1.

As a philosopher involved in the study of religion, I share with many of my colleagues in the latter field an interest in origin myths. One such myth, which I learned in college, concerned the inception of the analytic movement in British philosophy. It went something like this: during the age of primal chaos, when thought was fuzzy and obscure, the evil genie of Idealism wandered free on earth. Then, a hero named George waved his hands, recited the magic words, ‘here is one hand, and here another’, and, presto! the genie took flight to return forever to the black depths from which he had sprung. Victorious analysis was born!

Like many myths, however, this one proved to be false. By the time he delivered the 1939 lecture containing the famous argument about his hands, our hero, George E. Moore, had grappled with the genie for some four decades.Footnote ¹ In 1920 he had received a warm expression of gratitude in perhaps the greatest achievement of British idealism, McTaggart's The Nature of Existence, and he would remain sufficiently troubled by the problems posed by idealism to return to them periodically throughout his career. Other analytic philosophers shared his concerns: in 1946, for instance, John Wisdom and J.L. Austin published a symposium on the problem of other minds, and the same topic preoccupied A.J. Ayer soon after.Footnote ² By 1963, when Michael Dummett delivered his celebrated lecture on ‘Realism’, the direct realism of the early analytic school, and the revived correspondence theory of truth that accompanied it, were beginning to appear sorely frayed.Footnote ³ The genie, indeed, had never quite vanished. As he had been wont to do throughout the centuries, he merely morphed, and, while some philosophers, such as Timothy Sprigge, welcomed him back enthusiastically in something resembling one of his old guises, others appealed instead to a range of non- or anti-realisms.Footnote ⁴ Analytic philosophy, born as it had been from the puzzles posed by the effort to clarify the relations among language, knowledge, and the world to which they refer, could never wholly renounce the broad legacy of idealism.

But just what is ‘idealism’? Part of the unsatisfactory nature of Moore's and others' ‘refutations’ stems from the fact that idealism, like mind itself, is Protean – there is no one philosophical doctrine that corresponds in all cases to ‘idealism’.Footnote ⁵ Kant was already quite clear about this, and accordingly he sought to distinguish carefully among what he termed ‘skeptical idealism’ – which he identified with the cogito-argument of Descartes – ’dogmatic idealism’—typified by the formula esse est percipi attributed to Bishop Berkeley – and transcendental idealism, the doctrine Kant personally embraced, which holds in effect that all that we know, all that we can ever know, we know only in so far as its appears within the field of our knowing.Footnote ⁶ But many other species and subspecies of idealism may also be noted – Platonic idealisms, Leibnizian idealisms, Hegelian idealisms, and more – so that it probably makes better sense to think of idealism as designating a great and fecund philosophical clan than it does a particular doctrine. The distinction that I wish to accentuate at the outset, however, as it may prove useful in beginning to explore the Indian philosophical landscape, divides what we might term eliminative from non-eliminative idealisms. Eliminative idealisms, sometimes also called ‘immaterialism’,Footnote ⁷ are those that seek to remove all but minds and mental entities from their ontologies, their inventories of what is, while non-eliminative idealisms accept that the universe really does include at least some non-mental things, but insist nevertheless that mind (or spirit, or reason) in some sense takes precedence over those others; the universe is, if not mind through and through, nevertheless mind-dependent, mind-made, or determined by mind. Eliminative and non-eliminative idealisms as I conceive of them are not simple opposites; they cover broad bands along a spectrum of thought, and the boundary between them may blur around the edges of concepts such as ‘mind-made’. A second general distinction, complementing this, that may also prove useful to us here is that separating monistic from monadic idealisms, the former affirming reality to be in the final analysis but a single spirit or mind, the latter committed to a plurality of spiritual or mental entities.

2.

These rough parameters allow us to situate some of the problems posed by Indian, and particularly Buddhist, philosophical idealisms. These, like their counterparts in Western thought, have at times reigned supreme, or been challenged and supposedly refuted, but nevertheless, once present in the Indian philosophical landscape, never quite left the scene. Of course, there is no term in Sanskrit, or in any other pre-modern Indian language, so far as I am aware, that offers a neat equivalent to ‘idealism’, with all the peculiar messiness of that designation. The western term, however, was frequently taken over by early twentieth-century historians of Indian philosophy and applied to a variety of particular traditions therein.Footnote ⁸ Vedāntic thought, above all, was very often characterised as a type of idealism, most notably Śaṅkara's non-dualistic Vedānta, in which some found neo-Hegelian affinities.Footnote ⁹ Nevertheless, the better writers on the subject always recognised that, just as idealism covered a wide swath in the history of western thought, so it could be taken over only to refer to a similar sprawl in Indian philosophy, and not to a single school or tradition. From such a perspective Buddhism generally seemed to merit consideration as a tradition tending to idealism of what I am calling the ‘non-eliminative’ variety, for throughout the Buddhist scriptural corpus the primacy of mind was everywhere affirmed. Had not the Founder himself, in the famous opening lines of the Dhammapada, proclaimed that:

The Buddha's emphasis on thought as prime mover, moreover, was firmly maintained even among those early Buddhist philosophical schools that are often held to be ‘realist’ owing to their assertion of the real existence of the material world and of our veridical sensory knowledge of it. In his great synthesis of early Buddhist philosophy, the fifth-century thinker Vasubandhu, for instance, declares, ‘the variety of the world is born of karma’, that is, action.Footnote ¹¹ And, lest we imagine that ‘action’ might refer here primarily, as it does in Hobbes, to the movements of bodies, karman, in this context, is specfically defined as action impelled by cetana, forethought, intention or will. Actions may be physical, as is the motion of a chariot, but for early Buddhist thought it was the mind that held the reins.

Non-eliminative idealism of the sort that I am describing characterised not only early Buddhism, but pervaded much of Indian thought, though with many variants in accord with the nuances of the systems concerned. However, the same Vasubandhu to whom I have just referred was also one of the founders of a particular line of Buddhist philosophy, most often called Yogācāra, ‘liberative practice’, that adopted a type of eliminative idealism, arguing that the world of external, material objects can be erased from our ontology, and that what we perceive and think of as such a world is in reality wholly constituted by the activity of minds. It is Vasubandhu's philosophy, in fact, that is usually intended when one speaks of ‘Buddhist idealism’. This is accentuated in Indian philosophical parlance, too, which designates the teaching of Vasubandhu's school as vijñānavāda, the ‘doctrine of consciousness’, or cittamātra, ‘mind only’, Sanskrit terms that come as close as any to our word ‘idealism’, at least in certain of its significations. Throughout the discussion that follows, ‘idealism’ will refer in particular to the eliminative idealism attributed to Vasubandhu's tradition.

Within specialised scholarship on Buddhist philosophy it has recently been much discussed, however, as to whether the designation ‘idealist’ really fits Vasubandhu or not. It will not be necessary to rehearse all of this here, and the debate has been in my judgement sometimes more concerned with semantics than substance. Some of the discomfort with the characterization of Vasubandhu as idealist perhaps stems from the too hasty identifications that have sometimes been proposed between his thought and specific trends in Western idealism. Thus, he has been considered at one time or another as a type of ‘subjective idealist’, whose doctrine resembles Berkeley's, or as an ‘absolute idealist’, as is often associated with some streams of Hegelianism. I concur with the critics that neither of these rubrics is quite suitable here, although, as I have tried to make clear above, I do not think that our use of ‘idealism’ can be narrowed to refer to just these species of idealism alone.

It has been suggested, too, that the characterization of Yogācāra as idealist is primarily an artifact of the prominence of idealism in Western thought during the late nineteenth and early twentieth century, when Yogācāra began to be studied in the West; but, despite the evident fact that the precise term ‘idealism’ is a Western one, this line of criticism is, I think, flawed. For it is clear that Indian intellectual traditions considered the cardinal, puzzling doctrine of Yogācāra to be – as it was succinctly expressed by the Jain commentator Guṇaratna in the fourteenth century – that ‘this world is only consciousness. There are no external objects, for it is the nonduality of cognition that is real. And there are a plurality of continua of cognition’.Footnote ¹² Of course, the fact that non-Buddhist thinkers such as Guṇaratna attribute a type of idealism to Yogācāra by no means establishes that they were justified to do so, but this does make clear that the pertinent features of an eliminative, metaphysical idealism were well known to Indian thinkers long before they had our term for it. I believe, too, that, though Vasubandhu's final intentions remain open to dispute, the idealism attributed to him does indeed find its basis in his own work. To show this, I shall rapidly signal some of the key features of Vasubandhu's idealistic turn, while indicating in brief a few points about which I think that there has been legitimate contestation.Footnote ¹³

As I have shown in detail elsewhere,Footnote ¹⁴ Vasubandhu's reasoning for the elimination of material objects was expressed as a mereological argument to the effect that the logic of part-whole relations entails that no coherent conception of physical matter can be formed. In a nutshell, the argument holds that atomism is both necessary and false. That Vasubandhu held this much is quite certain. It is tempting to draw from the argument, further, the conclusion that Vasubandhu wished to exclude matter altogether, but it is perhaps also possible to read Vasubandhu as suggesting – for he is not quite explicit about this – that there may nevertheless be an external, physical reality whose constitution and nature are inscrutable to us. Personally, I favor the eliminationist over the skeptical interpretation, but both are suggested in the work of Vasubandhu's successors, as will be seen below. Some of the peculiar features of Vasubandhu's idealism I take to be the following:

It is important to stress, too, as should be evident from all this, that Vasubandhu's idealism was in no way monistic. This was clearly understood by the opponents of Yogācāra thought, such as the Jain Guṇaratna, cited above, or the Buddhist critic Candrakīrti, who in an interesting passage takes pains both to compare and distinguish Yogācāra from theism. He writes:

As we know from the Laṅkāvatārasūtra, a popular Mahāyāna Buddhist scripture, the adherents of Yogācāra were much concerned to distance their teaching from the suggestion that it involved the tacit affirmation of theistic or Upaniṣadic metaphysics in a Buddhist guise.Footnote ²⁰ But genuine Indian theists, such as those of the Nyāya school, were also not adverse to affirming the genuine reality of the world that their god had created.Footnote ²¹ Whereas the scandal of Yogācāra, in the view of most other Indian philosophical traditions, Buddhist and non-Buddhist alike, was crucially to be found in its denial of the external world. It is here, too, that the position of Yogācāra in the history of Indian thought resembles that of idealism in the West. Just how, once skepticism about external reality arises, do we regain our philosophical innocence and return to the unchallenged assurance that what appears to us indeed exists?

3.

Among Vasubandhu's leading disciples, Dignāga (active ca. 500) rivals his master for his broad importance in the history of Indian and Buddhist philosophy. The early Yogācāra school had been interested in the practice of debate, and Vasubandhu in several works had already sought to expound the rules of logic that were to be followed. On these foundations, and in critical dialogue with the major non-Buddhist philosophical traditions, Dignāga elaborated a comprehensive system of pramāṇa, a word referring to the measure or ‘criterion’ of knowledge and used to designate the fields of learning that we cover with the terms ‘logic’ and ‘epistemology’, i.e. the ‘theory of knowledge’.Footnote ²²

Dignāga's greatest work was his Pramāṇasamuccaya (‘Compendium of Logic and Epistemology’), in which he held that there are in fact just two criteria for knowing, veridical perception and sound inference, and that these, in turn, correspond to just two classes of objects: particulars that are characterised by their unique phenomenal characteristics (sva-lakṣaṇa), and concepts that are characterised by their universality (sāmanya-lakṣaṇa). In later Buddhist tradition, Dignāga's system as presented in the Compendium gave rise to both phenomenalistic and idealistic readings, though a number of his other writings suggest that, like his master, he was partial to idealism. While Dignāga held that the objects of our perceptions are particulars bearing unique characteristics, his concept of the particular becomes the point of departure for a number of difficult questions: Is the object that we perceive actually something that exists ‘out there’ in the world, just as we perceive it? Or is the object something that arises within our sensory field, perhaps corresponding to an external object that served as a stimulus, but not in fact identical to it? Or is the object exclusively an object of consciousness, on the basis of which we construct the idea of an external world that does not exist in reality? The first of these alternatives, the ‘direct realism’ that had been current in early Buddhism and several of the non-Buddhist philosophical traditions, Dignāga, like Vasubandhu, rejected decisively. The second, a type of ‘indirect realism’ (or, on some readings, ‘phenomenalism’), had been embraced by Vasubandhu in his realist writings, and was sometimes regarded as Dignāga's true position as well. The last, of course, is ‘idealism’.

Dignāga's short treatise, Ālambanaparīkṣā (‘The Examination of the Objective Referent’), is his foremost contribution to the problem we have just sketched out. Here he seeks to clarify the difficulties in this area by focusing upon the investigation of the ālambana, translated here as ‘objective referent’.Footnote ²³ Sanskrit has a number of distinct specialised terms where we use the one English word ‘object’ (think of the differences involved when we speak of ‘an object in the living room’, ‘the direct object of a verb’, ‘the object of her journey’, ‘an object of yearning’, etc.). It is often a challenge for translators of philosophical Sanskrit to find just the right way to modify ‘object’, so as to accord with the precise usage of the authors they translate.

The term ālambana had a long history in Buddhist thought before Dignāga. It had been used in the abhidharma literature of early Buddhism to refer to that to which any given mental act is directed; indeed, in some works the very having of an ālambana (sālambanatvam) is what defines a mental act.Footnote ²⁴ In this respect, it resembles what some philosophers in the West have called an intentional object. With this latter concept it also shares, as Dignāga's discussion makes clear, a peculiar, problematic status, stemming from the phenomenon sometimes called in Western philosophy ‘intentional inexistence’, the curious feature of mental objects that permits us to refer to them, without implying their real existence at all. I can desire to visit the Happy Land of Cockaigne, whether or not such a place exists or ever existed. The phenomenon of intentional inexistence by itself, of course, does not establish the general truth of idealism, but it does show that some of our minds' objects do not correspond in any clear way to what we suppose to be really so. For Dignāga, however, the problem of intentional inexistence arises not just in respect to the objects of our intentional attitudes such as desires but includes also perceptual objects, which, he holds, do not correspond to realities external to our perceptions of them.Footnote ²⁵ For taking, as his point of departure, the mereological puzzles posed by his master Vasubandhu, Dignāga shifts the ground subtly but significantly, by asking how our perceptions might correspond to an external world, particularly if it is an external world composed of atoms. The assumptions either that atoms possess the properties we attribute to the aggregations we perceive, or that they do not, cannot in his view be coherently maintained, and lead Dignāga to conclude that, because our perceptions do not correspond to the atoms that are their putative causes, the atoms need not be posited at all. But if we are warranted in jettisoning the atoms, what then of the atomic aggregations we had posited? We are left with a world of perceived properties that exist nowhere but in the minds that perceive them. And without external, physical objects, he maintains, apparent causal order can only be an attribute of the relations among our cognitive acts themselves. This pertains equally to the causal order lending coherence to our own sensory apparatus, which, being no longer physical, must be reinterpreted in terms of the potentiality of consciousness to experience specific types of apparent sensory phenomena.Footnote ²⁶

Dignāga's significant contribution to the discussion, in my view, was not so much in the details of his argument as it was in a key strategic step. With Dignāga, idealism in Indian philosophy became primarily a problem in epistemology, and the ontological concerns of Vasubandhu, though certainly not forgotten, were now subordinated to the inquiry into our knowledge of the external world.

This epistemological turn becomes fully explicit in the work of his greatest successor, Dharmakīrti, whose major treatise, the Pramāṇavārttika (‘Commentary on Logic and Epistemology’), presents itself as a commentary on Dignāga's Compendium. On the question of external objects – our central problem here – Dharmakīrti shares with his predecessor some measure of ambivalence, sometimes suggesting a robust commitment to idealist thought, while elsewhere appearing to acknowledge the possibility that external objects do exist, but are merely hidden from us. Moreover, in certain contexts Dharmakīrti definitely adopts a realist standpoint, to explore how things appear to us, whether or not, in the final analysis, they are as they appear. Later interpreters, therefore, tended to divide between those who favored a thoroughgoing idealist reading, considering the realist passages to be only the tentative expression of Dharmakīrti's view of conventional reality, and those who argued that Dharmakīrti's true intentions accorded more closely with those of the indirect realists, his idealism being a provisional bracketing of objects in order to better clarify the puzzles of reflexivity. Among the commentators, none was more committed to the defense of idealism than was Prajñākaragupta (ca. 800), and none more influenced later Indian philosophy, Buddhist and non-Buddhist, than he as well. We shall examine, therefore, some of Dharmakīrti's main arguments together with Prajñākaragupta's remarks on them. The latter introduces the problem, writing:

Prajñākaragupta restates Dharmakīrti's response (verse 321) as follows:

It is just here, I think, that it becomes clear that the position emerging among the Buddhist idealists more closely resembles what we would call transcendental idealism than it does subjective idealism. Dharmakīrti, on Prajñākaragupta's reading, is asserting not that objects exist just in being known; he is telling us, rather, that there is no way we could finally know either this or the opposite, for there is no position outside of our knowing that we can occupy, such as will permit us to judge just how to construe the relationship between awareness and its object. As Prajñākaragupta notes, ‘If one attempts to demonstrate an object ostensibly, it is not seen, for there is no “seeing of an object”’. This, I suppose, would be the essence of his answer to Moore.

Dharmakīrti introduces and responds to three principal objections to this line of thought (verses 322–26). The first, while conceding that there is no ‘view from nowhere’ that will permit us to assess independently the relationship between perceiver and object as such, nevertheless maintains that we might have good reason to suppose there to be a crucial resemblance between what we perceive and what exists in fact ‘out there’. But, following the doubts already raised by Dignāga, it may be asked why it is, in this case, that the material world of perception does not resemble that of theory: ‘If the atoms were the object, then, when there is the consciousness of a massive appearance, which does not resemble the atoms, in what sense are they the object of consciousness?’ And although, as Prajñākaragupta adds, one alternative might be to consider material things as partless, just like the perceptions in which they are given, the suggestion does not seem to have been seriously countenanced.Footnote ²⁸

The second objection (verse 324) seeks to reinforce the first by proposing that it is the concurrence of resemblance and causation that assures us that our perceptions are objectively founded. Prajñākaragupta presents the argument in these terms:

This will be clearer if we define some of the key terms: a ‘feature’ here is a phenomenon, or, more precisely, the phenomenal character of an act of awareness. The ‘immediate condition’, and indeed the entire issue of immediacy in this passage, refers to a peculiar view of consciousness that arose in early Buddhism, according to which the apparent unity of consciousness, synthesizing thoughts, feelings, and our various sensory inputs, was to be explained by the action of a subliminal conscious moment called the ‘immediate condition’ (samanantarapratyaya). But, as Prajñākaragupta notes here, the immediate condition, conceived in this way, might just as well be synthesizing only data supplied by prior consciousness in order to fulfill the conditions of resemblance and causation upon which the objection was based.

The third objection (verses 325–26) relates to the feeling of certitude that we so often attribute to our perceptions. Again, following Prajñākaragupta:

The doubts voiced here concerning the evidential value of the sense of certitude are by no means foreign to recent anglophone philosophy; Peter Unger's fine essay, Ignorance: A Case for Skepticism offers an excellent example.Footnote ²⁹ The assertion of certitude as evidence for our claims to knowledge is a protective strategy, often deployed to mask the unsettling possibility that we might be wrong. Though Dharmakīrti and his school by no means embraced skepticism and indeed affirmed that we can achieve a kind of certainty, they restricted its scope to the conclusions of reasoned investigations, once our doubts have been raised and resolved. In itself, however, it does not warrant the conclusions to which it might be ascribed.

Having dispensed with these objections, concerning causality, resemblance and certitude, what does Dharmakīrti propose as a positive account of our experience of apparently external objects? (For, as we have said, he is no skeptic.) His response is unequivocal:

4.

In introducing the final section of this essay, concerned with the Jain critics of the Buddhist idealists, I must begin with a modest disclaimer: I am not at all a specialist in the study of Jainism and am indeed a neophyte in this area. My current engagement in Jain Studies is due to ongoing research on Buddhist idealism and the resulting appreciation of the importance of Vasubandhu's legacy for later Indian philosophy generally. In exploring increasingly the works of non-Buddhist thinkers who were influenced by or critical of this tradition, I have become particularly impressed by the rich contributions of Jain philosophers to the debate, but I am still at the beginnings of this research, and so present here only preliminary observations.

In responding to the Buddhist idealists, the Jains, unlike both Buddhist and Brahmanical opponents of Vasubandhu's tradition, were not primarily concerned to elaborate decisive refutations. For the Jain approach to philosophy overall was based on the principle called literally ‘non-one-sided-ness’ (anekānta), or ‘non-absolutism’ in Satkari Mookerjee's phrase, that precluded by and large the search for philosophical finalities.Footnote ³⁰ Reality, for the Jains, cannot be reduced to any single viewpoint and the rich multiplicity of perspectives required to exhaust the reality of any phenomenon can be only within the purview of an omniscient mind. Accordingly, in their approach to idealism, as in their approach to other philosophical doctrines, the aim of Jain criticism was to point out that what they considered to be the inadequacies stemmed from its one-sidedness, or as we might put it, reductiveness.

Nevertheless, for Jain thinkers it was particularly important to resist the idealist challenge. More than for many other traditions in Indian religions, for the Jains the attainment of liberation, mokṣa, was not merely regarded as a spiritual problem, but equally as a physical problem. For the purgation effected by the strict asceticism that is characteristic of Jain spiritual disciple was intended not just to discipline body and mind, but to drive out and cleanse the impure karmic matter which subtly prevades the physical bodies of mundane beings. Hence, it appeared to the Jains that, by eliminating the reality of the physical world, Yogācāra was undercutting, despite its name, the very basis for the practice of yoga as conducing to salvation. It is not a matter of astonishment, therefore, that the Jains were particularly determined to overturn this problematical teaching.

At the same time, it must be stressed that, despite their disaffection with Yogācāra, this was the Buddhist tradition in whose study the Jains were probably most deeply invested. This was owing to the considerable influence of Dignāga and Dharmakīrti, whose advances in the fields of logic and epistemology were very largely taken over by Jains of the late first millennium as the methodological basis for refinements within their own system.Footnote ³¹ A very large proportion of the Jain philosophical literature written from about the seventh century on – that is to say after the age of Dharmakīrti – thus engages Buddhist pramāṇa to varying degrees. To date I have identified over a dozen sustained discussions in Jain philosophical works of the particular problem they regarded as central to the project of Yogācāra, concerning the assessment of arguments bearing on the existence of the external world.

For purposes of illustration, I will focus upon two particular Jain treatments of this subject-matter. The first, by the renowned and prolific eighth-century teacher Haribhadra Suri, is his critique of Vijñānavāda – the ‘consciousness doctrine’ – in his verse treatise, the Śāstravārtāsamuccaya, or the ‘Science News Digest’, as I like to call it, and its autocommentary. The second is the sprawling subcommentary on this, entitled Syādvādakalpalatā, the ‘Wishing Vine of the Doctrine of Tentatives’, by the remarkable philosopher Yaśovijaya Gaṇi (1624–88), who was a contemporary of the Mughal emperor Aurangzeb.Footnote ³² While I am of course concerned here with the content of their treatment of Yogācāra, I wish also to illustrate something of philosophical progress in Indian intellectual history, for during the period of almost a millennium separating Haribhadra from Yaśovijaya, great advances were made in logic and philosophical method and these are clearly reflected in Yaśovijaya's work. Indeed, much as Haribhadra represents a period in the history of Jain thought in which the lessons of Dharmakīrti and his early successors were being absorbed, Yaśomitra demonstrates an analogous movement with respect to developments within the Brahmanical Nyāya tradition, which specialised in logic and debate, culminating in the emergence of the ‘new logic’, or Navya-Nyāya, that flourished from about the thirteenth century on. It will be seen that the realism that was a fundamental characteristic of Nyāya thought contributed to Yaśomitra's critique of the Buddhist idealists.

Haribhadra begins his examination by raising a long-standing problem for Indian philosophers: just what are the conditions that permit us to affirm the non-being of anything?Footnote ³³ For what the Buddhist idealist needs to affirm is the non-being of the external world altogether. Adhering to Dignāga's reduction of the criteria of knowing to veridical perception and valid inference, Haribhadra asserts that because non-being is nowhere perceived, we can only have recourse in this case to inference. But what kind of inference will do? Only one that derives at least one of its premises from what Indian philosophers term ‘non-apprehension’ (anupalabdhi).

In order to explain this concept in brief, as well as why it is that it need not be, for Haribhadra, a tacit admission of a perception of non-being, let us consider an elementary example. (This one is in fact derived from Dharmakīrti.) We are worried that there might be a fire in a given place and, arriving there, exclaim, ‘there's no smoke here!’ Now, some philosophers may wish to hold that we have just affirmed a perception of the non-being of smoke in that place – indeed, there were Indian philosophers who did adopt just this approach – but Dharmakīrti will have none of it: for him, ‘there's no smoke here’, is the expression of a peculiar sort of inference. On reaching the place in question, we have no perceptions of non-being; we have, rather, perceptions of the place that, it so happens, do not include the characteristic properties we associate with smoke. And it is based on this assessment of the properties of our perceptions that we draw the conclusion that ‘there is no smoke here’. In short, we infer from our perceptions' not having particular properties that we regard as evidentiary to the non-being of the things that those properties are supposed to be evidence of, but non-being as such is never perceived.

Haribhadra, I think, accepts all of this and then asks how it could possibly apply in respect to external reality overall. For, the kind of instance in which an inference from non-apprehension may be supposed to be valid must conform, he says, to this principle:

Applied to Dharmakīrti's example of the smoke, this of course means that in such a case we perform the inventory of our perceptions as we do only because we had formerly perceived smoke, and so can now check our current perceptual field in relation to our memory of the experience. But none of this is applicable when the object whose non-being is asserted is external reality überhaupt. For in this case what could possibly count as the characteristic property, the not-having of which conduces to the inference of non-being? And, assuming that such an experience were possible, it could only be so if external reality had in fact been formerly perceived and were now recalled, in which case, the very denial of such reality entails its existence.

This argument, however, can have little force against the Buddhist idealist, because it does not really address the major strategies deployed by the Buddhists in seeking to demonstrate idealism. These involved attempting to demonstrate that the available theories of external reality were incoherent, to propose that the appearances of external reality could be as well or better explained by an alternative theory, one without commitment to external objects, and to show that we cannot in any case step outside of the circle of our own awareness. Such a strategy had led to Dharmakīrti's conclusion, as we have seen above, that ‘cognition experiences itself, and nothing else whatsoever’. Haribhadra's argument, in relation to these points, can at most be taken as indicating – and perhaps this is all he really intends to indicate up to this point – that, whatever the force of the Buddhists' arguments, they fall short of a proof of idealism. In the remainder of his discussion he then seeks to offer, if not exactly a proof of his own contra idealism, then at least a number of suggestions favoring the affirmation of something rather like commonsense realism. As an example, one of his more forceful arguments is the following:

His argument is something like this: Objects are not processed by us in isolation; indeed, their perception presupposes a rich nexus of background causes and conditions, in virtue of which the apparent occurrence of the thing in question is engendered. The mere existence of some such framework is of course not itself what is in question – even the idealist affirms that perceptions belong to complexes – but what Haribhadra suggests is that it is not plausible to regard the relevant frameworks as stemming wholly from within ourselves, for, if we posit perceptual frameworks at all, we must also posit their externality.

Yaśovijaya's unpacking of the argument provides an excellent illustration of his commentarial project as one of rational reconstruction of Haribhadra's work, something quite distinct from the composition of a mere gloss upon it. He characteristically employs here later advances in Nyāya, in this case reflecting the contributions of the eleventh-century philosopher Udayana in respect to the logic of non-apprehension, the knowledge of absence, and its entailments with respect to the absentee, called pratiyogin, which for Nyāya thought must be something real. The intuition here can be made clear if we consider a statement such as ‘there are no square circles in this room’. Now, from one point of view, this statement is an absurd bit of nonsense, much like ‘the Jabberwock is out to lunch’. For in both of these cases, because there is no referent, there is nothing that these sentences are about; they are just silly. But whereas it is immediately evident that the latter, concerning the Jabberwock, means nothing at all, the former may appear to affirm a true proposition; for it seems to be indeed that case that there are no square circles here or anyplace else. Does this not show that we can, in some cases, speak meaningfully of things that are unreal? The Nyāya philosopher would hold that this impression is due to a mistaken analysis. We have an idea of squares and an idea of circles, and these have real referents. It is our combination of them in the phrase ‘square circle’ that is an error; for ‘square’ excludes the idea of ‘circle’. The appearance of meaningfulness is due only to the meaningfulness of the terms that we have illicitly joined in composition. Language, if it refers at all, can only refer to realities, even when it speaks of absence.

Consider now Yaśovijaya's reconstruction of Haribhadra's argument. He expresses it thus:

Yaśovijaya's presentation of the argument here deploys the technical usage of late Nyāya discourse, though, in analogy to the technical usages of analytic philosophy in the anglophone world, the peculiar diction of Udayana and his successors was motivated by the desire to ‘make our ideas clear’. And, indeed, once one grasps the argument, it does seem clear. Let us consider an everyday example: I open the refrigerator in search of the mustard jar, and I perceive that there is no mustard. Now, some philosophers would hold that what one perceives is an ontologically peculiar type of entity, namely an absence. Adopting this approach, I find that my refrigerator is quite a populous place, for, besides the absence of mustard, it contains infinitely many other absences, including the absence of the Andromeda galaxy and the absence of Tyrannosaurus Rex. But how can it be that my fridge holds all these things and that I can perceive them? And why is it that, when I'm searching for mustard, that is the absence I find, instead of being overwhelmed by gazillions of absences that must be then laboriously sifted until I find the absence that pertains to my search? Some versions of the Nyāya realism, like the realism of Alexius Meinong, seem to entail just such problems in their treatment of absences.

In Yaśovijaya's construction, however, my refrigerator presents me with a nexus of conditions for my objectifying the mustard jar as absent, conditions that are not to be identified either with the mustard jar or whatever it comprises. It is indeed the mustard jar that is my object, but the object is in this case qualified by its absence, and it is this qualification that becomes known to me by reference to what I perceive to be present. Thus we are definitely distinguishing presence and absence, discerning their alterity, which seems not plausible if the present conditions and the absent object are thought to be indiscernible insomuch as they exist only as objects of of mind. Moreover, we must ask, just what are the present conditions that would permit me to know the absence of external objects altogether, as the idealist wishes to do? Given the radical realism of Nyāya, because all absentees must correspond to realia, just to entertain the external world as an absentee would seem now to entail its real existence.

For Dharmakīrti, the difficulties of establishing the independent reality of the object, together with the reflexivity, or self-presentation, of consciousness, incline to the conclusion that what we know is all and only what appears to reflexive consciousness. If my seeing is self-presenting for me, and my object of sight is known only insomuch as it appears within that seeing, then it would seem that the object is known only in relation to my self-presentating awareness. Haribhadra does not exactly say that this is wrong, so far as it goes. Rather, he holds (in verse 5.12) that this is in fact just what we mean by being aware of an object. We cannot, indeed, be aware of it any place except within the confines of our own awareness. It is only by taking too seriously an unwarranted skepticism that one would argue on this basis that objects do not exist, or that their existence must be bracketed. And to one who might retort that, given this much, we would have no basis for distinguishing genuine objects from hallucinations, Haribhadra replies:

We can perhaps get at what Haribhadra means here by a ‘mere awareness’ by considering what we mean when we say that a bit of knowledge is ‘merely theoretical’. Haribhadra is calling our attention, I think, to a distinction we commonly make between cognitions with or without practical value, and he is asserting that, in effect, idealism robs us of the force of this distinction but treating all knowledge as being in a crucial sense ‘merely theoretical’. Yaśovijaya makes this point by saying that, on this account, there is ‘nothing outside of our delimitations’. Our common notion of external reality makes sense to us not owing to a fully rationalised framework for our affirmation of it, but rather in virtue of its coherence with the network of practices in which we situate both ourselves and the objective world as it appears to us.

In concluding his critique of the Buddhist idealists, Haribhadra returns to the properly Jain refusal of one-sidedness and he applies this to his own critique. The idealists, he holds (verses 6.52–53), are wrong, but not one-sidedly so; for their doctrine, although subject to metaphysical rebuttal, is nevertheless one that may have propaedeutic value for those cultivating a spiritual path; for idealism undermines belief in the ultimacy of external and material goods, so that to entertain it may aid those who find it difficult to achieve equanimity and detachment in regard to worldly things. A provisional acceptance of its truth may therefore be, in some circumstances, practically warranted. But it must be rejected before it can corrode the Jain commitment to ascetic rigor. And about this he is quite decisive (verses 5.29–39): the real danger of idealism stems from its potential to undermine the distinction between bondage and liberation. For the Jains, our impoverished spiritual state stems from corruptions that can be purged only through long and painful ascesis. The fault of Buddhist idealism is, in the final analysis, more soteriological than metaphysical, for it seems to treat our spiritual predicament as a matter of illusion, to be removed, as if it were mere morning mist, by the workings of insight alone.

The fundamental Buddhist error, in the Jain view, therefore seems to have been a penchant for an exaggerated intellectualism, weakening the commitment to the unyielding spiritual discipline that Jainism holds to be the essential foundation for the realization of freedom. It goes without saying, of course, that the Buddhists would not concur with such an assessment. Where both parties agreed was in the proposition that epistemology and metaphysics were of importance for their contributions to the grounding of practical soteriological endeavours.Footnote ³⁴

In accord with my understanding of the theme of the present volume on ‘Philosophical Traditions’, it has been my aim here not so much to provide a detailed account of particular lines of argument, but rather to suggest something of the manner in which so-called Buddhist idealism unfolded as a tradition not just for Buddhists, but within Indian philosophy more generally. Seen from this perspective, Buddhist idealism remained a current within Indian philosophy long after the demise of Buddhism in India, in about the twelfth century, and endured in some respects at least until the Mughal age, when the Jain philosopher Yaśovijaya was active.Footnote ³⁵ Of course, I have omitted altogether from this discussion the considerable legacy of Indian Buddhist idealism in the philosophical traditions of China, Tibet, Korea and Japan, continuing down to our own time. All that, however, is subject-matter for another occasion.