Saturday, July 10, 2010

Reading Vasubandhu: Karma-siddhi-prakaraṇa

Rollin', Nice, 2010

Vasubandhu's Karma-siddhi-prakaraṇa,  "Discussion for the Demonstration of Action," is a bastard. As in, I had a great deal of trouble reading it, much more so than Vāda-vidhi or Pañcaskandhaka-prakaraṇa. Nevertheless, I think I've finally got some kind of hold on it, although it's bound to be even more superficial than for the previous two works. This is because it's a part of a grand philosophical debate about which I'm pretty much completely clueless. 

I remain a complete newbie to this whole show, so absolutely do not take anything I say as authoritative in any way. If you're only stumbling on my blog now and want to read this bit, I would suggest that you check out my pontification on Pañcaskandhaka-prakaraṇa first, though, because a good many concepts discussed here were introduced there.

The central question in Karma-siddhi-prakaraṇa is that of psychological continuity: if there is no abiding 'self,' how does the illusion of continuous consciousness arise? How do we remember and learn? If consciousness ceases through deep meditational attainment or more conventional means, how come the consciousness-streams resume where they left off, once you come out of the meditative state or just wake up? How do retributive effects—the "maturing of karma" that produces pleasant effects for beneficial acts and unpleasant effects for unbeneficial ones—make themselves felt?

This Discussion is written in the form of a dialog featuring a number of different speakers.

In the first section, we have Vasubandhu debating a Vaibhāṣika, an Āriya-Sāmmitīya, and a Saurodayika. The only group I had even heard of previously were the Vaibhāṣikas—these are the same guys as the Kashmiri Sārvastivādins Vasubandhu had studied with, for whom he had supposedly compiled his Kośa. In other words, they're all Buddhists with rather different ideas on the subtler points of doctrine. More such schools would be introduced in the second and third parts of the text. Vasubandhu seems to be perfectly fluent with all of their doctrines, and engages with each of them on their own terms. Unfortunately, for somebody who's less than familiar with the distinctions between Vaibhāṣika atomism and Saurodayika atomism, this richness of concepts and constant shifting in the conceptual framework makes the text damn hard to read.1

Vasubandhu's purpose with this Discussion is threefold: (1) to demonstrate that none of the competing schools of philosophy have a truly satisfying conceptual model of morally determinate action, (2) to introduce his own model, and (3) to demonstrate that his model resolves the problems inherent in the competing ones. He has his work cut out for him for (1) and (3), since he needs not only to debunk the competing models and provide a logically coherent justification for his own, but being an orthodox Buddhist, he also has to demonstrate that his model conforms to the Buddhist canon. He doesn't settle for anchoring it only in the Mahayana canon either; instead, he endeavors to show that Vaibhāṣikas, Saurodayikas, Ārya-Sāmmitīyas and the rest should adopt it too, based on the texts they consider canonical.

Since I'm not particularly interested at this time in the competing models and I'm certainly not competent enough to say anything remotely informed about them, I'm going to have to skip discussing his demonstrations and arguments, even though they make up the bulk of the text. Instead, I'll focus on what I understood of the model Vasubandhu is presenting.


I have an aversion to metaphysics. Simply put, I think it's a waste of time. I find many branches of philosophy extremely interesting and highly useful—those dealing with epistemological questions (how can we know if a proposition is true or false?), problems related to cognition, ethics, or general questions related to the mystery of being human. Metaphysics, however, invariably ends up constructing cloud cathedrals that very quickly lose sight of anything that's useful for any other purpose than validating the system itself, or debunking other similar systems. Basically, I think metaphysicians are wankers with way too high an opinion of themselves. I have a feeling Vasubandhu may share this view to an extent.
Saurodayika: If it is correct that color is not manifest action, is its arising in another locus manifest action?
Vasubandhu: Beloved of the gods! Though it can be seen that you are making efforts to the best of your abilities, what is the point in making an effort towards things that can't be demonstrated by any effort whatsoever?
Unfortunately, the first half of the text veers dangerously close to metaphysics, probably because Vasubandhu's philosophical opponents emphatically did not share his aversion, and he engages with them on their own terms. In the first two parts of the text, he takes each of the cloud cathedrals built by his opponents, demonstrates a complete mastery of them, and then pulls the foundation from under them, using only the rules set by the constructors of each of those cathedrals. No wonder the Vaibhāṣikas ended up hating his guts. That's a masterful performance, but unfortunately one I'm very ill-equipped to admire—and it left me with comparatively little gold after sifting through all this earth.

Despite my aversion to metaphysics, I'm going to have to point out two metaphysical propositions that are rather central to the text. One of them is something of a profound insight, and I still think the other one is somewhat flawed.

The flawed proposition is Vasubandhu's insistence on atomic cittas. These are treated as instantaneous and indivisible. His citta-streams look like chains of pearls rather than threads. Since his approach is introspective and psychological, these atomic cittas logically lead to atomic time and atomic space. A moment of time is the time associated with an individual citta, and a locus of space is a point in space associated with a citta. This leads to a whole bunch of metaphysical complications. For example, if time is discontinuous, how is motion possible? Vasubandhu's answer is that motion is an illusion: in reality, what looks like motion is actually a body being destroyed at one locus and another body arising at another one, with the body at the former locus being the cause for the arising of the body at the latter locus. This is a pretty good description of how, say, a motion picture works—a fast stream of separate still images blurring into an illusion of motion.

However, I still don't see any compelling reason to treat the universe itself this way—nor cittas. The universe can be thought of as neither atomic nor infinitely divisible. For example, you can cut a string at any point you like, even though it stops being string if you make two cuts close enough to each other, and you won't be able to precisely pinpoint the exact length you need to cut that marks the transition between string and not-string. Vasubandhu's psychological model would be stronger if his citta-streams were more like strings of thread than strings of pearls, I think, and he would avoid a whole mess of trouble he has to deal with throughout this text.

The insightful proposition is the idea that only the present exists. The past consists of phenomena that once existed by no longer do, and the future consists of phenomena that do not yet exist. Neither past nor future events can exert any kind of influence on the present. If this is the case, what happens to the basis of Buddhist ethical philosophy, karmic retribution? In treating these problems, Vasubandhu is at his most insightful, and the psychological theory he comes up with by way of explanation is every bit as relevant today as it was 1600-odd years ago.

The root-consciousness

Orthodox Buddhist psychological theory recognizes six consciousnesses, corresponding to the six senses: the visual, auditory, olfactory, tactile, gustative, and mental. In meditational attainments, each of these six consciousnesses ceases. If we accept that phenomena can only exist in the present, meaning that past events cannot exert any force on present ones, how, then, do they resume? Where do we keep memories of sights, sounds, smells, tastes, feelings, and concepts? How do we account for distortions of memory? Why do we forget? How come retributive effects still follow beneficial or unbeneficial acts, even if the consciousness has been interrupted through meditational attainment or by other means? When we speak of rebirth, what, exactly, gets reborn?

Vasubandhu's answer to these problems problem is to posit a seventh consciousness, which he calls the root-consciousness, seed-consciousness, retributive consciousness,or appropriating consciousness. Unlike the other consciousnesses, which can be directly apprehended through their matching senses, the root-consciousness can only be inferred from its effects. The root-consciousness consists of cittas Vasubandhu calls "seeds" or "impressions." Every time a citta arises in one of the six consciousnesses, it simultaneously creates a seed in the root-consciousness. These seeds then continue as citta-series, just like, say, a train of thought persisting for a while in the mental consciousness, or the series of sounds that make up a piece of music in the auditory consciousness. The citta-series in the root consciousness just tend to persist longer and change more slowly. Nevertheless, Vasubandhu makes it clear that the seeds are no more abiding or permanent than any other citta-stream. The root-consciousness is not a passive receptacle of memories and impressions; it is a living, constantly changing continuity.

How, then, does Vasubandhu infer the existence of a root-consciousness? By identifying some of its effects, and then arguing that there is no other plausible explanation for them. These effects are memory, the resumption of ordinary consciousness and karmic retributive processes after meditational attainment, and rebirth.2

Every so often, a citta-series changes in a way that causes it to kick off a citta-series in one of the other consciousnesses. A smell in the olfactory consciousness might trigger a seed-series to cause a flood of memories to break into the mental consciousness, the seed from a long-forgotten act of selfless generosity might mature into a blissful spiritual realization, or the seed from a long-forgotten act of selfish grasping might mature into a really bad bout of depression.

For Vasubandhu, the root-consciousness is, in fact, nothing more or less than a collection of these series of seed-cittas, which are fundamentally no different from cittas occurring in any of the other consciousnesses. Just like a citta in the olfactory consciousness can cause the idea of coffee to arise in the mental consciousness, cittas in the root consciousness arise from cittas in the other consciousnesses, and vice versa. He also emphasizes that because of this fundamental similarity, it is quite erroneous to conceive of the root-consciousness as an abiding "self"—it is neither abiding, since it's constantly changing just like the other consciousnesses, nor a self, since it's no different from the other citta-streams that aren't the self either.
Opponent: In that case, why not accept a self with existence as an entity, as the substratum of the six consciousnesses?
Vasubandhu: In what way is a self accepted? If it presents itself only as a series of moment-events, and transforms itself constantly through conditions, then what is the difference between it and the store-consciousness?
Opponent: But it is single and constantly devoid of transformations.
Vasubandhu: In that case, how can it be demonstrated that it is also influenced by the latent impressions left in it by the consciousnesses, etc.? It is the latent impressions which produce the special forces which make the consciousness-series continue … If there is no special characteristic which undergoes transformation, how, as there are no impressions possible in such a case, do there arise in time special memories, recognitions, passions, etc. …? In what way are the consciousnesses subject to it, through which it could be understood that the self is their substratum? … Accordingly, the conception that there exists a lasting independent entity "self," is a poor one.
However, unlike the other six consciousnesses, the root-consciousness can never be directly apprehended. Its existence can only be inferred. We are never directly aware of the streams of seed-cittas evolving in the root-consciousness; we can only become aware of their existence when they kick off cittas in the six other consciousnesses. Also unlike them, the citta-streams making up the root-consciousness persist even in the highest meditational attainments and in complete unconsciousness. In fact, Vasubandhu says that these streams persist even through death, subsequent to which they will appropriate another just-conceived body, as long as samsāra persists—they will only cease when an enlightened being attains nirvāna. This makes the root-consciousness nothing less than the glue that binds us to the wheel of karma—no small potatoes, that.3

In my opinion, Vasubandhu's root-consciousness with its citta-streams that flow ever on below our awareness is as good a model for memory, learning, and psychological persistence as any, and a good deal better than most, whatever you want to make of his views on its persistence after physical death. I was very interested to learn that a good many people back then proposed to account for memory and psychological continuity through the existence of a material basis. Vasubandhu specifically rejects this solution largely through metaphysical reasons, and, I think, because it could not account for the continuation of the karmic stream after physical death, which he took as axiomatic. Personally, I don't see any big philosophical issues with reconciling what we know of the functioning of the brain and Vasubandhu's root-consciousness—it's pretty easy to continue the chain of causation one notch, and describe Vasubandhu's seed-cittas reflected as physical structures in the brain. It's a rather a nice, parsimonious solution, in fact, with no need to posit a large number of complex, unknown and potentially unknowable entities (cf. Freud's subconscious, superego, ego, and traumas, or Jung's complexes.)

Intentionalist absolutism

There are two opposed ways of approaching ethical philosophy, often called intentionalism and consequentialism. A consequentialist holds that the ethical charge of an act is determined by its consequences, whereas an intentionalist holds that it is determined by the intention behind that act. Most of us are probably a bit of both: we would hold someone morally culpable if he carelessly ran over a child, but much less culpable than if he did it on purpose; we often treat attempted murder more leniently than accomplished murder. Personally, I feel that "I meant well" is usually a pretty lousy excuse when something goes disastrously wrong.

There was a lively debate going on about this topic at Vasubandhu's time too, with the Jains being near the consequentialist end of the scale, and old-style Buddhists near the intentionalist one. As a Jain sūtra quoted by Stefan Anacker uncharitably puts it:
If one thrusts a spear through the side of a granary, mistaking it for a man, or through a gourd, mistaking it for a baby, and roasts it, one will be guilty of murder according to our views. If one puts a man on a spit and roasts him, mistaking him for a fragment of the granary, or a baby, mistaking him for a gourd, he will not be guilty of murder according to our views. If anybody thrusts a spear through a man or a baby, mistaking him for a fragment of the granary, puts him on the fire, and roasts him, that will be a meal fit for the Buddhas to breakfast on.
Vasubandhu is an absolute intentionalist. For him, the only thing that matters is intention. He arrives at this conclusion through a pretty rigorous logical exercise, where he debunks the Vaibhāṣika notions of manifest and unmanifest action. He argues that what they call manifest bodily action—visible action, like waving your hand, or driving a spear through a gourd, or a baby—is morally entirely neutral, since there is no way to distinguish externally between a motion of the hand produced by a volition and the motion of the hand produced by an involuntary seizure. Instead, "bodily action" in the ethical context refers to the volition that initiated the bodily action.

However, I don't think Vasubandhu would have accepted the Jains' characterization of his views. A man thrusting a spear through a baby, mistaking it for a gourd, would not be guilty of murder, perhaps, but he would certainly be guilty of gross carelessness, to make such a horrible mistake. Karmic retribution for that would surely follow. On the other hand, he might argue, if someone accidentally dropped a baby into a fire-pit due to a completely involuntary and unpredictable muscular seizure as he was trying to keep her warm in a freezing wintry wind, it would certainly be absurd to consider him guilty of murder—even if the subsequently roast baby would probably not make a karmically acceptable breakfast, for any of a number of peripherally related reasons.

In Vasubandhu's model, "volition" is a great deal broader concept than the sense in which we usually use it. It's not just your stated or consciously formulated intention; it's also the whole set of cittas that cause it to arise—greed, ignorance, delusion, compassion, wisdom, generosity, and so on. "I meant well" really is a lousy excuse, most of the time, because if you dig a bit deeper, you'll usually find cittas of anger, delusion, greed, or fear behind the consciously-formulated "meaning well." And if you really don't, then it really is just bad luck.

Vasubandhu's excursion into ethics brings us right back to where we started our journey, self-knowledge and introspection. The logical conclusion is that only someone who has perfect understanding of himself is able to act in an ethically pure manner. In fact, such a one cannot help but act in such a manner, since acting in any other way would be absurd—it would be inflicting suffering on himself and others for absolutely no reason whatsoever. Someone feeding the poor in the slums of Calcutta might be completely corrupt ethically, if she is ignorant of her volitions, and it turns out that the primary driver of her actions is, say, a prideful desire to feel morally superior to others who don't perform such service.

This is very dangerous territory, with deep pits menacing with spikes in all directions. Even a little bit of delusion will make you believe that you are "such a one," which gives you the ethical license to do just about anything in the name of the greater good. The only way I can see to avoid that trap is to treat intentionalist ethics as personal and internal, not as a basis for public morality. Vasubandhu may well be pointing us towards nirvāna, but it would be very dangerous to mistake that path for a system of judging the ethical merits of other people. It would be very easy to twist these ethics into a system that lets delusional "enlightened" gurus run amok, and I think that that's precisely what has happened on a number of occasions. "I have only taught two things, suffering, and the end of suffering," the Tathāgata said, and it behooves the "sons of Śakya" to remember that.

The next text in the series is the famous "Twenty Verses on Representation-Only," with a commentary. I've already come across this, and was very impressed by them. I'm looking forward to what a more in-depth reading will reveal.

1I've no doubt that this debate would be a very interesting topic of study in its own right. It's remarkably similar in many ways to the great metaphysical debate between the Atomists and the Idealists in Greece some centuries previously, although it has a deep psychological dimension that was almost completely lacking among Greek philosophers.
2With "karmic retribution," Vasubandhu means an internal process where unbeneficial acts eventually cause cittas involving suffering to arise, whereas beneficial acts eventually cause cittas associated with the end of suffering to arise. I'm not going to get into a discussion of rebirth at this point. For Vasubandhu, though, it was taken for granted, and any valid psychological theory would have to account for it.
3I wonder if this is where the Tibetan concept of the Very Subtle Mind comes from? I hear they're pretty big on Vasubandhu up there in the mountains…


  1. Glad you pointed to (and moved beyond) the Freudian theories, even though they seem to give an inkling of this.

    What also stuck me was the theory of Nishijima Roshi. His concentration on the autonomic nervous system in particular. Some people have dismissed this due to his age but I think he may be pointing out a similar idea to yours regarding brain structure. Brains change with meditation.

    Interesting. And your thoughts on root consciousness also have me intrigued.

    Am really enjoying your reports of your encounters with Vasubandhu's work.

  2. Thanks for the encouragement! I'm enjoying the encounters too, and writing about them. I'm getting the feeling that I'll have to start over once I finish this round; there's such a density of material in there that most of it just flies right past and above me. But there you go, 1600 years of difference and a very different culture has got to mean something.

    Yeah, I've read some of Nishijima Roshi's thoughts on this too. I think he's expressing it in a rather a weird way; "balancing the autonomous nervous system" sounds like so much pseudoscientific gobbledegook. Also, if you're looking at the actual physical make-up of the body, I'm pretty sure that the interesting meditation-related changes do happen in the brain rather than the nerves.

    However, I also thought it is worth pointing out. The autonomous nervous system is like the conveyor belt between the brain and the plumbing. Like the root-consciousness, it functions beneath our awareness; we can only infer its functioning by its effects. Obviously meditation does affect the way it functions—for example, before I started doing zazen, I had occasional bouts of stress-related acid reflux; it hasn't happened even once since. That's clearly related to the autonomous nervous system.

    I'm also not quite sure about his categorical insistence on the half- or full-lotus. I have finally made it into half-lotus, although I still mostly sit in quarter-lotus, and I quite often have to sit in seiza. It is a little different in each of these postures, but not really any more different than the difference between riding a mountain bike, road bike, and fixed-gear bike; it's still riding a bike.

    But he's clearly an interesting guy, and the community that has coalesced around him is very interesting too. I probably wouldn't fit in that well (not enough tattoos, in a manner of speaking), but it's clearly exactly what one particular temperament needs.

  3. Extremely interesting stuff. I'm normally very put off by explorations into ideas that get so complex and convoluted that the original meaning of those ideas is long-forgotten by the time the middle of the argument is reached, but like all of the material you've presented from Vasubandhu, there's a great deal of underlying logic, cohesion and simplicity to the apparent complexity here.

    The argument between intentionionalism and consequentialism is particularly fascinating, if only because it's something that seems obvious but if thought about at all, instantly is not. I'm going to have to give some thought to these concepts, in particular, "volition" as you attribute it to V.'s posit of the root-consciousness. Like understanding the place where inhale becomes exhale, this perception of what seems random action or experience actually being willed into existence by a very different and completely ego-separated area of the person (and of course, coming from a very different realization of personhood than I'm used to) looks like a deceptively simple but very concentrated message-mission. Like a dense,tough-husked grain, some thorough chewing over here ought to be rewarding.

    Looking forward to the next installment, and thanks for the mental challenge.

  4. Fascinating discussion. I too have been bewildered by the problem of: If an instance of consciousness is atomic, how does it come complete with a past? From whence comes that past. Can we relate this to John Wheeler's view that the present creates its past? Should we relate it to "computational universe" concepts? Are we to accept Jaques Lacan's thesis that every world view is inescapably incomplete and incorrect?

    I am in awe of and grateful to the originators of these posts. Thank you.