Philosophy of Mind’s “Hard Problem” in Light of Buddhist Idealism

Dan Arnold

University of Chicago

For Steven Emmanuel, ed., Philosophy’s Perennial Questions: Comparing Buddhist and Western Approaches

I. Introduction: Buddhist idealism and “the hard problem”

Arguing that the Buddha’s ultimately true teachings are best understood in terms of philosophical idealism, the Indian Buddhist philosopher Vasubandhu (fl. 360 CE) could exploit central Buddhist commitments of long standing; for he recognized that some of the Buddhist tradition’s main commitments may already entail idealist conclusions. This makes it unsurprising that perhaps the most influential philosophers in the history of Indian Buddhism – Dignāga (c. 480-540 CE) and his successor Dharmakīrti (c. 600-660 CE), who are read to this day by traditionally trained Tibetan Buddhists – advanced cases for kinds of idealism. That many if not most late Indian Buddhist philosophers were

idealists makes it especially interesting that so many contemporary philosophers and scientists should think of Buddhists as philosophical allies. These contemporary exponents rightly recognize Buddhist thinkers as typically arguing (like themselves) for reductionist accounts of the person. However, many if not most of the contemporary thinkers who claim Buddhists as allies are physicalists (i.e., they think everything about the mind must be explicable in terms of brain events and the like, since only physical things are ultimately real); but the Buddhist Dharmakīrti, an idealist, argued lengthily that mental states cannot be reducible to physical events. This is, in fact, the thrust of Dharmakīrti’s celebrated proof of rebirth, which many in the Buddhist tradition take to be the definitive such argument. On some perennially debated issues in philosophy of mind, then, Buddhist philosophers thus stand at odd angles vis-à-vis contemporary debates. While many contemporary thinkers suppose that some sort of physicalism virtually follows from reductionism, among classical Indian Buddhists we find reductionists who yet argued for idealism. There is an opportunity, then, to uncover some unappreciated features of the various positions if we put them in play together. In that spirit, I will enlist some arguments from Dignāga and Dharmakīrti as a way to differently characterize one of the most vexed

issues in philosophy of mind – the one, indeed, that philosopher David Chalmers aptly and influentially christened the hard problem. Engaging these Buddhist arguments can help us see that the hard problem is not, after all, a problem just for physicalists; the hard problem may look different if we appreciate that something like the same problem arises in some of the Buddhist tradition’s arguments for idealism. As typically presented, “the hard problem” is that of why and how certain physical events – paradigmatically, those of a body’s sensory and neurophysiological systems – should happen to be accompanied by subjectivity. Why are certain neurophysiological events (objective events involving an environment’s impinging upon the sense faculties of an organism) also conscious events? How can it be that for the subjects of such events, there is “something it is like” for them to occur, some way these events are for the subject? As David Chalmers puts it, “The hard problem of consciousness is the problem of experience. Human beings have subjective experience: there is something it is like to be them. We can say that a being is conscious in this sense – or is phenomenally conscious, as it is sometimes put – when there is something it is like to be that being. A mental state is conscious when there is something it is like to be in that state.”1 The problem is that no matter how great the advances in scientific understanding of such neurophysiological goings-on as cause conscious mental events, no wholly third-personal understanding of these can capture what is for their subjects their most salient feature: that occurrences of consciousness are about something (they have content), that there is some way it “feels” to encounter that. Insofar as this feature – call it subjectivity2 – would go missing on any 1 David Chalmers, “Consciousness and its Place in Nature,” in S. Stich and F. Warfield, eds, Blackwell Guide to the Philosophy of Mind (Blackwell, 2003), p.103. A locus classicus for the “something it is like” criterion is Thomas Nagel, “What is it Like to Be a Bat?” Philosophical Review 83/4 (1974), pp.435-50. See, as well, Joseph Levine, “Materialism and Qualia: The Explanatory Gap,” Pacific Philosophical Quarterly 64 (1983), pp.354-61. 2 I beg the reader’s indulgence for my inconsistent alternation among “consciousness,” “experience,” “subjectivity,” and the like. All such terms are here intended in a minimal sense as commonly involving qualia (“felt” phenomenal qualities) and intentionality (the fact that they are about something).

wholly impersonal analysis of experience, it would seem it is irreducible to any interactions among objects; but how, in that case, could subjectivity be part of the natural world? How could any physical occurrences produce or entail there being (from the inside, as it were) “something it is like” for them to occur? A hard problem, indeed. Contemporary discussion of this problem may, however, mislead in suggesting the real difficulty centers on what kind of stuff existents are made of – that the hard problem just is that of how physical things or events could ever constitute or produce subjectivity. Chalmers, for example, characterizes the hard problem in the course of considering “consciousness and its place in nature,” where the problem is this: “On the most common conception of nature, the natural world is the physical world. But on the most common conception of consciousness, it is not easy to see how it could be part of the physical world.”3 Once we have engaged some arguments from Dignāga and Dharmakīrti, though, we may see that there is in the vicinity of Chalmer’s “hard problem” a more conceptually basic problem, one quite independent of whether or how consciousness can emerge from physical existents.

II. Buddhist dualism and Dharmakīrti’s proof of rebirth

Indian Buddhist philosophers were emphatically not physicalists; they tended, rather, to be committed dualists of a sort, and if there was pressure towards any kind of monism in their thought, it was clearly in the direction of idealism. We can, then, appreciate that the issue of physicalism may not be central to the “hard problem” if we consider how the same problem arises in one of the Buddhist tradition’s arguments for idealism. That a variation on the hard problem should bedevil Buddhist philosophers despite their strenuous rejection of physicalism surely tells us something interesting about what’s going on. 3 Chalmers, op. cit., p.102.

Now, some will be surprised to hear it said that any Buddhist philosophers were dualists; surely some will have heard it said that Buddhist thought (perhaps even “Eastern philosophy” more generally) is “non-dualist.” (This preconception informs the classic joke about the Zen Buddhist’s request of a hot dog vendor: “Make me one with everything.”) What makes it particularly tempting to accept that Buddhists were non-dualists is that Buddhists commonly affirmed, as a cardinal doctrine of the tradition, that persons are “without self” (anātma) – that we are not (or do not have) enduring and unitary “selves.” Perhaps especially because dualism is for modern readers most familiarly epitomized by Descartes – whose famous “I think, therefore I am” argued that I must denote a really existent self, a thinking thing (res cogitans) – it can seem that nothing could be more “nondualist” than to affirm, with Indian Buddhists, that there are no real “selves” or “souls,” and that persons ultimately consist just in causally continuous series of momentary events. Isn’t the denial that there is any self over and above one’s fleeting experiences precisely a denial of dualism? Many Buddhist philosophers do indeed exemplify a reductionist approach to persons, “reducing” apparently enduring wholes to the momentary events that alone really exist; for they hold that there is no enduring identity of “persons,” that the causal continuity of momentary events of experiencing sufficiently explains the illusion that there is. It is, perhaps, especially their reductionism that makes Buddhist philosophers look to some modern enthusiasts like veritable “mind scientists”; are not Buddhists much like contemporary cognitive scientists in showing that experience consists entirely in causal interactions among innumerable sub-personal events?4 While it’s perhaps tempting to think so, it matters that many Buddhists in fact affirmed a kind of dualism. It’s important to The literature on Buddhist thought vis-à-vis the cognitive sciences is vast; one can begin to see this by perusing the website of the Mind and Life Institute (https://www.mindandlife.org/). For an accessible and appropriately critical account more generally of the “Buddhism and science” meme, see Donald Lopez, The Scientific Buddha: His Short and Happy Life (Yale University Press, 2012).

appreciate that a Buddhist’s event-based ontology – a view, that is, on which the ultimately real existents are not “substances” but events – does not, after all, rule out there being essentially different kinds of events. It’s possible to affirm, as most Buddhists do, that no enduring substances really exist (that every candidate for the status of “substance” turns out, like the “self,” to be reducible to innumerable momentary events) – and at the same time to urge, without contradiction, that mental events are essentially different in kind from physical ones. Indian Buddhists had a strong stake in arguing just that; for it is because mental events are independent of physical ones that there can occur, even in the absence of enduring selves, something worth the name “rebirth.” Of course, the English word (with its prefix re-) suggests there must be some one thing that is repeatedly undergoing the same experience; Buddhists can, however, coherently affirm simply that a causally continuous series of mental events continues even after the manifest demise of the body. Indeed, if mental events do not essentially depend on a body, series of them can continue indefinitely. This is just as Dharmakīrti’s celebrated proof of rebirth aims to show. Dharmakīrti’s proof consists in a series of arguments showing that mental states cannot coherently be thought to depend on physical things, since any view on which they do entails numerous absurd consequences.5 Together with some other widely shared Buddhist views – such as that every mental event must have among its causes a previous mental event – these arguments are taken to show that a mental series cannot begin at birth; rather, what we take 5 For example, if everything there is to being a person were reducible to facts about the body, there would be no way to distinguish a living body from a dead one; after all, the latter still consists in precisely the same physical stuff as the former. Further, there is the fact that mental activity can continue even when some or all of the sense faculties are damaged; the sense faculties are useless, however, if no mental activity occurs, which shows that the latter is clearly basic. For more on Dharmakīrti’s proof of rebirth, see Richard Hayes, “Dharmakīrti on Rebirth” (in Egaku Mayeda, ed., Studies in Original Buddhism and Mahāyāna Buddhism, vol.1, pp.111–129; Kyoto: Nagata Bunshodo); and Roger Jackson, Is Enlightenment Possible? Dharmakīrti and rGyal tshab rje on Knowledge, Rebirth, No-Self, and Liberation (Ithaca, N.Y.: Snow Lion, 1993).

as the first mental activity occurring in “this” life must really be the effect of a previous such moment from another life. This is evident, Dharmakīrti’s commentators note, in the purposeful behavior exhibited by newborn infants; that an infant “knows” to suckle is due precisely to this mental continuity. Of course, this picture raises (among other things) all the problems that familiarly bedevil dualists. If, for example, mental events are really independent of bodily events, why are they only ever experienced as embodied? How could a mental series that is independent of a body work in such precise coordination with one? At the end of the day, though, Dharmakīrti may not much care about reconciling mental and physical; the chief point of his proof of rebirth is really to show that mental continuity is ineliminably basic – not the physical but the mental is what ultimately exists. Insofar as he is finally an idealist, it bothers him little if his arguments for this make it hard to understand how physical stuff fits in – indeed, it’s so much the better for his final view if that’s the case. In advancing, then, what the Buddhist tradition considered the best argument for the reality of rebirth, Dharmakīrti furthered a larger case for some kind of idealism – a family of views, I have said, towards which there was strong pressure in the Buddhist tradition. The tradition’s tendency toward idealism is a function not only of the commitment to the reality of rebirth, but indeed of the tradition’s basic diagnosis of the human situation; for integral to the possibility of Buddhist transformation is the extent to which minds create the kind of world we experience. (Through appropriate cultivation of the mind, then, one can change what kind of world is experienced; this is what it is to become a Buddha.) This idea is reflected in a traditional view of long standing – the view that it is the “actions” (karma) of sentient beings that produce the world, and that karma ultimately consists in thought. Taking this idea to its perhaps inevitable conclusion, the philosopher Vasubandhu argued that idealism is positively entailed by these cardinal Buddhist commitments; his arguments are different enough from those of Dignāga and Dharmakīrti that the latter may come more sharply into focus if we briefly consider Vasubandhu’s first. 7

IV. Interlude: Vasubandhu’s metaphysical argument for idealism

Vasubandhu’s influential corpus of works attests the working-out, over the course of his whole career, of conclusions that he takes to be entailed by basic Buddhist commitments – chief among them, the view that it is the karma of sentient beings that creates (our experience of) the world.6 In his Treasury of Abhidharma (Abhidharmakośa) – a massive text representing his engagement with the Abhidharma traditions of philosophy that prevailed in his day – Vasubandhu expresses this idea in the course of arguing against theism, which he says is (among other things) unnecessary since there is a ready explanation for everything that theism would explain: “This whole manifold world is produced by the karma of sentient beings.” Significantly, he immediately adds that karma essentially consists in thought (cetanā); for it is thought that finally impels action, and so it is thought that must be changed if we are to desist from producing worlds of suffering.7 If, however, karma creates worlds, and if karma is finally something mental, mightn’t it make most sense that the world itself is somehow mental? This would be a more 6 Note the shift that makes a move toward idealism seem natural: The intuitively plausible soteriological intuition here is that how we think shapes what experience is like; when that intuition is made the basis of a complete worldview, it can become, as well, an explanation of (what is a logically distinct matter) what there is to be experienced. On Vasubandhu’s project, see Jonathan Gold, Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy (Columbia University Press, 2014). 7 These passages from Vasubandhu center on Abhidharmakośa 4.1; for a translation, see Gelong Lodro Sangpo (based on the French translation by Louis de La Vallée Poussin), Abhidharmakośa-Bhāṣya of Vasubandhu (4 volumes; Delhi: Motilal Banarsidass, 2012). The equation of karma with “thought” (cetanā) should not be taken to imply that our actions are impelled particularly by deliberative thought; for “thoughts,” on this account, are themselves effects of habituated patterns, and include all manner of unconscious dispositions. For more on cetanā, see Karin Meyers, “Freedom and Self-Control: Free Will in South Asian Buddhism,” Ph.D. dissertation, University of Chicago (2010). See, too, Maria Heim, The Forerunner of All Things: Buddhaghosa on Mind, Intention, and Agency (Oxford, 2013). 8 ontologically parsimonious view; otherwise, one has the difficult task of explaining how thought could produce (say) essentially physical existents – eggplants and tables and rivers and whatnot. Vasubandhu exploits this point in one of his late works, a short treatise of 22 verses (the number is rounded so the text can be entitled Twenty Verses), which together argue that the best way to understand the definitive teaching of the Buddha is in terms of idealism – that, as he baldly puts it at the outset of the text, “All this is nothing but mental representations.”8 A pivotal move in the text comes early on, in a discussion of the hells so vividly imagined by Buddhists. Having argued that the beings who populate hells share experiences of suffering just on account of their shared karma (and not on account of their sharing any physical space), Vasubandhu entertains the possibility that perhaps their karma produces really existent beings (real guardians of hells) who inflict their torment. Why, that is, couldn’t the torturers in hell, for all that they are karmically produced, nevertheless be physically real beings? The problem with this picture, Vasubandhu argues, is that it is much more complicated than the view that naturally follows from embracing the mental character of karma. As he says in verse 6, “If you admit that it is material elements that arise in hells due to the karma of the beings there, why not allow that that is a transformation only of consciousness?” Why, that is, additionally posit material elements? The problem with doing so, he then says in verse 7, is that it saddles karma theory with too much to explain: “You imagine that the traces of karma are in one place, their result in another; why not just accept that the results occur in the very same place where the traces are?” Why not, that is, accept the more elegant view that the results of our actions occur (that what we experience 8 The text of Vasubandhu that I here have in mind is the Viṃśatikā; translations from this text are my own (and are fairly liberal), and are done from the 1925 edition of Sylvain Lévi (in Vijñaptimātratāsiddhi: Deux traités de Vasubandhu). For another widely available translation, see Stefan Anacker, Seven Works of Vasubandhu (Motilal Banarsidass, several editions), pp.161-75. See also, and especially, Birgit Kellner and John Taber, “Studies in Yogācāra-Vijñānavāda Idealism I: The Interpretation of Vasubandhu’s Viṃśikā,” in Asia 68:3 (): 709-756.

occurs) “in” the same “place” where the causes occur? On the view that the causes in question are “traces” of past actions (i.e., karmically habituated mental dispositions to act), why not accept that the effects, too, occur just “in” the same series of mental events in which these traces first emerged? Now, this is not itself Vasubandhu’s argument for idealism; he is at this point in his text just disarming obvious objections to his thesis, thereby showing that his counterintuitive claim is not, after all, implausible. The real argument for the claim only comes in verses 11-15, which advance an essentially metaphysical argument. By this I mean the argument does not involve a posteriori analysis of conscious experience; rather, it involves a priori analysis of the adequacy of our concepts – in particular, the concept of atoms, which are (Vasubandhu argues) necessarily presupposed by anyone who would affirm the reality of physical entities. His argument against them is that no coherent account of atomism can be given. This is because atoms that had spatial extension – atoms that, able to join with other atoms on all sides, can combine to form macro-objects – would, ipso facto, be reducible, since they would have (e.g.) left and right sides.9 (This can be expressed as the point that anything with spatial extension would be infinitely divisible, hence “reducible.”) On the other hand, nothing without spatial extension could constitute things having spatial extension; if really extensionless atoms combined, the combinations would be no larger than the atoms themselves, and could not add up to the macro-objects of ordinary experience. Vasubandhu thus argues that realism about external objects is fundamentally incoherent: the realist both must and cannot give a coherent account of atoms. This is, I said, a metaphysical argument. As such, it distinctively consists in its showing not just (as Dignāga and Dharmakīrti will) that only mental items are directly encountered in 9 It’s a fair question whether this sense of being “reducible” is the same as the sense in which persons, for example, are at least plausibly reducible to more basic parts. On the argument particularly of this part of Vasubandhu’s text, see Matthew Kapstein, Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought (Boston: Wisdom Publications, 2001), pp.181-204.

experience; rather, it aims to demonstrate the stronger conclusion that non-mental existents cannot coherently be conceived. The distinctive force Vasubandhu took the argument to have is evident in the question to which the argument gives an answer. As noted, Vasubandhu’s text begins by scouting obvious objections to idealism, showing them not, after all, to be disabling. Following the above-noted discussion of hell beings comes a line of basically hermeneutic interrogation: how, given Vasubandhu’s thesis, can we make sense of the countless places in Buddhist scripture where the Buddha is represented as referring to various things of the material world? Vasubandhu answers that such texts must, like all scriptures, be interpreted, and that his interpretation is preferable. According to that interpretation, the Buddha invariably said such things with a specific intention – that of introducing the doctrine to people not yet prepared for its full force. It is just this hermeneutical claim that occasions Vasubandhu’s argument for idealism; for as he anticipates, one might now ask: “How, though, can it be known that the Buddha taught the existence of things like form with just this intention, but that those things which individually become the objects of mental events representing things like form do not really exist?”10 Even granting, that is, that scriptures of course require interpretation, how can we be sure they should be interpreted in just such a way as to come out consistent with idealism? Why should Vasubandhu’s claim be thought to provide the interpretive key to all of the Buddha’s teachings? The distinctive force of the argument from atomism is clear if we appreciate why that argument counts as answering this question; the answer, we are clearly to understand, is that idealism must be what the Buddha ultimately taught, since that is the only coherent view to hold. It is clear, then, that Vasubandhu’s is a peculiarly strong argument to the effect that the reality of a physical world cannot be coherently conceived; and, writing as a Buddhist philosopher, Vasubandhu takes it as axiomatic that of course all 10 This is the sentence of Vasubandhu’s commentary that immediately introduces verse 11.

the views attributed to the Buddha must be coherent, and that the Buddha must therefore have intended that his teachings be interpreted in terms of idealism.

V. Dharmakīrti’s epistemic arguments for idealism

The distinctiveness of Vasubandhu’s argument was recognized by one of Dharmakīrti’s commentators (a certain Manorathanandin), who alludes to Vasubandhu’s argument in the course of clarifying what Dharmakīrti can and cannot reasonably claim to have shown. Dharmakīrti (we will see) aims to show that every act of cognition turns out to be immediately acquainted only with other mental items of some sort (mental representations or sense data, e.g.); it is only an inference that an external world is represented by these. To show, however, that we are not immediately aware of an external world is not the same as showing that no such thing exists. Clarifying how Dharmakīrti’s arguments work, the commentator Manorathanandin concedes as much; Dharmakīrti’s arguments, he says, can show only that what is present to us is just “appearings” to consciousness, and that these appearings are not themselves external – regardless of what really exists, things show up for us only “in” mental states. Manorathanandin allows, though, that if you want to argue for the strong claim that there cannot be physical entities – if, as he puts it, your desire to refute realism is “heavier” than can be borne by Dharmakīrti’s arguments – in that case “the arguments of the teacher should be considered.” Which teacher? The one, Manorathanandin clarifies, whose arguments concern “the refutation of atoms (by considering whether or not they have parts).”11 Dharmakīrti’s commentator thus distinguishes Vasubandhu’s a priori analysis of the concept of atoms from the conceptually different kind of case made by Dharmakīrti, who 11 I have discussed this passage from Manorathanandin, and also more generally Dharmakīrti’s epistemic arguments for idealism, in “Buddhist Idealism, Epistemic and Otherwise,” Sophia 27 (2008), pp.3-28.

followed his predecessor Dignāga in advancing essentially epistemic arguments for idealism – arguments, that is, based on analysis of what is (and what is not) immediately present to awareness. For purposes of this sort of argument, one can set aside the question of whether what is present to awareness represents anything outside that; for purposes of showing just that we are immediately aware only of mental items, nothing need be decided about the status of the world beyond. The approach of Dignāga and Dharmakīrti thus represents a route to idealism that follows from particularly privileging perception as the most exemplary way of knowing – which is to say this route to idealism runs through empiricism. This is an appealing route for philosophers like Dignāga and Dharmakīrti, who greatly advanced the cause of cross-traditional debate in Indian philosophy by first venturing epistemological arguments for commitments not unlike Vasubandhu’s.12 Before Dignāga and Dharmakīrti, Buddhist philosophical thought had been advanced predominantly in the discourse stemming from Abhidharma literature; this was largely intramural in character, always finally driven by such hermeneutical considerations as we saw in Vasubandhu’s celebrated argument for idealism. Arguing, instead, in ways meant to be at least in principle persuasive across party lines (which means, among other things, not framing one’s arguments as interpreting scriptures whose authoritativeness others deny), Dignāga and Dharmakīrti exploited the fact that an epistemic case for idealism starts from ideas that are much more intuitively plausible (and therefore much more likely to be admitted by their interlocutors) than Vasubandhu’s stronger claim that only mental events really exist. Indeed, their epistemic argument for idealism gets going from the seemingly innocuous thought that perception, uniquely among the ways in which we know things, will admit of a wholly causal description. 12 For further references on Dignāga and Dharmakīrti, see my “Philosophical Works and Influence of Dignāga and Dharmakīrti,” Oxford Bibliographies Online (<http://oxfordbibliographiesonline.com/obo/page/buddhism>).

Consider, then, that perceiving consists, in John McDowell’s felicitous phrase, in “impingements by the world on a possessor of sensory capacities”13 – in the world’s causing us (as when its rays of light strike our retinas) to have certain experiences. This is what Kant had in mind in characterizing our perceptual capacities as a faculty of receptivity; for according to an intuitively plausible causal picture of experience, perceiving is distinguished from other ways of knowing chiefly by its being, uniquely, caused by its object. That we are, according to this picture, acted on by something in the world is precisely the appeal of the empiricist thought that perception is the paradigmatic way of knowing; for what could be more objectively true than those cognitive moments in which we come up against the world as we do in perceiving? Perceptual cognitions alone, as Dharmakīrti emphasizes, are actually caused by what they are about. Now, though, the proponent of idealism can exploit the fact that once this picture seems natural, it can come to seem clearly right that the world, as J. J. Valberg puts it, is “possibly irrelevant”14; for it seems right that what is caused by our perceptual encounters with the world is just mental representations (e.g., the empiricist’s “sense data”) – that what is really present to the mind’s eye is just “pictures” inside the head. Once it is granted, though, that we are immediately acquainted only with mental representations of whatever sort, the status of the world outside us becomes essentially doubtful; our grasp on the world becomes only as sure as the inference that it must be a world of physical objects that causes the mental representations we experience – and Dharmakīrti’s wager is that there are other, preferable inferences to be drawn. Such is the slippery slope towards which Dharmakīrti aims to push his interlocutor when, in elaborating just such a causal account of perception, he introduces the idea of 13 John McDowell, Mind and World (Harvard, 1996), p.xv. 14 J. J. Valberg, The Puzzle of Experience (Clarendon Press, 1992), p.11. Valberg’s discussion of the philosophical havoc wrought by the intuitively plausible “causal picture of experience” is uncommonly good.

mental representations. Here, then, is what Dharmakīrti says in characterizing perceptible objects as against everything we can know only through inference: There are only two kinds of things, perceptible and imperceptible. Among these, the perceptible is what causes the content of cognition to track its own presence and absence. That – unique, having the nature of a thing – is a unique particular. But the imperceptible, lacking the capacity for projecting its nature directly into thought, is something with which immediate acquaintance is impossible.15 A real existent’s capacity thus to cause the content of cognition – evident in the cognition’s varying depending on whether or not that existent is present – is here contrasted with everything that lacks the capacity thus to “project” its own form “directly into thought.” A really existent object – a particular ruminant, say, located at an appropriate distance from an observer – can actually impinge upon our sensory capacities, producing mental representations whose content is uniquely vivid; abstract items like the concept ‘cow’ do not, in contrast, directly “produce” representations in anything at all like the same way. (An empiricist will say that is just why perceptual awareness is foundational.) To similar effect, Dharmakīrti says elsewhere that “there is nothing at all worth the name ‘being apprehendable’ apart from being a cause.” That is, to be the kind of thing that can be perceptually “apprehended” just is to be the kind of thing that can cause mental representations. Dharmakīrti recognized, however, that this raises a problem; he anticipates someone objecting that the sense faculties themselves are surely among the causes of awareness – why then are the sense faculties not also “apprehended” in experience? Indeed, many of the things that cause any moment of experience (firings of the optic nerve, for example) are not themselves what the experience is of or about; only some of the causes – the scene outside my window, for example, reflected light from which is striking my eyes – correspond to the content of the awareness. How, then, can we distinguish among the 15 For this and the following passages from Dharmakīrti, see my Brains, Buddhas, and Believing (Columbia University Press, 2012), pp.25-26. 15 different kinds of things that can be causes of cognition, or specify exactly how that one among the causes which is also what the cognition is of differs from all the other causes? Is there any principled reason why some causes must remain phenomenologically inaccessible to a subject (why we will never see our own neural activity), while others are precisely what we are experiencing? Simple, Dharmakīrti answers; it’s just that cause whose image the awareness bears that counts as the one we’re interested in: “Among the various causes of awareness, the one in whose image a thought arises is the one that is said to be ‘apprehended’ by the thought.” Having seen, though, that we may be talking about essentially different kinds of “causes” – the scene outside my window does not, it seems, cause my experience in just the same way that neurophysiological events in my body do – we may wonder whether Dharmakīrti’s answer is sufficient. We can press that question by turning to Dignāga, who closely considers the two conditions that, we have now seen, must be met by the objects of perceptual awareness: they must make sense both as causing perception, and as what is showing up therein.

VI. Dignāga, and the real root of the hard problem

It turns out that Dharmakīrti’s apparent confidence that we can pick out just that cause “in whose image thought arises” may be premature; it is actually hard to say how or whether anything can really be identified at the same time as the content of cognition and as a cause thereof. Among the problems, Dharmakīrti recognized, is that reference to the causes of cognition brings into play all manner of subpersonal events (various neurological goings-on, in a contemporary idiom) – and while there is something intuitively plausible about Dharmakīrti’s response to this problem (duh! we’re interested in the cause whose image is present to consciousness!), there is a more conceptually basic problem here that still eludes us. The thing is, it’s hard to see that we are even talking about the same kinds of

things when we include the phenomenal content of awareness along with sub-personal sensory events as alike being causes of awareness. Do the kinds of things we’re typically aware of (furniture, vegetables, light bulbs, dogs) really “cause” cognitions in anything like the same way that things like neurotransmitters do? Perhaps it will be allowed that essentially different kinds of causation are occurring – one kind involves interactions among really existent basic entities (among synaptic discharges, on a contemporary reckoning, or among momentary mental states, for Dharmakīrti); and one kind involves such temporally enduring macro-objects as typically figure in experience. Could both these kinds of causation be real? Or must the former finally explain the latter? And can either kind of causation explain the other if there is no way so much as to correlate these different levels of description? These are among the questions engaged in a concise text by Dharmakīrti’s predecessor Dignāga, and with this we arrive at the argument I’ve said may shed some light on “the hard problem.” The text, which consists in just eight verses along with Dignāga’s own commentary thereon, is called “Critical Analysis of the Ālambana” (Ālambanaparīkṣā), where the Buddhist technical term in the title denotes precisely the concept at the intersection of the foregoing questions. Thus, of the many sophisticated taxonomies of mental events and factors posited by Buddhist philosophers, one of the most basic enumerates kinds of “causal conditions” (pratyaya) that figure in the occurrence of moments of consciousness. Authors of the Abhidharma literature generally agreed there are four such conditions; a moment of (say) seeing some autumn trees must have as its causes: (1), a properly functioning ocular sense faculty (this is the adhipati-pratyaya, or “predominant condition”); (2), a previous moment of ocular experience (the samanantara-pratyaya, or “immediately preceding condition” – a category that explains how a series of fleeting moments of experience can seem, phenomenologically, to be a continuous flow; the reason we don’t experience each new moment of seeing as having just popped into being is that the moments occur in continuous series); (3), a collection of other causes (the hetupratyaya, or

“causal conditions which are causes,” where hetu refers to another list of causes)16; and, (4), the autumn trees themselves, insofar as they are among the causes of my seeing them. The latter is the ālambana-pratyaya, which we might translate as the percept.17 Dignāga’s concise text analyzes this last category, which was defined by Buddhist philosophers as meeting just the two conditions on being an object of perceptual awareness; thus, the “percept” denotes that one among the causes of a cognition which is at the same time what the cognition is of. Characterizing the difficulty in thinking that anything that could cause a cognition could also be the content thereof, Dignāga’s analysis provides the occasion for an argument for idealism; for he particularly aims to show that physical objects could not satisfy both these conditions, and so could never be present to awareness. As we will see, though, it is not entirely clear that Dignāga’s proposed alternative – that perceptual content is produced only by previous moments of awareness – fares any better. Dignāga’s idealist alternative, I suggest, remains stuck with something like the hard problem – a problem, we can appreciate by considering its arising for an idealist, that has less to do with what kind of stuff experience is made of than with the more basic problem of reconciling first-person and third-person perspectives on mental content. Dignāga argues, then, that the kinds of things we typically take cognitions to be of cannot, in principle, be among the causes of cognitions. That “in principle” may make this argument more akin to Vasubandhu’s metaphysical argument than meets the eye; Dignāga, 16 The relatively indeterminate character of the category hetu-pratyaya made that available when Yogācāra philosophers introduced the “storehouse consciousness” (see note 18); thus, in Yogācāra appropriations of Abhidharma (e.g., Asaṅga’s Abhidharmasamuccaya), it was claimed that the hetu-pratyaya refers precisely to this unconscious repository of “seeds.” 17 Percept is the translation favored in a forthcoming volume on Dignāga’s text (and many commentaries thereon) that I have seen in manuscript: Douglas Duckworth, et al, eds., Dignāga’s Investigation of the Percept (Ālambanaparīkṣā) and its Philosophical Legacy in India and Tibet. I have also benefited from the translation given in Fernando Tola and Carmen Dragonetti, “Dignāga’s Ālambanaparīkṣāvṛtti,” Journal of Indian Philosophy 10 (1982): 105-34. Translations from this text (again fairly liberal) are my own, from the edition given by Tola and Dragonetti.

too, argues from analysis of a concept, and the problem he identifies is closely related to the problem demonstrated by Vasubandhu’s analysis of atomism. The argument is simple, and presupposes these premises: To be real, for most Buddhists as for most empiricists, just is to be capable of causally interacting with other existents; anything without this capacity doesn’t ultimately exist. (Consider the thought that since you can stub your toe on a brick but not on the concept ‘brick’, the former really exists in a way the latter doesn’t.) But on a Buddhist account, only irreducible things could count as ultimately real – and if that’s right, then for someone who affirms the reality of physical objects, it can really be only spatiotemporal atoms that are doing any “causing.” Dignāga’s opponent would thus have to allow, we can say, that that chair isn’t causing my cognition of a chair; it’s only the chair under a completely different description that is really “causing” anything – as described, for example, in terms of the structural properties of its parts, or their material composition, or whatever (their molecular or even sub-atomic make-up? it’s not clear how we would know just how fine-grained to make our identification of a chair’s ultimately existent parts). All that could be thought to cause any cognition, then, is atomic sensible particulars (fleeting occurrences of shape and color and solidity and whatnot), since only these are ultimately real. The problem, though, is that it is not irreducibly fleeting sensory atoms (not momentary occurrences of shape and color, etc.) that are present to us in experience; rather, our experience typically concerns what J. L. Austin memorably referred to as “mediumsized dry goods” – school buses and carpets and dogs and all the other stuff of everyday experience. In terms, then, of the Buddhist category that Dignāga is analyzing, the problem is that the kinds of things that can meet the causal condition (atoms, momentary sensa) do not meet the content condition; for experience is manifestly not of such things. Conversely, the kinds of things that meet the content condition (the temporally enduring wholes, the recognizable types, that show up in experience) do not, on a Buddhist view, ultimately exist – which is just to say they cannot cause anything.

Let us here recall Dharmakīrti’s confidence that we can specify that one among the causes of any thought “in whose image the thought arises.” Dignāga, we can now see, gives us reason to think there is a basic conceptual problem with this: The things present to us in consciousness cannot be the same kinds of things that figure in the causation of consciousness; for mental content (we might most generally say) typically involves wholes (relations, continuities, types), while causation is essentially a matter of relations among irreducibly basic existents. Dignāga concludes that it therefore cannot be anything basically “atomic” – cannot, in particular, be really existent external objects – that is present to awareness. Does that leave any way to salvage the idea that perceptual content can be characterized in terms of its causes? Dignāga proposes that an idealist account can circumvent the problems here scouted. Thus, he takes the foregoing considerations to recommend concluding that only something intrinsic to cognition, only something that is itself “mental,” could be at once a cause and the content of any moment of consciousness. Dignāga laconically says as much in verse 6 of his brief text: “What appears as though it is external is something whose nature is to be known internally; only this satisfies the criteria of the percept, since this has the nature of cognition itself, and since it is also a causal condition thereof.” This needs some unpacking. Dignāga gives two reasons why only something intrinsic to cognition – something itself of the type “mental” – could meet both the relevant conditions. The first reason (“since it has the nature of cognition itself”) is that as something that is itself mental, Dignāga’s “something whose nature is to be known internally” – he has in mind such things as “mental representations” or “sense data” (ākāra), but also the mental artifacts of one’s own deep psychological past (e.g., the “traces” of karma, the “latent dispositions” to which we are habituated by past actions) – makes sense as developing “in” one’s mental continuum. Among the ideas here is that the content of cognition has to make sense as something that is somehow “in” cognition; there is no problem in explaining how mental

content is related to mental events if the former turns out to be ontologically of the same kind as the latter. The second reason why only “something whose nature is to be known internally” is given by Dignāga’s laconic “since it is also a causal condition thereof” – and here things get tricky. In order make good on his claim that “something whose nature is to be known internally” makes sense as causing perceptual content, Dignāga ends up having to grant a couple of different senses of “being a cause.” The first way of “being a cause” is the one that makes sense if we adopt an objective, third-personal perspective on the occurrence of mental events. From this perspective – which need make no reference to what the subject of a cognition herself takes its content to be – Dignāga’s “something whose nature is to be known internally” is temporally prior to the mental event it causes. (Indeed, temporal priority would seem to be a sine qua non for being a cause; effects never come before their causes.) In particular, what causes the mental content we experience is things like mental “seeds” (bīja) or “latent disposition” (vāsanā) – beginninglessly habituated capacities and tendencies to experience the world in certain ways. (Consider, as comparable, the kinds of psychological causes that moderns might say figure in “the unconscious” – the kinds unearthed in psychoanalysis.) I say that causes like these are identified from a third-person perspective because even though there is indeed a sense in which the nature of these is “to be known internally” – specifically, they are carried forward “within” a mental continuum, passed down until they “ripen” within the same series – they are nevertheless phenomenologically inaccessible. That is, the envisioned process of long-term mental development is not transparent to the subjects thereof, and the kinds of things Dignāga has in mind – long-transmitted dispositions to act this way or that – are not available to introspection. From the subject’s perspective on any moment of experience, it is not seen that his or her experience represents the fruition of unconscious mental processes.

Of course, this is to say that moments of experience are typically not of things like “habituated dispositions” or other past mental events; but wouldn’t that mean that Dignāga’s appeal to these various psychological artifacts cannot, after all, satisfy the content condition?  Dignāga will answer that the artifacts of past mental events that he has in mind are not, so long as their capacities are dormant, part of experiential content; they become contentful only when they “ripen,” at which point they show up as part of this or that experience.  The Buddhist image of seeds is supposed to make sense of this.  Just as a seed’s capacity remains dormant until it is no longer a seed but a sprout, so, too, subconscious processes can continuously transmit latent dispositions until some moment when they somehow burst into consciousness.18  However, since it is thus our own psychological pasts that cause our mental states, it must be that our mental states mislead us in seeming instead to be about a present world; everything about how they seem is really a function only of one’s psychological past.  Suppose we grant that it makes sense for there to occur moments of experiencing that just intrinsically seem to their subjects to be about something, even though they are really produced only by the series that precedes them.  Even if we agree, however, that a real world then becomes (as Valberg says) “possibly irrelevant” to any experience’s seeming as it does, there is still a problem; for now the mental item that figures as the content of experience – a latent disposition or “seed,” let’s say, at just the moment that it blossoms as an occurrent sense datum – has to be understood, it seems, as part of (as “in”) that experience.  But if the phenomenal content of any cognition must be understood as a part thereof, it becomes hard to see how that could still make sense as a cause of the cognition; for it doesn’t make sense that a presently integral part of anything be at the same time a 																																																								18 The relevant subconscious processes are called by Buddhists “store-house consciousness” (ālayavijñāna), on which see William Waldron, The Buddhist Unconscious (RoutledgeCurzon, 2003); and Paul Griffiths, On Being Mindless: Buddhist Meditation and the Mind-Body Problem (Open Court Press, 1986). 

cause of the whole that comprises it – how could any part precede, as cause, the effect of which it is presently part? It seems, then, that when mental content is identified from a first-person perspective – when we attend, that is, not to an experience’s psycho-genesis, but to what is most salient for the subject thereof (namely, that it seems to be an experience of something) – it is hard to retain the same notion of causation that figures in the account that involves “seeds” and “habituated dispositions” and the like. Thus, in order to salvage his claim that “something whose nature is to be known internally” can make sense as both content and cause of awareness, Dignāga has to allow that when considered from a first-person perspective, experiential content meets the causal condition – it counts, that is, as also a cause of the cognition whose content it is – only on a different understanding of what “being a cause” consists in. He appeals, in particular, to an alternative understanding of “cause” according to which anything co-occurrent with some event – as, for example, anything’s defining characteristic is, ipso facto, occurrent along with the thing itself – can be called one of that event’s “causes.”19 So, the phenomenologically accessible content of any cognition – what the cognition seems to its subject to be about – can be reckoned as one of the causes of the cognition only in the limited sense that whenever a cognition occurs, it is in the nature of the case for it to have that content. Clearly, though, that is a very different sense of “cause” than when we entertain the idea that moments of experience are “caused” by a long and complex psychological past. Indeed, I suggest that Dignāga’s difficulties here show that when we try to reconcile first-personal accounts of the occurrence of contentful mental events with thirdpersonal accounts thereof, there’s an important sense in which we are just not referring to the same things. 19 Dignāga can here appeal to the Abhidharma category of sahabhūhetu, which denotes the kind of “cause” (hetu) which simply goes “with [something’s] being” (sahabhū); see Abhidharmakośa 2.50c-d (see note 7, above), where Vasubandhu gives the relation between “characteristic” and “characterized” as among the examples of this.

VI. Conclusion: Back to the hard problem

I suggest that this argument clarifies something at the root of “the hard problem”; the conceptually basic question that now comes into view is how or whether first-person perspectives on cognitive events can ever be reconciled with third-person perspectives on the same event. Perhaps, that is, what’s really intractable about the problem of getting subjectivity out of objectivity isn’t whether or how one kind of stuff can emerge from another kind; perhaps, rather, the real problem is in reconciling two essentially different perspectives, each of them involving distinct temporalities. Recall, then, that much discussion of “the hard problem” has (I noted at the start) centered on questions like whether physicalism leaves any room for consciousness at all. The foregoing line of argument from the Buddhist philosopher Dignāga, in contrast, argues for idealism – despite which, there arises for Dignāga, as acutely as for physicalists, what I take to be a conceptually more basic form of “the hard problem.” The more basic problem, I have suggested, is that in setting out to reconcile first-person perspectives on cognitive events with what can be identified from a third-person perspective on what is putatively the same event, it’s hard to see how we could know we are identifying the same events. In one contemporary idiom, this is a problem that bedevils the project of identifying the supposed “neural correlates” of experience; how could we know that when we observe this or that objective occurrence (say, heightened neural activity in some brain region), we have identified the “outsides” (as it were) of the very same events whose “insides” are conscious experiences? Our consideration of Dignāga’s argument gives us reason to think we could never be sure of this. He argues that there can be no relating physical objects even to those mental states that are supposedly about them; for the causal efficacy of physical objects could only be the causal efficacy of their atomic constituents – and there’s no explaining how manifold

momentary existents could cause the images of unitary wholes that are present to experience. But Dignāga has, we saw, some difficulty getting clear on his own, idealist alternative; he, too, is at pains to explain why it makes sense for such mental items as “habituated dispositions” to explain (as physical objects, he has shown, cannot) both the occurrence and the content of cognitions. This alternative works, he has to allow, only insofar as we admit two totally different senses in which anything can be a “cause” of cognition: The sense that’s relevant for a third-person perspective (from which we can identify complex psychological factors as “causing” present moments of experience), and the sense that’s relevant for a first-person perspective – the perspective from which what is most salient about any moment of experience is not its psychogenesis, but its content. Being itself a part of the experience, however, mental content cannot at the same time be a “cause” of the very event that contains it – so, the features of experience that show up for a firstperson perspective can only count as “causes” in a fundamentally different sense than do things like the functioning of sensory capacities. Whether one is an idealist or physicalist, then, may not finally matter for getting clear on what is so elusive about consciousness. The really hard problem, I suggest, is not how or whether subjectivity can be gotten out of objects; rather, it is how or whether we could ever be sure that third-personally identified cognitions are in any sense the same things that are first-personally experienced by the subjects thereof – a problem that can, as engagement with an argument by the Buddhist philosopher Dignāga shows, arise just as intractably for an idealist as for a physicalist. To see this as the more conceptually basic version of the hard problem is to see how really intractable the problem is; for before we address the question of how subjectivity can be instantiated in objects, we have to get clear on what subjectivity is – and Dignāga’s argument affords one perspective on why there may be nothing that makes sense as explaining that.

Bibliography

Anacker, Stefan. Seven Works of Vasubandhu: The Buddhist Psychological Doctor. Delhi: Motilal Banarsidass, 1984.

Arnold, Dan.  “Buddhist Idealism, Epistemic and Otherwise,” Sophia 27 (2008): 3-28. 
——.  “Dharmakīrti and Dharmottara on the Intentionality of Perception: Selections from Nyāyabindu (An Epitome of Philosophy).” In Buddhist Philosophy: Essential Readings, ed. William Edelglass and Jay L. Garfield, 186–196. Oxford: Oxford University Press, 2009. 
——.   Arnold, Dan.  Brains, Buddhas, and Believing: The Problem of Intentionaliy in Classical Buddhist and Cognitive-Scientific Philosophy of Mind.  New York: Columbia University Press, 2012. 
Chalmers, David.  “Facing Up to the Problem of Consciousness.” Journal of Consciousness Studies 2/3 (1995): 200–219. 
——.  The Conscious Mind: In Search of a Fundamental Theory.  New York: Oxford University Press, 1996. 
Dennett, Daniel, Consciousness Explained.  Boston:  Little, Brown, 1991. 
Dreyfus, Georges.  Recognizing Reality: Dharmakīrti’s Philosophy and Its Tibetan Interpreters. Albany, N.Y.: SUNY Press, 1997. 
Dunne, John.  Foundations of Dharmakīrti’s Philosophy.  Boston: Wisdom Publications, 2004. 
Eltschinger, Vincent. “Dharmakīrti.” Revue internationale de philosophie 64, no. 253 (2010): 397–440. 
Franco, Eli.  Dharmakīrti on Compassion and Rebirth. Vienna: Arbeitskreis für tibetische und buddhistische Studien Universität Wien, 1997. 
Gelong Lodrö Sangpo, trans. (based on the French translation by Louis de La Vallée Poussin), Abhidharmakośa-Bhāṣya of Vasubandhu.  4 volumes.  Delhi: Motilal Banarsidass, 2012. 
Gold, Jonathan.  Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy.  New York:  Columbia University Press, 2014. 
Griffiths, Paul.  On Being Mindless: Buddhist Meditation and the Mind-Body Problem.  LaSalle, IL:  Open Court Press, 1986. 
Hattori Masaaki. Dignāga, On Perception, Being the Pratyakṣapariccheda of Dignāga’s Pramāṇasamuccaya from the Sanskrit Fragments and the Tibetan Versions. Cambridge, Mass.: Harvard University Press, 1968. 
Hayes, Richard.  Dignāga on the Interpretation of Signs.  Dordrecht:  Kluwer, 1988. 

26 Jackson, Roger. Is Enlightenment Possible? Dharmakīrti and rGyal tshab rje on Knowledge, Rebirth, No-Self, and Liberation. Ithaca, N.Y.: Snow Lion, 1993.

Kellner, Birgit, and John Taber.  “Studies in Yogācāra-Vijñānavāda Idealism I: The Interpretation of Vasubandhu’s Viṃśikā,” in Asia 68:3: 709-756. 
Kapstein, Matthew.  Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought.  Boston: Wisdom Publications, 2001. 
Levine, Joseph.  Purple Haze: The Puzzle of Consciousness. New York: Oxford University Press, 2001. 
Lopez, Donald.  2008b. Buddhism and Science: A Guide for the Perplexed. Chicago: University of Chicago Press. 
——.  The Scientific Buddha: His Short and Happy Life.  New Haven: Yale University Press, 2012. 
McDowell, John.  Mind and World: With a New Introduction.  Cambridge:  Harvard University Press, 1996. 
McMahan, David L.  “Modernity and the Early Discourse of Scientific Buddhism.” Journal of the American Academy of Religion 72/4 (2004):897–933. 
Nagel, Thomas.  “What is it Like to be a Bat?” “Philosophical Review 83/4 (1974): 435-50. 
——.  The View from Nowhere.  New York:  Oxford University Press, 1986. 
Schmithausen, Lambert. On the Problem of the External World in the Ch’eng wei shih lun. Tokyo: The International Institute for Buddhist Studies, 2005. 
Siderits.  2009. “Is Reductionism Expressible?”  In Mario D’Amato, et al, eds., Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, pp.57–69.  New York: Oxford University Press. 
Taber, John.  “Dharmakīrti Against Physicalism.” Journal of Indian Philosophy 31 (2003): 479–502. 
Tola, Fernando, and Carmen Dragonetti.  “Dignāga’s Ālambanaparīkṣāvṛtti,” Journal of Indian Philosophy 10 (1982): 105-34. 
Valberg, J. J.  The Puzzle of Experience.  Oxford:  Clarendon Press, 1992. 
Waldron, William.  The Buddhist Unconscious:  The Ālaya-vijñāna in the Context of Indian Buddhist Thought.  New York: RoutledgeCurzon, 2003. 
Williams, Paul.  “On the Abhidharma Ontology.” Journal of Indian Philosophy 9 (1981): 227–257. 

Source

[[Category:]]