IDEALISM AND YOGACARA BUDDHISM

IDEALISM AND YOGACARA BUDDHISM

Over the last several years, there has been a growing controversy about whether Yogacara Buddhism can be said to be idealist in some sense, as used to be commonly thought by earlier scholars. On one side of the debate, writers such as Jay Garfield, Jeffrey Hopkins, Paul Williams, and others maintain the idealism label, while on the other side, Stefan Anacker, Dan Lusthaus, Richard King, Thomas Kochumuttom, Alex Wayman, Janice Dean Willis, and others have argued that Yogacara is not idealist.

In this short essay, I shall first try to clarify the different senses of idealism that might be pertinent to the debate. I shall then focus on some of the works of Vasubandhu, limiting myself to his Vimsatika, Trimsika, and Trisvabhavanirdesa. My attempt will be to try to suggest that classical Yogacara Buddhism, at least as found in these works of Vasubandhu, is closer to epistemic rather than metaphysical idealism or absolute idealism, as I understand these. However, there are undoubtedly some important differences between Vasubandhu’s Yogacara and Western epistemic idealism that cast doubt on the idea that Yogacara is simply to be lumped together with Western epistemic idealism.

I. Different Senses of Idealism

One of the most striking things in the debate about whether Yogacara is in fact idealist is that few writers are clear about the different senses of idealism, and few agree about what sense of idealism is said by some to apply to Yogacara. As Mario D’Amato has recently pointed out, there is need to define idealism in this debate so that the terms of the debate will be clear.

I want to distinguish here very briefly between at least three senses of idealism: (i) metaphysical idealism; (ii) epistemic or epistemological idealism; and (iii) absolute idealism. Metaphysical idealism is the view that the external world and objects in it exist in a mind-dependent way, and it is opposed to metaphysical realism, the view that the world and material objects in it exist in a mind-independent, objective way. I take Berkeley to be the prime example of a Western metaphysical idealist. I cannot outline Berkeley’s views in detail here, but, in a nutshell, Berkeley claims that material objects are nothing but collections of mind-dependent ideas or secondary qualities, the existence of which consists in their being perceived, their essence being what is perceived. There is a further division amongst metaphysical idealists between subjective idealists, who claim that things depend for their existence on my mind, and objective idealists such as Berkeley, who claim that the existence of things depends not just on my mind or your mind but on a more objective or greater mind, God’s mind, as Berkeley claims. Moreover, some metaphysical idealists are solipsists who are extreme subjective idealists and claim that the only thing that exists is me, my mind, and ideas and things in it.

Epistemic idealism, in contrast, makes not an ontological claim but rather the claim that we know things not as they really are, as claim epistemic realists, but rather as they are given to us by our ideas, our concepts, and categories. To put things differently, epistemic idealists claim we know things not as they are but rather as we are. Kant is the prime example of a Western epistemic idealist, and I cannot here present his position in great detail. Very briefly, however, Kant claims that things-in-themselves, the noumena, are forever beyond our ken, and things as they appear to us, the phenomena, are given to us as conditioned by our concepts and categories, which we cannot step out of. To use a well-worn analogy, it is almost as if our knowledge and perception are conditioned and framed by unremovable conceptual glasses tied to our heads. It is interesting to note that one could be a metaphysical realist and an epistemic idealist; I take Kant to be one. One could also be an epistemic idealist, while being above the battle, so to speak, in the debate between metaphysical realism and metaphysical idealism. Such a metaphysically agnostic reluctance to make ontological pronouncements, combined with something like epistemic idealism, just might be Vasubandhu’s position if he doubts not external objects themselves but externality, that is if he doubts not external objects but instead whether our ordinary consciousness can say anything about objects outside its acts of cognizing them.

Absolute idealism, very roughly, is the view that what exists ultimately (and, in some versions of absolute idealism, creates all that exists) is one overarching mentalistic or spiritual thing or principle or force, whether the Absolute or Mind or Brahman. Advaita Vedanta as well as Hegel and later Hegelians such as Bradley provide examples of this variety of idealism.

Having clarified these three senses of idealism, let us return to the debate between Yogacara and idealism. While I cannot here offer a comprehensive overview of what sense of idealism different authors impute or deny to Yogacara in the debate about Yogacara and idealism, it is important to note that different authors disagree about what sense Yogacara might be said to be idealist in. For example, Garfield sees Yogacara as metaphysically idealist at least as does Ornan Rotem, while Williams sees Yogacara in terms closer to absolute idealism. Anacker, Lusthaus, and Willis deny Yogacara is absolute idealist and Lusthaus also denies Yogacara is solipsist; Lusthaus also claims that Yogacara is closer to what I have called epistemic idealism and perhaps closest even if not identical, claims Lusthaus, to Western phenomenologists such as Husserl, a claim I will not delve into here. Interestingly, Lusthaus also understands idealism in the sense of that term in which it is opposed not to realism which is the contrast that I have mainly been concerned with, but rather to materialism, the view that all that exists is ultimately constituted only by matter or stuff. Idealism in this anti-materialist or immaterialist sense, a fourth sense of idealism, is the view that all that exists is ultimately only made up of minds or ideas or something mentalistic, and Lusthaus denies that Yogacara is idealist in this sense. Note that one could, of course, be idealist in both the anti-realist (or irrealist) and the anti-materialist (or immaterialist) senses, and Berkeley is, once again, a good example of this.

II. Vasubandhu’s Yogacara

I now turn to explaining what I take to be the main claims in Vasubandhu’s formulation of Yogacara. Though I do not offer here a translation of and commentary on the well-known Vimsatika, Trimsika, and Trisvabhavanirdesa, I shall first try to summarize what I take to be Vasubandhu’s main claims in these works. All too often, scholars and Buddhologists get caught up in the details of Yogacara texts and their translation and thus miss the forest for the trees, and I hope that what I offer next is a broader overview of Vasubandhu’s views. I shall then look closely at specific passages from these three works of Vasubandhu.

As I see it, Vasubandhu’s classical Yogacara emphasizes the use of Yoga and meditation, hence the name Yogacara, meaning the practice or conduct of yoga. This use of meditation is meant to overcome our ordinary, dualistic consciousness that imposes subject-object and other distorting dualities on the world, dualities which prevent us from seeing reality as it is. Yogacara wants to understand our ordinary consciousness, which conditions all our ordinary knowledge and perception with its subject-object dualities, and stresses the use of meditation to transcend our dualistic consciousness so that we see things as they really are. To borrow and vary somewhat Dan Lusthaus’ memorable formulation, for Yogacara, our ordinary consciousness is not the ultimate reality or solution, but rather the root problem. One might thus see Vasubandhu’s Yogacara teachings as providing some sort of preparatory school before one practices meditation, and indeed as providing the theoretical framework and rationale behind the practice of meditation. As Alex Wayman puts it, “Yogacara put its trust in the subjective search for truth by way of a samadhi. This rendered the external world not less real, but less valuable as the way of finding truth.”

To see things as they really are, claims Vasubandhu, is to see them in meditation without the distorting dualistic mentations and imputations of our ordinary consciousness. Instead, they are seen in an ineffable meditative experience as being dependent and always changing, as being part of a flow of things that have no essences or fixed natures or own-beings (nihsvabhava). This, I suggest, is what Vasubandhu means by the claim that the perfected or fulfilled aspect (parinispannasvabhava) of things is their dependent aspect (paratantrasvabhava) without the imagined or constructed aspect (parikalpitasvabhava). Indeed, such an understanding of Vasubandhu makes his Yogacara continuous with Nagarjuna’s Madhyamaka school and with earlier Buddhism, which stress the basic Buddhist insight of dependent origination. This is not to deny that there are important differences between Yogacara and Madhyamaka, most notably in that while Madhyamaka sees emptiness as absence of inherent or independent existence, Yogacara sees emptiness as absence of subject-object dualities and emphasizes the use of meditation to see this. One might say, then, that Vasubandhu’s Yogacara extends some of the basic insights of Madhyamaka, and adds its own distinctive teachings to them, just as other schools of Indian thought (e.g. Vedanta) often expound old teachings in new ways and through new interpretations. Moreover, it is possible to argue, though I shall not do so here, that absence of subject-object dualities (is a metaphor that) means that one should not see the subject or knower and the object or known as two (radically) separate, inherently existing independent entities instead of as interacting, interdependent entities, and thus as non-dual and “one”, in some sense.

It might be objected here that absence of subject-object dualities means only that the object vanishes in the transcendental awareness of meditation, wherein one has objectless knowledge whereby the object no longer appears as an object set against an experiencing subject. But against this, it must asked where specifically Vasubandhu says the object vanishes, and one must also wonder if to say that the object vanishes and (presumably) the only thing that is found to exist in meditation, as Williams claims, is Mind (or Subject) is not a claim closer to absolute idealism rather than the Berkeley-style straightforward metaphysical idealism that Garfield, for one, sees Vasubandhu as espousing. Moreover, even if it be claimed in response here that Vasubandhu’s view is metaphysically idealist in that the existence of an object is ontologically dependent upon a deluded subject, I must point out that such a view is not metaphysical idealism, or at least not metaphysical idealism as found in Berkeley. For Berkeley does not appeal to deluded subjects nor claim that the existence of objects depends on ordinary, deluded subjects such as ourselves as opposed to being dependent on the mind and perception, ultimately, of God. Instead, such an understanding of Vasubandhu - which I take to mean that our ordinary consciousness cannot reveal the nature of things and instead reifies them in a deluded manner as substantialist objects - is in some ways closer to Kant’s epistemic idealism, though it also differs from Kant in the ways I note at the end of this paper. Moreover, “the object vanishes” might only mean that objects do not appear in meditation as they appear to our ordinary consciousness, as reified and substantialist, not that they do not exist at all.

To return to Yogacara, why is this school of Mahayana Buddhism also sometimes called Cittamatra, which means Mind-Only or Consciousness-Only, and suggests to many that the school is idealist in some sense of that term? I suggest that this is because this school of thought, as found in Vasubandhu, claims that representations of things, as given to us by our minds, are merely that, to wit, representations only or cognitive representations only or perceptions only or conceptualizations only (vijnapti-matra). Things as they appear to us, and as they are experienced by us, are due to the dualistic constructions and imputations of the mind or consciousness alone (cittamatra). What things are really like is something that we cannot know through our dualistic ordinary consciousness alone, which we must transcend through meditation. External things and externality is something the true nature of which is not revealed by our ordinary consciousness, which only gives us a picture of things as shaped by and limited to our dualistic mental constructions.

But how about the standard textbook Yogacara notions of the eight consciousnesses and their role in Vasubandhu’s Yogacara? The first six consciousnesses, including the five external senses and the mind, and the seventh consciousness, the tainted mind (klista-manas) which imposes distorting subject-object dualities on things, are all intentional and directed outwards. The eighth consciousness, the storehouse consciousness (alaya-vijnana), is non-intentional and is the underlying repository which contains various karmic influences as well as the seeds (bijas) of future actions as well as experiences. One might wonder here where the very first seeds come from if not from an experience of something external, and even if it be said in reply that all ordinary experiences come from the defiled seeds which the storehouse consciousness has always been full of, that still does not show objects do not exist, as an idealist interpretation of Vasubandhu might claim. For not only does one still need textual support in Vasubandhu for the claim that objects do not exist at all, but also, more importantly, while reified, substantialist objects and our ordinary experiences of them may come from the seeds, that shows at best that these constructions are the creations of our ordinary consciousness, not that objects themselves must be identical with such imagined constructions or be otherwise non-existent.

With that brief summation of Vasubandhu behind us, I turn to crucial passages in the Vimsatika, using Anacker’s translation throughout (unless specified otherwise), with some changes of my own:

All this is consciousness-only, because of the appearance of non-existent objects, just as someone with an optical disorder may see non-existent nets of hair. (Verse 1)

Here, for comparison, is Alex Wayman’s translation of this opening verse:

This just amounts to representation, as the sight of unreal hair, moon etc. of one with an eye-caul - because being the (subsequent) manifestation of an unreal artha (external thing).

This opening verse suggests, as I take it, that all phenomena as they appear to us, as we know and experience them, are due to the representations and constructions of our consciousness alone (vijnapti-matra). It does not mean, as is sometimes thought, that external objects themselves are created by the mind or are somehow mind-dependent. For while what is experienced in an optical disorder or illusion or a mirage may be created by our consciousness, the content of what is given to us in such experiences by our ordinary consciousness and its constructions does not say anything about things themselves. It is important to note here that the phrase “appearance of non-existent objects” (avabhaasanaat) in Anacker’s translation of this opening verse should not be taken literally or out of context to mean that objects themselves are non-existent. Instead, it means only that objects do not exist as they appear to us, as given to us by our representations. To see this, consider again the simile with optical disorders (or eye-cauls or cataracts). The one with optical disorders sees non-existent hairs and moons, but that does not mean that outside her acts of perception, there are no hairs and moons at all. Likewise, I suggest Vasubandhu is saying that our ordinary consciousness represents objects to us in a (reified, substantialist) manner in which they do not exist, but that need not mean that objects do not exist at all outside our ordinary acts of perception.

...the teaching of perception-only is entry into the selflessness of events, when it becomes known that this perception-only of visibles etc. arise, and that there is no experienced event with the characteristics of visibles etc......But it’s not because there isn’t an event in any way that there is entry into the selflessness of events...through this determination of perception-only, there is entry into the selflessness of all events, and not by a denial of their existence. (Vasubandhu’s Commentary on Verse 10; italics mine)

Here I take it Vasubandhu is saying that all experienced events and appearances as experienced are shaped by ordinary consciousness only. This Yogacara teaching is one way to convey the fundamental Buddhist idea of dependent origination as applied to events, i.e. that events are not independently existing entities with a fixed essence but instead are dependent and always changing. It is important to note that Vasubandhu is not denying the existence of events, as metaphysical idealists as well as absolute idealists might do. Note also that Vasubandhu is not denying events the way an anti-materialist idealist might do in saying they are solely mental creations. Also, while Vasubandhu is clearly trying to refute atomism in the Vimsatika, he may well be doing this to refute the atomism of the contemporaneous Hindu school of Nyaya-Vaisesika rather than to deny matter. For to deny atomism, one need not deny matter, as Vasubandhu must have been smart enough to realize, and while the text clearly says (e.g. in Verse 11) that atoms cannot be demonstrated, it does not say that matter does not exist, which is what an idealist interpretation of Vasubandhu would need to show. It might be said in response here that verse 10 of the Vimsatika has a statement that there are no objects, so let us look at that verse, first in Sanskrit and then in English:

Tathaa pudgalanairaatmyapravesho hi anyathaa punaha Deshnaa dharmanairaatmyapraveshaha kalpitaatmanaa

And now here is the English from Anacker, with some changes of mine:

And in this way, there is entry into the selflessness of personality And in another way, this instruction is entry into the selflessness of events in regard to a constructed self.

Instead of finding a statement here that there are no objects, all this verse says, as I understand it, is that the self, events, and so forth are not substantialist. And it might also be said in response that verse 16 of this work by Vasubandhu claims that perception makes no sense because it has no object at all, so let us look at the Sanskrit and then the English of this verse as well:

Pratyakshabuddhihi swapnadau yathaa sa cha yadaa tadaa Na sohortho drishyate tasya pratyakshatvam katham matam

Here is the English for this verse from Anacker, with some changes of my own:

Cognizing by direct perception is like in a dream, and so on And when it happens, the object is not seen, so how can it be called direct perception?

As I understand it, this verse is saying only that we do not ordinarily and directly see objects as they really are but instead perceive directly only a dream-like constructed or imagined object. This does not mean that perception has no object or that objects do not exist, but only that the object of our ordinary, direct perception is something constructed by our ordinary consciousness. The object of our ordinary perception is in fact a reified, substantialist object as presented to us by our ordinary consciousness and the constructed aspect of things. I turn now to conduct a similar exercise with the Trimsika, picking out some crucial passages in that work that cast doubt on the idea that Vasubandhu is a metaphysical idealist a la Berkeley as well as on the idea that Vasubandhu is an absolute idealist in the manner of an Advaita Vedantin or a Hegelian:

Whatever is discriminated is just a constructed essential nature and doesn’t really exist. The dependent underlies these constructions, and the perfected is the dependent without the constructed. (Verses 20-1)

Here I take Vasubandhu to be saying that our conceptual imputations and constructions are just imagined and need not have a basis in reality, which is always in flux. To see reality as it is, i.e. to experience the fulfilled or perfected aspect of things in meditation, is to see things as being dependent and in flux, without imposing dualistic constructions on them. It is very important to note here that Vasubandhu is not denying that reality exists, as an idealist might do, but instead is only denying that our conceptual constructions, as presented to us, correspond to something out there. A simple and familiar example will make this point clear and indeed show some of the continuities, despite other differences, between Vasubandhu and earlier Buddhists including Nagarjuna.

Take the notion of Atman or a permanent, separate self found in Hindu philosophy and criticized heavily by Buddhists over the ages, beginning with the Buddha himself. Suppose you believe you have an Atman. Vasubandhu would say, I suggest, that the notion of an Atman is only an imagined conceptual construction, with nothing corresponding to it out there. However, this does not mean that you do not have a self at all out there that other selves such as mine can interact with and discuss Buddhist philosophy with, say. Rather, Vasubandhu would say that your self is a dependent, ever-changing thing, dependent on its

five psychophysical aggregates (skandha) or processes as well as on other things. You will see this dependent, fluctuating nature of your self in meditation, when you will be rid of your distorting conceptual constructions such as the notion of an Atman. I suggest similar claims would apply to other things. Vasubandhu would not deny their existence, as idealists might do, but instead claim they are revealed in meditation to be dependent entities in flux, not reified substantialist entities as presented to us by the constructions of our ordinary consciousness.

Let us now look at the Trisvabhavanirdesa to see if we can find any crucial passages that shed light on the idealism debate concerning Vasubandhu’s Yogacara:

The constructed, the dependent, and the perfected These three own-beings are the most profound things known by the knowledgeable. That which appears is the dependent, how it appears is the constructed The former develops subject to conditions, the latter is construction-only. (Verses 1-2)

Here I take Vasubandhu to identify the three aspects of things straightforwardly in the opening verse. The second verse (which I will return to below) characterizes the nature of reality in the familiar Buddhist terms of dependent origination and flux, while how reality appears to us, how it is represented to us by our cognition, is purely constructed. Note again that Vasubandhu does not deny that reality exists. Instead the only thing that is constructed or imaginary, according to Vasubandhu, is how reality is presented to us, how we experience it ordinarily, that is, as reified.

The perfected is the absence of the constructed in the dependent. What appears? A construction of what was not. How does it appear? Through dualities. But non-duality is the perfected. What is a construction of what was not? An awareness by which the dependent becomes constructed in such a way that the object which it constructs cannot be completely found in that way. (Verses 3-5; italics mine)

Vasubandhu here stresses ideas presented above in my brief discussion of verses 20-1 from the Trimsika. Constructed dualities are only apparent, and things do not exist in the way they are presented to us by the constructions of our cognition. Note that he is not denying the existence of things, nor is he is saying that no objects we experience exist outside our minds. Instead he is only denying that they exist in the manner constructed and given to us by our ordinary consciousness. For Vasubandhu, things may be presented to us in certain ways by our consciousness but no matter how they are presented, they exist as dependent, dependent at least on other things even if not necessarily on the mind. Indeed, such an understanding of things as existing and dependent is continuous with the dependent origination, anti-eternalism, and anti-nihilism found in Buddhist thinkers, including Nagarjuna, before Vasubandhu.

The constructed is perceived as existent, but it is non-being; so it has both an existent and a non-existent character. The dependent exists but not as it appears, so it also has an existent and a non-existent character. (Verses 11-12; italics mine)

Again, Vasubandhu is not denying the existence of things, but is only claiming that they do not exist in the dualistic manner in which our constructions present them. Indeed, things exist as being part of a dependent flow, a claim continuous with fundamental Buddhist ideas.

III. Is Vasubandhu’s Yogacara Idealist?

In this section, I will look closely at arguments and passages cited in support of the claim that Vasubandhu’s Yogacara is idealist in some sense. For several reasons, I will limit myself here to claims made in this vein by Garfield. Why Garfield? For starters, Garfield says explicitly that Yogacara is metaphysically idealist a la Berkeley, which view is my main target, whereas others see Yogacara as idealist in different senses, as explained briefly in the first section of this paper. Moreover, I focus on Garfield also as he is a contemporary analytic philosopher, as am I, and a very fine one at that who has made significant contributions to Buddhist philosophy. Also, I cannot do a comprehensive literature survey (which is not my purpose anyway) in a short paper such as this one, and singling out Garfield gives the paper a tight focus, I believe. I do not mean any disrespect to those who I cannot discuss at length here.

To begin with, Garfield writes that the Trisvabhavanirdesa “is an idealistic treatise. As far as Vasubandhu is concerned, to be a phenomenon is to be an object of a mind.” Now phenomena or things as they appear to us, and as they are experienced ordinarily, as reified and substantialized, are certainly mental objects for Vasubandhu, represented to us by the mind’s constructions. But, contra Garfield, it is not clear that Vasubandhu would say that objects themselves are phenomena only, as an idealist might say. As Richard King points out, for Vasubandhu we cannot go from phenomena to externality, for objects may or may not exist independent of the mind. This makes Vasubandhu’s view one that would have no ontological commitments such as metaphysical realism or metaphysical idealism at least in the domain of our ordinary consciousness. Rather, talk about external objects would not be appropriate on this understanding of Vasubandhu, for it goes beyond what is experientially given at least to our ordinary consciousness. To put it differently, Vasubandhu’s view is metaphysical agnosticism plus something like epistemic idealism, at least as far as our ordinary cognitive operations are concerned. We cannot infer external causes from our ordinary experiences, for to do so causes delusion and suffering due to the dualistic distortions of our ordinary consciousness. What is in doubt is not external objects but externality, that is whether our ordinary consciousness can say anything about external objects outside its acts of cognizing them.

Garfield also translates the second verse of the Trisvabhavanirdesa, from Tibetan, as follows:

Arising through dependence on conditions and existing through being imagined, It is therefore called other-dependent and is said to be merely imaginary.

Compare Garfield’s translation of this second verse (which I have also discussed above) with the original Sanskrit, and then with Anacker’s translation from Sanskrit with my minor changes:

Yatkhati paratantrohosau yatha khyati sa kalpitaha Pratyayadhinavrittitvat kalpanmatrabhavataha

And now here is Anacker’s translation:

That which appears is the dependent, how it appears is the constructed The former develops subject to conditions, the latter is construction-only.

As I hope this comparison shows, unlike Anacker, Garfield here fails to distinguish clearly and adequately between the dependent aspect and the constructed or imagined aspect of things. Instead, he conflates the two, and then proceeds to claim: “For anything to exist as an object, its objective existence depends upon mental causes and conditions. This is a straightforwardly Kantian point, that there are conditions on the side of the subject that make it possible for anything to exist as an object.” Such an idealistic interpretation of Vasubandhu differs from one which, following Anacker, sees the objective existence of things as dependent on causes and conditions, which need not be mental. While our representations of things are mental, objects themselves need not be mental. This is what I take Vasubandhu to hold. Moreover, contra Garfield, it would be closer to Kant to see him as saying not that for anything to exist as an object there are mental causes and conditions, but rather that for anything to appear as an object of a subject’s knowledge, as a cognized object that is, there are mental causes and conditions such as our concepts and categories. To put it in Vasubandhu’s terms, the imagined or constructed aspect of things as having subject-object and other dualities depends on our minds and conditions how things appear to us. But the dependent aspect of things, things as they are, is not conditioned by the mental constructions of our ordinary consciousness, and is to be seen in meditation without the constructed aspect, as being dependent and without subject-object and other dualities.

Garfield writes: “...all external appearances are merely ideal and originate from potentials for experience carried in the mind.” Here I submit that in Vasubandhu’s scheme of things, appearances and how things appear to us as substantialized and reified may be mental and ideal, being due to the constructions of our minds. But it need not follow that objects themselves are mental, which is what someone who believes Vasubandhu is a metaphysical idealist or an absolute idealist, or even an anti-materialist idealist, would need to show.

Garfield also cites verses 1 and 16 from the Vimsatika in support of his claim that Vasubandhu is an idealist. As I have discussed the first verse from this work by Vasubandhu above, I shall quote below Garfield’s translation and comment on verse 16 :

Perception is like a dream That is, when it occurs The object it distinguishes does not appear. So, how can one call this perception?

Garfield then writes: “Vasubandhu in this text explicitly asserts that the entire phenomenal world is in fact ‘consciousness only’...What is apparently the perception of external objects is actually more like a dream - the confusion of mere phantoms of the mind with physical objects.” Here, contra Garfield, I suggest that Vasubandhu means that phenomena, or objects as they appear to us, are representations only and due to the constructions of our mind or consciousness, which gives us the imagined or constructed nature of objects. This need not mean that objects themselves are non-existent or are solely mental, which is what an idealist might claim. Vasubandhu means that perception is like a dream, but this need not mean that objects themselves are dream-like. The act of perception as well as the direct object of perception, which is what is directly before our minds, may be due to the constructions of the mind, but this need imply nothing about what objects themselves are really like out there. This is why it is claimed sometimes, as I have suggested above, that what is in doubt for Vasubandhu is externality, not external objects themselves. A comparison with Anacker’s translation of this verse might help make this point clearer:

Cognizing by direct perception is like in a dream etc. And when it occurs, the object is already not seen, so how can it be considered a state of direct perception? (Italics mine)

To return to Garfield, let us now look at verse 20 of the Trimsika (which I discussed earlier and) which Garfield also cites to support the claim that Vasubandhu is an idealist:

Whatever is an object of conceptual thought That is thoroughly imaginary. Without any entity, it does not exist.

Here I submit that what is imaginary and does not exist is what is constructed by the mind, not necessarily objects themselves. For example, suppose this constructed aspect or nature of things is presented to us by our ordinary consciousness in terms of subject and object being dual, being separate and independently existing. Nothing corresponding to such a construction exists, on Vasubandhu’s view, for there are no separate, independently existing entities. But this need not mean that nothing exists whatsoever, or that whatever exists is purely mental as an idealist might say. The constructed aspect of things whereby our ordinary consciousness substantializes and reifies things and ourselves as separate and dual is purely imaginary. But things and our selves still exist and have a dependent aspect or nature, which we see in meditation as the perfected aspect of things.

Garfield also cites verse 5 from the Trisvabhavanirdesa (a verse I discussed earlier) to defend the idealist thesis:

What is the imagination of the nonexistent? Since what is imagined absolutely never exists in the way it is imagined It is mind that constructs that illusion. (Italics mine)

He then goes on to claim that: “In the Trisvabhavanirdesa as well, Vasubandhu emphasizes the illusory character of external objects and the reality of the mind as the source of that illusion.” Here again I am not convinced that Garfield is right. As I understand Vasubandhu, and as I hope the phrase I italicized above in Garfield’s own translation shows, Vasubandhu only means that things as represented and perceived are only imagined. They do not exist that way, and are only constructed by our ordinary consciousness to be in that manner. However, contra Garfield, this need not mean that things themselves do not exist or are illusory.

So much for Garfield’s use of various passages from Vasubandhu. I turn now to other passages from Garfield himself where he expounds his view that Yogacara is idealist. Garfield writes: “...for Vasubandhu insofar as any phenomenon is ideal, its status as an external object is merely imagined. We see physical objects...as existing external to us. But that status is illusory...Second, each phenomenon is asserted by Vasubandhu to have an other-dependent nature. That is, for an object to be ideal is for it to exist in dependence upon the mind...Third, each object of consciousness has a consummate nature. This is the nature a thing is seen to have when it and its ideal status are completely understood. The consummate nature is the absence of the imagined nature...” Here again I am unpersuaded that Garfield has Vasubandhu right. For Vasubandhu, the representation of objects is mental, not necessarily the status of objects themselves. Our cognitive representations of things depend on the mind, not necessarily objects themselves. Second, for things to be other-dependent means only that they do not exist as separate, independently existing entities but as part of a flow of things. They are dependent on other things, which may or may not include the mind. Even if they are not dependent on the mind, they could be dependent on other things and each other. Third, contra Garfield, the consummate nature is the dependent nature without the constructed nature, and we see this as we meditate and see the flow of things as being without subject-object dualities.

Garfield also writes: “...from the standpoint of the consummate nature, it [the imagined nature] reveals the nondifference of object from mind, by virtue of its nonexternality. The consummate nature hence reflects a complete understanding of objects qua ideal and an abandonment of the subject-object duality apparent in the imagined...things as they really are are empty of the subject-object duality and are empty of any real distinction from the mind through which they are imagined.” Once again, such an idealist understanding is debatable. As I understand Vasubandhu, to give up subject-object duality means only that one sees the flow of all things, seeing all things as dependent instead of as independent, separate, substantialist entities. It does not mean, as Garfield contends, that we see things as empty of any real distinction from the mind. Indeed, if Vasubandhu is idealist in this sense, a Berkeleyan metaphysical idealist sense, then he is saying that things themselves are merely mental creations, that they are merely imaginary and illusory, a charge made by Kant against Berkeley. If so, then Vasubandhu would be some sort of nihilist, quite contrary to Nagarjuna and earlier Buddhists who do not deny the existence of reality in claiming that it is dependently originated and between the two extremes of eternalism and nihilism. This alone should cast at least some doubt on whether Vasubandhu would want to make so radical a departure from basic Buddhist ideas that idealism entails.

Some brief, final remarks about Garfield before concluding this paper. Garfield seems to think that claims by Berkeley and Kant map on to Vasubandhu’s imagined nature and dependent nature. However, pace Garfield, it is not clear that the mappings are as neat as he suggests. For Vasubandhu’s project is quite different from those of Berkeley and Kant. This should not surprise us given that Vasubandhu lived more than a millennium before these two Western philosophers, in a different culture that had little contact with the West (barring the presence of some Greeks such as Menander or Milinda after Alexander the Great’s invasion of northern India), and Vasubandhu had Nagarjuna, Asanga and others, not Locke, as predecessors in his own tradition. In particular, in Vasubandhu’s terms, Berkeley and his brand of metaphysical idealism are claiming more than warranted. For the epistemic idealism that Vasubandhu bears some similarities to does not, strictly speaking, entail metaphysical idealism, even if it is compatible with it.

IV. Conclusion

I hope it has emerged from the discussion above that Vasubandhu’s Yogacara has some similarities with epistemic idealism. Most notably, Yogacara thinks like Western epistemic idealism that our ordinary mental processes are so tied up with our ways of knowing that they condition the latter. Our knowledge of things is determined by what concepts and mentations we bring with us to our experience of them. In particular, according to Yogacara, our ordinary consciousness brings our subject-object and other dualities to our experience of things.

But we should also note differences between Vasubandhu’s Yogacara and Western epistemic idealism, despite various similarities between them, for Vasubandhu’s Yogacara is not identical to the latter; and while it is very easy to get carried away by similarities when doing Comparative Philosophy and see other traditions through concepts most familiar to us, it is at least as important to note the many genuine differences too. Most importantly, while for Western epistemic idealists such as Kant we cannot know noumena or things as they are, in the Yogacara scheme of things we can know the world as it is once we transcend our ordinary dualistic consciousness and see things through meditation as being part of a flow of things. This is precisely why Vasubandhu stresses meditation and Yoga and his position is known as Yogacara, for in common with many other schools of Indian thought (Yoga, Advaita Vedanta etc.), Yogacara stresses that our ordinary waking consciousness cannot give us knowledge of reality and our selves as they truly are. Consequently, while there is a basic Buddhist notion of enlightenment in Yogacara, consisting of seeing reality as it is, there are no equivalent soteriological notions in Western epistemic idealism. Moreover, Western epistemic idealists such as Kant posit the existence of noumena or things-in-themselves as what lies behind phenomena and causes them and is the source of freedom, even though we have no access to noumena. In contrast, at least as far as ordinary consciousness is concerned, Yogacara makes no such ontological posits or assumptions about reality - or about metaphysical realism or idealism, for that matter - for our ordinary consciousness is limited to its own constructions, and its cognitive objects never appear outside acts of cognition. It is only when we transcend this ordinary consciousness in meditation and see the dependent flow of things that we see reality as it is, according to Yogacara.

By way of conclusion, imagine the following thought-experiment. Suppose Germany had been invaded by the Scandinavians in the late nineteenth century, and German scholars had escaped en masse to France in the wake of the destruction of German universities and texts. Extant works by German philosophers such as Kant and those after him were then translated into French. Kant was now studied not in the context of philosophers who preceded and influenced him such as Leibniz, Hume, Wolff, and others, but rather in French and through the eyes of later German philosophers such as Fichte, Schelling, and the like, as well as through the eyes of French philosophers such as Sartre and Camus. While such a post-Kantian take on Kant may be novel and interesting and very valuable in itself, it cannot, I submit, substitute for reading Kant in his own context. I suggest the situation with Buddhist Studies today is somewhat similar, as this remark from Garfield amply illustrates: “Sthiramati’s commentaries became the standard entre into Vasubandhu’s work for Tibetan scholars, following Tsong Khapa’s extensive use of them...” While it is lamentable that Buddhist scholars were driven out of India into Tibet by Islamic invaders and that Buddhism more or less died out in India, to see classical Buddhism and Yogacara solely through later interpretations (whether Tibetan or Western) would be to ignore classical Buddhism in its own context.

One might wonder here what gives us, almost two thousand years on, the right to see ourselves as having a superior understanding of these texts from our contexts than do Tibetan exegetes, say. The reply to that is that we may know more about the broader overall cultural, historical, and philosophical context of Vasubandhu (including earlier Buddhist philosophy as well as other contemporaneous schools of Indian philosophy), and thus have more privileged access than Tibetan exegetes. Note also that to bracket the Tibetan commentarial tradition is not necessarily to appeal to a hermeneutic of authorial intent, with all of its pitfalls. For one might take the anti-intentionalist position that the meaning of texts is determined only by the public meanings of their constituent words and word-sequences; or else one might take the hypothetical intentionalist view, very roughly the view that the meaning of texts is determined by reading them and their constituent words in their broad cultural, historical, and philosophical contexts of origin, a position I have defended elsewhere. Also, I submit that Sthiramati and Candrakirti are not so much a part of the cultural, historical, and philosophical context that Vasubandhu himself was writing in as they are part of the post-Vasubandhu context, just as Fichte and Schelling are part of the post-Kantian context rather than Kant’s own context. Thus to read Vasubandhu through Sthiramati or Candrakirti would be to see Vasubandhu through a later perspective rather than in his own context, just as to read Kant through Fichte and Schelling would be to adopt a post-Kantian take on Kant.

When seen in its own context, Vasubandhu’s classical Yogacara at least is, I contend, not idealist in any sense, even if it has some affinities with epistemic idealism. I have not said anything here about later Yogacara developments and whether these are idealist; that topic is beyond the scope of this paper and it might well be the case, as Willis suggests, that later Yogacara thinkers such as Dharmapala and Hsuan-tsang give an idealistic cast to Vasubandhu.

Notes

1 - See, for example, Jay Garfield, Empty Words (New York: Oxford University Press, 2002); Jeffrey Hopkins, Emptiness in the Mind-Only School of Buddhism (Los Angeles: University of California Press, 2003); Paul Williams and Anthony Tribe, Buddhist Thought (London: Routledge, 2002); Stefan Anacker, Seven Works of Vasubandhu (Delhi: Motilal Banarsidass, 1984); Dan Lusthaus, Buddhist Phenomenology (London: Routledge Curzon, 2002); Richard King, “Early Yogacara and its Relation with the Madhyamaka School”, Philosophy East and West, 44 (1994): 659-86; Richard King, “Vijnaptimatrata and the Abhidharma Context of Early Yogacara”, Asian Philosophy, 8 (1998): 5-18; Thomas Kochumuttom, A Buddhist Doctrine of Experience (Delhi: Motilal Banarsidass, 1982); Alex Wayman, “A Defense of Yogacara Buddhism”, Philosophy East and West, 46 (1996): 447-76; Janice Dean Willis, On Knowing Reality (Delhi: Motilal Banarsidass, 1982).

2 - All of these are translated in Anacker, op. cit. 3 - See Mario D’Amato, “Review of Jay Garfield’s Empty Words”, Philosophy East and West, 53 (2003): 136-40. 4 - See George Berkeley, Three Dialogues Between Hylas and Philonous, ed. C. Turbayne (Indianapolis, Bobbs-Merrill, 1954). 5 - See Immanuel Kant, Critique of Pure Reason, ed. and trans. Norman Kemp Smith (New York, St. Martin’s Press, 1965). 6 - Compare Willis, op. cit., p. 59, n. 79: “...Yogacara’s ‘conceptualization-only’ (vijnaptimatra) theory...declares that ordinary cognitions do not actually ‘reach’ outside objects. Rather they are just mental reflections, object-like mental images (vijnapti).” 7 - See Ornan Rotem, “Vasubandhu’s Idealism: An Encounter Between Philosophy and Religion”, Asian Philosophy, 3 (1993): 15-28. 8 - Dan Lusthaus, “Buddhism, Yogacara school of”, in Routledge Encyclopedia of Philosophy (New York, Routledge, 1998). 9 - Wayman, op. cit. , p. 470.

10 - Compare Walpola Rahula, “Vijnaptimatrata Philosophy in the Yogacara System and Some Wrong Notions”, The Middle Way, 47 (1972), p. 118: “[[[Yogacara]] thinkers]...were expounding the old teaching with their own new interpretations...Their contribution to Buddhism lay not in giving it a new philosophy but providing, in fascinatingly different ways, brilliant new interpretations and explanations of the old philosophy.” 11 - Compare Dan Lusthaus, “Response for Yogacara Seminar American Academy of Religion 1998”, http://www.uncwil.edu/p&r/yogacara/External_Objects/lusthaus_response.html, p. 4: “Dualism is not the imagining of metaphysical objects opposed to metaphysical subjects. That is a linguistic embellishment, and, according to Sthiramati, comes later. The appearance of duality, Sthiramati says, is the rending apart of the sensorial act in such a way that eyes are given essential (svaruupa) existence apart from their apprehending visibles, and visibles are given essential (svaruupa) existence apart from eyes that envision them.”

12 - Wayman, op. cit., p. 457. 13 - Garfield, op. cit., p. 117. 14 - King, “Vijnaptimatrata and the Abhidharma Context of Early Yogacara”, loc. cit. 15 - Garfield, op. cit., p. 117. 16 - Ibid., p. 136. 17 - Ibid., p. 129. 18 - Ibid., p. 157. 19 - Ibid., p. 157. 20 - Ibid., p. 159. 21 - Ibid., p. 160.

22 - It is noteworthy that there are passages even in Asanga that cast doubt on the idea that classical Yogacara is metaphysically idealist. For instance, Asanga does not deny that rupa or matter is one of the five aggregates of the Buddhist notion of the self, nor does he reduce it to something purely mental as an idealist might do. See Asanga, Abhidharmasamuccaya, trans. Sara Boin-Webb (Fremont: Asian Humanities Press, 2001), p. 1 ff. 23 - Garfield, op. cit., p. 112. 24 - See my “....”...Journal...Issue...Year... 25 - See Willis, op. cit., pp. 21-2. 26 - Many thanks to Jay Garfield for prompt and detailed comments and discussion via email on an earlier version of this paper, despite our disagreements. His vigorous opposition has greatly enriched this paper.