Version classiqueVersion mobile

La métaphysique du temps : perspectives contemporaines

Alexandre Declos
Claudine Tiercelin

I- Intuitions, sciences, métaphysique : les images du temps

Memory, McTaggart and the Descriptive Metaphysics of Time

Robin Le Poidevin


In the debate over the question whether time passes in some objective, mind-independent sense, as it has developed over the last half century or so, those who have argued against objective passage have typically conceded that our ordinary conceptual scheme represents the presentness of events as non-perspectival, while insisting that in reality presentness is purely perspectival. To employ P.F. Strawson’s distinction between descriptive and revisionary metaphysics, we could express this position as proposing a discrepancy between the descriptive and the revisionary metaphysics of time. Deniers of time’s passage have also typically (though not invariably) appealed, as a key part of their case, to McTaggart’s contention that the idea of non-perspectival presentness, together with the associated ideas of non-perspectival pastness and futurity, lead to contradiction. This combination of views, however, implies that our ordinary conceptual scheme is incoherent in this respect. The implausibility of such a conclusion should prompt a reconsideration of how our ordinary conceptual scheme does represent time. This paper uses an exploration of episodic memory as a way of motivating the view that we naturally represent presentness and pastness as perspectival. It goes on to ask whether this temporal case study has implications for other instances where there are alleged discrepancies between descriptive and revisionary metaphysics.

Texte intégral

1. Two kinds of metaphysics

1In his 1959 book Individuals, P.F. Strawson famously distinguished between two kinds of metaphysical enquiry:

Descriptive metaphysics is content to describe the actual structure of our thought about the world, revisionary metaphysics is concerned to produce a better structure (...) But (…) [r]evisionary metaphysics is at the service of descriptive metaphysics. (Strawson 1959: 9).

2Strawson’s own project, as the subtitle of the book intimated, was one of descriptive metaphysics. The resorgimento of revisionary metaphysics had to wait another decade or so. But the great philosophers of the past provided Strawson with some (perhaps somewhat ambiguous) examples of the two approaches. Kant, for example, he classed as a predominantly descriptive metaphysician: the exploration of the phenomenal world (to which Kant restricts us) is an exploration of our ordinary conceptual scheme. The world as it is in itself is largely unknowable. Berkeley, in contrast, counts as a revisionary metaphysician, since we do not ordinarily conceive the world to be solely composed of minds and ideas. Both of these philosophers, nevertheless, were apparently prepared to work both sides of the divide. Kant’s First Antinomy of pure reason (Kant 1789/1929) was intended to show that the world as it is in itself could be neither temporal nor spatial: that is a significantly revisionary thesis! And Berkeley professes to ‘side in all things with the Mob’ (Berkeley 1707: 405), that is, to defend commonsense. (Though not everyone is convinced by what they see to be a disingenuous statement, one can represent his arguments for idealism as demonstrations of what our ordinary understanding of such concepts as perception, ideas and representation commit us to.) As for revisionary metaphysics being at the service of descriptive metaphysics, the idea perhaps is that contemplating the sometimes very strange products of revisionary metaphysics may bring into sharper focus aspects of our ordinary thinking about the world.

3Strawson’s distinction is a very useful one. We should note, however, that not all non-descriptive metaphysics is necessarily revisionary. Admittedly, a survey of metaphysics over the last half-century will reveal some strikingly revisionary ideas which various writers have wanted to defend: there are no ordinary objects like rocks, there are no composite objects, there is no such thing as numerical identity through time, there is no such thing as the passage of time, other possible worlds are as real and concrete as our own, the characteristic ‘feel’ of conscious states have no causal consequences, mind exists in every object, and so on. But metaphysicians concerned to explore the actual nature of reality sometimes attempt to vindicate our ordinary conceptual scheme against the revisionists (concerning, for example, the persistence of persons through time, the existence of composite objects, and so on). So perhaps ‘normative metaphysics’ is a better title: much of contemporary metaphysics is concerned to show us what we ought to think the world is like, what exists and what its nature is. In some cases, we may already be following the dictates of reason; in other cases, perhaps not.

4In this paper, my concern is, like Strawson’s, with descriptive metaphysics, and in particular with the descriptive metaphysics of time. Unlike Strawson, however, I want ultimately to put descriptive metaphysics in the service of normative metaphysics. At least, I suggest that getting clear on how we ordinarily think about time has consequences for some of the moves made in the normative metaphysics of time. I want to question a widely-held assumption about the way we ordinarily conceive of time, and in particular, about the way we conceive of the presentness and pastness of events. This then feeds into a long-standing debate in normative metaphysics over whether time passes in reality.

5So, what do we ordinarily suppose time’s nature to be?

2. Time and our ordinary conceptual scheme

6Consider the following comments from various writers on the subject of what we might called the tensed aspect of time, as we conceive and experience it, that is, the division of events into past, present and future, and the fact that any event is only fleetingly present, before it is banished to the past, making such tensed statements as ‘It is now snowing’ only transiently true:

(...) it is only by tensed statements that we can give the cash-value of assertions which purport to be about ‘time’ (...) (Prior 1967: 198).

The essential egocentricity of time is reflected in the ineliminability of tenses. (Lucas 1973: 280).

Tense is so striking an aspect of reality that only the most compelling argument justifies denying it (...) (Mellor 1981: 4-5).

(...) there is no doubt that strong intuitions support the tenser’s view that that present is objective. Here are three. We more readily take as real the presently existing objects, wherever they are, than the objects that are at some time located here (...)We are more apt to give sentences time-variable truth-values than space-variable ones (...) And we think of ourselves as sharing a common, albeit ever-changing, now, while we each have a different here (...) If all and only presently existing objects (but no objects that are sometime here) are in some sense real, the present is surely objective in a way in which the local is not. Similarly the second and third intuitions suggest that the present and its ‘movement’ are in some way objective. (Butterfield 1984: 161).

There is hardly any experience that seems more persistently, or immediately given to us than the relentless flow of time. (Schlesinger 1991: 427).

We experience the reality of tense in a variety of ways that are so evident and so pervasive that the belief in the objective reality of past, present and future is a universal feature of human experience. (Craig 2001: 127).

7These passages converge on the thought that, as we ordinarily conceive of time, tense is essential. That is, to think of time is to think it terms of presentness and related properties. Tense is a pervasive aspect of our experience (perceived events are perceived as present, time is experienced as passing), but it also features in our intuitive beliefs: we suppose that we all share a common now, though not a common here, and, relatedly, we think that the present moment is somehow special, perhaps even to the extent that only what is present can be real. We are natural tensers, in Butterfield’s terminology. But this consensus is over the descriptive metaphysics of time. Not all the writers above would agree that our ordinary conception reflects reality as it is. They are, in fact, divided on this issue. Prior, Lucas, Schlesinger and Craig think that time as it is in itself is tensed; Mellor and Butterfield think that it is not: for them, a description of reality which included no tensed expressions would not thereby be incomplete. For Mellor and Butterfield, then, the descriptive metaphysics of time (on the issue of tense, at least) is in conflict with the correct normative metaphysics, a detensed one. For the other writers, there is no such tension.

8I want to focus on two specific notions about time which appear in this picture of our ordinary conception of time:

(i) The division into past, present and future is objective, not merely perspectival.

(ii) The presentness of an event or state of affairs is a transient feature of it.

9A representation is ‘perspectival’ if it reflects the representer’s own point of view. A feature is merely perspectival if the world would lack that feature in the absence of anyone to represent it. Consider a world which contains no beings capable of representing anything – a mindless world. There are nevertheless ordinary objects in that world, such as rocks and rivers, and there are also changes – rainstorms, for example. In such a world, would anything count as ‘here’? Intuitively not, as hereness is clearly a merely perspectival property: an object counts as here relative to someone if it is in the immediately or very close vicinity of that person (the obviously vague boundaries of the vicinity being settled by context and convention). This hereness depends on someone to represent the proximity of the object, and in the absence of any such representation (or the potential for one), nothing could count as here in a mindless world.

10But what of now? Could any event (a leaf being caught in one of the eddies in a river, for example) count as now, or present in a mindless world? If those writers quoted above are correct, then the intuitive answer to this question is very definitely yes! The presentness of an event does not depend on the observer, or any kind of mental representation. It is something intrinsic to the event. And this is connected to the second component noted above in our ordinary conception of time (as it has typically been represented): belief or awareness of the transient nature of presentness. Present today, past tomorrow. This is the constant passage of time. Presentness cannot be merely perspectival because the passage of time is not merely perspectival. A further reason for thinking that presentness is objective is the idea that reality itself is confined to the present, though whether this further idea is part of our ordinary conceptual scheme has been disputed.

11If, prior to your first engagement with the philosophy of time, you would have been inclined to believe that time passes in a mindless, though changing, world, and that some time in that world would be present, independently of anyone to note it, then you may well agree that the above writers have characterised the descriptive metaphysics of time accurately. There is, however, a problem.

3. McTaggart’s paradox as a problem in descriptive metaphysics

12There is a long tradition, certainly dating back to the early years of the twentieth century, and perhaps considerably earlier, of denying the reality of tense. Of denying, that is, the objective passage of time, and the non-perspectival nature of presentness. There have been a number of sources of motivation for that denial, some to do with a posteriori arguments from physics, and the Special Theory of Relativity in particular. But one key motivation for many writers has been an a priori argument, due to J.E. McTaggart (1908 and 1927). When this argument is appealed to, and endorsed (in one version or another), it is nearly always in order to establish a position in normative metaphysics: time, as a matter of objective fact, does not pass, and presentness is merely a matter of temporal perspective. And when it is so appealed to, it is also typically accompanied by a recognition that the proposal is a revisionary one, representing a challenge to the way we ordinarily think of time. But my present purpose is somewhat different: it is to see McTaggart’s argument, which takes the form of a paradox, as a challenge to what is often said about what we ordinarily think of time. I want to present it, in other words, as a problem in descriptive metaphysics.

13McTaggart’s paradox is a part of a wider argument for the unreality of time. To set it out in full: McTaggart begins by distinguishing between two ways in which we order events in time: the A-series and the B-series. The A-series is that series of events which runs from the distant past through the nearer past, through the present, the proximate future, and on to the distant future. With the passage of time, positions in the A-series are subject to constant change. What was future becomes present and then ever more distantly past. Besides ‘past’, ‘present’ and ‘future’, other tensed expressions, such as ‘now’, ‘today’, ‘yesterday’, ‘tomorrow’, ‘last year’s Spring’, ‘next year’s Summer’, ‘a week on Wednesday’, and so forth, also serve to locate events, objects or states of affairs in the A-series. In the B-series, events (states of affairs, etc.) are located in time by means of their relations to each other, in particular their simultaneity or temporal priority relations to each other. To say that one event is earlier than another is to locate that event in the B-series. The B-series relations between events do not change over time: they remain constant. It is impossible for two events to swap their B-series positions.

14So far, we have defined the two series as two methods of temporal ordering. But if we now think of these two orderings as picking out fact in the worlds, A-facts and B-facts, we can go on to ask what the relation between the two is. Are they logically independent? Or is one set of facts dependent upon, derivative from, the other? They are clearly not independent. For example, from the fact that x is past and y future it follows that x is earlier than y. And if x is simultaneous with y, then, if x is present, then so is y. So the two orderings constrain each other. But there is this asymmetry: from complete information about the B-series positions of events, we cannot deduce their absolute A-series orderings, but only their conditional A-series orderings. For example, if we are told that x is wholly earlier than y, then we cannot deduce which, if either of them, is present. We can only deduce that, if y is present, then x is past. It seems, then, that B-facts supervene on A-facts, but not vice versa. And that suggests that the A-facts are fundamental.

15This conclusion is further reinforced by McTaggart’s argument that there can be no genuine change unless there is an objective A-series. Consider a poker which is hot at 4 o’clock and cold at 6 o’clock. There has certainly been a change here, but the B-facts do not change, only the A-facts do. That the poker’s being hot is earlier than the poker’s being cold is a fact which remains constant over time. And the fact that the poker is hot at 4 and cold at 6 (on that particular day) also does not change. What does change is that the poker’s being hot is present. This becomes past. And the poker’s being cold, which was future, then becomes present. McTaggart assumes that time requires change, so the reality of time requires A-facts.

16From our point of view, the key component of McTaggart’s argument for the unreality of time is his demonstration that the A-series cannot be real (there are, that is, no A-facts), because it contains a contradiction. This part of the argument I shall refer to as ‘McTaggart’s paradox’. The contradiction arises from the combination of the following two propositions:

(i) The A-series positions are mutually incompatible: if an event is past it cannot also be present, etc.

(ii) Every event has them all.

17One might think that extended events provide a counterexample to (i). Take an event which has been going on for some time, such as a financial crisis. It is both present and past. To avoid this issue, I shall assume that ‘past’ means ‘wholly past’: that is, every part of the event in question is past. As for (ii), this is a consequence of the passage of time: every event which is present becomes past. But (i) and (ii) together form a contradiction: no event can be past, present and future, yet with the passage of time, every event is exactly that.

18The obvious solution to the conundrum, concedes McTaggart, is to say that no event is simultaneously past, present and future, but rather successively: if an event is present at a moment of present time, it is future at a moment of past time, and past at a moment of future time. But equally, it is present at a moment of past time, for the present will be past. The only way to prevent further contradictions which arise as we iterate the tenses is to fix the various A-series positions of an event, and so prevent them from conflicting with each other, by relativising them to permanent B-series positions: the event of writing these words is present at 1300 Central European Summer Time on June 20th, 2020, future at 1200, and past at 1400 on the same date. But with this relativising of A-series series positions, we have destroyed their mostg characteristic feature: the fact that they change. The event’s being present at that time on that date is, in contrast, a static, unchanging feature of reality. What we have done, in effect, is to reduce the presentness of an event to the simultaneity between that event and another.

19Take a spatial parallel: the reason we think that one and the same object can at one and the same time be truly described as ‘here’, and also as ‘over there’, without contradiction, is that we recognise that such terms merely reflect the describer’s spatial relation to that objects, and that an object can be ‘here’ from one spatial perspective, and ‘over there’ from another. If, however, we thought that objects were ‘here’ or ‘there’ in some free-floating sense, independently of any representing mind, then we would be quick to insist that objects could not exhibit both hereness and over there-ness. Yet if we refuse to relativise A-series positions to permanent times, it seems that we would be treating presentness, pastness, etc., precisely as such free-floating properties.

20McTaggart thought that, without an objective, non-relativised A-series, there would be no time. No A-series, no change; no change, no time. We do not have to follow him in this: it is possible to consider the central paradox, the contradiction in the A-series, independently of the first part of the argument, linking the reality of time to the reality of the A-series. The point is that, if McTaggart has indeed correctly identified a contradiction in the idea of an objective A-series, and we suppose that reality is not itself contradictory, then the passage of time is not real.

21McTaggart’s paradox has typically been addressed as a problem in normative metaphysics. What should we say, if McTaggart’s paradox is genuine? That, whatever we might intuitively be inclined to think, time does not in fact pass, in any substantial sense. Pastness, presentness and futurity are just as perspectival as their spatial counterparts. Many, if not all, of those who deny the reality of tense – detensers, or ‘B-theorists’ (who take their name from the series whose fundamentality they champion) ─ take McTaggart to have shown, on a priori grounds, that time does not pass. B-theorists, however, distance themselves from McTaggart’s more radical inference from this, that time and change are not real. Time and change, say the B-theorists, can be explained in terms of the B-series alone.

22But let us now consider it from the perspective of descriptive metaphysics, the attempt to describe our ordinary conceptual scheme, independently of whether that scheme accurately reflects the structure of reality. Since McTaggart’s argument against the A-series is an a priori one, it shows (if it is successful) that there is a contradiction at the heart of our ordinary conceptual scheme. We are making a conceptual error, not merely a factual one, in supposing (if we do) that there is an objective passage of time, and that presentness, etc., are non-perspectival. That is a serious allegation: it suggests a confusion at the heart of a cognitive system on which we rely our daily lives, as tensed beliefs are essential for timely and effective agency.

23In normative metaphysics, a move that is sometimes made to avoid McTaggart’s paradox whilst retaining a non-relational, irreducible A-series is to adopt a presentist position (see, e.g., Bourne 2006). For presentists, only what exists now is real, the past and future being, as Prior put it, species of unreality (Prior 1970: 245). For, it is suggested, an implicit assumption in McTaggart’s argument is that all times are equally real. Making this explicit provides a simple way of representing the paradox:

(1) It is a present fact that p.

(2) It is a past fact that Not-p.

(3) Both present and past facts are real.


(4) Both p and Not-p obtain in reality.

24(1) and (2) can be established by the fact of any change: the poker’s having been hot and now being cold, for example. The contradiction cannot be avoided by suggesting that what obtains in reality is the perfectly consistent

it is the case that p and it was the case that Not-p,

25for this is simply a present fact. It is the reality of past facts which, together with present facts, leads to contradiction. The presentist, however, can avoid the problem by denying (3). Now, supposing this to be a successful strategy in normative metaphysics, we may wonder whether there is a counterpart in descriptive metaphysics. Presentists sometimes aver that presentism is actually the intuitive view of the ontology of time: that it is part of our ordinary conceptual scheme that the past is unreal, as well as the future, that the passage of time consists of things slipping out of reality. If so, then perhaps our conceptual scheme is not incoherent after all: we believe in the reality of tense, but avoid contradiction by being natural presentists.

26Unfortunately, it is also sometimes asserted by growing block theorists (those who posit the reality of the past and present, but unreality of the future, and describe the objective passage of time in terms of the accretion of past facts) that theirs is the natural, intuitive view, one that corresponds to our ordinary way of conceiving time. But the growing block theory has to accept all three premises of the argument above.

27So what should the descriptive metaphysics of time look like? Do we naturally accept both the reality of tense and the reality of the past? If so, then, if the defenders of McTaggart are right, we are committed to the following (where ‘p’ stands for the allegedly contradictory view of time):

(i) Our ordinary conceptual scheme takes p to be true.

(ii) p is incoherent (in virtue of being self-contradictory).

From which it immediately follows that:

(iii) Our ordinary conceptual scheme is incoherent (in this respect).

28But (iii) is a deeply problematic thesis. Of course, we cannot expect that the whole of our conceptual scheme would be accurate in every respect. But that our conceptual scheme views the world as it does is not a matter of mere accident: we think of the world in a certain way because there is a point to doing so. This is the reasonable inference to draw from the fact that natural selection is at work in shaping the most general features of our conceptual scheme. Now it is generally conceded that false hypotheses about the world may have a use: they may even confer a significant natural advantage. Our tendency to perceive causal connections in the world, for example, may sometimes get things wrong, but it is important that we have this (perhaps hyperactive) causal sense. But what is the use of an incoherent world view?

29Considering McTaggart’s paradox as a problem in descriptive metaphysics, that is, as showing us that supposing p to be a feature of our ordinary conceptual scheme implies the incoherence of that scheme, should lead us to reconsider exactly how our ordinary scheme does represent time. I suggest that a key to our ordinary understanding of time is afforded by looking at the nearest source of knowledge of the past: memory.

4. The structure of episodic memory

30The kind of memory that is particularly pertinent to our enquiry is episodic memory. It is typically contrasted with semantic memory, which is characterised by a that clause: you remember that you boarded the 0806 train to work yesterday, you remember that you bought a newspaper shortly before doing so. That information may or may not be accompanied by the recall of certain experiences. If you remember not just that you did these things, but also remember experiencing the events, the sense of hurry, wondering whether you had time to buy a paper first, feeling relief as you got on to the train as the whistle was blown, the discomfort, physical and social, of squeezing between two other passengers...then you are accessing episodic memories. This kind of memory gives us access, not just to the past, but to our own experiential pasts. Or at least to our current reconstruction of that experiential past. Whether such memory tends to be accurate or not need not concern us here. What matters is not its epistemic status, but rather its structure.

31The difference between (mere) semantic memory and episodic memory can be captured in terms of the structural difference between the simple past tense statement:

It was the case yesterday that I caught the 0806,

32and an admittedly less idiomatic one involving iterated tenses:

It was the case yesterday that (it is the case that I am catching the 0806).

33In the second of these statements, ‘it was the case that’ shifts the temporal context to one in which some event is present. The content of a semantic memory is captured by the simple tense, but the content of an episodic memory is, I suggest, more perspicuously captured by the compound tensed statement. What the episodic memory of catching the train, the events leading up to it, and the affective states accompanying all these, provides is a present-tensed perspective on those events. It is this that gives us the impression of reliving those events.

34We are, I suggest, presented with two temporal perspectives on the same event in episodic memory: the event is presented as past (for this is recognisably a memory), and it is represented as present (for we view the event as it was experienced). We are not at all confused by this, wondering which of two competing perspectives can be correct. For the present-tensed perspective is embedded, as it were, in the past-tensed perspective, just as the present tense operator in the second sentence above is embedded in the past tense operator. The memory invites us to shift the context of contemplation to a past moment, and, having done so, we can contemplate the event as present – in our mind’s eye and ear we see and hear it as it was, and feel it too.

35That we can negotiate this psychological shift without difficulty or confusion argues in favour of the hypothesis that we naturally appreciate that these different perspectives are indeed perspectives. We keep the event in view as we shift the angle at which we view it. An analogy is the way in which we can continuously track a single object while changing our spatial perspective on it. In doing so, we do not correct or reject our earlier perspectives. Those spatial perspectives are just as legitimate, just as revelatory of some of the object’s features as our current perspective. Similarly, when we mentally travel back in time through an episodic memory, we appreciate that the perspective we then had on the event was a legitimate one; we are not inclined to correct it, simply because memory also presents the event as past.

36Now, if we have a natural, pre-philosophical appreciation that the different tenses indicate different perspectives on temporal reality, then our ordinary conception scheme need not after all, be dogged by McTaggart’s paradox. For we can appreciate that the poker’s being hot is present relative to one time and past relative to another.

37But now perhaps we have over-reached ourselves. Does this represent the thin end of the wedge of an attack on revisionary metaphysics itself? Let me explain.

5. How revisionary can revisionary metaphysics be?

38One moral one might draw from the discussion of memory is that our descriptive metaphysics of time does not, after all, conflict with the supposedly revisionary proposal that there is no such thing as an objective, mind-independent passage of time. But if that is the right moral to draw, then why should so many people find that proposal so surprising? Isn’t that precisely because it does challenge ordinary intuitions?

39Moreover, if, in this instance, positing a wide divergence between descriptive and normative metaphysics – between what we think and what we ought to think – has the unacceptable consequence that our ordinary conceptual scheme is incoherent, won’t this carry over to other revisionary theories in normative metaphysics? If these theories are to be believed, then we are radically mistaken about a whole host of things: the persistence of the self through time, the very existence of a self, the existence of ordinary objects, the reality of colour, our capacity for genuinely free action, the contingency of objects and events, the possibility of physical contact, the impossibility of two objects occupying the same space at the same time, and so on. Since the arguments for these theories are typically based on the ultimate incoherence of the supposedly intuitive position, doesn’t that imply – absurdly – that we spend our days in a fog of confusion? Recall the earlier argumentative schema:

(i) Our ordinary conceptual scheme takes p to be true.

(ii) p is incoherent (in virtue of being self-contradictory).


(iii) Our ordinary conceptual scheme is incoherent (in this respect).

40Substitute for p any feature of our descriptive metaphysics against which revisionary metaphysics can launch a reductio, and the cumulative result is a deeply incoherent ordinary conceptual scheme. If the revisionary metaphysics of time is under threat, then revisionary metaphysics as a whole would appear to be so. But now that looks too strong a conclusion. Surely, metaphysics is permitted to aim at improving our understanding of reality, and that might well include jettisoning ideas we took for granted.

41So what conclusion should we draw from McTaggart’s paradox, considered as a problem in descriptive metaphysics? Should we conclude that there cannot, after all, be such a gap between the descriptive and normative metaphysics of time? And, if so, should we generalise that conclusion to apply to all putative gaps between descriptive and normative metaphysics, with respect to such topics as persistence, identity, composition, and so on?

42McTaggart’s paradox appears to show that there is a contradiction in the very idea of the A-series. The contradiction boils down to the observation that the world contains such contradictory facts as the poker’s being hot and the poker’s not being hot – these having an equal claim to reality. It is not a contradiction that is buried very deep, such as a contradiction in some mathematical theorem, which it would require a demonstration of considerable complexity to unearth. But if our ordinary concept of time is incoherent, as a result of a conflict between tenses, this is not something that should escape our notice until pointed out by philosophers. It might be proposed that we do not ordinarily confront the contradiction because much of the time our consciousness is filled with present-tensed beliefs. We can afford to be, as one might put it, instrumental presentists – that is, in agency we only attend to the present. But that ignores both memory and anticipation. Each of these confronts us with states of affairs which did or will obtain, but which do not obtain now. That we are able to handle these potentially contradictory perspectives suggests that we readily appreciate that they are perspectives, appropriate to different times. It does not follow from this, however, that our descriptive metaphysics effectively contains within it a denial of the passage of time, a view of time as essentially space-like. We are, I think, aware that our perspective on time is not of our choosing, not within our control, but rather is imposed on us by some aspect of time itself. But to take that as implying that tense is a non-perspectival aspect of reality is perhaps to read too much metaphysics into our ordinary outlook. This is, indeed, a danger philosophers are prone to when identifying their own favoured theory as ‘intuitive’: the perception of the correctness or warrant of the theory is read back into the picture of our pre-theoretic outlook. Admittedly, we do say such things as ‘time passes’, and admit further that this is something inevitable, and not under our control, but this may mean no more than that change happens whether we want it to or not. When confronted with a worked-out metaphysics, that goes beyond anything which common-sense tells us, we are bound to find it novel. So normative metaphysics can indeed be revisionary, even surprising. But if the dialectical starting point is that there is a contradiction on the surface of our ordinary conceptions, then we should pause before conceding it. McTaggart’s paradox, I suggest, does not show us that our ordinary conception of time is incoherent in this way, but rather that a particular attempt to embellish that ordinary conception runs into incoherence.

43What of normative metaphysics in general? Are there limits as to how revisionary it can be? Take the theory which tells us that there are in reality no composite objects, just simple, partless objects. This is mereological nihilism. According to mereological nihilism, however, we do not need to give up ordinary talk of composite objects such as rocks, rivers, tables and chairs, as long as we realise that what makes such talk true is facts about the arrangement of groups of simples. A chair is nothing more than a collection of simples arranged in a certain way. When we say that composite things are going out of existence, what makes this true is that certain arrangements of simples are replaced by other ones. For all its accommodation of our ordinary ways of talking, though, it does not accommodate our ordinary ways of conceiving the world. And this might put us on our guard. Is the suggestion that our conception of ordinary objects is confused, or incoherent? And if so, wouldn’t we have the right to object? Well, the nihilist is not, or not quite, saying that our ordinary conception of composite objects is confused, or incoherent. The central argument for nihilism is that a number of propositions that we might intuitively wish to assent to are in fact inconsistent. Let’s take our paradigm composite object to be a rock. Then the inconsistent set of intuitively true propositions will go something like this:

(1) There exist rocks.

(2) A rock is a composite object, composed of a multitude of very small parts, none of which is a rock.

(3) Removing just one of these parts from any given rock would not result in the rock’s going out of existence.

44Where is the contradiction? Consider a specific rock, and imagine removing a tiny part. Given (3), you still have a rock. Now remove another such part. Again, given (3), you still have a rock. However many times you repeat this process, you should still end up with a rock. But eventually, you will end up with just one tiny part, which, given (2), is not a rock. So (2) and (3) together describe something which could not possibly exist. If there existed a rock, then you could subject it to a process which, given (2) and (3), would result in something that is both a rock and not a rock. Since it is impossible for anything to be both a rock and not a rock, it would appear that no rocks actually exist. Indeed, no rocks could exist. (See, e.g., Unger 1979 for a fuller statement of the argument).

45Supposing the argument to be cogent, and that we are therefore justified in supposing there to be (despite appearances) no real rocks, this is clearly a revisionary position. What we in fact think and what we should think in this case have come fairly dramatically apart. That does not show that the ontology of our ordinary conceptual scheme, containing, as it does, rocks, is thoroughly confused. It shows, at most, that one belief we hold (that rocks exist) conflicts with other propositions we either already hold true, or would assent to if put to us. Quite a bit of revisionary metaphysics proceeds by juxtaposing propositions which, if put separately we would assent to, and then showing that they are in fact inconsistent. The argument for mereological nihilism (if successful) thus shows that we should revise our conceptual scheme without showing that we were simply confused before. Moreover, what is really at issue in the case of the argument against the reality of rocks is whether they are conventional objects, or things which exist quite independently of our conventions, an issue on which our ordinary scheme might take no definite view.

46So when McTaggart’s paradox is presented as a challenge to our ordinary conceptual scheme, is this the same kind of challenge as is presented by the argument for nihilism above? That is, are we presented with a contradiction between propositions to which we might be inclined to assent? Or is it being suggested that our ordinary grasp of tenses is somehow confused? If the latter, we would be right to insist that there is no such confusion, and that more is being read into our scheme by metaphysicians of time than is actually there. What McTaggart’s paradox does, I suggest, is to bring into focus issues on which we are ordinarily agnostic, but which, when questioned, we may be prepared to take a view on.

47That, I think, makes a difference to the dialectic in normative metaphysics over the reality of tense (that is, of non-perspectival pastness, presentness and futurity). The typical starting point is an assertion by the supporters of real tense that theirs is the intuitive view, an assertion which the detensers typically concede (as illustrated by the passages quoted in section 2 above). So the detensers start on the back foot, with the onus on them to justify their supposedly highly revisionary position. But if the descriptive metaphysics of time turns out, as the structure of memory suggests, to be less obviously pro-tense, but rather more agnostic on what is at issue in the normative metaphysics of time, then the onus of argument is rather more evenly distributed than the history of the debate would suggest.


Berkeley, G. (1707/1989). Philosophical Commentaries (Luce, A. A. & Thomas, G. H. eds). London: Routledge

Bourne, C. (2006). A Future for Presentism. Oxford: Oxford University Press.

Butterfield, J. (1984). Seeing the present. Mind, 93(370), 161-76.

Craig, W. L. (2001). Time and Eternity: Exploring God’s Relationship to Time. Wheaton, Illinois: Crossway.

Kant, I. (1787/1929). Critique of Pure Reason (2nd ed., Smith, N. K. trans.). London: Macmillan.

Lucas, J. R. (1973). A Treatise on Time and Space. London: Methuen

McTaggart, J. M. E. (1908). The Unreality of Time. Mind, 17(68), 457-74.

McTaggart, J. E. (1927) The Nature of Existence (Vol. II). Cambridge: Cambridge University Press.

Mellor, D. H. (1981) Real Time. Cambridge: Cambridge University Press.

Prior, A. N. (1967). Past, Present and Future. Oxford: Clarendon Press.

Prior, A. N (1970). The Notion of the Tresent. Studium Generale, 23, 245-8.

Schlesinger, G. N. (1991). ‘E pur si muove’. Philosophical Quarterly, 41(165), 427-41.

Strawson, P. F. (1959). Individuals: An Essay in Descriptive Metaphysics. London: Methuen.

Unger, P. (1979). ‘There Are No Ordinary Things’, Synthese, 41, 117-54.


University of Leeds

© Collège de France, 2021

Conditions d’utilisation :


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search