YogAcAra Strategies against Realism: Appearances (Akrti) and Metaphors (upacAra)

Jonathan C. Gold* Drew University

Abstract

This paper describes and elucidates two philosophical parallels: the first between two early Yog-c-ra Buddhist positions, and the second between these positions and a contemporary discussion in analytic philosophy. The two Yog-c-ra positions are (1) that reality is a mere appearance (AkRti); and (2) that all language is metaphorical (upacAra). First, the distinctive Yog-c-ra belief in the reality of mere appearances is clarified by carefully distinguishing the terminological uses in Vasubandhu’s Three Natures Exposition (TrisvabhAvanirde8a) from the terms used to make parallel conceptual moves in a very similar passage from the Ornament to the MahAyAna Sutras (MahAyAnasutrAla kAra). Second, the theory of language advocated by Sthiramati in his commentary to Vasubandhu’s Thirty Verses (Trim8ikAkArikA) is explained as a counter to arguments that Yog-c-ra non-realism is self-defeating. This latter view shows how Yog-c-ra may be fruitfully brought into dialogue with Hilary Putnam’s famous antiskeptical Brain-in-a-Vat argument. The paper finds basic accord within Vasubandhu’s ontology and theory of reference, and shows how they represent a new, coherent Yog-c-ra response to Putnam’s realist challenge.

Buddhism today is better known as a tradition of meditators than as a tradition of critical thinkers. Given this, students often find it difficult to free themselves from the expectation that Buddhist doctrines are insights derived from, and confirmed by, meditation. This common assumption makes Buddhist philosophy (along with much of Indian philosophy) a form of spiritual insight or even divine revelation, and thereby places it beyond rational reflection. This misunderstanding has many causes, not least of which is that what is best known of Buddhism in the West remains within the fantasy-laden circle of Western Romanticism. Of course Buddhist traditions also tend to present themselves as coming into being through a series of great meditative discoveries – starting with the story of the Buddha’s enlightenment, and continuing throughout the hagiographies of subsequent teachers. Such descriptions have tended to unjustly undermine Western philosophers’confidence in the reasonableness of the views articulated by these traditions, and in their right to the name “philosophy.” For

Religion Compass 1 (2006): 10.1111/j.1749-8171.2006.00014.xAs Johannes Bronkhorst has noted in rejecting such biases, many Indian philosophical positions that appear bizarre or confounding to outsiders are best explained not as revelation, but as in fact reasoned, philosophical attempts “to deal with difficulties arising from shared assumptions” – assumptions, that is, that Indians share, but Western philosophers do not (Bronkhorst 2001, p. 479). But this by no means prevents Indian thought from being “philosophy.” Every tradition has its shared assumptions, and as anyone who has gained insight from reading Aristotle knows, there is much to be learned from studying great philosophers even if their concerns are only adjacent to, but not identical to, one’s own. Sometimes, in fact, the difference in assumptions itself reveals a new direction for one’s own thinking.

One such case in Buddhist philosophy is the tradition’s continuing need to deal with the “shared assumption” of linguistic and ontological non-realism – that is, the Buddhist belief that we are deluded in our most basic beliefs and concepts. This philosophical need has issued in a range of doctrinal formulations that cleverly evade many sophisticated criticisms of non-realism, including some under current circulation in Western philosophy. In particular, I hope to show here that the great Buddhist philosopher Vasubandhu (320–400)1 presents a view – the Yog-c-ra view of the “three natures” (trisvabhAva) – that constitutes a sophisticated response to at least one issue of current concern among philosophers working in the Western tradition. Before turning to the comparison with Western thinkers, however, it will be necessary to gain a clear view of Vasubandhu’s notoriously tricky Yog-c-ra doctrine, which I will present through a reading of his succinct ontological statement in his Three Natures Exposition (TrisvabhAvanirde8a).

Vasubandhu on Ordinary Reality: Not Real or Unreal, but Appearance (-kRti) The particular passage under examination here has been widely discussed and variously interpreted (simply compare, for instance, the English translations of Anacker 1998; Garfield 2002; Kaplan 1990; Kochumuttom 1989; Nagao 1991; Tola & Dragonetti 1983; Wood 1991). Academic scholarship on Yog-c-ra Buddhist philosophy has little to agree on, even as regards its most basic doctrines and terminology. This may be thought of as a sign of the health (if also the youth) of this area of Buddhist philosophical research. It may be simply a natural result of the diversity of perspectives that appear within the Indian texts themselves, grouped although they are under the common heading, “Yog-c-ra.” In any case, I would be amiss if I failed to mention here, in a forum for non-specialists, that within a contentious field my interpretation is only one, somewhat unconventional approach (I would say “controversial,” but no controversy has yet erupted).

I have argued my perspective on historical and philological grounds elsewhere (Gold 2006), and I am extending that general view here. But the reader is advised to seek his or her own perspective through examining the full range of sources in the list of references.2

The passage in question contains three versions of an analogy of a magical illusion (mAyA). The general idea, widely used as a thought experiment in this philosophical arena, is that there are magicians who can produce illusions, using magical incantations (mantras) and geometric designs (yantras), so as to make their audience members think that they are viewing something that is not there. Vasubandhu refers to the magician generating an illusory elephant, but other related passages refer to a horse, a pot of gold, a herd of wild animals, a crowd, a battle, or an illusory man. Vasubandhu and his Yog-c-ra contemporaries seemed to believe that the magical illusion depends upon some causes outside the mind – such as, perhaps, a piece of wood shaped like an elephant – but the illusory image itself is somehow conjured so that it overrides the audience member’s normal sensory impressions, appearing directly in the mind (Gold 2006, pp. 4–14).

For Vasubandhu, ordinary reality is like this kind of an illusion, and the “three natures” of reality in his philosophy are each comparable to a specific aspect of the magical illusion. That is to say, different aspects of the illusion represent different vantage points from which to illustrate one and the same reality, which is our reality, the reality of ordinary sentient beings. Reality is like a magical illusion because ordinary beings like us (as opposed to extraordinary beings like Buddhas) are generally deluded as to the true nature(s) of existence. According to Buddhist doctrine, all beings continue to suffer the eternal round of rebirth (samsAra) until they are able to overcome their fundamental ignorance about the nature of reality – most importantly, their false belief in the reality of their “self ” (Atman) or independent soul. Vasubandhu, articulating what was to be a common Yog-c-ra position, describes this belief in the self as a distinct reality, separate from what is non-self, as the basic error of “duality.” In truth, self and other are unified in a single causal stream of mental events (which is why the Yog-c-ra are sometimes called the Mind Only, or citta-mAtra school). The impression of “duality” is the impression that I am one thing, and the things I experience – such as my perceptions of the world, and my thoughts – are something else (not me). Vasubandhu says that the impression of “duality” is a mental “construction” – in the sense that it is made to appear, but does not exist as it appears. The goal articulated in Yog-c-ra texts is to purify one’s mental stream so that the false constructions cease, and one gains a direct apprehension of the actual causes of experience – a liberating achievement called the “transformation of the basis” (A8raya-parAvRtti). Like everything else, the “basis” that is transformed is the mind. The three natures, then, as I say, are three vantage points on this reality. Very briefly, the “dependent nature” (paratantra svabhAva) is the process viewed from outside – a causal description of the arising of an erroneous duality. This is likened to the magical production of an illusory appearance of an elephant. The “constructed nature” (parikalpita svabhAva) is simply the appearance itself – ordinary, erroneous (dual) experience. This is likened to the elephant that appears to exist. The “perfected nature”

YogAcAra Strategies against Realism . 3(parinirpanna svabhAva) is a fact about the constructed nature: that it is “empty” of reality, that it does not exist as it appears. This is likened to the non-existence of a real elephant in the locus of its appearance. In order to best display Vasubandhu’s distinctive use of the illusory elephant analogy in elucidating his Yog-c-ra position with the conceptual mechanism of the three natures, I will be juxtaposing Vasubandhu’s eight independent verses from the Three Natures Exposition on the illusory elephant with a very closely related, but far less well known passage: three verses from the Ornament to the MahAyAna Sutras (MahAyAnasutrAlamkAra) together with their commentary that is attributed to Vasubandhu. First, notice now the following two Ornament verses are paralleled by Vasubandhu’s four independent verses below them: (Ornament XI.15–16 with Vasubandhu’s commentary)3 15. The construction of what does not exist is like a magical illusion. The error of duality is like the appearance of a magical illusion.

The construction of what does not exist, the dependent nature, should be understood to be just like a magical illusion (mAyA) in which the cause of error such as a piece of wood and a lump of earth are surrounded by a yantra. The appearance of the constructed nature, the construction of what does not exist’s dual error, appears to be grasper and grasped [i.e. sense organ and sense object], and should be understood to be just like the appearance of a magical illusion, in which the illusion’s appearance of an elephant, horse, gold, etc. appears to exist. 16. The ultimate [[[truth]]] is said to be like its existence which is not there.

The nature of conventional truth is like perceiving it.

The ultimate truth, there in the dependent, is the non-existence of the constructed that has a dual characteristic, and it is said to be like its existence which is not there, [i.e.] the non-existence of elephant-ness (hastitva) etc. in the appearance of the magical illusion. The nature of conventional truth is perceiving the construction of what does not exist, which is like perceiving the appearance of the magical illusion as existing as an elephant, etc. (Three Natures Exposition 27–30)4 27. It is just as [something] made into a magical illusion with the power of an incantation (mantra) appears as the self of an elephant. A mere appearance (AkAramAtra) is there, but the elephant does not exist at all. 28. The constructed nature is the elephant; the dependent is its appearance (AkRti); and the non-existence of an elephant there is the perfected. 29. In the same way, the construction of what does not exist appears as the self of duality from the root mind. The duality is utterly non-existent. A mere appearance (AkRtimAtraka) is there. 30. The root-consciousness is like the incantation. Suchness (tathatA) is understood to be like the piece of wood. Discriminative construction (vikalpa) should be accepted to be like the appearance of the elephant. The duality is like the elephant.

Vasubandhu’s independent verses appear to be a systematization of the Ornament passage, explaining it (nirde8a) by means of the three natures (trisvabhAva). The two texts quite evidently say very similar things about reality and about its likeness to a magical illusion. Nonetheless, the three natures are not the central rubric of the Ornament passage (especially if we focus on its verses without the commentary attributed to Vasubandhu). They only appear in the first verse (XI.15) above and even here the verse only mentions two of the three: the “constructed nature” and the “dependent nature.”What Vasubandhu calls the “perfected nature” in the Three Natures Exposition appears in the second verse of the Ornament section (XI.16) as “ultimate” (paramArtha) as opposed to “conventional truth” (samvRtisatya) – thus, under the rubric of the “two truths” instead of the “three natures.” In the third verse of the Ornament section (XI.17), which I will discuss below, there is no reference to the three natures per se, yet Vasubandhu’s independent verses will nonetheless reiterate the Ornament’s basic point through the rubric of the three natures. These substantial changes are key to understanding Vasubandhu’s distinctive philosophical position, but we need to work our way through the verses and develop the basics of the elephant analogy before we can see exactly what he is doing. The basic distinction made in the first Ornament verse above (XI.15) is, to me, one of the most important in Yog-c-ra philosophy, a distinction that once understood provides a foundation for understanding everything else that follows: it is the distinction between an appearance (AkRti) described causally (the dependent nature) and an appearance described as what it appears to be (the constructed nature). What appears to be there is (for instance) an elephant. From a causal perspective, however, we know that a magical illusion of an elephant comes about through various means – based on a piece of wood and a magician’s yantra or mantra. This is likened to the causal description of experience that is the dependent nature. From this perspective, we can speak of reality as “the construction of what does not exist.” The mind constructs an illusory appearance, called a “dual error” which “appears to be grasper and grasped” but is not, in fact, caused by the duality of sense organs and sense objects, to which these terms refer (we could say “subject and object,” as long as we remember that we’re referring to objects of sense such as appearances and tastes, not objects of the world, such as nebulae and gumdrops). Vasubandhu makes this point by saying that, rather than coming from “grasper and grasped,” the “self of duality” comes about “from the root mind.”The evident irony is that each mistaken self-apprehension appears as though based on the twosome of sense organs and their sensory objects, but in fact comes about through a single mental cause. Just what is that mental cause? This “root mind” or “root consciousness” is also often called the “storehouse consciousness” (Alaya-vijñAna). This concept has a complex pedigree in the early development of the Yog-c-ra, for which I will simply refer the reader to its detailed, divergent, analyses by Schmithausen (1987) and Waldron (2003). Perhaps the best way to put

YogAcAra Strategies against Realism . 5it is to use Vasubandhu’s analogy and say that the storehouse consciousness is the magician’s magical incantation (mantra) that brings about the illusion of reality as we experience it. More precisely, although, it is a subconscious “mental” storehouse that contains the karmic results of past deeds in the form of seeds of future experiences. Buddhists of course believe that beings are reincarnated after death in such a way as follows appropriately from their morally significant actions (karma). Yog-c-ra Buddhists believe that all of sentient beings’ experiences are unreal, merely experienced results of karmic acts. When the conditions are right, the seeds stored in the “root mind” come to fruition – the root grows and produces a fruit – in the form of a specific experience of reality commensurate with the past deed(s). The Yog-c-ra worldview describes all experiences as constituted by a stream of mental events with causes quite different from the (sensory) causal story of experience as we ordinarily conceive it. Readers familiar with The Matrix will not be misled by the comparison – although for the Yog-c-ra, our false reality is shaped by our own past actions, and not by an external, evil machine intelligence.

The crucial distinction between this causally “dependent nature” of reality and its “constructed nature” is likened to the distinction between a “magical illusion” and an “appearance of a magical illusion.” If there is an illusion, it must appear as something. Therefore, an inseparable part of the causal description of the illusion is the appearance itself, as it appears to exist: in this case, it appears to be a living, breathing elephant, “the self of an elephant.” What I wish to emphasize is that the constructed nature is an inseparable aspect of the dependent nature. The dependent nature could not be a “construction of what does not exist,” were there no “constructed” appearance that appears to exist. This appearance is called “the self of duality” – i.e. sensory reality in which I appear to exist, be in the world, have a self that is separate from the world, use my senses, etc. By making the constructed an aspect of the dependent – and analogically pointing out that every illusion has a particular appearance – the Yog-c-ra tradition provides an explanation of mere appearance (AkAramAtra) that is not reducible to a distinction between appearance and reality. This is a fact that is rarely understood, but is crucial to Vasubandhu’s three natures presentation, and is unambiguous once we attend to it. Neither the dependent nor the constructed nature needs to be thought of as more “real” than the other.

Both can be understood properly or improperly. My ordinary experience of reality is false, in the sense that my self and the world don’t exist as they appear to, but that same experience is real as an appearance. This is easily shown if we notice how Vasubandhu restates the distinction in the second Ornament verse (XI.16) between “ultimate” and “conventional” truth. Ultimate truth here refers to a particular aspect of the illusion, but one that has already been noted – specifically, the fact that it is illusory. We do not learn anything new about the dependent nature or the constructed nature when the Ornament defines the ultimate as “the non-existence of the

constructed which has a dual characteristic.” If we did not already understand that the constructed did not exist, we could hardly have understood the earlier definition of the dependent nature as “the construction of what does not exist.” The ultimate truth is therefore just another aspect of the dependent nature – in particular, the fact that the appearance that it constructs does not exist. This is why I describe the three natures as “vantage points” on one and the same reality.

It is only with the Ornament’s description of “conventional truth” that we finally come to something that is not in itself an inherent part of the dependent nature – that is, something unreal. It is, specifically, a deluded, relative perspective: “The nature of conventional truth is perceiving the construction of what does not exist.” Recall that the constructed nature itself, the duality, was likened to “the appearance of a magical illusion, in which the illusion’s appearance of an elephant, horse, gold, etc. appears to exist.” Here the conventional truth, what is ultimately false, is likened to “perceiving the appearance of the magical illusion as existing as an elephant, etc.” (emphasis added). The constructed nature is simply the misleading content. What defines the conventional truth and makes it something over and above the constructed nature of reality is that it is the perception of that content as real. This may seem like a minor terminological distinction, but it has consequences that become clear if we run the concepts through the analogy of the magical illusion. Knowing the erroneous content of an illusion – experiencing it – is not the same as perceiving, grasping or apprehending it as real. I can look at a mirage and see the appearance of water without being fooled into thinking that it is really water. The figure of the magical illusion, similarly, describes a real appearance, that is not what it appears, but can be correctly known. The elephant in the theater does not exist as it appears. There is no reason to call the police, or the zoo. But as an appearance the elephant is real – as real as the piece of wood! If there were no real appearance, the illusion would be a failure, the magician out of a job. Thus, the Yog-c-ra position retains the reality of the appearance, while claiming that the constructed appearance does not exist as it appears.

Vasubandhu’s Three Natures Exposition takes the two pairs of terms in the first two Ornament verses (dependent and constructed, ultimate and conventional), drops the fourth, and describes the first three as the three natures. Vasubandhu’s presentation of the three natures thus eliminates the “conventional,” which was simply the “perception” of the constructed as real. This purges from the discussion the only inherently unreal aspect of the “three natures” reality he is describing. It also eliminates the deluded perceiver from the description of the illusory elephant metaphor. Once conventional truth is eliminated from the discussion, all three natures can be described as legitimate, real and ultimately unified aspects of reality. Vasubandhu’s transformation of the illusory elephant analogy into a three natures analogy provides a brilliant opportunity to “speak” about © Blackwell Publishing 2006 Religion Compass 1 (2006): 10.1111/j.1749-8171.2006.00014.x YogAcAra Strategies against Realism . 7cognitive illusions without falling into the common trap of skeptics – namely, the use of language that is self-undermining. For without this transformation, one could ask as follows: If Vasubandhu is indeed stating that all of reality is an illusion, is not his own presentation simply part of that illusion? Is not the distinction between “ultimate” and “conventional” truth ultimately only conventional? Is not the distinction between a deluded subject and a true perceiver of reality “dualistic”? I submit that Vasubandhu eludes such criticisms by creating a “unity” out of the three natures.

Table 1 Parallel Terminology in the Ornament and the Three Natures Exposition Ornament XI.15–16 Three Natures Exposition Dependent nature Dependent nature Constructed nature Constructed nature Ultimate truth Perfected nature Conventional truth (perceiving the -x-constructed) This is best illustrated in the next four verses, when he turns to describe the “transformation of the basis” – the achievement of realization – without relying upon either of the “dualities” of conventional and ultimate or “subject” and “object” which are still present in the Ornament: (Ornament XI.17 with Vasubandhu’s commentary)5 17. Just as when it no longer exists, the visible appearance of its cause becomes manifest, So when there is a transformation of the basis, [the appearance of the cause] of the construction of what does not exist becomes manifest.

Just as when the magical illusion’s appearance no longer exists one perceives the existing thing (bhutArtha), the appearance of the piece of wood, etc. that [was] its cause, so when there is a transformation of the basis one perceives the existing thing (bhuto ’rtha) [that was the cause] of the construction of what does not exist, because the error of duality does not exist. (Three Natures Exposition 31–34)6 31. When there is understanding of how things really are (arthatattva), then at the same time there is knowledge, abandonment, and obtaining, [which relate] in proper order to the three characteristics. 32. It is acknowledged that here knowledge is non-perception; abandonment is non-appearance; and obtaining is direct and full awareness, apprehension without cause. 33. With the non-perception of duality, the dual appearance goes away; as a result of its removal one arrives at the perfected, which is the absence of duality. 34. So also in the case of the magical illusion: simultaneously there is the non-perception of the elephant, the removal of its (the elephant’s) appearance, and the apprehension of the piece of wood.

Here, instead of simply saying that at the moment of enlightenment “one perceives the existing thing (bhuto ’rtha),” Vasubandhu describes three purifying actions taken with respect to the three natures: knowing, abandoning, and obtaining. This description ingeniously avoids implying that there is a person who “perceives” incorrectly first, and then later “perceives” correctly. It avoids saying that there is a “real thing” that is correctly perceived, as opposed to the mere “visible appearance” that goes away. Of course, one can speak metaphorically – about “perceiving” the piece of wood, as the goal. But Vasubandhu’s Three Natures presentation eliminates the perceiver from the description of the three natures, and eliminates the subject from the final “apprehension” of ultimate reality – saying instead that it is “apprehension without cause.” And in place of the Ornament’s statement that “one perceives the existing thing (bhuto ’rtha) [that was the cause] of the construction of what does not exist,” he says, more in line with his view that enlightened knowledge is non-conceptual (nirvikalpa) and objectless, that “there is understanding of how things really are (arthatattva),” and that as a result of this “one arrives at the perfected, which is the absence of duality.” Subject and object are subsumed within the unified “three natures.” At the same time, the three natures view allows Vasubandhu to describe the illusions of ordinary reality without retaining the “duality” of ultimate truth vs. conventional truth. This is accomplished by embedding the two alternative perspectives – the constructed and the perfected – within the causal unity of the dependent nature. Of course, instead of two, we now have three natures – which hardly simplifies the two truths. Yet the simultaneity of the three natures – that each implies the others – allows for a kind of unification of dualities, making for a powerful weapon pointed at, in particular, the main dualistic targets of the Three Natures Exposition and this passage from the Ornament: appearance vs. reality, or existence vs. non-existence.

The magical elephant figure (and the theory of the three natures it elucidates) is thus very carefully designed to undermine our belief in the reasonableness of these commonsense distinctions. In short, these passages seem to argue for a view of reality as simply appearance, but not as opposed to reality. The only “reality” expressed here is the perfected nature, which is in fact the non-existence of what appears to exist – so, it is only “reality” within the illusion, and with respect to the illusion. The view of reality that we need to reject, on the other hand, is the view that we base on trust in our senses – namely, a view that what we experience is “dualistically” either real or unreal.

One important point to note is that there is a clear distinction here between the realm of the magical illusion and the realm of the three natures. Although readers often mistake the elephant illusion for an example of our ordinary deluded state, in fact it is only a figure. Of course, when an illusion appears we are fooled by it, and we imagine it to be real. But when YogAcAra Strategies against Realism . 9it goes away, we do not transcend “duality.”We are still a viewing subject, believing finally in the reality of a piece of wood, instead of an elephant. We are enlightened as to the illusory nature of the magical illusion of the elephant, but we are still deluded as to the figurative “magical illusion” of our own self. The realm in which we are deluded by a magician is basically a realistic realm, with a few mistaken perceptions – which is fundamentally incommensurate with the realm in which we are deluded by the root mind into believing in our own subjective reality. Such an interpretation seems confirmed by Vasubandhu’s removal of the viewing subject, the “perceiver” from the equation with respect to the transformation of the basis. When an enlightened Buddha engages in perception, there is no “perceiver,” there is only “apprehension without cause.” Vasubandhu likens the piece of wood that remains when the illusion disappears to the ineffable “suchness” (tathatA). Ultimately our reality is only like a magical illusion. If it were really one, then there would, indeed be a deluded viewing subject. But there is no subject, no self, and therefore, ultimately, no illusion.

Metaphor (upac-ra) as an Answer to Realist Philosophy of Language I said above that I took Vasubandhu’s “three natures” view to be relevant to contemporary issues in Western philosophy, and it is now possible to be more specific. Vasubandhu’s position should to be heard, for instance, in connection with the discussion that springs from one of philosopher Hilary Putnam’s most famous and well-read papers, entitled “Brains in a Vat” (Putnam 1981). In this paper, Putnam claimed to prove that, contrary to the worries of skeptics, you are not really just a brain in a vat connected to a computer that simulates your reality (think once more of The Matrix). This is Putnam’s imaginative variation on Descartes’ basic worry in the Meditations on First Philosophy that his experience of the world might be the result of the deceptions of “some evil genius” (Descartes 1996, p. 62). Putnam calls upon recent philosophy of language – specifically, his own causal theory of reference – to show that such radical skepticism is self-refuting.

Under a causal theory of reference, if you wish to refer to something, you need to be clear about just how that something might be causally connected, in a meaningful way, to your ability to utter its name. What Putnam says is that if you are a brain in a vat, you can’t refer to the vat you are in, because it is not a part of your world of experience. There is no proper causal connection between the real-world vat and your isolated brain-in-a-vat-world use of the word “vat” such as might allow you to refer to the real vat. All you can ever refer to with your use of the word “vat” is an image of a vat created by a computer. Therefore, whenever anyone says that he or she is or might be a brain in a vat, the statement fails to refer properly, and so cannot possibly be true as it stands.

As brilliant as this argument is (and I can only caricature it here, but see Ben-Menahem 2005 for a contemporary summary of Putnam’s views on

it tends to provide little comfort. It is difficult to shed the feeling that one is being subjected to a kind of sleight of hand. For those who are sufficiently skeptical as to take the possibility of the brain-in-a-vat scenario seriously (even without accepting the likelihood of vats per se), it is hard to escape the feeling that even if I can’t refer to my vat, I might be (am?) still in one. After all, Putnam’s scenario does not preclude my developing a machine that simulates a false reality plugged into someone else’s brain, such that they would become incapable of referring to real vats, even as I retain my proper belief in real-world reality. But if I could do this to someone else without their knowing it, it stands to reason that someone else could do this to me without my ever knowing it. What this means is that even without being able to say directly that I am a brain in a vat (because such words couldn’t refer properly), I can speak hypothetically of this happening, and I can also speak analogically. I can say that my reality today might be permeated with mistaken conceptions, and my language permeated with terms that fail to refer properly, in a way that is like someone else’s that is a brain in a vat. Such hypothetical and analogical talk would seem to be sufficient to return us to a radical scepticism, since it undermines our confidence in the reality we experience. Okay, such a sceptic might say, we’re not literally brains in vats, but we still might be in a brain-in-a-vat-like situation (a statement which, if we are to trust current psychological research, would be close to true – see, entertainingly, Gilbert 2006).

Of course, such a strategy undermines ordinary language, because it accepts that my best reference to reality might need to be analogical rather than direct. It will require work, too, because I must explain what it means to be referring analogically – generally considered a kind of “secondary” use of meaning – if I do not allow any possibility of a real,“primary” meaning.

But if I am a radical sceptic, it is conceivable that I might be willing to reconfigure my understanding of how language works, since I am already committed to rejecting a commonsense understanding of reality. Yuval Steinitz, a critic of Putnam, makes this point well when he says that Putnam’s brain-in-a-vat argument privileges realism when he assumes his causal theory of reference. A strict ontological skeptic, he says (using the Western idealist Berkeley as evidence), simply rejects the traditional referential causal story on the belief that “the best method of overcoming scepticism is to neutralize it by giving up realism and then readjusting the theory of reference accordingly.” (Steinitz 1994, p. 216.) Even purged by Steinitz’ critique, however, Putnam’s argument has highlighted an important connection between reference and realism. A causal theory of reference, in particular, does seem to imply at least a minimal degree of realism. Remember, a causal theory of reference requires that there be a clear causal story connecting your words to the objects to which they refer – and it would appear that non-realism at least potentially undermines that causal connection. Conversely, if you want to reject realism,

YogAcAra Strategies against Realism . 11you need to rework your theory of reference so that it has no such causal requirements for meaningfulness. Let’s call this very general observation the Putnam/Steinitz Hypothesis.

So, what does this all have to do with the Yog-c-ra view of the reality of appearances? As I understand it, Vasubandhu’s philosophy of language parallels his “three natures” view with a theory of reference that in fact lives up to the “reworking” requirements of the Putnam/Steinitz Hypothesis. That is to say,Vasubandhu is able to reject realism without giving up meaningful reference. The referential strategy that Vasubandhu employs appears at the start of his seminal Thirty Verses (Trim8ikAkArikA), and it is to declare that all language is figurative (upacAra).

Whereas we ordinarily think of language as literal, it cannot be, he says, since (given that all things are appearance only) there is no literal reality to which such words might refer. Instead, as we see in Vasubandhu’s opening verse, as interpreted by his great commentator Sthiramati (470– 550), all language refers figuratively to the transformation of consciousness – to the illusory play of mind itself: A varied figurative use of “self ” and “things” is what sets things going – that is to say, in the world and in treatises. It is with regard to the transformation of consciousness.7 What this means is that whenever we use terms that refer to either “self” or “things” (and in this system all terms do one or the other), we are in fact using figures – metaphors – wherein the illustrative terms are these self and things and the topic illustrated by the metaphor is the transformation of consciousness. What makes them say that words can only be only figurative?

Sthiramati’s commentary is very clear:“This is so because selves and things do not exist outside of the transformation of consciousness.”8 Since there are no real things (only illusions), all reference to things must be merely figurative – and to prove this Sthiramati need only restate the traditional Sanskrit grammarians’ definition of metaphorical reference: “[A word] is used figuratively with regard to something which is not there, as when [one calls] a Bahikan [[[person]]] an ox.”9 Clearly, even with a xenophobic remark such as this, the intention is not to say that that person is, literally, an ox. You say that, but there’s no real ox there. That is what makes the reference figurative. By this definition, then, any predication where the thing you are referring to is not there (in the locus of reference) is a metaphor (or a figure). In such a situation Sthiramati says that one is referring by way of “qualities only” – in the case of the “ox” figure, the quality is (perhaps) stupidity.10 Sthiramati is saying, though, that our words are just as figurative when we speak in ostensibly literal terms, as in the sentence, “There is an umbrella.”This is a linguistic view in which most of the time we literally do not know what we mean – but this is part of its impeccable internal logic.

Imagine you are a brain-in-a-vat and the computer makes you see an umbrella on the table in front of you. You do not know that it is not real, of course, and you say, “There’s an umbrella.” What Sthiramati is saying is that whether you like it or not, since there is no umbrella there, you can only really refer to the qualities of the umbrella that you experience, and not its “real” identity. So what your statement means is something like, “Umbrella-like qualities are present.” If Sthiramati really believed in the brain-in-a-vat scenario, he would say that your sentence figuratively expresses something like,“The computer is producing umbrella-like qualities, making it appear to my brain as though there is an umbrella.” What he actually believes is that the basis of experience is the transformation of consciousness alone. So this sentence is saying something like, “The transformation of consciousness is in a state such that it’s like there’s an umbrella.” Such a theory implicitly accepts Putnam’s so-called linguistic “exter-nalism,” which understands our words to have meanings that are not determined exclusively by our beliefs and intentions (so we can be wrong about what we mean). It also seems to share Putnam’s belief that reference applies only to what is real (which is why self and things cannot be objects of reference) – although of course it differs from Putnam on what is included in the real (basically, only mental qualities attributable to the transformation of consciousness). At the same time, however, this view reasserts the meaningfulness of a secondary form of reference – metaphor – which takes advantage of the implicit “reality” of qualities-only discourse.

This is how I take Sthiramati to be, in a sense, answering Putnam (more below). My point here, then, is not merely to bridge East and West in an abstruse area of philosophy, or to suggest that these ingenious Buddhist intellectuals might appear of greater consequence because they have stumbled upon and dealt with a doctrine articulated by a prominent Western thinker such as Putnam. These are rather inappropriate reasons for studying Buddhist philosophy and engaging in comparison – in the first case, useless and arbitrary; in the second, ethnocentric and paternalistic.

Instead, I believe that the Putnam/Steinitz Hypothesis is almost a matter of common sense (which is perhaps why Steinitz considers Putnam’s analysis “disingenuous in places,” 1994, p. 213), and viewed as such, it can be seen to sit unvoiced behind a range of developments in Buddhist philosophy. To look at Vasubandhu and Sthiramati from this perspective highlights what I take to be common philosophical instincts across intellectual traditions. If the Putnam/Steinitz Hypothesis seems obscure, perhaps it is only because we are not accustomed to a philosophical world in which non-realism – especially non-theistic non-realism – is taken seriously at the same time as a causal theory of reference. But since Buddhist philosophers do commonly hold these views, we have a right to expect that they will be under regular threat from the Putnam/Steinitz Hypothesis challenge. This assumption, I submit, turns out to be exegetically quite fruitful.

YogAcAra Strategies against Realism . 13From this perspective,Vasubandhu’s otherwise confounding declaration that all language is metaphorical (confounding because it might seem unnecessarily defeatist) comes to take on a brilliance fitting to his philosophical greatness by positing a theory of reference that parallels his distinctive Yog-c-ra “three natures” ontology. This view quite elegantly enacts the “analogical” strategy that I mentioned above as a response to the realist challenge of the Putnam/ Steinitz Hypothesis. To say that reality is an illusion would be self-defeating, since the statement itself would have no true (non-illusory) reference; but Vasubandhu says instead that reality is like an illusion, and accepts that the only true reference will have to be metaphorical. In fact, properly understood, this linguistic theory also circumvents the criticism that appeals to “primary” and “secondary” meanings implicit within talk of “metaphor.” For, as Sthiramati argues, all language must operate by appeal to experienced “secondary” qualities, and never by direct apprehension of some original “primary” substance.

But how can there be secondary qualities without primary references in which the qualities inhere? Recall that, according to Sthiramati, to say that all language should be understood as essentially figurative (upacAra) is to say that it operates by way of qualities alone. The things that seem to have qualities do not exist as they appear. Their apparent qualities reside only in the transformation of consciousness itself. Notice, now, that this theory reiterates the relation between the dualistic appearances (AkRti) that the dependent nature produces, and the causal reality out of which, and in which, that appearance resides. Just as the non-reality of the existence of the elephant does not prevent the reality of the elephant’s appearance, so also the failure of direct reference does not devalue or undermine the reality of the expressed qualities themselves. Figurative meaning is simply a linguistic version of the three-nature theory of appearances. We have no need of a real umbrella if we know that “umbrella” refers to a set of experiences from the past. We refer to “umbrellas” just as easily as we refer to (for instance) our hunger – which we name correctly based on experienced qualities alone, not on some real object in which those qualities inhere.

For Vasubandhu and Sthiramati, all words then refer to the transformation of consciousness itself. These thinkers might say to Putnam, then, that if we are brains in a vat then all of our words must refer directly to the brain in a vat itself – the ultimate reality – and only metaphorically predicate of that brain-in-a-vat the presence of whatever appearance appears. Such a response might initially appear to support, rather than undermine, Putnam’s case; the trapped brain cannot even refer beyond its own stream of consciousness. Yet their problem with Putnam would not be his theory of the limitations of reference, but his decision to rather arbitrarily reject the brain’s isolation on the basis of the need for true direct reference – especially when the nature of reality itself is in question. That is, Putnam bases his ontological claims on the referential needs of his philosophy of language, and he assumes that we cannot be really engaging in a conversation at all without true direct reference.

Yet if the whole discussion was supposed to begin with skepticism, then (with Steinitz) Vasubandhu and Sthiramati might wish to suspend the decision as to the true nature of linguistic entities right along with the ontological ones. They simply rework the theory of reference in line with their ontology, and not the other way around. The meaningfulness of language does not need to rely on the reality of the objects of linguistic reference, because just as reality itself is neither real nor unreal, but only appearance, terms need no “reality” to refer meaningfully as appearances. The Yog-c-ra position would thus be to accept no terminology that determines the reality or the unreality of the objects of perception and reference – for all that is real in things, they say, is the appearance of things as if real.

The following table places the terminology from this paper under such a rubric.

Table 2 Distinguishing Between Dualism and Non-Dualism Terms that decide about actuality Terms that suspend the decision Constructed/perfected Dependent Reality/illusion Appearance (Akrti) Elephant/wood How the elephant appears Example Figure/metaphor (upacAra) Thing/transformation of consciousness Quality Terms in the first column are “dualistic” in that they take sides – they determine and value the reality of the appearance. Terms in the second column allow for a non-dual conception of appearance as mere appearance, although not thereby necessarily unreal. I began this paper by suggesting that non-Buddhist traditions of thought might have something to learn from the results of centuries of creative energies directed by Buddhist philosophers against a recurring self-referential problem – namely, the basic Buddhist belief that until enlightenment we are all deluded in our most fundamental concepts and ideas. Philosophical traditions of course wish to declare the appropriate doctrines in the appropriate forms; yet Buddhists who do so are subject to the criticism that their own doctrines suggest that anything in language is probably wrong. Many Buddhist thinkers have had to deal with this issue, though here is not the place to discuss them. One early Yog-c-ra response, as I read it through Table 2, is to define terminology that allows one to remain suspended between “dualistic” and therefore false opposites, either of which would be the horn of a dilemma. To suspend the decision about reality, they claim, is a necessary first step toward recognizing how things really are, and might someday keep these dilemmas from ever appearing again. Acknowledgement I am grateful to Daniel Arnold, Mario D’Amato, Paul J. Griffiths, and Parimal Patil, all of whom made extremely helpful comments on an early version

YogAcAra Strategies against Realism . 15of part of this paper. I am especially grateful to Paul J. Griffiths for introducing me to Yog-c-ra, and for allowing me to study and use his draft translations of the TrisvabhAvanirde8a and the Trim8ikAbhArya. All errors that remain are, of course, my own. Short Biography Jonathan C. Gold (BA, Amherst College; MA and PhD, University of Chicago) is Assistant Professor of Asian Religions in the Department of Religion at Drew University in Madison, New Jersey. Before arriving at Drew, Gold taught at the University of Vermont (Burlington, VT). He specializes in Indian and Tibetan Buddhist intellectual traditions, with a focus on theories of language and learning. His research attempts to bring an understanding of Buddhist thought and its insights for comparative religions to bear on current issues in Western thought. His book The Dharma’s Gatekeepers: Sakya Pandita on Buddhist Scholarship in Tibet (in press at State University of New York Press) addresses issues of central intellectual concern to today’s academic community – including translation and translatability, theories of reading and authorship, the connections between religious values and academic institutions, and theories of language and literary aesthetics – from the unique perspective of a thirteenth-century Buddhist philosopher. Gold’s research has also appeared in Asian Philosophy and the Journal of Indian Philosophy.

Notes

• Correspondence address: Department of Religious Studies, Faulkner House, Drew University, Madison, NJ 07940, USA. 1 Dates are from Nakamura 1980 (pp. 264–286), where the reader can also find references to all of the major studies of the Indian works mentioned in this paper and both major and minor works in Japanese, up to its publication date. 2 The scholar Charles Muller maintains a useful “Bibliography of Yog-c-ra Studies” on his own website: http://www.acmuller.net/yogacara/bibliography/yogacara-bib.html (checked August 31, 2006). Lusthaus 2002, in particular, provides quite a different presentation of the issues discussed here. 3 Sanskrit source in Lévi 1907, pp. 59.2–17. 4 Sanskrit source in La Vallée Poussin 1933, p. 156. 5 Sanskrit source in Lévi 1907, pp. 59.2–17. 6 Sanskrit source in La Vallée Poussin 1933, p. 156. 7 Lévi 1925, p. 15: AtmadharmopacAro hi vividho ya[[[pravartate]] / loka8Astrayor iti vAkya8era[ / vijñAnaparinAme ’sau// 8 Lévi 1925, pp. 15–16: dharmAnAmAtmanar ca vijñAnaparinAmAd [16] bahir abhAvAt/ 9 Lévi 1925, p. 16: yac ca yatra nAsti tat tatropacaryate/ tad yathA bahike gau[/ 10 Lévi 1925, p. 17: sarva evAyam gauna eva na mukhyo ‘sti/ (“All [[[Wikipedia:expressions|expressions]]] are related to qualities only, not literal meaning.”)

References

Anacker, S, 1998, Seven Works of Vasubandhu, the Buddhist Psychological Doctor, Corrected edition, Motilal Banarsidass, Delhi, India. 16 . Jonathan C. Gold © Blackwell Publishing 2006 Religion Compass 1 (2006): 10.1111/j.1749-8171.2006.00014.xBen-Menahem, Y, 2005, “Putnam on Skepticism,” in Y Ben-Menahem (ed.), Hilary Putnam, pp. 125–55, Cambridge University Press, Cambridge, UK. Bronkhorst, J, 2001, “The Peacock’s Egg: BhartRhari on Language and Reality,” Philosophy East and West, vol. 51, no. 4, pp. 474–91. Descartes, R, 1996, Meditations on First Philosophy, in D Weissman (ed.), Discourse on the Method and Meditations on First Philosophy,Yale University Press, New Haven, CT.

Garfield, J, 2002, Empty Words: Buddhist Philosophy and Cross-Cultural Interpretation, Oxford University Press, New York. Gilbert, D, 2006, Stumbling on Happiness, Knopf, New York. Gold, JC, 2006, “No Outside, No Inside: Duality, Reality and Vasubandhu’s Illusory Elephant,” Asian Philosophy, vol. 16, pp. 1–38. Kaplan, S, 1990, “A Holographic Alternative to a Traditional Yog-c-ra Simile: An Analysis of Vasubandhu’s Trisvabh-va Doctrine,” The Eastern Buddhist, vol. 23, no. 1, pp. 56–78.

Kochumuttom, TA, 1989, A Buddhist Doctrine of Experience: A New Translation and Interpretation of the Works of Vasubandhu the YogAcArin, Motilal Banarsidass, Delhi, India. de La Vallée Poussin, L, 1933, “Le petit traité de Vasubandhu-Nagarjuna sur les trios natures,” Mélanges chinois et bouddhiques, vol. 2, pp. 147–61.

Lévi, S, 1907 –11, MahAyAna-SutrAlamkAra: exposé de la doctrine du Grand Véhicule selon le système YogAcAra, H. Champion, Paris. Lusthaus, D, 2002, Buddhist Phenomenology: A Philosophical Investigation of YogAcAra Philosophy and the Ch’eng Wei-shih lun, RoutledgeCurzon, New York. Nagao, GM, 1991, Madhyamaka and YogAcAra: A Study of MahAyAna Philosophies: Collected Papers of G.M. Nagao, LS Kawamura and GM Nagao (eds.), State University of New York, Albany, NY.

Nakamura, H, 1980,Indian Buddhism: A Survey with Bibliographical Notes, KUFS Publication, Osaka, Japan. Putnam, H, 1981,“Brains in a Vat,” in Reason, Truth and History, pp. 1–21, Cambridge University Press, Cambridge, UK. Schmithausen, L, 1987, ?layavijñAna: On the Origin and the Early Development of a Central Concept of YogAcAra Philosophy, International Institute for Buddhist Studies,Tokyo, Japan.

Steinitz,Y, 1994, “Brains in a vat: different perspectives,” The Philosophical Quarterly, vol. 44, no. 175, pp. 213–22. Tola, F & Dragonetti, C, 1983,“The Trisvabh-vak-rik- of Vasubandhu,” Journal of Indian Philosophy, vol. 11, pp. 225–66. Waldron,WS, 2003, The Buddhist Unconscious: The ?laya-vijñAna in the Context of Indian Buddhist Thought, RoutledgeCurzon, New York. Wood, TE, 1991, Mind Only: A Philosophical and Doctrinal Analysis of VijñAnavAda, University of Hawaii Press, Honolulu, HI.