The fist number of L. refer to the standard CtP edition published in 1987. The following number shows correspondence between letters in the new 2010 edition. Note that on this website CtP is available only 1987 edition with minor additions.

[L. 6 | 12] 4 March 1964

Dear Mr. Wijerama,

Many thanks for your admirably detailed letter. The attitude you speak of, that of cursing the world and oneself, is, in a sense, the beginning of wisdom. Revolt is the first reaction of an intelligent man when he begins to understand the desperate nature of his situation in the world; and it is probably true to say that nothing great has ever been achieved except by a man in revolt against his situation. But revolt alone is not enough—it eventually contradicts itself. A man in blind revolt is like someone in a railway compartment trying to stop the train by pushing against the opposite seat with his feet: he may be strong enough to damage the compartment, but the damaged compartment will nevertheless continue to move with the train. Except for the arahat, we are all in this train of samsāra, and the problem is to stop the train whilst still travelling in it. Direct action, direct revolt, won't do; but something, certainly, must be done. That it is, in fact, possible to stop the train from within we know from the Buddha, who has himself done it:

I, monks, being myself subject to birth, decay, and death, having seen the misery of subjection to birth, decay, and death, went in search of the unborn, undecaying, undying, uttermost quietus of extinction (nibbāna), and I reached the unborn, undecaying, undying, uttermost quietus of extinction. <M. 26: i,167>
Revolt by all means, but let the weapons be intelligence and patience, not disorder and violence; and the first thing to do is to find out exactly what it is that you are revolting against. Perhaps you will come to see that what you are revolting against is avijjā.

Now for flux. I see that you make a certain distinction between physical objects and mental states: let us therefore consider first physical objects. You say 'The idea of continuous change or that everything is continuously changing seems to me to be correct. But the difficulty arises when the idea is extended and it is stated that this object is not the same object. The chair that is in front of me being of matter is undergoing change. In that sense it will not be the same chair. But in another sense but much more real is the idea that the chair is there and till it breaks down it will be so. This is still valid in spite of the changes that are taking place which may or may not be perceptible so long as the chair could be used as a chair'.

The distinction that you make here between 'the idea of continuous change' and 'the idea that the chair is there' is of the greatest importance, since it marks the distinction between the scientific view and the existential (or phenomenological) view. The question arises, Are these two views compatible, or if not, which is correct?

In spite of the fact that you say 'The idea of continuous change is a matter of observation and it accords with the scientific view that matter is subject to continuous change', I wish to suggest that the idea of continuous change is not a matter of observation (I shall discuss this later), but is purely and simply a theoretical consequence of the scientific claim to achieve complete objectivity. (Science aims at completely eliminating the observer—or individual point of view—from its results, thereby attaining complete generality. As soon as the observer is reinstated, as in quantum theory, change once again becomes discontinuous. The existential view, on the other hand, is that for an existing individual the world necessarily presents itself in one perspective or another. No individual can possibly see the world as science claims to see it, from all points of view at once. See Preface (f).)

You say 'The chair that is in front of me being of matter is undergoing change'. This sounds as if you are deducing continuous change from the fact that the chair is of matter, and I suggest that what you are doing is to apply an abstract notion that you have learnt about theoretically to your concrete experience (i.e. to the 'much more real idea that the chair is there'). The fact that you speak of 'changes that are taking place which...may not be perceptible' also gives the impression that you are making theoretical assumptions about the nature of change—how do you know anything about changes that you cannot perceive? (Here is Sartre speaking about material objects that are there in front of him:

Of course someone will object that I merely fail to see changes.... But this is to introduce very inappropriately a scientific point of view. Such a point of view, which nothing justifies, is contradicted by our very perception.... [B&N, p. 205])
You say 'the difficulty arises when... it is stated that the object is not the same object'. Quite true; but you yourself show the way out of the difficulty when you say 'When it is said that the infant is not the same as the grown up man... it is correct. When it is said that it is the same infant who has grown up it is also correct...'. When an infant grows up into a man, we perceive that the infant has changed, and we express this by saying that the infant both is and is not the same as the man (we are taking the infant and the man only as physical objects, not as 'selves', which is a different question). Clearly, then, in order for us to be able to say 'this has changed' two things are necessary: (i) sameness, and (ii) not-sameness, or difference. Unless there is something that remains the same, we cannot say 'this'; and unless there is something that is different, we cannot say 'changed'.

Take your mango tree. Ten years ago it was a small plant, now it is a big fruit-bearing tree, and in virtue of this difference you say it has changed; but both the small plant and the big tree are mango, and both are in the same place (the small mango plant has not grown up into a jak tree, nor is it now in another part of your garden), and in virtue of this sameness you say that it is not another tree. Or consider a leaf that changes colour—first it is green, then when it dies it becomes brown, but it is still the same leaf. What remains the same is the shape, and what is different is the colour, and so we say 'this leaf has changed'. This is quite simple owing to the fact that vision is a double sense, giving us perceptions both of shape and of colour, and it often happens that one remains constant while the other varies.

But let us take a more difficult case, and consider a change of colour alone. Suppose I have some blue curtains, and after a time I notice that 'the blue has faded'—how are we to understand this? Obviously, if I look at the curtains one day and find that they are crimson I shall not say 'the blue has faded' for the good reason that crimson is not blue at all—it is a different colour altogether. So I shall say simply 'the curtains have changed their colour' (just like the leaf). But if I say 'the blue has faded' I am saying that the curtains are still blue, but a slightly different blue, a lighter blue. What remains the same here is the general determination 'blue', and what is different is the particular shade of blue.

Take another case. I am looking at a spoon on the table in front of me. First I fix my attention on the bowl of the spoon and see the handle less distinctly out at one side; then I fix my attention on the handle and see the bowl less distinctly out at the other side. The spoon, as a whole, remains unchanged—in both cases it is exactly the same spoon. What is different is the particular aspect of the spoon within the general experience called 'seeing a spoon'. (Cf. CETANĀ.) Two points arise here.

1. Leaving aside the cases where one sensible quality varies while another remains constant (the leaf, for example) and considering only the more fundamental cases where the change takes place within one and the same sensible quality or characteristic, we notice that it is always the more general feature that remains invariable while the subordinate or more particular feature varies. This suggests that there may be a certain structure of change that must be taken into account whenever we consider the question of change; and if this is so, it will mean that the statement 'everything is changing' needs strict qualification. (In the last part of the Notes I have tried to give a formal account of this fundamental structure within which change takes place, but I expect that you have perhaps not been able to make very much of it. No matter.)

2. If it is possible, in any given change, to make a clear-cut distinction between those features that do not vary and those that do, it will follow that the distinction between sameness and difference is absolute: in other words, that we cannot say 'approximately the same' or 'approximately different'. (So long as we use the word 'approximate' at all that will be an indication that we have failed to make the distinction properly clear-cut, since 'approximately the same' means 'the same but with a difference' and 'approximately different'—i.e. 'somewhat different' or 'rather different'—means 'different but partly the same'.) If this is so, it will follow that all change takes place discontinuously; for if 'same' means 'absolutely the same' and 'different' means 'absolutely different', there can be no intermediate category between sameness and difference.

Perhaps you will object that it is ridiculous to speak of one's curtains 'fading discontinuously', and from the commonsense point of view I would agree with you. But the fact remains that we do not 'see our curtains fading'; what happens is that one day we 'notice' that the curtains 'have faded'; and this is a sudden perception. No doubt, after a few more weeks, we shall notice that the curtains have faded still more, and we shall infer that all this time the curtains have been gradually fading 'without our noticing it'. 'But' you may say 'do we not sometimes actually see things in process of changing—as when, for example, the lights are quickly lowered at the cinema and fade in five or ten seconds?' We do: but observe that, in the first place, the change is from 'steady light' to 'fading light' and then from 'fading light' to 'darkness'. In other words, 'fading light' is perceived as a thing distinct from both 'steady light' and 'darkness', and the change from one to another of these things is discontinuous. In the second place, there are reasons for supposing that what we actually perceive when we see a 'fading light'—which has the same essential structure as a 'flying arrow'—cannot be properly described as 'continuous change'.

A. The 'Gestalt' school of psychology has specialized in experimental investigation of perception of change, and has reported that every change that we perceive takes place suddenly and absolutely. (See the passage from Sartre translated in ANICCA [a].) Whenever a perceived change is described as 'taking place continuously' it is to be presumed either that the necessary analysis of a complex experience is beyond the power of the perceiver, or else that, unwittingly, rationalization has taken place. (That we do, in fact, have experience of movement and other such changes is, of course, not to be denied; but these experiences are notoriously difficult to describe, and the problem of motion has puzzled philosophers from time immemorial.)

B. It can be shown by argument that the notion of continuous change is self-contradictory (in other words, that it contains a short circuit somewhere). There are two ways of doing this.

(i) The first is to show that all experiences that we might be tempted to describe as 'continuous change' (motion of material objects, fading [or brightening] of lights and colours, decay of matter, and so on) can be adequately and completely described in terms of discontinuous changes at different levels of generality. I am satisfied that the dialectic outlined in FUNDAMENTAL STRUCTURE is capable of doing this (which is one reason why I have included it in the Notes), but unless you have understood this note I cannot hope to make myself intelligible to you here. I have summed up this argument against the idea of flux in PATICCASAMUPPĀDA: 'The contradiction [involved in the definition of flux or continuous change] 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'. (The starting-point of any discussion of motion must always be Zeno's Eleatic arrow. Some account of this celebrated paradox is given by Bertrand Russell—M&L, pp. 79-83—but the problem is not so easily solved as Russell likes to think.)[a]

(ii) The second way of dealing with the notion of flux is to discuss it directly, and to show that it cannot be defined without en-countering a self-contradiction. This, in fact, is what I have tried to do in the briefest possible way in PATICCASAMUPPĀDA [c], with the definition borrowed from Poincaré: A = B, B = C, A ≠ C. Let us, however, consider the notion of flux in more detail. The word itself means a flowing, and the idea it conveys is that of smooth transition, that is, continuous change. This is evidently opposed to discontinuous change, but without implying no-change or fixity.

My dictionary defines it as 'a continuous succession of changes', which we can use as a starting point. A succession of changes clearly means one change after another, and a continuous succession of changes will mean that there is no interval of time between these changes. But how much time does a single change take? Either it takes some time, in which case we are obliged to say that each individual change is a continuous change, and therefore itself a flux; or it takes no time and is instantaneous, in which case we have to conclude that a flux is itself instantaneous, since the individual changes take no time, and there is no time between the changes. The second alternative at once raises the objection that you cannot have a succession of changes—one change after another—if no time is involved. The first alternative—that every individual change is a flux—makes the definition circular: 'a flux is a continuous succession of fluxes', and we still do not know what a flux is.

Perhaps, then, we are wrong in thinking that 'a continuous succession of changes' is the same as 'continuous change'. If these two are not the same, and 'continuous change' is the truth, then we must deny the existence of separate individual changes: there will be change, but not changes or a change. In other words we must renounce all attempt at defining flux in terms of individual changes, and must seek, rather, to take a sample of flux, of continuous change, and describe it. Here, then, is a flux—continuous change. Let us take a slice of this flux and divide it into three consecutive sections, calling them A, B, and C (note that we cannot take three consecutive instants in the flux without falling into contradiction, since instants, which are of no time, cannot be consecutive, i.e. both contiguous and successive—if two instants are contiguous both are of no time and have no time between them, and there is still no time and therefore no succession; if they are successive both are of no time and have some time between them, therefore they are not contiguous).

We have to ignore for the moment the fact that each of these three sections itself consists of continuous change, and we regard each section as a whole, without inquiring what is going on inside. We are expressly forbidden to introduce the idea of an individual change, and so we must say that 'A is the same as B' (A = B) and that 'B is the same as C' (B = C); for if we postulate that A and B (or B and C) are both contiguous and different we thereby automatically define a discrete individual change—there is 'a change' at the junction of A and B, where A changes to B. So far so good. But a flux is, in fact, change; and so we must introduce the idea of difference into our description. Let us therefore say that 'A is different from C' (A ≠ C). Since A and C are not contiguous we have not defined any discontinuous change between them, and all is well—between A and C there is change but not a change. So our description—A = B, B = C, A ≠ C—does, in fact, agree with the notion of flux as continuous change. And we can take each individual section (A, B, and C) in turn and divide it into three lesser sections (a, b, and c) and describe it in the same way (a = b, b = c, a ≠ c). In this way our description can be seen to apply to any sample of the flux that we like to take. But, alas! our description contains a self-contradiction: B = C (or C = B) and A ≠ C; therefore A ≠ B; but also A = B; therefore both A = B and A ≠ B; and this outrages the Law of Contradiction, 'A is not both B and not-B'.

Regarding states of mind, which you differentiate (quite rightly) from physical objects in that they do not come within the sphere of science (though I cannot agree that they are 'not objects': they are mental objects),—you seem to think, and again you are right, that the notion of flux or continuous change does not apply to them. I have a slight impression that one reason why you do not apply the notion of flux to mental states is, precisely, that they are not in the sphere of science; and this, in its turn, suggests to me that you do apply the notion of flux to physical objects because they are in the sphere of science—in other words, out of 'superstitious regard for the prestige of contemporary science' (see Preface to Notes). It is quite possible that I am doing you an injustice here, but this is a matter that you must decide for yourself—in any case, I am only recording the impression that I get from your letter.[b] But though I say that you are right in thinking that the notion of flux cannot be applied to states of mind, you will have gathered from what has gone before that I maintain that the notion of flux also cannot be applied to physical objects. Once the notion of flux is ruled out entirely, it becomes clear that the structure of change of mental states (or mental objects) has much more in common with that of physical objects than might appear at first sight. (You say that mental states such as pleasure and grief 'appear, vanish, and reappear'—but is this not true also of physical objects? Do we not have familiar sights, sounds, smells, tastes, and bodily contacts?) It is necessary to remember that the three characteristics (Notes, ANICCA), namely arising, disappearance, and change while persisting, apply to all experience, whether of physical objects or of states of mind. (The last characteristic, thitassa aññathattam, I understand as expressing the combination of absolute sameness and absolute difference that I suggested earlier in this letter was the essential structure of all change.)

As I understand your last paragraph, I gather that you consider that all mental states cease when one becomes arahat. This is not so (except in the particular sense of 'cease' of A NOTE ON PATICCASAMUPPĀDA §22 & VIÑÑĀNA). There are still mental states for the arahat so long as he continues to live, but these states are now wholly free of lust, hate, and delusion. In other words, there is still consciousness for the arahat until his body breaks up in death. See also PHASSA [b].

Perhaps you will be wondering why it is that I am so anxious to destroy the notion of flux—or at least to eliminate it from the context of the Dhamma (I have nothing to say against its use in the context of science, nor have I anything to say against science itself in its proper place; but its proper place is not the Dhamma: scientific thinking and Dhamma thinking belong to two quite different orders, as I hope to have made plain in the Preface to the Notes). The reason is to be found in your letter itself. You say 'The word flux means continuous change. If this idea is applied to everything it would be correct to say that what I see now, e.g. a tree, is not the same as I continue to watch it as it is subject to continuous change' and also 'I have heard as an extension of the same idea, Buddhist monks saying, pointing to an object, that the object is not there'. This doctrine is a complete misunderstanding and is wholly misleading. And, as you quite rightly point out, it is based on the notion of universal flux. In order, therefore, to undermine this false doctrine, it is necessary to point out that the notion of flux, at least as applied to experience, is a self-contradiction.

But why, if it is false, is this doctrine taught? The answer is, because it provides a conveniently simple interpretation of the Suttas, easily learned and easily preached. The Buddha has said that 'What is impermanent, that is suffering; what is suffering, that is not-self'. This is understood (or rather, misunderstood) in the following way.

Impermanence is taken to mean continuous change (flux), and (as you have said) if this notion is correct, the idea of a thing's continuing self-identity cannot be maintained—what appears to be the self-same tree persisting in time is not really the same since it is continuously changing. In consequence of this, the idea of self is an illusion; and it only persists on account of our avijjā, or ignorance of the truth of universal flux. If we remove this ignorance, we shall see that what we formerly took to be a lasting (or existing) selfsame tree ('A = A', the Principle of Self-identity) really has no abiding self at all—it does not really exist. And this explains why 'what is impermanent, that is not-self'. And what is wrong with this? What is wrong with it is—as perhaps you have noticed—that it does not explain why what is impermanent is suffering, and what is suffering is not self.

Suffering (dukkha) is the key to the whole of the Buddha's Teaching,[c] and any interpretation that leaves suffering out of account (or adds it, perhaps, only as an afterthought) is at once suspect. The point is, that suffering has nothing to do with a tree's self-identity (or supposed lack of self-identity): what it does have to do with is my 'self' as subject (I, ego), which is quite another matter (see PARAMATTHA SACCA §6). As I point out (ATTĀ), 'With the question of a thing's self-identity (which presents no difficulty) the Buddha's Teaching of anattā has nothing whatsoever to do: anattā is purely concerned with "self" as subject'. But this is very much more difficult to grasp than the misinterpretation based on the notion of flux, so flux inevitably gets the popular vote (like the doctrine of paramattha sacca, of which it is really a part). The misinterpretation is actually of Mahāyānist origin; and in one of their texts (Prajñāpāramitā) it is specifically stated that it is only on account of avijjā that things appear to exist, whereas in reality nothing exists. But the fact is that, even when one becomes arahat, a tree continues to have a self-identity; that is to say, it continues to exist as the same tree (though undergoing subordinate changes on more particular levels—falling of leaves, growth of flowers and fruit, etc.) until it dies or is cut down. But for the arahat the tree is no longer 'my tree' since all notions of 'I' and 'mine' have ceased.

I don't know whether all this discussion will make my criticism of the notion of flux any clearer to you, but it may at least make you aware that there are serious objections to the introducing of this notion from scientific contexts into Dhamma contexts. If this letter raises any fresh difficulties, please let me know.

P.S. If you do not want to keep this letter when you have finished with it, I would suggest that, rather than destroy it, you might give it to Mr. Samaratunga to put in his file.


[6.a] The solution described by Russell solves the problem by leaving it out. The problem is: What is time? [Back to text]

[6.b] It is perhaps worth noting in passing that the current 'orthodox' interpretation regards mental states as no less of a flux than physical objects. Here is an example: 'The stream of self-awareness that the uninstructed conceive to be a soul is made up of point-moments of consciousness, each of which has no more than a momentary duration.' This is pure speculation, with no relation at all to actual experience. [Back to text]

[6.c] 'Both formerly, monks, and now, it is just suffering that I make known and the cessation of suffering.' <M. 22: i,140> [Back to text]