02 May 2014

Sarvāstivāda Approach to the Problem of Action at a Temporal Distance.

In recent weeks I've become a bit more involved in a distributed discussion about  the twin Buddhist doctrines of karma & rebirth. This has been in response to apologetics defending traditional articles of faith with respect to karma & rebirth.

Of course I have blogged about karma & rebirth (together and separately) quite often, mainly exploring the challenges that 400 years of empiricism raise for traditional belief. But one of the other topics I write about is the nature of religious belief and I have become increasingly aware that the discussion about karma & rebirth was in danger of becoming bogged down. It's all too easy to see the discussion as a contest between pejorative and polemical accounts of fideism and scientism. The two sides are already talking past each other. 

So I began to explore a new tack. I was aware that the Buddhist tradition itself had a history of modifying these doctrines and some explorations of this have appeared as essays on this blog (see e.g. How the Doctrine of Karma Changes). I'd also been exploring some of the metaphysical problems in early Buddhism. I realised that it might be fruitful to dive into the history of Buddhism and develop this a bit more. I wrote an essay for our Order journal which can be found on my static website: Some Problems With Believing in Rebirth. In that essay I briefly outlined eight problems that people who believe in karma & rebirth ought to have thought about and tried to resolve. These problems are not reasons to disbelieve in karma & rebirth, but they are quite serious problems most of which have historically troubled Buddhists and resulted in doctrinal innovations. 

In discussing karma & rebirth our frame of reference is usually quite narrow: for most of us circumscribed by what is available in our bookshops. As a result our discussion of the history of Buddhist karma & rebirth seems to me to be rather constrained. This essay will attempt to broaden it out a little. In addressing the problem of karma & rebirth I've tried to show that it is (at least) three sided:
1. Inconsistent Early Buddhist accounts.
2. Later Buddhist adaptations and innovations.
3. Knowledge from 400 years of empiricism.
Buddhists themselves found the earliest received versions of karma & rebirth unsatisfactory and changed them. Ignoring this aspect of Buddhism results in a lopsided discussion. The equivalent would be like discussing British history in terms of the Celts and the Industrial Revolution, but missing out the Romans, Saxons, Vikings and Normans. Importantly, the Pali suttas cannot solve the problems we encounter, because they (along with their counterparts in other scriptural languages) are the source of the problem as I will try to show.

cetanāhaṃ, bhikkhave, kammaṃ vadāmi

One of the key issues related to karma & rebirth that unsettled Buddhists is what I call the problem of Action at a Temporal Distance: the ability for short lived mental processes to have consequences spanning multiple lifetimes. This problem has two main aspects:
  1. Karma, according to the Buddha, is cetanā (AN 6.63) and cetanā is a short-lived mental event. 
  2. Pratītya-samutpāda requires that when the condition ceases the effect must also cease (imassa nirodhā idaṃ nirujjhati).
Thus, on face value karma cannot coexist with pratītya-samutpāda because it requires the possibility of an effect long after the cessation of the condition, usually with no effect in the intervening time - in other words the effect only arises long after the condition has ceased. Ancient Buddhists noticed this and the result was a raft of doctrinal innovations attempting to reconcile the two, usually by artificially prolonging the action of conditions long after they cease i.e. Buddhists adjusted pratītya-samutpāda to accommodate karma. Of the many responses, this essay will focus on the Sarvāstivāda. 

The Sarvāstivāda School has a far better claim to be representative of early Indian Buddhism than does the Theravāda School. It dominated the North Indian Buddhist scene for several centuries while the Theravada School was relatively isolated in Sri Lanka: having little influence and being little influenced. The Theravādin Kathāvatthu, which is an account of the Vibhajyavādin's dispute with the mainstream of Buddhism, does not engage with the arguments found below (Bronkhorst 1989).  There were of course other schools, but they all seem to have defined themselves at least to some extent in opposition to the Sarvāstivāda, or have subsequently been found to be part of the same movement (like the Sautrāntikas). We tend to ignore the Sarvāstivāda because of Mahāyāna polemics, and because their texts have yet to be translated into English. However, the Sarvāstivādins were very much alive to this problem and canny in their response to it.

With respect to the problem of Action at a Temporal Distance, Dundee philosopher David Bastow (1995) believes that he has discovered the earliest argument for the existence of dharmas in the past, future and present - the characteristic idea that gave the Sarvāstivāda School its name. The argument is found the Vijñānakāya, a Sarvāstivāda Abhidharma text dated to perhaps 200 BCE and available only in Chinese translation.

One Citta at a time

Consider a mental moment of greed, a "greed citta". It is axiomatic (for all Buddhist schools) that there can be only one citta at a time, though it may be accompanied by mental factors (cetāsika) such as attention, volition and so on (each citta and cetāsika being a "dharma"). This imposes a temporal sequence on experience. Cittas arise one after another in sequence, each lasting a fraction of a second.

imasmiṃ sati idaṃ hoti
imass' uppādā idaṃ uppajjati
imasmiṃ asati idaṃ na hoti
imassa nirodhā idaṃ nirujjhati

Knowing that what we are experiencing is "greed", is itself a citta. So the knowledge that a citta was greed can only follow after the fact of the greed. Knowledge follows from experience. If we know we have experienced a greed citta then that greed citta cannot be non-existent, since, imasmin sati, idam hoti. Sati is a present participle from √as 'to be' while hoti is a dialectical variant of bhavati from √bhū 'to be, to become'. The phrase says "while this exists; this exists" (the pronoun in both cases is the deictic idam which is conventionally translated as 'this' and indicates something present to the speaker).

sense object + sense organ + sense discrimination = contact

It is also axiomatic in Buddhist psychology that for vijñāna to arise there must be a sense object (ālambana) and sense faculty (indriya). Thus the greed citta must "exist" in some form (imasmin sati). We don't need to get bogged down in defining in what way it exists, only to acknowledge that like any sense object it functions as a condition for vijñāna to arise, so it cannot be non-existent. And since the greed citta must sequentially precede the knowledge citta, the greed citta must exist (in some way) in the past. The same is true of all cittas.

Furthermore, it is essential to both Buddhist ethics and karma that a greed citta has future consequences. The classic pratītya-samutpāda formula informs us that if the condition has ceased then the effect ceases. The corollary is that if there is a future effect, then a future condition must exist. Thus, in our example, the greed citta must exist (in some form) in the future or it could not have future consequences. In order for karma to work as advertised the citta must potentially continue to exist over several lifetimes. The same is true of all cittas.

Thus, dharmas exist in all three times: present, past, and future.

To summarise: a citta "exists" (in some form) in the present, but in order for us to have knowledge of it, the citta must also "exist" (in some form) in the past; and in order for it to have consequences at a later time it must "exist" (in some form) in the future. Minimally "exist" means that it must at least be able to function as a condition for the arising of mano-vijñāna (i.e. as a dharma); it must be consistent with the imasmin sati formula. Thus, cittas (i.e. dharmas) exist in all three times: present, past, and future. And this, according to the Vijñānakāya, is what sarva-asti means. 

This view, and developments of it, dominated the first phases of sectarian Buddhism in Indian from around the 2nd century BCE until it was replaced by the metaphysical speculations of the Yogacārins in about the 5th century CE. The sarva-asti view emerges from the application of standard Buddhist axioms to karma and, unlike the Yogacāra view, it does not introduce further speculation or further axioms. It is a plausible solution to the problem of Action at a Temporal Distance, certainly no less plausible than suggesting the existence of seeds in a storehouse. Thus, we should not dismiss the Sarvāstivāda view lightly. If we are going to dismiss it, then it ought not to be on the basis of further metaphysical speculation. More importantly, we ought to offer a better solution to the problem of Action at a Temporal Distance.


Bastow, David. (1995) 'The First Argument for Sarvāstivāda.' Asian Philosophy 5(2):109-125. Text online
Bronkhorst, Johannes. (1993) 'Kathāvatthu and Vijñānakāya'. Premier Colloque Étienne Lamotte. Bruxelles et Liège 24-27 septembre 1989). Université Catholique de Louvain: Institut Orientaliste Louvain-la-Neuve. 1993. Pp. 57-61)


Dhivan Thomas Jones said...

Sadhu Jayarava! What a nice clear exegesis. Probably I'm like most western Buddhists (who take an interest in these things) that Sarvāstivāda sounds implausible, but you've presented in such a way that it begins at least to make sense.

It strikes me that the Indian Buddhist attempts to make sense of karma are quite like Christian attempts to prove the existence of God – theology rather than philosophy, though often disguised as philosophy.

Jayarava Attwood said...

Thanks Dhīvan,

A lot of the credit goes to Collett Cox whose work I will more directly reference next week. And Barstow sets it out nicely in his article. I think between them they ought to have had more influence on how we see early Buddhism. My contribution is only to define the problem and give it a name.

The Sarvāstivāda seen in this light are far more interesting than how they are generally presented. Turns out the Pudgalavādins were looking at the same problem.

Theology indeed :-)

greg222222 said...

Interesting work, thanks!

"The Sarvāstivāda School has a far better claim to be representative of early Indian Buddhism than does the Theravāda School. It dominated the North Indian Buddhist scene for several centuries while the Theravada School was relatively isolated in Sri Lanka: having little influence and being little influenced."

That is tangential but perhaps worth noting: we often forget about the Pudgalavāda because their literature is mostly lost, but they have a claim to be among the representative of early Indian Buddhism, considering Xuanzang reports that half of all the śrāvakas in India belonged to the Saṃmitīya by the 7th century. And they of course were willing to toss out the anātmavāada almost explicitly to preserve the karmavāda!

Jayarava Attwood said...


Not too tangential. I haven't had a chance to look closely at the Pudgalavāda, but yes it does seem that they too were trying to solve the problem of Action at Temporal Distance and were dissatisfied with early Buddhist accounts of how pratītyasamutpada worked with karma. And yes they also modified pratītyasamutpāda to preserve karma.

I still find this astounding, having been inculcated with the idea that pratītyasamutpada is at the heart of everything!


me123 said...

Would it be correct to say that the Sarvastivada doctrine is that time doesn't exist in the way which we experience it?

That the past, present and future all exist equally and our experience of the flow of time could be compared to the illusion created by a film reel, a projector and a screen - the scenes of a film look like they are in constant motion, but whatever scene we see at any given moment is a chunk of still frames in the film reel that have passed and are passing in front of the flickering light of the projector.

Any experience at a given moment - the present - is a group of cittas, the past is a different group of cittas, the future a different group of cittas. No different to an array of particles in space with differing functions and characteristics. There is no enduring, unchanging, permanent entity among or behind any of them - just causal connections or interconnections between them. Causality, including temporal causality, is a vast web, not just the linear chain that is only available to our ordinary perception, which we experience as past>present>future.

If I'm not mistaken, the alaya vijnana concept posits a storehouse for the "seeds" planted by past karmic actions, because it accepts that our experience of time is an absolute reality - things of the past have really vanished out of existence, things in the present really are coming into existence and enduring before they vanish, and the future doesn't exist at all in the present. Thus the need for a storehouse for past karma which hasn't had a result yet. This only seems necessary without the sarvastivada doctrine of time, but I was under the impression that this was the only buddhist view of time.

Jayarava Attwood said...

To be honest I don't know the sarva-asti doctrine well enough to absolutely rule out the idea that time doesn't exist in the way we experience it. But nothing I've read would make me think this was their view. Indeed it seems to me that the opposite must have been behind the sarva-asti innovation, ie. that they saw time in much the same way that we do. They were trying to understand how condition and effect could be separated in time, particularly how an effect could manifest at a time when the condition had ceased. If the flow of time was an illusion then they would not have had the problem of action at a temporal distance.

As I recall, something like what you suggest is Nāgārjuna's solution to the problem. A solution that appears to have convinced very few Buddhists.

The idea of an ālaya-vijñāṇa comes from a metaphor popularised by the Sautrāntika sect. I'm not aware that the Sautrāntika or the Yogācārins though of time as an absolute reality. Indeed I'd be very surprised indeed if the Yogācārins did think like that as it seems to cut across the way they thought.

The problem of action at a temporal distance arises from a view of time which is fairly conventional from our point of view.

Related Posts with Thumbnails