pham wants. The only alternative is that he considers the nonconceptual ye shes, as nonconceptual, not to come within the range of absence of inherent existence and for that very reason to be truly ultimate. However the key to understanding Mi pham's position lies in his appeal to true nonconceptuality. Thus it is a condition of the existence of a truly nonconceptual entity that one cannot say that it exists. Nevertheless it is also a condition of praxis and therefore the philosophical system which supports certain sorts of praxis that one can, one must, indicate that the nonconceptual entity exists. Thus Mi pham, in denying that the true ultimate can be a mere negation, and denying that it can be a nonaffirming negative, in fact indicates a true ultimate which really -- dare I say, inherently -- exists, exists in the fullest possible sense. However because it is nonconceptual it must be beyond both existence and nonexistence. The view that Mi pham does not hold to a gzhan stong-type position is I suspect largely the result of Mi pham's criticisms of the Jo nang pa gzhan stong Absolutism in his bDe gshegs snying po'i stong thun chen mo Seng ge'i nga ro. The problem with any gzhan stong position which is based on the nonconceptuality of the Ultimate is that there is a sense in which it contradicts itself in speaking. The Ultimate has to be shown to exist while at the same time the claim to existence has to be withdrawn in order that one is not accused of falling within the conceptual dichotomy of existence and nonexistence. There are perhaps radically different types of gzhan stong position in Tibetan thought, and indeed not all those who hold such a position would want to use this expression for it. The Jo nang pa tradition had for centuries provided Tibetan thinkers with an example of a philosophy which apparently contradicts nonconceptuality by affirming the existence of the ultimate. Irrespective of whether other thinkers in fact hold substantially the same position on ontology or ontological primacy as the Jo nang tradition, any thinker who wishes to clarify his (or her) adherence to true nonconceptuality will in Tibet need to refute someone for falling into the wrong view of existence, and someone for falling into that of nonexistence. The Jo nang tradition supplies the first, and for many critics Tsong kha pa et al. provide the second. The Jo nang pas have contradicted themselves in speaking of their nonconceptual Ultimate. The more they speak and systematise, the more they contradict themselves. Thus it is perfectly possible to argue that the problem for the Jo nang pas is not what they say, but the fact that they are saying it at all (plus -- I would want to add -- any difficulties which might result from the very coherence of nonconceptual entities, let alone mentioning them). The problem for their critics however, is that short of accepting a third value which would equal the fourth member of the catu.sko.ti (i.e., the ultimate neither is nor is not), to deny that the ultimate is a mere negation, mere absence of inherent existence -- and indeed to relativise that negation -- must logically entail that the ultimate, if it is to be an actual ultimate reality at all, is something which does inherently exist -- even if one is then going to retreat by saying that since it is beyond all conceptuality we cannot actually say that it exists. In refuting both the Jo nang tradition on the one hand for adhering to existence, and Tsong kha pa et al. on the other for adhering to nonexistence, however, it seems to me that Mi pham also in true ris med fashion wants to be able to incorporate both their contributions. Thus Tsong kha pa, for example, is right in that anything which can be conceptualised, i.e., can be spoken about, can be analysed with ultimate analysis and shown to be empty of inherent existence, merely conventional. Any ultimate which can be spoken about also comes within the range of refutation through ultimate analysis. Thus it is also correct to say that for Mi pham properly speaking one cannot refer to anything which resists ultimate analysis. But of course, anything which can be spoken about for Mi pham comes within the range of conceptualisation (= rnam grangs), and its ultimate also is a rnam grangs pa'i don dam. The rnam grangs min pa'i don dam is precisely unutterable and thus cannot be refuted by ultimate analysis. How could it? And the Jo nang pas are, I suggest, right for Mi pham in indicating that the true ultimate is beyond any emptiness which is a mere negation, and cannot be refuted by conceptual (or any other) means. In the dBu ma'i lta khrid zab mo Mi pham refers to the praxis which finally understands the rnam grangs min pa'i don dam as he has explained it as the "Great Madhyamaka yoga" (dbu ma chen po'i rnal 'byor yin). Mi pham frequently refers to the true position in such contexts as the Great Madhyamaka (see, for example, Williams 1992a, and Williams 1998a; cf. Khenchen Kunzang Palden 1993, p. 114 following Mi pham's commentary to BodhicaryĀvatĀra 9:140 exactly: "To abide in that Primal Wisdom in which all concepts have subsided -- this is indeed the Great Madhyamaka, the Great Middle Way free from all assertions."). The expression "Great Madhyamaka" has a very venerable ancestry in Tibetan thought (see van der Kuijp 1983, pp. 37 ff; Hookham 1991, passim, but esp. pp. 157-9). Basically it has one of two meanings. On the one hand it can be used to refer simply to PrĀsa"ngika Madhyamaka, or the particular understanding of Madhyamaka held as finally correct by thinker X. This way of using the term tends to be earlier and with reference to earlier thinkers. But the term had for centuries, and especially in more recent centuries, been used particularly as a technical term for the gzhan stong position -- very often associated with the Jo nang pas -- and it had become a key expression among ris med thinkers precisely for an approach which, in its stress on a nonconceptual Ultimate which thereby transcends ultimate analysis was, I suggest, important to the ris med project of harmonising traditional doctrinal rivalries. Van der Kuijp refers in this context to Kong sprul blo gros mtha' yas (1813-99), an older contemporary of Mi pham and very much part of the same circle and movement: "it is useful to realise how one of the greatest founders and exponents of the nineteenth century 'non-partial' (ris-med) movement, which was the direct cause for a renewed interest in this 'Great Madhyamaka,' thought about the antecedents of a way of thinking he felt so close to" (p. 40). It is indeed possible that Mi pham intended by his use of this expression nothing more than the PrĀsa"ngika position of absence of inherent existence as the only ultimate. But I very much doubt it, particularly in the light of the discussion above. The term "Great Madhyamaka" was very much part of his ris med environment as a term for a form of gzhan stong. Had Mi pham wanted to use it in a completely different way he would surely have made the point very clear. Mi pham was no fool.
Note that I am perfectly aware of Mi pham's criticisms of the Jo nang pa position, and in spite of what Pettit says I do not want "to suggest that his [Mi pham's] thought stems more or less directly from the Jonang gZhan stong tradition." Moreover, I am also prepared to accept Pettit's point that "'gZhan stong pa' is probably not a label with which Mi pham himself would have been comfortable," at least if gzhan stong pa is seen to be all but identified with Jo nang pa. Perhaps at least part of the disagreement here between Pettit and myself is one of approach. Pettit is rightly very interested in an accurate understanding of how Mi pham would describe his position. My interest is one of philosophy. I am interested in philosophical unpacking -- in what a philosopher must mean in order to be coherent, and in the understanding and analysis of supporting arguments. I hasten to add that it does not follow from this that I would be happy with a description which could in no way be supported from the texts themselves, or which the philosopher himself would reject as alien, quite contrary to both word and intention. But there is a difference of nuance between these two approaches. I argue on grounds of logical coherence that Mi pham's position appears to imply some form of gzhan stong, even some form of absolutism, whether or not Mi phan would be happy with that characterization. Note, incidentally, that as regards my association of Mi pham with gzhan stong (albeit not of a Jo nang sort) I doubt very much that Pettit will agree. But at least it can be seen that since writing my earlier paper I have become more sensitive to the complexity of this issue, and I am willing to defend my interpretation.
It is with Pettit's treatment of my lengthy fifth chapter -- a chapter which takes up about half of the book -- that our disagreements are perhaps most acutely felt. Again, some of this disagreement may relate to approach. I am interested in a particular argument given by "SĀntideva in chapter eight of his BodhicaryĀvatĀra (vv. 101-3). I take it that he is giving an argument intended to imply a conclusion. It certainly looks that way, and if "SĀntideva had intended it to be taken in some other way (as an emotional exhortation to be kind, for example) he should have expressed himself more clearly. Anyway, it seems clear to me that it is not taken in some other way in the context of the BCA, nor by any commentary that I am aware of. Thus, if "SĀntideva is arguing for a conclusion, using reasons, I take it that given the sort of topics we are dealing with he is likely to be doing philosophy. Since he does appear to be doing philosophy, it seems that we are entitled to examine the premises and the argument to see if they work or not. Whether the argument works or not is quite independent of who is stating it. Thus whether "SĀntideva is "one of the most compelling (and presumably coherent) of Indian Buddhist philosopher-saints" (Pettit) is irrelevant. No doubt he has been (and still is) compelling. Whether he is coherent remains to be seen, but either way the attack is not on "SĀntideva per se. I much admire Aristotle, but I do not hold that all Aristotle's positions and arguments are either compelling or coherent. This is how it is with all philosophers. It would be most surprising if that were not the case with Buddhist philosophers too.
Thus my interest in "SĀntideva here is in the coherence of an argument, and an argument given in the BodhicaryĀvatĀra by "SĀntideva, not a variant argument given, for example, centuries later by a dGe lugs pa. One would not defend Plato by discussing the views of Plotinus, or at least, one could not do that without considerable prior justification. I am not sure why we should treat Buddhist philosophers any differently from the way we treat Greek, Medieval Christian, Scottish, or German philosophers. In discussing "SĀntideva's argument at BCA 8:101-3, I deal with a number of central issues in Buddhist philosophy relating in particular to personal identity and ethics, and I argue against a bundle theory of personal identity and also in favor of a necessary relationship between pain and the person, the subject, who is in pain. But in order to focus here more closely on my disagreements with Pettit let me first outline what my central thesis concerning "SĀntideva's argument is.
I contend that "SĀntideva's argument at BodhicaryĀvatĀra 8:101-3 will not work since, in order for an implication from no Self to no selfishness to have any plausibility, no Self would have to entail no conventional self as well as no ultimate, unchanging metaphysical Ātman. My contention here is one of logical implication. It seems to me obvious that selfishness is a matter of conventional selves. It involves giving preferential treatment to this person -- the one who is me -- rather than that person, the one who is other. No selfishness simply does not follow from no Self (no metaphysical Ātman) per se. I can quite consistently accept that I have no metaphysical Ātman -- say, an independent unchanging Cartesian self -- and be perfectly selfish. I suspect many philosophers and scientists do and are. Thus if "SĀntideva's argument at BCA 8:101-3 is to be plausible it involves making some sort of ontological distinction between the conventional self and du.hkha such that we can consistently deny the conventional self and hold that it still makes sense to speak of there being du.hkha. Thereby it will also still make sense to exhort the removal of du.hkha, but a du.hkha which must be stripped of any reference to subjects because stripped of any reference to conventional selves. "SĀntideva chooses to do this in terms of a model which would be perfectly acceptable to other Buddhists as well, a model familiar from the Abhidharma. There are no selves (even conventionally) because persons are fictions, conventional constructs superimposed upon a spatio-temporal series (of momentary dharmas).
But I argue this strategy will not work. Take the case of physical pain, the actual sensation felt when treading on a drawing-pin. Such pains form a subclass of mental events coming under du.hkha. Thus if the bodhisattva is to remove the du.hkha of others he or she has to remove such physical pains. But I contend that in removing the conventional self in order to argue for altruism "SĀntideva has himself removed the very basis for making sense of pain-statements. I argue, using as alternatives an adverbial analysis and an event analysis of pain-statements, that physical sensations of pain have a necessary dependence on subjects. "SĀntideva's argument requires that we can make sense of pains not anchored to subjects, and the concept of pains not anchored to subjects seems to me to be meaningless. Thus I argue that since granted "SĀntideva's argument the bodhisattva could not remove pains, the bodhisattva could not remove all du.hkha. Thereby "SĀntideva has destroyed the bodhisattva path.
Pettit seems to consider that my problem is one of how psychologically to bring about altruism from the position of no self. He indicates this inter alia in his observation that "Williams is certainly correct in arguing that to negate the philosophical misconception of a True Self will not do much to undermine our tendency to selfishly prefer ourselves over others. However…." My problem, however, is not psychological, but rather conceptual, logical. I find certain inferences "SĀntideva wants to make simply do not follow. Whether psychologically a person as a matter of fact becomes altruistic as a result of the Buddhist arguments against the Ātman is an empirical matter to which I do not know the answer. Meditating has nothing to do with what actually interests me here. If an argument does not logically follow, repeated familiarization with it in meditation does not bring it about that it logically follows, even if it might lead to psychological conviction. It is no use arguing that repeated familiarization in meditation with the idea of square circles will make the idea of square circles coherent. It might well be the case psychologically for all I know that a particular person, contemplating repeatedly that we lack an Ātman, comes to be less selfish. But that would not be as a matter of logical implication, and there is in "SĀntideva's argument no logical reason why this altruism should come about. One could construct alternative scenarios, where repeatedly contemplating that one has an Ātman brings about altruism, or where repeatedly contemplating that one does not have an Ātman brings about selfishness. I leave my reader to think of the examples. It simply does not follow that lack of Ātman logically entails selflessness. It is perfectly logically possible both to hold that one lacks an Ātman and to be selfish. People do, and are. Many modern psychologists would accept that we lack an Ātman, but are also be selfish. They cannot be accused of logical contradiction.
Contrary to what Pettit says, I do not hold that "[t]he negation of the True Self…will not suffice to eliminate our everyday conception of 'myself,' nor presumably our instinctive preference for that self over other selves…because in Gelugpa Madhyamaka the ordinary conception of 'John' as a person is distinct from 'Paul' and so forth is held to be a valid distinction…." My interest in the logical coherence of "SĀntideva's argument has nothing to do with presupposing a much later dGe lugs analysis, nor does it have anything to do with the psychology of how we bring about from the supposed truth of no self the desired altruism. I argue on logical grounds that the negation of the Ātman will not eliminate selfishness because denying the Ātman is compatible with being selfish. This is because selfishness proceeds from giving preference to this person (the person one is) over that person, and privileging this person has nothing to do with whether we have an Ātman or not. The distinction I am making is indeed recognized in the dGe lugs distinction between the Ātman and the pudgala (gang zag), where the former is denied even conventionally while the latter exists as the conventional self, the person, marked by subjectivity. In making this distinction it seems to me dGe lugs pa theorists are philosophically right, but they adopt one horn of the dilemma I have constructed for "SĀntideva and in so doing render "SĀntideva's argument void. They could have adopted the other horn of the dilemma, and argued that no Ātman entails no conventional self. This, I contend, is what "SĀntideva needs for his conclusion and it appears to me to be what "SĀntideva himself actually holds. But it has many paradoxical implications, including the incoherence of pain-statements and thus the bodhisattva's activity in removing pain. Therefore, quite contrary to what Pettit says, I precisely do not hold that "SĀntideva's problem has anything to do with adopting the dGe lugs distinction between the Ātman and the conventional person. I have presented "SĀntideva (or rather, his argument) with a dilemma. The dGe lugs distinction is a strategy which "SĀntideva does not adopt. It would not have helped his argument here if he had (although its implications would certainly have been less destructive to the bodhisattva path). And it is the argument here which interests me.
Pettit considers that the key to exonerating "SĀntideva of an argument that does not follow lies in the distinction between the theoretical misconception of a True Self (Ātman) and the innate misconception of self. I am said to be "confused, or at least underinformed" about this crucial distinction. Actually, I am quite aware of this distinction, which is common in dGe lugs writings. I often refer to it in discussing dGe lugs Madhyamaka. But I am not quite sure what relevance this distinction is in discussing "SĀntideva's actual argument. In other words, I am not sure what relevance it is to the actual task I set myself in my paper. Whether the Madhyamaka (Tibetan or otherwise) can introduce strategies to escape the issues for which I criticized "SĀntideva is quite different. If the Madhyamaka can, it might satisfy those who now follow Madhyamaka, but it will not exonerate "SĀntideva. Pettit states that "what "SĀntideva is most concerned with is the innate misconception of self," and he then proceeds to talk about "the dGe lugs tradition of analytical meditation," sliding from this to "Madhyamika analytical meditation" as if they are identical. But as far as I know "SĀntideva does not hold to the distinction between the theoretical misconception of a True Self (Ātman) and the innate misconception of self, at least as a self-conscious formalized category. I suppose the Sanskrit for "innate (mis)conception of self" would be something like sahajĀtmagrĀha, but this would be a back formation from the Tibetan bdag 'dzin lhan skyes. I have not found the Sanskrit expression in the standard dictionaries of classical or Buddhist hybrid Sanskrit. I have also not found so far the expression used in Indian Madhyamaka sources (it may be there -- I have not looked everywhere). I have examined all the Indian commentaries on BCA 8:101-3, and a range of Tibetan commentaries from all schools, including the standard dGe lugs commentaries and a commentary by Mi pham's pupil Kun bzang dpal ldan. Not one makes any reference to the innate conception of self, or the relevance to the distinction between the innate conception of self and theoretical conception of Self, in considering these verses.
Commentators like Praj~nĀkaramati (whose commentary has the advantage of surviving in Sanskrit) make it quite clear that the argument here concerns simply the Buddhist case against the Ātman. Thus Praj~nĀkaramati states that "because there is no existence of any Ātman and so on which is the owner, the one of whom -- the subject -- there is pain does not exist" (evam ĀtmĀde.h svĀmina.h kasyacid abhĀvĀd yasya sa.mbandhi du.hkha.m sa nĀsti). The reason for there being no Ātman is the composite nature of the person. Praj~nĀkaramati begins his commentary on BCA 8:101 with "There does not exist any unitary ultimate reality called a 'continuum.' This is of the nature of a stream which is the ordered succession of moments occurring under the aspect of cause and effect" (sa.mtĀno nĀma na ka"scid eka.h paramĀrthasan sa.mbhavati / ki.m tarhi kĀryakĀra.nabhĀvaprav.rttak.sa.napara.mparĀpravĀharuupa evĀyam). Note the association of absence of Ātman with the lack of a "unitary ultimate reality" (eka.h paramĀrthasan). This is nothing to do with any innate conception of self. Praj~nĀkaramati is saying that because, as the Buddhist holds, there is no real Ātman (a True Self), therefore there is no subject for pain. Thus, he will want to add (following "SĀntideva) we cannot distinguish between my pain and the other's pain. Therefore in removing pain -- it is a given that we all do as a matter of fact remove pain -- it is irrational to remove my pain and not the pain of others.
I have argued this reasoning will not work. Pettit has yet to convince me that the argument here is not about the Ātman which Buddhists throughout history have attacked, the True Self, and is actually about the innate conception of self. But even if it were about the innate conception of self, it is not clear to me how introducing this notion will help "SĀntideva's argument. What is the innate conception of self? Pettit refers to the common dGe lugs pa explanation where the innate conception of self is identified when a sense of "I" or "me" arises as one is falsely accused or verbally abused. But there are major philosophical and psychological unclarities here which need unpacking. The claim is that this sense of "I" felt in cases of false accusation is the same sense of "I," differing only in degree, as that possessed by all unenlightened sentient beings. The sense of "I" possessed by a cat, qua unenlightened cat, differs only in degree from the sense of "I" which manifests when falsely accused. Now, I am not at all sure this is correct. Do I always have the same sense of "I" as I have when falsely accused, differing only in degree? Even when asleep, or in coma, or in ecstasy? I do not know. But supposing I did. What sense of "I" could that be? Well, the only sense of "I" which seems to me to be always present is the sense of subject, subjectivity. It is what Kant referred to when he spoke of a contentless "I think" accompanying all my experiences (Critique of Pure Reason B131-2). But if it is this sense of "I" which is the innate sense of self, and if it is this sense of "I" which is being criticized in "SĀntideva's argument at BCA 8:101-3, then Pettit is claiming that "SĀntideva's argument is directed against subjectivity as such. That I can well believe. I think "SĀntideva's argument will indeed only follow when directed against subjectivity as such. Unlike Pettit, however, I argue that "SĀntideva simply seeks to draw his conclusion from the regular Buddhist refutation of the Ātman, without making any distinction between an acquired conception of Self and the innate conception of self. And in directing his argument against subjectivity as such -- for Pettit, the innate sense of self -- "SĀntideva falls onto the other horn of my dilemma. For I argue at great length in my book that pains without subjects (subjectivity) are incoherent. I also indicate several moral problems which would follow for "SĀntideva from such a position.
Thus I can put to Pettit's position a version of the same dilemma that I used originally for that of "SĀntideva. If Pettit were to hold (which he does not) that "SĀntideva's argument is directed solely against the acquired, theoretical misconception of Self, then altruism would not follow. However if "SĀntideva's argument were to be taken as directed against the innate sense of self then we need to be clear what the innate sense of self is. If the innate sense of self is something more specific than subjectivity as such (say, something associated with my own self-importance, or even an innate apprehension of inherent existence -- whatever that might be; see here my comments in the book under review, pp. 245-6) -- then it does not seem at all obvious to me that it would be always present in all unenlightened mental states. I do not know if a cat or someone in a coma always has a sense of his own self-importance, or inherent existence. Nor does it seem obvious to me that there is only a difference of degree in such cases between something felt at times of false accusation and something always present. In other words, even if Pettit is right that "Madhyamaka analytical meditation should inculcate a profound certainty that the sense of self that is present in moments of emotional duress is a fabrication and does not exist in reality," I am not sure what this has to do with what would be an actual innate (that is, always present) sense of self, or with the specific argument used by "SĀntideva. But if the innate sense of self is indeed subjectivity as such -- the subjectivity that means being the subject of fears, desires, pains and so on -- then without subjectivity (for example, pain) becomes incoherent. A pain without a subject seems to me to be meaningless. Without subjectivity there could be no conventional self, no person. Moreover, I am by no means clear what would be an intensified sense of subjectivity as such at times of false accusation. If that is what Pettit takes to be the object of "SĀntideva's refutation then he has not succeeded in exonerating "SĀntideva.
But Pettit considers he has, and considers that his exoneration will allow "SĀntideva to continue with the conventional sense of person. Thus Pettit states that to "eliminate that selfishness would not have to mean, as Williams seems to believe, that one would no longer favor oneself in a naturally quotidian fashion." But here we seem to have lost the details of "SĀntideva's actual argument altogether. "SĀntideva's argument is intended to show that we cannot establish rationally a distinction between removing our own pain and refusing to remove the pains of others. Pettit now tells us that on his interpretation of "SĀntideva we can continue to favor ourselves (if not our Selves). If we are talking about "SĀntideva's actual argument, however, this could only mean that having removed the innate sense of self we can still distinguish this person from that person. But if that were right then removing the innate sense of self is still compatible with removing the pains of this person ("oneself") and refusing the remove the pains of the other. "SĀntideva's argument would not lead to the desired conclusion. It would not give us the rational reasons required.
Likewise in any state of meditative aftermath. It is not clear to me that a nondual meditative absorption on emptiness would entail no (innate) sense of self, if the innate sense of self were interpreted as subjectivity. And I also have great difficulties with the coherence of a nonconceptual experience. But let me put all that to one side (for more on some of these issues see Williams 1992b). Still, in the state of meditative aftermath one either does have the sense of oneself as different from others or one does not. If one does have that sense, then it is still the case that one can consistently (logically) favor removing the pains of oneself, and not of others (even if psychologically, religiously etc. one might not. Remember, "SĀntideva is supposed to be giving arguments for a conclusion). On the other hand if in the state of meditative aftermath one does not have the sense of oneself as different from others then among the many paradoxes that would follow it would make no sense to speak of removing the pains of others because one removes the pains of oneself, or indeed to give others pleasure, or enlightenment, because oneself has pleasure or enlightenment. Without subjectivity there can be no pain, pleasure or -- I would imagine -- enlightenment. And to repeat something I said earlier, to bring in personal practical experience will not make an incoherent argument coherent. Before meditation takes place the topic of that meditation must at least be coherent. If "SĀntideva's argument does not work then it is no use appealing to meditation in the hope that meditation will in some mysterious way bring about a valid argument.
It should be clear that particularly on the subject of the last chapter of my book Pettit and I have some radically different views and approaches. He considers that my arguments are "fundamentally misguided," but in the very intelligent and challenging way he has engaged with the arguments, and developed a defense of "SĀntideva, he shows I think that while he has not been convinced by my arguments nevertheless the arguments are worth taking seriously. We are frequently told nowadays that Buddhism is a religion which values critical investigation. In the dGe lugs tradition one is constantly exhorted to engage in minute examination as a means of bringing about insight. I hope and trust that when dGe lugs pa monks are debating with such vigor, often all night, it is these issues -- the coherence of arguments -- which are being debated. One can but hope and trust they are doing philosophy. Thus in spite of Pettit's reservations about my criticisms of "SĀntideva, is it not the case that such philosophical assessment is precisely what we are told we should be doing? Pettit may find my arguments unconvincing, but I am truly grateful to him for the debate.
References
Hookham, S.K. 1991. The Buddha Within: Tathagatagarbha According to the Shentong Interpretation of the Ratnagotravibhaga. Albany, N.Y.: State University of New York Press.
Khenchen Kunzang Palden and Minyak Kunzang Sönam. 1993. Wisdom: Two Buddhist Commentaries. Translated by the Padmakara Translation Group. Peyzac-le-Moustier: Editions Padmakara.
Mi pham. 1973. dBu ma'i lta khrid zab mo. Tibetan text contained in Lama Mipham: Calm and Clear. Translated from the Tibetan with commentary by Tarthang Tulku. Emeryville: Dharma Publishing.
Mi pham. 1975. Spyod 'jug shes rab kyi le'u'i tshig don go sla bar rnam par bshad pa Nor bu ke ta ka. Contained in the Collected Writings of 'Jam mgon 'ju Mi pham rgya mtsho, vol. 13. Gangtog: Ngagyur Nyingmay Sungrab Series.
Mi pham. 1976. bDe gshegs snying po'i stong thun chen mo Seng ge'i nga ro. Contained in the Collected Writings of 'Jam mgon 'ju Mi pham rgya mtsho, vol. 3. Gangtog: Ngagyur Nyingmay Sungrab Series.
Samuel, G. 1993. Civilized Shamans: Buddhism in Tibetan Societies. Washington and London: Smithsonian Institute Press.
"SĀntideva. 1960. BodhicaryĀvatĀra of "SĀntideva with the commentary Pa~njikĀ of Praj~nĀkaramati. Edited by P.L. Vaidya. Darbhanga: Mithila Institute.
Van der Kuijp, Leonard W.J. 1983. Contributions to the Development of Tibetan Buddhist Epistemology: from the eleventh to the thirteenth century. Wiesbaden: Franz Steiner Verlag.
Williams, P. 1983. "A note on some aspects of Mi bskyod rdo rje's critique of dGe lugs pa Madhyamaka." Journal of Indian Philosophy 11: 125-45.
Williams, P. 1992a. "On prak.rtinirvĀ.na/prak.rtinirv.rta in the BodhicaryĀvatĀra: A study in the Indo-Tibetan commentarial tradition." Asiatische Studien/Études Asiatiques XLVI:I: 516-550. Reprinted in P. Williams. 1998a.
Williams, P. 1992b. "Non-conceptuality, critical reasoning and religious experience. Some Tibetan Buddhist discussions." In Michael McGhee [ed.], Philosophy, Religion and the Spiritual Life. Philosophy Supplement 32. Cambridge: Cambridge University Press.
Williams, P. 1998a. Altruism and Reality: Studies in the Philosophy of the BodhicaryĀvatĀra. Richmond: Curzon.
Williams, P. 1998b. The Reflexive Nature of Awareness: A Tibetan Madhyamaka defence. Richmond: Curzon.
