by Ted Honderich

What is the connection between beforeness and the like and pastness and the like? Are the temporal relations of beforeness and so on the real facts about time? This is still the main philosophical problem about the nature of time -- at any rate outside of the arcane sub-department of philosophy that is the Philosophy of Science. Are the somehow obscure matters of pastness and the like somehow reducible to the plain facts of beforeness and the like? Many empirically-minded philosophers have thought so, but it can seem, as the paper below argues, that there is a certain disproof of the project.  

Let us have in mind three visits to the earth of a recurrent comet, say Halley’s, and one other event, the falling of a leaf. In speaking of a visit of the comet, we understand what we need not define closely, a movement of the comet in which it is near to the earth. We have, let it be supposed, some way of distinguishing each of the three visits which does not have to do with its time, and hence not with its relations in time to the other two visits. Let us suppose that we identify it by perhaps its particular visibility from England - perhaps the particular degree to which it was obscured by clouds. Thereafter we refer to one visit so identified as the first mentioned visit or, with exactly the same meaning, the first visit, and refer to the other two in the related ways. 

We have certain beliefs about the four events. The first-mentioned visit of the comet, we believe, is before the second, the second is simultaneous with the falling of the leaf, and the third is after the second. Each of the events, then, has what may be called a temporal relation to another event. This is to say, only, that it is before, simultaneous with or after another event.[1] Taking into account only the four events and no others, each has but three temporal relations, one with each of the other events. 

It would be more natural to make use of ordinary tensed verbs and thus to say, for example, that the first-mentioned visit was before the second, or is happening before it, or will be happening before it. However, what is in question here is a relation which seems to be independent of tenses of past, present and future. Despite the fact that the relation can be asserted in tensed sentences, and ordinarily is, they also carry further meaning. Hence, it will be best to stick to a tenseless usage. 

There is also something else which we believe of our four events, something which does have entirely to do with tenses. Each event has happened, is happening, or will happen. Each one is past, present or future. Each, then, has what may be called a temporal property. This is only to say, in one way or another, that it is past, present or future. 

Temporal relations are unchanging. Given that the first-mentioned visit of the comet is before the second, then it always stands in just that relation of precedence to the second. An event’s temporal properties are otherwise in that they do change. The falling of the leaf, if we suppose that it is happening now, was once going to happen and soon will have happened. The event was future but ceased to be. Is present, and will be past. 

Here and hereafter, it will be important to keep in mind that when temporal properties or temporal relations are mentioned, terms of convenience are being employed. They are merely invented abbreviations for more ordinary usage. To say an event has a temporal property is merely to say it is past, present or future. Similarly, to say that temporal properties of events change means only whatever we ordinarily mean when we say, in one of several possible ways, that events are in turn future, present and past. It is not being suggested that events possess properties in one of the several different manners in which items of one kind and another possess properties. It is not being suggested that temporal properties are like, for example, colour or location. So with the term ‘temporal relations’. It is just an abbreviation of ordinary talk. 

The fact that events have both temporal relations and also temporal properties enters into one of the fundamental disputes about time in Western philosophy. Given the nature of this book, my primary intention is to report on this controversy. My exposition of the competing doctrines, for several reasons, will be cursory and partial.[2] To advance the dispute, rather than report on it, would require a kind of close argument which here is out of place. My lesser purpose is to offer some informal argument.[3] 

The problem which we shall consider involves what is meant by saying that an event is earlier than another, or stands in some other temporal relation, and what is meant by saying that an event is past, or has some other temporal property. The questions are compelling in themselves. Moreover, the answers to them are highly relevant to other mysteries about time. 

A cautious analysis of temporal relations 

With respect to the two temporal relations of antecedence and subsequence. as distinct from the relation of simultaneity, a part of our understanding is beyond doubt: (a) if the first-mentioned visit of the comet was before the second, then the second was not before the first; (b) if the first was before the second and the second before the third, then the first was before the third; (c) finally, it would be absurd to say that the first visit was before the first visit. There are three related truths about our understanding of the statement that one event was after rather than before another. 

The two relations in question, then, are understood by us to have certain formal properties: those of asymmetry, transitivity and irreflexivity. In these three properties they are unlike certain other relations. The relation of equality, for example, is symmetrical and reflexive, and the relation of fatherhood is intransitive. 

Our understanding of the temporal relations of antecedence and subsequence, however, certainly has more in it than that. This follows from the fact alone that other relations, having nothing to do with time, do have the same formal properties. We may say of items in space, perhaps the furniture in a drawing-room, that from a certain point of view one piece is to the left of another. Here too we have a relation which is asymmetrical, transitive and irreflexive. 

Given the gravity with which some philosophers produce these formal properties, it needs making plain how very little we know about temporal relations in knowing that they have the properties. Take asymmetry. What we know about the statement that the falling of the leaf is before the third visit of the comet, given a grasp only of the asymmetry, is only that the falling is so related to the visit that the visit cannot be related in that way, whatever it may be, to the falling. We know virtually nothing of the nature of the relation. All we know is explicitly given in the statement that since the visit stands in relation R to the falling, the falling does not stand in relation R to the visit. We do not know how to understand R. Given that we have a grasp only of asymmetry, indeed, we cannot even distinguish antecedence from subsequence. We do not understand the difference between saying that the falling is before the visit rather than after it. 

A second part of our understanding of the two temporal relations, it may be said, is simply that in talking of items being before or after one another, we are talking of relations which are taken to connect items of a certain kind: events. These relations are not understood to connect such items as locations or numbers. How far does this take us? We may understand an event to consist, in part, in a thing’s having a certain feature. Do we, in seeing that the relation of antecedence has events as its terms, see more about the relation? The answer to that question, despite the fact that a full definition of an event may include a reference to time, seems to be yes. We could not be allowed to understand the relation of fatherhood at all, for example, if we did not understand that its terms could not be locations. Exactly what it is that we understand, however, when we see that temporal relations hold between events, is unclear. 

Thirdly, talk of temporal relations may be said to have to do with change. That is, to know that our four events stand in their various temporal relations is to know of different states of the world. One state includes the first visit but none of the other three events. Another includes the second visit and the leaf’s falling, but not the first or the third visit. Another state includes the third visit, but none of the other three events. 

To say the first visit is before the second, then, is to talk of two different states of the world and hence of what can be called change. 

This is likely to seem to contribute more to the understanding of temporal relations. However, to mention but one point of relevance, it is essential that the particular notion of change be made clear. What we are understanding by 'change’ is no more than a matter of different states. In the sense in which change is involved in temporal relations, according to the suggestion, there is precisely the same kind of change in the fixed line of hills we see on the horizon. In one part the line is smooth and in another it is broken. The line has ‘successive’ segments or 'states’. We are not entertaining the idea of change which seems to have to do, as we may obscurely say, with a thing’s coming into existence. 

Is there a fourth distinct thing involved in talk of the two temporal relations? Is there something over and above their formal properties, the fact that they hold between events, and the fact that they involve a kind of change? It seems that there must be, principally for the reason that we still seem to have no distinction between antecedence arid subsequence. What has been suggested is that these relations, and hence, a series of events defined in terms of these relations have a direction. Our own series, it may be said, has direction from the first-mentioned visit of the comet to the third. Here we come upon one of many terms of a metaphorical kind, which are to some philosophers the very guide to reflection about time. To other philosophers they are its very curse. 

Those who speak of the direction of the temporal relations, and of the resulting series of events, are presumably not concerned with the fact that we commonly think of precedent events before we think of subsequent ones. We can in our thinking go the other way, and we often do. Rather, they are concerned to assert that the relations and the series themselves have a direction. In this, such relations and series are unlike the spatial relation of one thing’s being to the left of another, or of a series of positions in space defined on the basis of that relation. 

Can it be that this intuition of direction, despite its evocative nature, actually has to do only with something relatively unmysterious? Some physical processes, notably increase in entropy, are regarded as irreversible. That is, they are processes such that an initial state of a given kind is not followed by another state of the same kind. Hence, if one passes in thought one way along a series of events, one comes upon certain sequences or orderings of events which do not occur if one passes along the series the other way. What we have here, it appears, is a further characterization of temporal relations got from a feature of the series of terms defined by these relations. Much must go unsaid, but it may seem that this irreversibility will do the job of distinguishing antecedence and subsequence in the way we do. 

Can it be that a search for a grander thing, in so far as the direction of temporal relations is concerned, is no more than the result of allowing one’s mind to wander from the business at hand? Certainly, there seems to be a pretty impressive 'direction’ in events when they are considered as future, present and past. The same is true when we simply think of events, in an ordinary and unanalytical way, as in time. However, we are not now concerned either with the temporal properties of events or with general talk that has to do with time. We are concerned only with events taken as before or after one another. 

I have, as some will be aware, been guided in most of these remarks by an awareness of one tradition of philosophical thought about temporal relations. For want of a better name, and for reasons to which we shall come, it may be called the tradition of caution. We shall return to the question of temporal relations and to another tradition of thought about them. For the moment, we have the hypothesis that our statements about the precedence and subsequence of items have as their content (a) an awareness that the terms of the relations are events, (b) certain beliefs about formal properties, (c) a belief having to do with what is called change, and (d) some belief or other about direction. 

This sketch of one possible analysis of statements about temporal relations is lamentably unfinished. It raises many questions which must go unanswered. However, it is worth noticing that it has a further feature so far unmentioned. What I have in mind is (e) that this analysis of statements about temporal relations does not have recourse to temporal properties. It does not attempt to explain the belief that one visit of the comet is before or after another by bringing in familiar notions of past. present or future. The same is true, incidentally, of a closely related analysis of the temporal relation of simultaneity, which we have left untouched in all of this. 

A cautious analysis of temporal properties 

Let us turn now to our beliefs about the temporal properties of our four events. Let us, for the purpose of reflection, suppose that the second visit of the comet is now happening, the first is past, and the third is to come. If we believe these things, what is it that we believe? If we are guided by the philosophical tradition of caution already mentioned, we shall answer that the meaning of such statements about temporal properties has to do with temporal relations conceived in the way that we have just conceived them. 

The answer, in one clear version, can be given very quickly. It is that my statement, which we may call S1 , that the second-mentioned visit is now happening. amounts to this: the second-mentioned visit is simultaneous with my utterance of the statement S1. We shall say that the statement, S2, that the first visit is now past, amounts to this: the first visit is before my utterance of the statement S2. As for the statement S3, that the third visit is to come, or is in the future, it amounts to a claim about the third visit’s being after my utterance of S3. All this, at least after a moment’s reflection, may seem natural and persuasive. 

If we bring together the analyses of temporal relations and temporal properties, we have a view of our talk of time. Let us notice two principal features of the view. Past, present and future are reduced to the relations of precedence, simultaneity and subsequence. We have, in all of our talk about time, no more than ideas of the kind supplied with respect to temporal relations. All of our talk about time, including talk of temporal properties, is reduced to the particular ideas which we supposed ourselves to have in speaking of events as before other events, simultaneous with them, or after them. The relations which we are said to have in mind in speaking of past, present and future are different from others of their kind only in one way: they have as one term a linguistic event, and thus an event dependent upon the existence of consciousness. 

There is then a second principal feature of the view which we are entertaining. Given the particular understanding of temporal properties, there would be no such properties in a world without consciousness. The point is not, of course, that the properties would exist, but that there would be no awareness of them or beliefs about them. There would in fact be no such properties, since the existence of such properties depends on the existence of consciousness. 

More explicitly, the view we are entertaining has the consequence that to say an event is now happening is to assert or presuppose the existence of consciousness. Given that view, what we believe in believing that the second-mentioned visit of the comet is happening now, would not be true if no one were here to make statement S1 or some version of it. If the world were without conscious inhabitants, the second-mentioned visit of the comet would indeed be after the first, simultaneous with the falling of the leaf, and before the third visit. If what we mean by know’ and our tensed utterances is what we are supposing we mean, it would not be true that the second visit was happening now, or, as we ordinarily say. happening. It would not be true that the first-mentioned visit had happened or was past, or that the third visit was to happen or was in the future. 

There is no doubt that when this second feature of the analysis before us is made clear, some people will be quick to conclude that the analysis is unutterably mistaken. They may also conclude that arguments against it can easily be discovered. There is something to be said for restraint, however. Let us turn to the alternative tradition, and first to some reflection on temporal properties rather than relations.  

An affirmative analysis of the temporal property of presentness  

When we depart from our ordinary thoughts and utterances, in which items are taken to be past, present or future, and direct our attention to the substantive term, ‘the present’, we come upon a host of quite familiar usages. It is natural to say, for example, that the present passes, or passes away. We may indeed become enthusiastic and declare, with one philosopher, that the present has 'the jerky or whooshy quality of transience’. Such utterances and declamations, no doubt, are distantly related to some truth. They may also seduce one into confusion.  

Some philosophers, it appears, have been led by them to conceive of the present as a thing, and indeed a thing distantly related to the rolling stock of British Rail. The principal difficulty with such ideas, however expressed, is that all movement in the end consists in, or at least requires, a thing’s being in different places at different times. If ‘the present’ is a kind of moving thing, in what time does it do so? We started with the present, or, as we can say, present time, and we must now have another time in which it makes its progress. Does that further time also have a moving present? 

Let us leave all that without further ado. Let us not forget, however, our familiar usages about 'the present’. We are not so well supplied with kinds of data about time that we can ignore any we have. Let us keep the familiarities at hand, forget about time as a moving thing, and return to the statement that the second visit of the comet is now occurring. What does it mean? 

Perhaps the truism that the second visit is neither future nor past suggests a line of reflection. It makes some sense or other to say that the future does not yet exist, and the past no longer exists. Can we then attach sense to the statements that future events do not yet exist and that past events no longer exist? If so, we may move to the obscure speculation that to say that the second visit of the comet is now occurring is to say that it now exists. 

Whatever else we have, we now appear to be in line with our familiar usages. We need not take the step of ignoring our notion that the present passes. We can attempt to understand it as having to do with such matters as a thing’s coming into existence, being in it, and going out of it. 

Still, we are in difficulty. If we consider one of the things just said, that past events are events which no longer exist, one of our embarrassments may be that it seems to amount to this: past events do not exist now. We shall not have made much progress if we explain our use of ‘past’ by introducing, in the explanation, a use of ‘now’. We wish, somehow, to give an explanation of all the terms by which we ascribe temporal properties. We do not wish merely to make connexions between the terms, to ‘define’ one by means of the other, but rather to get outside the ring of terms. Similarly, we have just supposed that to say the second-mentioned visit of the comet is occurring now is to say that it, the visit, now exists. How much good is that? The difficulty is plain circularity. ‘Now’ turns up once too often. 

Also, and differently, there must be some doubt that we have made a clear advance by moving from talk of the visit’s happening to talk of the visit’s existing. We need not be too perturbed by the response of the plain man, or the plain philosopher, that it is things which exist, rather than events. We can admit that the usage is unusual, and attempt to justify it. Let us look at this difficulty first and then return to the matter of circularity. 

How does it come about, as it does, that philosophers are led to assertions about time like the one we are considering? The explanation often has a great deal to do with the cautious analysis of temporal properties at which we have looked already. That analysis issues in a particular understanding of the statement that the second visit of the comet is now happening. The happening of the visit is made into a matter of a simultaneity relation. This, it is felt, is wonderfully inadequate. It is true that when we say truly that the second visit is now happening, our utterance is simultaneous with the visit. Surely, however, that is not what we are saying, or all of it. Something other than that is being said. 

In order to attempt to express what seems to be meant by the statement that the visit is happening now, the philosophers in question have recourse, as we have had, to the declaration that it now exists. We need not look at several variations on this theme. 

The trouble is that in so far as analysis is concerned, we get very little from this talk of existence. Until more is said, the clear content that it has seems to amount to two implications. It is implied that to say the second visit is happening now is not to say something, or not only to say something, about a simultaneity relation. It is implied that the present happening of events would continue if we were not here to be aware of events. 

What are we to understand that the statement about the second visit states? The only answer I can see, one to which we might as well come sooner rather than later, is that in saying of the second visit that it is now happening, we are saying something of the visit which we all understand but which is not open to further analysis. We here have an unanalysable or primitive notion, expressed by such descriptions of something as ‘happening now’. 

If we take up this view, we shall certainly have to put up with the standard retort made to anyone who asserts about a controversial notion that it is unanalysable or primitive. The retort is that saying the notion is primitive is no more than an admission of failure in the task of analysis. It is to be replied, perhaps, that sonic such ‘failures’ consist in the perception of truth. The existence of certain primitive notions in our conceptual scheme is undeniable and perhaps has never been denied. Is there adequate reason for denying that we have another one here? Certainly time is in some sense basic to our perception and reflection. What more likely place to find an unanalysable concept?  

What we have, then, is the hypothesis that to say the second visit is happening now, or simply happening, is to say something about the visit which is understood by all and which is in a certain sense simple or fundamental and thus not open to reduction. And, since we have not got an analysis of statements about present events, we have not got a circular analysis of them. That objection is no longer a possibility. The circularity that was apparent in our attempted analysis, however, is not without importance. It, like the many circularities which emerge in discussion of time, is an indication that we are dealing with a primitive concept. As for circularity that may arise in analyses of statements about past and future, we shall return to the matter. 

One of two questions that remain about our hypothesis as to statements about present events has to do with the formulation of the hypothesis. Shall we persist in saying that the hypothesis is that such assertions as the one about the second visit ascribe existence to events? Perhaps there is no harm in this so long as we know what we are up to. The hypothesis, if we choose to assert the implications mentioned above, consists in three propositions: (a) statements about temporal properties do not have to do with, or only to do with, simultaneity relations; (b) such statements would be true if there were no consciousness; (c) such statements rest on a primitive notion of existence. 

The other question is this one: are we in fact engaged in an absurd enterprise of inventing a notion that is simply redundant? ‘We all agree’, someone may say, ‘that the second visit of the comet is now happening. That is, the comet is near the earth. The comet has that feature. Can we really suppose sensibly that on top of that, something else is importantly true, something which you refer to by saying that the visit has "existence"?’ The proper answer must be that we are not adding anything. What we are doing is supposing that to say the second visit is now happening, or that it is happening, or that the comet is (now) near the earth, or that the comet (now) has a feature, is to make use of a notion for which there is no analysis.  

We have another account, then, of one class of statements about temporal properties, those that have to do with present events, It is an account in accordance with the second tradition of philosophical thought about time. The first tradition, that of caution, has the distinction among others of being informed by a reluctance to take up anything about which clear things cannot be said. The second tradition expresses, among other things, an attitude of affirmation. We may call it the tradition of affirmation. What is affirmed is principally one conviction about the present, and it is affirmed in spite of the fact that nothing of an analytical kind can be said about it. 

An affirmative analysis of past and future and of temporal relations 

To continue with temporal properties, we must have some view of past and future. It seems sometimes to be supposed by philosophers of the second tradition that all of our talk of temporal properties can be explained, if that is the right word, by seeing that it rests on a notion of 'becoming’, or 'passage’ or 'real transiency’. Here we appear to have bundled up into one bag all of past, present and future. One may wonder, of course, if these terms are intended to have to do only with talk of events being present. 

That is, one may wonder if they are merely other ways of referring to the primitive notion with which we have been concerned. If so, their intended use is entirely analogous to our use of the term 'existence’. The suspicion that more is intended, that the terms in question arc thought to enlighten us about all of past, present and future, is sometimes reinforced by the fact that nothing at all is said explicitly about past and future. No separate account whatever is given of past and future. 

Be that as it may, and to come to something clearer, it would be eccentric to claim that we have a single primitive notion of past, present and future. After all, we do distinguish these three things. It would also be unacceptable, surely, to call into being two more primitive notions, one for past and one for future. In so far as there is an argument for the assertion that a notion is primitive, it must consist partly in the fact that no tolerable analysis for it can be found. Given that we have what we seem to have. including a notion of events as present, this turns out to be untrue of past and future. 

One possibility about past and future that must come to mind is that they be analysed in terms of the notion of the present and the notions of precedence and subsequence. That is, we make use of those two temporal relations, along with our concept of the present, in order to explain past and future. What we consider, obviously, is that to say an event is past is to say that it is before present events, and to say an event is future is to say that it is after present events. 

Let us reserve judgement on that possibility and turn our attention directly to temporal relations. Let us reconsider their analysis. Philosophers in the tradition of affirmation have usually been unsatisfied with the kind of analysis given of temporal relations in the tradition of caution. They take it to be mistaken, or anyway inadequate, to describe the relations of antecedence and subsequence by way of the nature of their terms, formal properties, and, in certain limited senses, change and direction. 

Their move at this point is also predictable enough. It is said that part or indeed all of what we mean in saying the second-mentioned visit of the comet is after the first, is that when the first visit is present the second is future, and that when the second is present the first is past. A similar line is followed with the relation of antecedence and that of simultaneity.  

Whatever else is to be said of this analysis of temporal relations, it is clear that it has one consequence. If we adopt this analysis of temporal relations, we must give up the possibility noticed a moment ago with respect to the temporal properties of past and future. If we adopt both things, we shall again have a circularity. We shall be offering an analysis of past and future which depends in part on the notions of antecedence and subsequence, but those latter notions will themselves be explained partly in terms of past and future. 

There seems to be a way out of this difficulty, although one cannot follow it with great confidence. 

If, like the philosophers of the tradition of affirmation, we are inclined to think the cautious analysis of temporal relations is mistaken or inadequate, what is the root of our inclination? It is, in a sentence, that what is said of temporal relations makes them too much akin to spatial relations. It would be mistaken, for obvious reasons, to claim that the cautious analysis of temporal relations does not make them at all different from spatial relations. None the less, one feels there is not enough difference. 

To put the question one way, what can be added in order to get time into the analysis of temporal relations? Let us first recall that items in temporal relations form a series. This fact, which is obvious enough, can be explained partly in terms of the formal properties of the relations. The four events with which we have been concerned are members of this series, which also includes events before the first-mentioned visit of the comet and after the third-mentioned visit. We are inclined to regard the series as in a way infinite. Be that as it may, and to come to the essential point, let us say of this series that some one event in it, and all events simultaneous with that event, are occurring or are present. That is, they are such that the predicates which express our primitive notion are true of them. In short, then, we add to our account of temporal relations that they hold between events which form a series of which some member-events are present. 

This additional characterization of temporal relations is to be distinguished from something we have considered already: the characterization of them as holding between events. At any rate, it can and is to be distinguished from anything that might be meant by the latter characterization within the cautious tradition. There, presentness is restricted to a matter of simultaneity relations. Whatever may there be added to the analysis of temporal relations by specifying that they hold between events, it cannot be our current idea. All that talk of events being present is allowed to mean in the cautious tradition is something about simultaneity relations. Nothing would be gained by saying, in explanation of the relations of antecedence, simultaneity and subsequence, that they hold between terms which can enter into certain simultaneity relations. 

One advantage of the suggestion that temporal relations be tinder-stood partly as relations between items which form a series having a certain feature is the advantage that we can now complete our analysis of temporal qualities without circularity. We are enabled to say that past events are to be understood as events before present events, without the embarrassment that the notion of antecedence has been explained partly in terms of past events. Similarly, we seem to be enabled to say that future events are to be understood as events after present events, without the embarrassment that the notion of subsequence has been explained partly in terms of future events. 


Let us sum up the accounts of temporal relations and temporal properties. The account of temporal relations, taken from the tradition of caution, is to the effect that such a statement as this one, that the first-mentioned visit of the comet is before the second, conveys four things. The related terms are events, the relation has certain formal properties, implies the existence of a kind of change, and has what is called direction. The account of temporal properties taken from the tradition of caution is to the effect that such a statement as this one, that the second visit of the comet is now occurring, conveys that the visit is simultaneous with the utterance of the statement. 

Temporal relations, given the account suggested a moment ago, one within the tradition of affirmation, are to be understood in terms of the four features plus the further one that the terms of the relations form a series such that some of its member-events are present. As for temporal properties, the statement that the second visit is now happening is taken to involve the use of an unanalysable notion. The statement that the third visit is to come is taken to mean that it is after present events, and the statement that the first visit has occurred already is taken to mean that it is before present events. 


To repeat something said in the beginning, I do not suppose that a decision can be made between these doctrines on the basis of what has been said here. We have before us only hurried propositions and inconclusive reflections. None the less, if a verdict cannot be reached, we may at least speculate about one. 

The cautious account of temporal properties seems unpersuasive. Indeed, it seems open to something like a disproof. Take one example, the cautious analysis of the statement S3, that the third visit has not yet happened, or is future. The analysis must surely be mistaken. If I make the statement, what it means according to the analysis is that the third visit is after my utterance. The statement means, in short, that two events are in the temporal relation of subsequence. Temporal relations, as we know, do not change. Hence, when the third visit is happening, and when it is past, it will remain true of it that it is after my utterance of statement S3. But surely it cannot be that what is meant by S3 is something that will be true when the third visit is present or past. 

We may notice at this point, if only in passing, a related doctrine which is not open to this objection, whatever else may be said of it. What it amounts to, in terms of our example, is that S3 conveys that the third visit of the comet is after a present mental event of mine, one associated with my utterance. Similarly, the statement S1, that the second visit is now happening, conveys that the second visit is simultaneous with a present mental event of mine, one associated with my utterance of S1. Such mental events, it is allowed, are present in a sense in which physical events are not present. What can be meant by this? It is fair to observe, I think, that no answer to this question is given. One is driven to suppose, in the absence of any instruction to the contrary, that when mental events are allowed to be present, what is being relied upon is a primitive notion. 

There are several reasons why this particular view does not command attention. To mention but one, the admission that mental events have a temporal property, conceived in a certain way, seems to lead inevitably to the admission that physical events also have this property. This is so, essentially, because of the connexion between mind and body. The proponents of the view have struggled manfully to avoid this further admission. In my opinion, they have not succeeded. Hence the view, although encumbered with propositions about simultaneity, comes to rest in the opposing tradition of affirmation. 

Putting aside temporal properties, the tradition of caution also seems open to question in its analysis of temporal relations. The objection here must be less clear-cut. Philosophical inquiries of the kind in which we have been engaged are directed, importantly, by pre-philosophical convictions, perceptions, guesses and so on. Analyses, in the end, are assessed against these things. No one will suppose, on reflection, that all such convictions and what-not are sacrosanct. Some of them can be rejected for the reason that they do conflict with the outcomes of philosophical inquiry. Still, to put the point far too quickly, it is difficult to escape the feeling that the cautious analysis of temporal relations is the loser in its real enough conflict with pre-philosophical commitments. One persists in the view that we do mean a good deal more in our temporal-relation utterances than is allowed by the analysis. 

To turn to the tradition of affirmation, and more particularly to the particular affirmative doctrine at which we ourselves arrived, it appears that it is not open to refutation, at least in any economical way. Also, it appears not to be involved in a losing conflict with pre-philosophical commitments. That is all very well, but it is not all that matters. The foundation of the doctrine, obviously, is the claim that there exists a certain notion that is beyond analysis. After the claim was first introduced, we did notice the possible reply that such claims about primitive notions are merely confessions of failure. There is, as we saw, the possibility of the brave rejoinder that the ‘failure’ may be a happy one since ‘success’, or the production of an analysis, would simply be a mistake. However, a brave rejoinder is not to be confused with coercive argument. There remains the possibility that an analysis, something deserving of the name, can be given of statements about temporal properties. 



1. Throughout the essay, I speak of the three principal temporal relations only. However, there are obviously more than three such relations. An event is also just before another, long before another, two minutes after another, a century after another, and so on. I ignore these further relations, which raise no special problems for me. 

2. More can be learned of these doctrines from The Philosophy of Time, a Collection of Essays (London, 1968), edited by Richard M. Gale. The book contains essays by different philosophers and also an extensive bibliography. 

3. I am grateful to my colleagues, Dr Malcolm Budd and Dr John Watling, for comments. 


This piece was first published as 'Temporal Relations and Temporal Properties' in the book Time and Philosophy, edited by Paul Ricoeur.