For a fuller discussion of some of this, see A NOTE ON PATICCASAMUPPĀDA.

  In spite of the venerable tradition, starting with the Patisambhidāmagga (or perhaps the Abhidhamma Pitaka) and continued in all the Commentaries (see Anguttara V,viii,9 <A.iii,107,§4>), paticcasamuppāda has nothing to do with temporal succession (cause-and-effect). Precedence in paticcasamuppāda is structural, not temporal: paticcasamuppāda is not the description of a process. For as long as paticcasamuppāda is thought to involve temporal succession (as it is, notably, in the traditional 'three-life' interpretation), so long is it liable to be regarded as some kind of hypothesis (that there is re-birth and that it is caused by avijjā) to be verified (or not) in the course of time (like any hypothesis of the natural sciences), and so long are people liable to think that the necessary and sufficient criterion of a 'Buddhist'[a] is the acceptance of this hypothesis on trust (for no hypothesis can be known to be certainly true, since upon the next occasion it may fail to verify itself). But the Buddha tells us (Majjhima iv,8 <M.i,265>) that paticcasamuppāda is sanditthiko akāliko ehipassiko opanayiko paccattam veditabbo viññūhi. ('immediate, timeless, evident, leading, to be known privately by the wise.') What temporal succession is akālika? (See CITTA [a].) For an ariyasāvaka, paticcasamuppāda is a matter of direct reflexive certainty: the ariyasāvaka has direct, certain, reflexive knowledge of the condition upon which birth depends. He has no such knowledge about re-birth, which is quite a different matter. He knows for himself that avijjā is the condition for birth; but he does not know for himself that when there is avijjā there is re-birth. (That there is re-birth, i.e. samsāra, may remain, even for the ariyasāvaka, a matter of trust in the Buddha.) The ariyasāvaka knows for himself that even in this very life the arahat is, actually, not to be found (cf. Khandha Samy. ix,3 <S.iii,109-15> and see PARAMATTHA SACCA [a]), and that it is wrong to say that the arahat 'was born' or 'will die'. With sakkāyanirodha there is no longer any 'somebody' (or a person—sakkāya, q.v.) to whom the words birth and death can apply. They apply, however, to the puthujjana, who still 'is somebody'.[b] But to endow his birth with a condition in the past—i.e. a cause—is to accept this 'somebody' at its face value as a permanent 'self'; for cessation of birth requires cessation of its condition, which, being safely past (in the preceding life), cannot now be brought to an end; and this 'somebody' cannot therefore now cease. Introduction of this idea into paticcasamuppāda infects the samudayasacca with sassataditthi and the nirodhasacca with ucchedaditthi. Not surprisingly, the result is hardly coherent. And to make matters worse, most of the terms—and notably sankhāra (q.v.) —have been misconceived by the Visuddhimagga.  

It is sometimes thought possible to modify this interpretation of paticcasamuppāda, confining its application to the present life. Instead of temporal succession we have continuous becoming, conceived as a flux, where the effect cannot be clearly distinguished from the cause—the cause becomes the effect. But this does not get rid of the temporal element, and the concept of a flux raises its own difficulties.[c]  

The problem lies in the present, which is always with us; and any attempt to consider past or future without first settling the present problem can only beg the question—'self' is either asserted or denied, or both, or both assertion and denial are denied, all of which take it for granted (see NA CA SO). Any interpretation of paticcasamuppāda that involves time is an attempt to resolve the present problem by referring to past or future, and is therefore necessarily mistaken. The argument that both past and future exist in the present (which, in a certain sense, is correct) does not lead to the resolution of the problem.  


[a] To be a follower of the Buddha it is certainly necessary to accept on trust that for one who is not rid of avijjā at his death there is re-birth, but it is by no means sufficient. What is sufficient is to see paticcasamuppādaYo paticcasamuppādam passati so dhammam passati ('He who sees dependent arising sees the Teaching') (Majjhima iii,8 <M.i,191>). For those who cannot now see the re-birth that is at every moment awaiting beings with avijjā, the dependence of re-birth on avijjā must be accepted on trust. They cannot get beyond temporal succession in this matter and must take it on trust that it is a question of dependence (and not of cause-and-effect)—i.e. that it is not a hypothesis at all, but (for the Buddha) a matter of certainty. But accepting this on trust is not the same as seeing paticcasamuppāda. (Past and future only make their appearance with anvaye ñānam [see NA CA SO [a]), not with dhamme ñānam. 'As it is, so it was, so it will be.' Paticcasamuppāda is just 'As it is'—i.e. the present structure of dependence.) [Back to text]  

[b] So long as there are the thoughts 'I was born', 'I shall die', there is birth and death: so long as the five khandhā are sa-upādānā, 'somebody' becomes manifest and breaks up. [Back to text]  

[c] The notion of flux can be expressed thus: A = B, B = C, A ≠ C, where A, B, and C, are consecutive (Poincaré's definition of continuity). This contradiction can only be concealed by verbal legerdemain. (The origin of this misleading notion, as of so many others in the traditional interpretation, seems to be the Milindapañha, which, to judge by its simile of the flame, intends its formula na ca so na ca añño to be understood as describing continuous change.) The misunderstanding arises from failure to see that change at any given level of generality must be discontinuous and absolute, and that there must be different levels of generality. When these are taken together, any desired approximation to 'continuous change' can be obtained without contradiction. But change, as marking 'the passage of time', is no more than change of aspect or orientation: change of substance is not necessary, nor is movement. (See ANICCA [a], CITTA [a], & FUNDAMENTAL STRUCTURE.) Kierkegaard (op. cit., p. 277) points out that Heraclitus, who summed up his doctrine of universal flux in the celebrated dictum that one cannot pass through the same river twice, had a disciple who remarked that one cannot pass through the same river even once. If everything is changing, there is no change at all. The assumption of a single absolute time, conceived as a uniform continuity (or flux) of instants, leads at once to a very common misconception of the Dhamma:


Even if I now perceive things as self- identically persisting in time, my present perception is only one out of a flux or continuous succession of perceptions, and there is no guarantee that I continue to perceive the same self-identities for two successive instants. All I am therefore entitled to say is that there appear to be self-identities persisting in time; but whether it is so or not in reality I am quite unable to discover.

The Buddha's teachings of impermanence and not-self answer this question in the negative: In reality no things exist, and if they appear to do so that is because of my ignorance of these teachings (which is avijjā).

But we may remark: (i) That A is the result of taking presumptively the rational view of time, and using it to question the validity of direct reflexive experience. But the rational view of time is itself derived, ultimately, from direct reflexive experience—how can we know about time at all, if not from experience? --, and it is quite illegitimate to use it to dig away its own foundations. The fault is in the act of rationalization, in the attempt to see time from a point outside it; and the result—a continuous succession of isolated instants each of no duration and without past or future (from a timeless point of view they are all present)—is a monster. The distinction in A (as everywhere else) between 'appearance' and 'reality' is wholly spurious. (ii) That since our knowledge of time comes only from perception of change, the nature of change must be determined before we can know the structure of time. We have, therefore, no antecedent reason—if we do not actually encounter the thing itself—for entertaining the self-contradictory idea (see Poincaré above) of continuous change. (iii) That, whether or not we do actually perceive continuous change, we certainly perceive discontinuous changes (so much is admitted by A), and there is thus a prima-facie case at least in favour of the latter. (iv) That the experiments of the Gestalt psychologists indicate that, in fact, we perceive only discontinuous changes, not continuous change (cf. Sartre, op. cit., p. 190). (v) That if, nevertheless, we say that we do at times and in the normal way have intuitive experience, distinct and unambiguous, of continuous change, and if we also say that continuous change, in accordance with B, is what is meant by the teaching of impermanence, then it will follow that at such times we must enjoy a direct view of 'reality' and be free from avijjā. Why, then, should we need a Buddha to tell us these things? But if we reject the first premiss we shall have no longer any grounds for having to assert a uniformly continuous time, and if we reject the second we shall have no longer any grounds for wishing to assert it. (On the question of self-identity, see ATTĀ.)

   Our undeniable experience of movement and similar things (e.g. the fading of lights) will no doubt be adduced as evidence of continuous change—indeed, it will be said that they are continuous change. That movement is evidence of what it is, is quite certain; but it is not so certain that it is evidence of continuous change. We may understand movement as, at each level of generality, a succession of contiguous fixed finite trajectories (to borrow Sartre's expression), and each such trajectory, at the next lower level, as a relatively faster succession of lesser trajectories, and so on indefinitely. But, as discussed in FUNDAMENTAL STRUCTURE [h], our ability to perceive distinctions is limited, and this hierarchy of trajectories is anomalously apprehended as a series of discrete continuities of displacement—which is, precisely, what we are accustomed to call movement. In other words, it is only where our power of discrimination leaves off that we start talking about 'continuous change'. (Consideration of the mechanism of the cinematograph—see the foregoing reference—is enough to show that continuous change cannot safely be inferred from the experience of movement; but it must not be supposed that the structure of movement can be reduced simply to the structure of the cinematograph film. See also FUNDAMENTAL STRUCTURE [m].) [Back to text]