Realism about tense and atemporality

Abstract

Realists about tense, or A-theorists of time, believe that some of the facts that fundamentally constitute reality are tensed, and most of them seem to think that those tensed facts are to be understood as fixing the way things are, absolutely speaking, or simpliciter. But there is a simple yet powerful argument, the argument from atemporality, to the effect that realists should reject the absolutist conception of reality’s constitution by facts because, despite appearances to the contrary, that conception is in fact inconsistent with realism about tense. After presenting the argument, I investigate why the relevant inconsistency has not been recognised so far and point to one possible explanation.

1 Introduction

In his recent pioneering work on the metaphysics of time, Fine (2005b), 2006) argues that the debate between the A-theory and the B-theory of time turns on whether or not the following thesis is true:

REALISM

Reality is fundamentally constituted (at least, in part) by tensed facts.

Thus, A-theorists endorse REALISM, whereas B-theorists reject it.

Further, Fine (2005b: 271) contends that REALISM is incompatible with the joint acceptance of the following three theses:¹

NEUTRALITY

No time is privileged, the tensed facts that constitute reality are not oriented towards one time as opposed to another.

ABSOLUTISM

The constitution of reality is an absolute matter, i.e. not relative to a time or other form of temporal standpoint.

COHERENCE

Reality is not contradictory, it is not constituted by facts with incompatible content.

According to Fine, one can distinguish between different types of realism about tense on the basis of which one of these three principles is rejected. In the Finean terminology, views that hold on to NEUTRALITY are versions of non-standard realism: ABSOLUTISM is given up by external relativists, while fragmentalists abandon COHERENCE. But the most popular form of realism is what Fine calls standard realism, the view that rejects NEUTRALITY and holds on to the other two.² Indeed, most A-theorists seem to endorse, along with the idea of an ‘objectively’ privileged present time, some version of the thesis that the tensed facts they believe in are to be construed as fixing the way things are, absolutely speaking, or simpliciter. To cite but one example, consider the following passage by Zimmerman (2005: 431):

The essence of the A-theory is the objectivity of the distinction between past, present, and future. What is presently true is true, simpliciter, not merely true relative to a time or utterance or situation. One source of temporary truths is the fact that there are temporary properties that some things exemplify, simpliciter – in other words, properties that are not merely exemplified relative to a time, utterance, situation, or anything else. The doctrines of temporary truth, simpliciter, and of temporary exemplification, simpliciter, seem likely (and closely interrelated) ways of making the A-theorist’s point: namely, that present things and events are objectively special.

If this is right, then Fine’s external realism is not a version of the A-theory after all. But let us put taxonomy aside. My main aim in this paper is to show that there is a simple yet powerful argument, the argument from atemporality, to the effect that all realists about tense should reject ABSOLUTISM because, despite appearances to the contrary, REALISM and ABSOLUTISM are in fact inconsistent. Thus, the argument targets any absolutist conception of realism about tense, or the A-theory, including, but not limited to, standard realism.

The paper is organised as follows. In the next section, I articulate the basic idea behind the argument from atemporality. In Sect. 3, I discuss a general strategy for resisting each of the two premisses of the argument, and then, in Sects. 4 and 5, I defend these premisses in turn. Finally, in Sect. 6, I close with some comments on the implications of the argument.

2 The argument from atemporality

Imagine a world in which there are only three instants and one single persisting spatiotemporal entity: Polymorph. Polymorph is a shape-shifter: it starts out as a triangle, then becomes square, and finally ends up circular. Apart from the change Polymorph undergoes with respect to its shape, nothing else happens in this world.

The realist and the anti-realist will give quite different accounts of how temporal reality fundamentally is in Polymorph’s world. The anti-realist’s account will list the following three (putative) tenseless facts: that Polymorph is³ triangular at t₁, that Polymorph is square at t₂, and that Polymorph is circular at t₃.⁴ Given its tenseless content, it seems clear that this account has an atemporal status. The tenseless facts it appeals to do involve a temporally located entity and its temporal locations themselves, but they are meant to constitute the way that this temporal object atemporally is at those locations: they make up the way Polymorph’s world would be seen to be if one looked upon it from an atemporal perspective—from a standpoint, that is, that lies beyond time and temporality.⁵ For the anti-realist, reality is temporal merely in the sense of comprising times and things located at them; in other words, it is temporal roughly in the same sense in which it is spatial. On the other hand, the constitution of reality by facts that involve these temporally and spatially located things is an entirely atemporal (and aspatial, for that matter) affair. In other words, anti-realists endorse the following thesis:

ATEMPORALISM

All of the facts that fundamentally constitute reality are ‘visible’ from the atemporal perspective; they all constitute the way reality atemporally is.

Contrast this with the realist’s account of Polymorph’s world. Let us focus on standard realism here, as it is the most widespread version of realism. Given her rejection of NEUTRALITY, the standard realist thinks that all the tensed facts that constitute reality in Polymorph’s world are ones that obtain at a single unique time, the one that is objectively present. Assuming that t₂, say, is present, the standard realist’s account of Polymorph’s world will feature the following (putative) tensed facts: that, presently, Polymorph is square, that Polymorph was triangular, and that Polymorph will be circular. Evidently, this account does not purport to describe the way reality atemporally is in Polymorph’s world. On the contrary, it is supposed to represent how things are from within a temporal standpoint—that is, from within the standpoint of the present time. Realists have not been exceptionally vocal about this, but consider the following passage by Ross Cameron (2017: 775):

According to the B-Theory, the God’s-eye perspective on reality is an atemporal one … According to the … A-Theory … the God’s-eye perspective on reality is a temporal one: it describes how things are now.⁶

Indeed, ATEMPORALISM seems incompatible with REALISM. For the ‘visibility’ of a fact from the atemporal perspective seems to require that fact’s constituting reality be entirely independent of any temporal factors whatever. But this is not how tensed facts are supposed to work: whether or not a tensed fact constitutes reality is supposed to depend on time. Tensed facts relate to time in a peculiar, twofold manner: they not only concern temporal goings-on, but also concern them in a distinctively temporal fashion. The (putative) tensed fact that Polymorph was triangular, for instance, is directed at a certain time, just as the (putative) tenseless fact that Polymorph is triangular at t₁. Unlike the latter, however, the past-tensed fact does not constitute how things atemporally are at the time at which it is directed, but rather constitutes how things are at that time from within the standpoint of a later time at which it is anchored: the past-tensedness of the fact does not only signal that what it concerns is in the past, but also that reality’s being constituted by this very fact occurs at a later time, in the present.⁷ It is therefore difficult to see how the way reality is as constituted by a tensed fact could fail to be a temporal way it is.

But consider now how closely ATEMPORALISM is linked to ABSOLUTISM. Given ABSOLUTISM, any putative fact, independently of whether or not it involves temporally located entities and/or times, either constitutes reality, full stop, or it does not, full stop—there is no room here for any temporal qualification. Given that constitution is entirely insensitive to time and absolved from all temporal affairs, there is no need to first presuppose a particular temporal standpoint when we ask what facts constitute reality. Whatever temporal standpoint you assume, the facts that constitute reality will be the same—but more importantly, no temporal standpoint is required for constitution to occur and hence for the facts that constitute reality to be ‘visible’. Hence, whatever facts constitute reality in this absolute manner, they will also constitute the way reality atemporally is—they will precisely be the facts that are ‘visible’ from the atemporal perspective. And conversely, it seems that an atemporally obtaining fact will precisely be one that obtains temporally absolutely—for, given that it is already acknowledged that such a fact is part of the way reality atemporally is, how could its obtaining be somehow sensitive to temporal factors?

In short, ABSOLUTISM seems to entail, and be entailed by, ATEMPORALISM. But we have already established that REALISM is not compatible with ATEMPORALISM; given that the latter entails ABSOLUTISM, we should conclude that REALISM is incompatible with ABSOLUTISM as well.

The argument from atemporality, then, goes like this, in summary form:

(P1) If REALISM is true, then ATEMPORALISM is false.

(P2) If ABSOLUTISM is true, then so is ATEMPORALISM.

(C) If REALISM is true, then ABSOLUTISM is false.

The argument is formally valid and attractively straightforward; the premisses, however, are quite controversial. In the next section, I describe a general response strategy that draws on a particular interpretation of the Finean framework and challenges both premisses of the argument.

3 Resisting the argument: tense and constitution

Of all the four Finean principles, only ABSOLUTISM appears to concern constitution itself, while the other three seem to characterise the facts that do the constituting. However, it has recently been argued by Loss (2017: 212–215) that NEUTRALITY, too, ought to be understood as a thesis about constitution proper: accepting it means taking constitution to be tenseless, and rejecting it means taking it to be tensed. To see more clearly how this proposal is supposed to work, consider Fine’s (2005b: 318–320) distinction between claims about reality’s form and claims about its content. The question of whether some of the facts that fundamentally constitute reality are tensed concerns the content of reality—and, while anti-realists say “No” to this, all realists answer it positively. But there is also the question of how the constitution of reality by facts is to be understood, which is a question about the form of reality—and realists of different stripes are supposed to disagree amongst themselves on this: accepting ABSOLUTISM commits one to the absoluteness of constitution, and accepting NEUTRALITY to its tenselessness.⁸  

There is indeed some textual evidence for this reading of NEUTRALITY. Consider how Fine (2005b: 276–277) describes standard realism, for instance: “[O]n this view, there is an absolute notion of constitution, but it is tensed; and the tensed facts that constitute reality are those that presently obtain.” Fine (2005b: 271) also seems quite explicit that whether the facts that constitute reality are tensed or tenseless, and whether constitution itself is tensed or tenseless, are two separate issues:

[T]he absolute notion of constitution that figures in the other assumptions can be taken to be either tensed or tenseless. Thus in saying that a given fact constitutes reality, one can either be speaking about the present constitution of reality or about its eternal composition. It is natural to suppose that tensed facts constitute reality in a tensed fashion and that tenseless facts constitute it in a tenseless fashion but there is no reason, in principle, why the tense-theoretic status of the fact and of the form of constitution should not come apart.

Similarly, assuming that NEUTRALITY, understood this way, and ABSOLUTISM are independent from each other, the distinction between absolute and non-absolute constitution crosscuts the distinction between tensed and tenseless constitution.

Here is how this conception of NEUTRALITY can be employed to respond to the argument from atemporality. As discussed above, ATEMPORALISM is something that standard realists and anti-realists disagree on: it is rejected by the former and endorsed by the latter. Given (P2) of the argument from atemporality, they ought to also disagree on ABSOLUTISM, as this latter claim entails ATEMPORALISM. But the above conception of NEUTRALITY suggests a different picture: for given that NEUTRALITY implies that constitution is tenseless, it is plausible that ATEMPORALISM is simply a consequence of NEUTRALITY, which is accepted by anti-realists and rejected by standard realists. It might therefore seem that the argument from atemporality mistakenly associates ATEMPORALISM with ABSOLUTISM—the former thesis is entailed by NEUTRALITY, not ABSOLUTISM, which is why (P2) is to be rejected.

Further, given that whether constitution itself is tensed or tenseless is an issue that is independent of the question of whether the facts that do the constituting are tensed or tenseless, the fact that ATEMPORALISM is entailed by (and entails) a tenseless conception of constitution is neither here nor there as far as the debate over REALISM is concerned. Indeed, according to the present approach, non-standard realism is precisely a strategy that seeks to exploit this apparent gap between the nature of constitution and the nature of facts by construing constitution as tenseless but the facts that do the constituting as tensed. In fact, Fine (2006: 401) himself seems to associate non-standard realism with the atemporal perspective:

The nonstandard position will … suppose both that reality is tensed and that it is not oriented towards one time, the present, as opposed to another. Thus even though we may stand outside of time, as it were, without adopting any particular temporal standpoint, we may still think of temporal reality as being constituted by tensed facts.

Thus, what entails the falsity of ATEMPORALISM is not REALISM, but rather the falsity of NEUTRALITY; and given that it is possible to combine REALISM both with NEUTRALITY and its rejection, (P1) seems to fail as well—ATEMPORALISM is consistent with REALISM, because NEUTRALITY is. The question of whether the way things fundamentally are is the way they are from the atemporal perspective is orthogonal to the debate over REALISM.

For lack of a better term, I shall call this resistance strategy the tenselessness of constitution response, or the TOC-response, for short. My defence of the argument against this challenge will proceed in two parts, the first focusing on (P1), and the second on (P2). I shall argue that, while the case against (P1) is mistaken about the implications of the tenselessness of constitution, the case against (P2) is mistaken about the implications of its tensedness.

4 Defending the argument – part I

The point of departure for the TOC-responder’s case against (P1) is interpreting NEUTRALITY as saying that constitution is a tenseless affair, implying the following equivalence:

(T1) NEUTRALITY is true iff constitution is tenseless.

As intimated above, it is independently plausible that the following equivalence holds as well:

(T2) ATEMPORALISM is true iff constitution is tenseless.

(T1) and (T2) entail:

(T3) NEUTRALITY is true iff ATEMPORALISM is true.

We know from Fine’s work that there is a strong case for the following thesis:

(T4) REALISM is compatible with NEUTRALITY.

But from (T3) and (T4), it follows:

(T5) REALISM is compatible with ATEMPORALISM.

And (T5) contradicts (P1) of the argument from atemporality.⁹

I agree with the TOC-responder that ATEMPORALISM entails, and is entailed by, a tenseless conception of constitution and that REALISM and NEUTRALITY are indeed compatible—in short, I agree that (T2) and (T4) are true. However, as I shall argue in this section, (T4) is in fact incompatible with (T1) (and hence (T3), given (T2)): if NEUTRALITY is interpreted as implying a tenseless conception of constitution, it cannot be something that is acceptable by realists. Thus, if one wants to hold on to (T4), the TOC-responder’s interpretation of NEUTRALITY has to be replaced with an alternative construal of that principle.

4.1 Tensing the facts vs. tensing constitution

The thought that (T1) and (T4) are compatible seems primarily to rest on the following thesis:

INDEPENDENCE

Whether reality’s constitution by facts is a tensed affair or not is independent of whether the facts that constitute reality are tensed or not.

Within the Finean framework, INDEPENDENCE can be elucidated as follows. According to Fine’s official metametaphysics, claims about the fundamental constitution of reality are to be parsed in terms of a primitive sentential operator, ℜ, such that ^┌ℜ(φ)^┐ is read as ^┌In reality, it is the case that φ^┐.¹⁰ Given this, it seems that claims about the form of reality, about the manner in which reality is constituted by facts, will concern ℜ, while claims about the content of reality will be about the sentences that ℜ operates on. Hence, whether some of the facts that fundamentally constitute reality are tensed will depend on whether some of the sentences that ℜ operates on are tensed. As Loss (2017: 213) points out, the question of whether constitution itself is tensed or tenseless can then be understood as the question of whether sentences of the form ^┌ℜ(φ)^┐ can themselves serve as arguments for tense operators.

Here is one drawback of this ‘official’ picture, though. Since being tensed or tenseless is usually understood as a feature of truth-bearers such as sentences or propositions, any theory that ascribes those features to reality itself had better offer some explanation as to what the relevant tensedness or tenselessness amounts to. But if one’s answer to that question is to be couched in the official Finean idiom, it is unclear how informative it can ever be. We are being told that one takes some of the facts that fundamentally constitute reality to be tensed, if one believes that, in some sentences of the form ^┌ℜ(φ)^┐, the embedded sentence is tensed. But this leaves unanswered the question of what feature of reality itself, of whatever is represented by the relevant embedded sentence, is supposed to compel us to use tensed language when stating how reality fundamentally is. One might prefer to appeal to ℜ, instead of talking of tensed or tenseless facts’ constituting reality, but this terminological choice hardly obviates the need to say more about the notion of a tensed or tenseless reality.¹¹

In fact, given INDEPENDENCE, the question of what the tensedness or tenselessness of reality, or of facts constituting it, consists in becomes all the more pressing. For one straightforward answer to that question would precisely be that there is nothing more to the status of facts regarding tense than the manner in which they constitute reality:

CONSTITUTIONALITY OF TENSE (cot)

What it is for a fact to be tensed (tenseless) just is for it to constitute reality in a tensed (tenseless) fashion.

To be sure, this is not very satisfactory, until one hears more about tensed and tenseless constitution of reality by a fact. But given cot, we can at least say that questions about tense as it applies to reality properly concern one core phenomenon instead of two: what it is for reality to have tensed ‘content’ just is for it to have whatever content in a tensed ‘form’. cot goes hand in hand with the intuitive picture of tensed and tenseless facts that we have already touched upon—with the idea that tensed facts seem to have, and tenseless facts seem to lack, a certain temporal ‘shape’. The difference between the anti-realist’s putative fact that Polymorph is triangular at t₁ and the realist’s putative fact that, presently, Polymorph is triangular seems to consist, at least in part, in the latter’s implying a temporal perspective, that of t₁, upon reality. If this tensed fact, rather than the tenseless one, is amongst the ones that fundamentally constitute reality in Polymorph’s world, then t₁ is not just a temporal location at which facts that constitute reality may be directed. Rather, it serves as an anchor for the fact that, presently, Polymorph is triangular—it is the locus where the constitution of reality by this fact takes place. While constitution of reality would be an atemporal matter altogether as far as tenseless facts are concerned, tensed facts would constitute reality in a distinctly temporal manner.

But cot is unavailable if INDEPENDENCE is accepted: proponents of INDEPENDENCE have two distinct and independent phenomena to explain, and it is unclear how they can make sense of the idea that tensed facts have a temporal ‘shape’ that tenseless facts are supposed to lack. For the tenselessness of constitution seems precisely to imply the temporal unanchoredness of the relevant facts: the facts that CONSTITUTE reality may well be directed at what is going on at some temporal location, but they are not supposed to do so from within a temporal standpoint. But if INDEPENDENCE is true, then genuinely tensed facts may very well CONSTITUTE reality, and hence be temporally unanchored, lacking any sort of temporal ‘shape’. This is also why INDEPENDENCE threatens to trivialise the distinction between tensed and tenseless facts, and hence the distinction between realism and anti-realism. For, given INDEPENDENCE, whatever exactly the difference between tensed and tenseless facts might amount to, it will not be of a magnitude to inevitably lead to a difference in the respective ways in which tensed and tenseless facts constitute reality: a realist and an anti-realist may fully agree about the ‘form’ of reality. If cot is true, however, this sort of ‘constitutional’ common ground between the realist and the anti-realist is impossible.

INDEPENDENCE makes the task of explaining what reality’s or its facts’ being tensed or tenseless consists in much more challenging, to say the least. But the central difficulty with the thesis is that it allows illicit mix-and-match. Let us focus here on the case in which reality is supposed to be CONSTITUTED by genuinely tensed facts.¹² In the official Finean idiom, this comes down to the view that, while ℜ can operate on complex constructions involving tense operators, it cannot itself be embedded under such operators. What exactly would one be saying about temporal reality if one held this view? Since one takes sentences of the form ^┌ℜ(φ)^┐, which are used to express how things fundamentally are, to be tenseless, one must be supposing that reality fundamentally is a certain way; and that way is precisely specified by the sentences ℜ operates on. But then, given that any sentence that ℜ operates on is a (partial) specification of how things fundamentally are, those sentences themselves must be understood as tenseless (or at least as reducible to such)—otherwise they would be unsuitable for specifying the way reality fundamentally is. Genuinely and irreducibly tensed sentences, it seems, cannot be in the business of perspicuously characterising the way reality fundamentally is. Switching to fact-talk: Facts that CONSTITUTE reality do so in a manner that is completely insensitive to time; and what gets thusly constituted is supposed to be the way things just are—without any temporal qualification, stripped of all temporal dependences, and immune to all temporal influences. But any genuinely tensed fact comes with temporality imprinted on it: a way that things presently are, have been, or will be is a way that they temporally are, have been, or will be—a way that is anchored at a certain time and directed at, from within the standpoint at which it is anchored, some other (either distinct or not) time. Any genuinely tensed fact presupposes a temporal perspective upon reality, whereas reality’s CONSTITUTION by facts is an entirely atemporal affair. The former, therefore, cannot take part in the latter.

The point can perhaps be made somewhat more tangible by focusing on an example. Suppose that, in Polymorph’s world, the fundamental constitution of reality is tenseless; hence, there must be certain facts about Polymorph’s shape that CONSTITUTE reality. But given that it is a tenseless matter whether or not any such putative fact CONSTITUTES reality, how could it not be a tenseless matter whether or not Polymorph has any given shape property? Since, for any putative fact f concerning Polymorph’s shape, the constitution of reality by f does not depend on any temporal factors, if f constitutes reality, then Polymorph simply has whatever shape property f involves, which means that f itself is tenseless. Given that there are no temporal influences involved in the constitution of reality by f, how could there be anything temporal going on when it comes to Polymorph’s having some particular shape?¹³ Given (T2), the tenselessness of constitution entails ATEMPORALISM; and the tenselessness of constitution in Polymorph’s world promises an atemporal characterisation of how Polymorph is, one that settles what shape it atemporally has. However, once one attempts to actually look at the specifics of how Polymorph is, one inevitably ends up leaving the atemporal perspective and assuming a temporal one, on account of the tensedness of the facts concerning Polymorph, which merely specify how it presently is, has been, or will be, rather than how it simply is.¹⁴

It therefore seems that a metaphysics in which constitution is taken to be tenseless, but facts that do the constituting are regarded as genuinely tensed is inherently unstable: the tenselessness of constitution pulls into one direction, and the tensedness of facts into the opposite one. If we follow Fine in associating constitution with the form of reality and the facts themselves with its content, we can say that theories that seek to combine a tenseless conception of constitution with REALISM result in a picture in which there is a fundamental mismatch, a fundamental discordance between the form of reality and its content.¹⁵

There may be ways of making sense of the bare idea of tenseless constitution by tensed facts. Suppose, for instance, that Polymorph’s world is one in which one single time has the primitive property of being present. Now, one could argue that what it is for the fact that, presently, Polymorph is triangular to CONSTITUTE reality is for it to be case that the following two facts CONSTITUTE reality: the fact that some time t is present, and the fact that Polymorph is triangular at t. If one wishes to accommodate all the relevant tensed facts in this way, one could suppose that any time whatsoever has the primitive property of being present. This might strike one as extremely far-fetched, but that is not the issue here. The crucial point is that this view is not one that combines a tenseless conception of constitution with REALISM, for it explains the (tenseless) constitution of reality by tensed facts in terms of the (tenseless) constitution of reality by certain tenseless facts.¹⁶

I think that these considerations make sufficiently clear why INDEPENDENCE and the idea that (T1) and (T4) are compatible have to be rejected. If NEUTRALITY were to be interpreted as claiming that constitution is tenseless, then (T4), the thesis that NEUTRALITY may be endorsed by realists, must be false; and without (T4), the TOC-responder’s case against (P1) collapses. I believe, however, that the correct response to the objection is to reject (T1) and (T3)—to deny, that is, that NEUTRALITY implies a tenseless conception of constitution and, therefore, ATEMPORALISM. There is in fact a much more satisfactory understanding of NEUTRALITY that is perfectly compatible with (T4), which, however, does not support (T5), the TOC-responder’s conclusion that REALISM is compatible with ATEMPORALISM.

4.2 Tenselessness vs. non-mono-anchoredness

As originally formulated by Fine, NEUTRALITY appears to dispute two things: the idea of a privileged time and the idea of facts’ featuring temporal orientation. We can associate the latter with what we have called temporal anchoring: a fact’s being oriented towards a time in the relevant sense appears to be the same as its being anchored at that time. The obtaining of such a fact is a temporal affair—it CONSTITUTES reality from within some temporal standpoint, rather than constituting it atemporally. Given this, it might seem plausible that accepting NEUTRALITY commits one to the tenselessness of constitution and therefore to ATEMPORALISM. Fine himself is no stranger to this notion of tenselessness. Consider, for instance, how he (2005a: 322; 323) contrasts tenseless sentences with tensed ones that happen to be always true:

There is a … distinction between sempiternal and eternal truths, a sempiternal truth being a tensed sentence that is always true and an eternal truth being a tenseless sentence that is true simpliciter … [A]n eternal truth will be true regardless of the time, i.e. regardless of how things are at the time, while a sempiternal truth will be truth whatever the time, i.e. however things are at the time. In the former case, there will be no genuine engagement with how things are at each time while, in the latter case, there will be.¹⁷

Above we have seen that, if NEUTRALITY is taken to express the tenselessness of constitution in this strong, atemporal sense, then it is not compatible with REALISM. But on closer inspection, NEUTRALITY can be read as taking issue not with the very idea of temporal anchoring or orientation per se, but rather only with its combination with the idea of a privileged time, as Fine’s (2005b: 271) following comments suggest: “What [NEUTRALITY] means … is that there should be no privileged time t for which the totality of tensed facts constituting reality are ones that obtain at t.” To be sure, even if every tensed fact must be anchored at, and hence oriented towards, some time, it does not follow that there is a single unique time at which all the tensed facts are anchored. What NEUTRALITY is supposed to rule out, then, is the supposition that one single time enjoys the privilege of serving as the unique locus of obtainment, as the sole temporal anchor, for all the tensed facts there are. To make this reading more salient, we can reformulate NEUTRALITY as follows:

NEUTRALITY*

No time is privileged as the unique locus of obtainment of the tensed facts that constitute reality.

Let us say that rejecting NEUTRALITY* commits one to the view that the constitution of reality by facts is mono-anchored, and accepting it commits one to the view that constitution is non-mono-anchored. (T1) is also to be reformulated accordingly:

(T1*) NEUTRALITY* is true iff constitution is non-mono-anchored.

If constitution is a fundamentally tenseless, atemporal matter, then it is also non-mono-anchored. But crucially, non-mono-anchoredness does not entail unanchoredness or tenselessness—for, constitution may be non-mono-anchored in virtue of being multi-anchored. There may be no single unique time at which constitution exclusively takes place, but it does not follow from this that constitution is tenseless in the sense of being a matter that is entirely insensitive to time. It is perfectly consistent to think that constitution is a tensed, temporally anchored affair, even though non-mono-anchored: it is something that takes place at more than one single temporal locus, and hence not ‘regardless’ of time.¹⁸ On the current approach, it is this gap between a mono-anchored and a multi-anchored conception of tensed constitution that non-standard realism exploits, rather than the alleged gap between the tense-related nature of constitution and the tense-related nature of the facts that constitute reality.¹⁹

I have argued above that the reading of NEUTRALITY encapsulated by (T1) conflicts with (T4), the claim that one may consistently accept both REALISM and NEUTRALITY. However, the conflict disappears once NEUTRALITY is understood along the lines of (T1*), as implying that constitution is non-mono-anchored rather than tenseless. For the combination of REALISM with NEUTRALITY* does not necessarily imply a view that attempts to cast tensed facts into a role that they are inherently unsuited for—it does not necessarily require them to be involved in the tenseless, atemporal constitution of reality. The fundamental constitution of reality by facts cannot be a uniformly unanchored matter if some of those facts are genuinely tensed. But there is nothing in the latter thesis that necessitates a mono-anchored conception of constitution, that is, the rejection of NEUTRALITY*.

Once (T1) is replaced by (T1*), however, the TOC-responder’s case against (P1) collapses once again. Given that NEUTRALITY is understood as NEUTRALITY*, (T3) is to be reformulated as follows:

(T3*) NEUTRALITY* is true iff ATEMPORALISM is true.

But (T3*) cannot be derived from (T1*) and (T2). Worse still, (T3*) is clearly false: Granted, if ATEMPORALISM is true, then so is NEUTRALITY*—if constitution is unanchored, it is ipso facto non-mono-anchored. But the possibility of a multi-anchored conception of constitution shows that the converse implication does not hold.

Let us briefly take stock of the discussion in this section. (P1) of the argument from atemporality claims that REALISM and ATEMPORALISM are not compatible. The TOC-responder’s case against this is premised on a particular reading of NEUTRALITY that construes this principle as ruling out tensed, temporally anchored constitution. But on that reading, (T4), the claim that realists may consistently accept NEUTRALITY, is no longer defensible. On the alternative reading of NEUTRALITY as NEUTRALITY* defended above, the principle rules out mono-anchored constitution only, rather than temporal anchoredness in general. Given this reading, (T4) is unobjectionable, but now the TOC-responder’s (T3) comes out as false. In either case, the objection to (P1) fails: there is no interpretation of NEUTRALITY on which both (T3) and (T4) can be true.

5 Defending the argument – part II

The TOC-responder’s most general and direct strategy against (P2) would be to claim that this premiss mischaracterises ATEMPORALISM as a consequence of ABSOLUTISM, while it is actually a consequence of NEUTRALITY. This simple line of attack is not tenable if NEUTRALITY is to be understood as NEUTRALITY*, as was argued in the previous section. But the TOC-responder need not in fact dispute the reasoning above in order to challenge (P2). For regardless of how one interprets NEUTRALITY, the following is clearly true:

(T6) If NEUTRALITY is false, then it is not the case that constitution is tenseless.

Recall now (T2) above from the case against the first premise:

(T2) ATEMPORALISM is true iff constitution is tenseless.

From (T6) and (T2), it follows:

(T7) If NEUTRALITY is false, then so is ATEMPORALISM.

The Finean framework seems to give us prima facie reason to assume that the following is true as well:

(T8) The falsity of NEUTRALITY is compatible with ABSOLUTISM.

From (T7) and (T8), however, it follows:

(T9) ABSOLUTISM is compatible with the falsity of ATEMPORALISM.

And (T9) contradicts (P2) of the argument from atemporality, which claims that if ABSOLUTISM is true, then so is ATEMPORALISM.

The argument is clearly valid, and both (T2) and (T6), and therefore also (T7), seem indisputable. (T8), however, is where the TOC-responder’s reasoning goes astray. As I argue below, neither a mono-anchored conception of constitution that is entailed by the falsity of NEUTRALITY nor indeed a multi-anchored conception of constitution is in fact compatible with ABSOLUTISM.

5.1 Tenselessness, mono-anchoredness, absoluteness

(T8) is the key premiss of the argument sketched out above, and since it is standard realists who simultaneously reject NEUTRALITY and endorse ABSOLUTISM, this particular strategy against (P2) is especially significant in the context of standard realism. But some philosophers seem to hold (T8), or something along those lines, because they already believe something much stronger: they seem to think that a theory that rejects NEUTRALITY is not just compatible with ABSOLUTISM but is rather committed to ABSOLUTISM precisely in virtue of its rejection of NEUTRALITY. For a particularly lucid articulation of this line of thinking, consider what Sider (2011: 240–241) has to say about presentism, the most widespread ontological variant of standard realism:

The presentist’s fundamental sentences describe reality from the current perspective, so to speak, rather than from an atemporal perspective on the whole of temporal reality. … When the presentist wishes to speak of the past, she prefixes her sentences with tense operators, as in: “It was the case that there are dinosaurs”, P∃xDx. This can still be regarded as being from the current perspective since from that perspective, dinosaurs are in the past. But this is not to say that presentist regards their pastness as a relative matter. Their pastness is absolute, since the current perspective is the only perspective. It is true simpliciter—not relative to a “perspective”—that P∃xDx.²⁰

Note that Sider fully acknowledges here that the presentist, or more generally, the standard realist, rejects ATEMPORALISM. This is our (T7) above. But Sider appears to think that the mono-anchored conception of constitution is ipso facto an absolute conception: the past-tensed fact he mentions, despite constituting reality from a temporal perspective, gets to constitute it in an absolute manner simply because the temporal perspective from which it constitutes reality is the only temporal perspective there is.

But this is clearly a mistake—that the temporal perspective from which the past-tensed fact obtains is the only temporal perspective there is does not make the manner in which it obtains any less temporally perspectival. In a metaphysics like standard realism in which constitution is mono-anchored, there is no perspectival variation in reality’s constitution by facts—it is not the case that, from one temporal standpoint to the next, the facts that constitute reality vary, for there is but just a single temporal standpoint.²¹ Such perspectival variation is also ruled out by absoluteness of constitution. However, absence of perspectival variation does not entail absence of perspectival relativity: the relevant notion of perspectival relativity is entirely compatible with there being no perspectival variation. Whether a fact constitutes reality in a temporally absolute fashion or rather only relative to some temporal perspective is one question, whether such temporally perspectival constitution occurs only from the perspective of a single unique time or rather from the perspectives of multiple times is quite another. Consider the thesis that some facts constitute reality in an irreducibly first-personal manner, in virtue of obtaining from the perspective of a person.²² Now imagine a world in which there is only one single person. Does this fact about the number of persons rule out the relevant thesis about the nature of constitution? I cannot see why it should. Whatever vices or virtues it may have, a mono-anchored conception of reality is one in which some of the facts that fundamentally constitute reality are irreducibly temporally anchored; and this sort of picture just does not seem compatible with ABSOLUTISM, let alone entailing it.

But Sider is not the only one who thinks that a mono-anchored, or more generally, tensed conception of constitution can consistently be taken to be absolutist. As already mentioned, most A-theorists, or realists about tense, seem to endorse a similar picture; indeed, above we have seen that Zimmerman (2005) believes such a commitment to be the trademark of the A-theory. More recently, Cameron (2015: 67–68) has also argued that this conception is central to presentism:

So I am now 6ft tall, and this is not a matter of me bearing some relation to the present time, or a matter of my having this property a certain way: I have this property simpliciter. That is how I am now, but it is also how I am simpliciter, for how things are simpliciter just is how things are now ….

Cameron (2015: 132) also maintains that this latter identity thesis is something that “all A-Theorists, presentists and non-presentists alike, should accept”.²³ Indeed, he (2015: 131) believes that endorsing it is “what it is to accept that the present is privileged in the way the A-Theorist wants.”

Recall the standard realist’s account of Polymorph’s world featuring the facts that, presently, Polymorph is square, that Polymorph was triangular, and that Polymorph will be circular. On the present approach, an account like this purports to represent how things temporally absolutely are in Polymorph’s world, despite offering a temporal perspective onto it. It is intended as a description of how things are, simpliciter, or, how they are, full stop; and we are told that this holds because, according to the relevant metaphysics, how things are, full stop, is identical to how things presently are.

One question these claims (and similar ones advanced by others) give rise to is how the absolutist conception of constitution is to be squared with the standard realist’s representational practice.²⁴ For suppose that, after being given the above account, we wait a moment and then ask the standard realist once again how things fundamentally are in Polymorph’s world. The standard realist will deliver a different account this time, one that cites the facts that, presently, Polymorph is circular, that Polymorph was triangular, and that Polymorph was square. But if this is the way in which the standard realist proceeds, then how is it that the initial account represents the way things are, full stop? Would not a comma, or perhaps a semicolon at best, be more appropriate given that the standard realist, immediately after she is finished with the first story, goes on to tell a different one? What the standard realist is prepared to say, at any given moment, about how things fundamentally are in Polymorph’s world comes with an expiration date—and a maximally short one at that. But how things temporally absolutely are cannot be just a temporary way they are; indeed, it cannot even properly be said to be permanent. For, on a natural understanding, temporariness and permanence are ways of ‘responding’ to time; the temporally absolute way reality is (the way it is, simpliciter, full stop), on the other hand, is not supposed to be something that ‘responds’ to time in any way at all.²⁵

Let us take the above reasoning one step further. Tenseless constitution, we have said above, is constitution that occurs (or fails to occur) regardless of time, without any temporal qualification whatsoever, independently of all temporal matters. But this is exactly what could reasonably be meant by saying that constitution is temporally absolute: constitution’s being free of any temporal qualification seems to be the core idea behind both ABSOLUTISM and the tenseless conception of constitution. It is difficult to see what else there could be to either of these claims about the nature of constitution such that they turn out to be distinct or even independent of each other. It therefore seems plausible that the tenselessness of constitution consist in nothing more than its temporal absoluteness:

ABSOLUTENESS OF TENSELESSNESS (AOT)

What it is for a fact to constitute reality in a tenseless fashion just is for it to constitute reality in a temporally absolute fashion.

Given AOT (and cot above) how things are, simpliciter, temporally absolutely, just is how they are. It is therefore difficult to see what to make of Cameron’s identity thesis.²⁶ How things are now is, at a minimum, how things are from a temporal perspective. But how things are, simpliciter, is not supposed to be how things are from any temporal perspective at all. Irreducibly tensed constitution, be it of the multi-anchored or of the mono-anchored kind (to which the standard realist commits herself by rejecting NEUTRALITY), implies that talk of what facts constitute reality is subject to temporal qualification—for what is tense supposed to do to constitution, if not temporally qualify it? The standard realist appears to say something like this: “As far as the present constitution of reality is concerned, things are this way, simpliciter.” But here a temporal qualification seems to be already in place by the time the standard realist gets to pronounce what she says about reality to hold absolutely.²⁷

It seems, then, that (T8), the thesis that those who reject NEUTRALITY may consistently endorse ABSOLUTISM proves untenable. But without (T8), the TOC-responder does not have a case against (P2), and (P2) of the argument from atemporality, which claims that ABSOLUTISM entails ATEMPORALISM, is vindicated as well.

Are we home free, then? Not exactly. The incompatibility between the tensed conception of constitution and ABSOLUTISM is manifest to such a degree that it just is not plausible that so many philosophers, including Zimmerman and Cameron, may have simply overlooked it—we need a more credible explanation of what is going on here. There is, I think, such an alternative explanation but before I can spell it out, we will have to take somewhat of a short detour first.

5.2 Qualitative change and temporal qualification

Suppose we accept ABSOLUTISM and wish to characterise the way things fundamentally are in Polymorph’s world. Given ABSOLUTISM, the facts we shall appeal to will be those that constitute reality without any temporal qualification whatsoever; indeed, given AOT, they will be those that CONSTITUTE reality. Our task is thus to articulate how Polymorph is, what shape it has. But here we are faced with an immediate challenge. Since Polymorph constantly changes its shape, there are three candidate shape properties that we can ascribe to Polymorph. There seems to be no basis for supposing that ascribing only one of these properties to Polymorph at the expense of the other two would settle the matter, in absolute terms, once and for all. But provided that we wish to preserve COHERENCE, neither can we simply claim that Polymorph has each of the candidate properties.

The issue is perfectly general. If something changes over a period of time, then there are at least two incompatible ways it is within that period. But once we take an atemporal perspective upon reality, we realise that each of those incompatible ways has a claim to be part of the way things are. Thus, the basic fact that qualitative change occurs in Polymorph’s world compels us to introduce some sort of temporal qualification into our account, as long as COHERENCE is to be retained. Typically, this is done by switching to talk of things’ having their properties at times: Polymorph is triangular at t₁, square at t₂, and circular at t₃. This, however, leaves it unclear exactly what metaphysical work the temporal qualification introduced in this way is supposed to do. We know that it is not constitution that thereby gets temporally qualified—for, ex hypothesi, the relevant facts are ones that CONSTITUTE reality. But then what in Polymorph’s world do we qualify by means of modifiers like “at t₁”?

One answer is that what gets qualified is none other than Polymorph itself. What is triangular or square is not the persisting object Polymorph; rather, Polymorph-at-t₁, an instantaneous entity, is triangular, and another distinct instantaneous entity, Polymorph-at-t₂, is square. Thus, the facts that fundamentally CONSTITUTE the way Polymorph’s world is involve three distinct instantaneous objects with distinct temporal locations, each of which has a single property different than the ones that the other two have, rather than a single persisting entity that has three different properties at different temporal locations.²⁸

Alternatively, one may opt to temporally qualify Polymorph’s properties instead. One straightforward way of implementing this idea involves increasing the adicity of properties and relations from n to n + 1 by augmenting them with an extra argument place for times. Thus, monadic properties are turned into dyadic relations, dyadic relations into triadic relations, and so on. On this view, then, a single persisting entity, Polymorph, is the bearer of various shape properties, but what it is for Polymorph to be triangular at t₁, for instance, is for it to stand in the relation of being-triangular-at to t₁; and mutatis mutandis for the other two shape properties.²⁹

What these two strategies have in common is that they both involve a more or less revisionary picture of either the subjects of change or the properties with respect to which they change. Some philosophers have argued, however, that there is no need to tinker with objects and their properties in order to make sense of the temporal qualification involved in the facts underlying an episode of change. There may well be a single persisting object that is a component of both the fact that Polymorph is triangular at t₁ and the fact that that Polymorph is square at t₂, and it may well be that each of the properties in question are monadic shape properties. What the temporal qualification should target is neither Polymorph nor its properties, but rather Polymorph’s having those properties.

There are, however, two fundamentally different ways of making sense of what is being proposed here. On one reading, this approach abandons ABSOLUTISM and switches to a temporally non-absolute notion of constitution: if Polymorph’s having triangularity is to be understood as genuinely temporally qualified, then it can no longer be claimed that triangularity is a property that Polymorph has. On an alternative reading, the proposal is supposed to be compatible with ABSOLUTISM. But in that case, it becomes questionable whether what is being offered constitutes a genuine alternative. For given ABSOLUTISM, it seems inevitable that the explanation of what it is for Polymorph’s having triangularity to be temporally qualified falls back on the tools favoured by one of the first two strategies presented above.

To illustrate, consider adverbialism.³⁰ On this view, Polymorph is t₁-ly triangular, t₂-ly square and t₃-ly circular. The adverbialist seems to leave Polymorph and its properties alone and qualify instead the having of those properties by Polymorph. Yet, on a natural interpretation, the adverbialist is simply doing to the instantiation relation that ties Polymorph to its properties what property-qualifiers are doing to properties themselves: increasing the adicity by one, introducing an argument place for times.³¹ Thus, Polymorph’s being t₁-ly triangular is simply a matter of its standing in the instantiation relation to monadic triangularity and t₁; and so understood, the adverbialist employs the same strategy as the property-qualifier, only more selectively.

The view Sally Haslanger (2003) calls SOFism is also worth mentioning.³² According to the SOFist, the obtaining of the fact that Polymorph is triangular at t₁ consists in there being a type of state of affairs involving Polymorph and triangularity, and this type’s being tokened at t₁. This immediately raises the question of how to understand the temporal modifier in the latter atemporally obtaining fact. One possible strategy mimics the property-qualifier by saying that the relevant state of affairs stands in the obtainting-at relation to some time. Another one suggested by Haslanger (2003: 349) takes instantaneous token state of affairs as basic, and thus appeals to the sort of tool employed by those who replace persisting things as property bearers with instantaneous entities.

Adverbialism and SOF-ism, as well as other views along similar lines, are clearly distinct from standard object-qualifying and property-qualifying approaches. But as long as they remain committed to ABSOLUTISM, the proponents of these views will be forced to explain the relevant kind of temporal qualification ultimately in terms of facts that obtain in a temporally unqualified manner—either ones that involve instantaneous entities at least partly individuated by the single time at which they exist or else ones that comprise predicables with an additional argument place for times.³³ ABSOLUTISM must be abandoned if it is really, irreducibly, things’ having some property that is to be temporally qualified, in a manner that is not explicable in terms of things’ being a certain way.³⁴

The upshot of all this is as follows. The mundane fact that things undergo qualitative change compels those who accept both ABSOLUTISM and COHERENCE to introduce some form of temporal qualification into their metaphysics. Although there are various ways how this can be achieved, the required temporal qualification will have to be associated with the components of the facts that underlie the relevant episodes of change: either with objects or with predicables. In other words, given that things change and COHERENCE holds, ABSOLUTISM cannot be combined with the following principle.

COMPONENTIAL ABSOLUTENESS (CA)

The components of the facts that fundamentally constitute reality do not feature any form of temporal qualification.

But CA becomes compatible with change and COHERENCE, once ABSOLUTISM is dropped—indeed, given the rejection of ABSOLUTISM, CA is a very attractive principle. For, as we have seen above, the absolutist strategies to accommodate qualitative change end up complicating their account of objects and predicables; and being able to avoid any such complication is a clear advantage of rejecting ABSOLUTISM. On this sort of view, the fact that Polymorph is triangular, for instance, may be a fundamentally temporal matter, even though it involves no componential qualification whatsoever: the fact components are simply the persisting entity Polymorph and monadic triangularity. But the fact fundamentally constitutes reality in a temporally qualified way, without this temporal qualification itself being understood in terms of some putative atemporal fact that has the fact Polymorph is triangular stand in the relation of constituting-at (or obtaining-at) to t₁.³⁵ In other words, non-absolutists can construe the temporal qualification required to accommodate qualitative change as a constitutional, rather than componential, matter.

5.3 Componential vs. constitutional qualification

If the argument from atemporality is sound, realism is inconsistent with ABSOLUTISM, which in turn means that it is compatible with CA—indeed, it is compatible with CA precisely because it entails a non-absolute, temporally qualified conception of constitution. Unlike anti-realism, realism already comes equipped with a form of temporal qualification in the guise of tensed facts; this is why realists need not introduce any further componential qualification in order to preserve COHERENCE in the face of qualitative change.

Return now to the question we have raised earlier: why do many philosophers think that something along the lines of (T8) is true, while the incompatibility of a tensed conception of reality’s constitution by facts with ABSOLUTISM seems rather obvious? Here is one hypothesis: they confuse constitutional absoluteness with componential absoluteness, or perhaps conflate these two distinct notions into a single one. This is already suggested by the relevant passages quoted above, but here are some further examples. Zimmerman (2005: 444) believes, for instance, that commitment to the claim that “[t]here are genuinely monadic properties exemplified by things that do not always have them” is a distinguishing feature of the A-theory, which he (2005: 446) elucidates as follows:

What is being insisted upon by the A-theorist is the monadic nature of some temporary properties … What is being ruled out is the need for further ‘completion’ of properties … in order for a thing to exemplify the property; in particular, nothing like a time need be ‘added’ to the property and the thing in order to make the proposition that the thing has the property ‘complete’.

Similarly, Cameron (2015: 5–6) writes:

I am 6ft tall, but I used to be 4ft tall. But this does not force us into saying that nothing is 6ft tall simpliciter, merely 6ft at some times and 4ft at other times. We need not reject the dyadic relation of instantiation that holds between an object and a property in favor of a triadic one that holds between an object, a property, and a time. All we need say is that I am 6ft tall, simpliciter, but that I was 4ft tall simpliciter. … Being true is just another property, one that can be had or lacked by propositions. And when it is had/lacked, it is had/lacked simpliciter. … [W]e do not need to reject the monadic notion of a proposition’s being true in favor of the dyadic notion of a proposition’s being true at a time; we simply need to allow that whether a proposition has that property is one of the features of reality that is subject to change.

In these passages, the focus is clearly on the nature of properties rather than the nature of constitution. Indeed, if the above hypothesis is on the right track, then at least some of those philosophers, such as Zimmerman and Cameron, who claim that realists about tense can, or even ought to, accept a tensed and absolute conception of constitution really mean by this something like the (true) thesis that realists can, or even ought to, combine a tensed conception of constitution with CA. In other words, when using such locutions as “simpliciter”, “full stop”, “absolutely”, and the like in the relevant contexts, these philosophers seem to be concerned with the temporal unqualifiedness of the fact components, rather than that of the manner in which constitution occurs. If this is right, Cameron’s identity thesis can be taken to mean that the facts that presently constitute reality do not involve any componential temporal qualification; and understood this way, it is unobjectionable. Given that things undergo qualitative change, either COHERENCE has to be given up or one of the two forms of temporal qualification, in some way or another, has to be implemented. It is true that, unlike anti-realists, realists can endorse COHERENCE while rejecting componential temporal qualification, but this comes at the cost of temporally unqualified constitution, and hence ABSOLUTISM.

6 Conclusion

The foregoing has hopefully gone some way towards making the case for the soundness of the argument from atemporality. There might be other ways in which the argument can be resisted, but here I want to conclude by briefly addressing the question of what implications the argument has for the overall debate between realism and anti-realism about tense.

If successful, the argument demonstrates that there are not in fact as many consistent versions of realism as the official Finean taxonomy suggests. What we seem to have instead are two general metaphysical ‘packages’. According to the first, atemporalist package, the facts that fundamentally constitute reality do so in a tenseless, temporally absolute manner, and are, precisely for that reason, tenseless: REALISM is false, whereas ABSOLUTISM and ATEMPORALISM are true. According to the second, temporalist package, at least some of the facts that fundamentally constitute reality do so in a tensed, temporally qualified manner, and are, again, precisely for that reason, tensed: REALISM is true, while ABSOLUTISM and ATEMPORALISM are false. Assuming that things change qualitatively, proponents of the first package must endorse some form of componential temporal qualification if they want to uphold COHERENCE; if, on the other hand, that principle is to be given up, then no form of temporal qualification is required.³⁶ The second package, by contrast, comes equipped with constitutional qualification, which is why no further componential qualification is necessary in order to square the view with COHERENCE. Thus, there are, generally speaking, three main contenders on the market: the standard B-theory or standard anti-realism, anti-realist fragmentalism, and non-absolutist realism.

A final point is worth mentioning. Note that the conclusion of the argument is not formulated as the claim that REALISM entails that constitution is temporally relative. This is no accident. Strictly speaking, the argument demonstrates only that, given REALISM, constitution cannot be absolute; thus, for all the argument shows, it could be argued that the choice between absolute and relative constitution is a false dilemma, and that there is a third conception of constitution that is neither absolute nor relative, but rather primitively tensed. I myself believe that tensed constitution can and should be analysed in terms of temporally relative constitution, correctly understood.³⁷ But it is often claimed that A-theorists have tense as a primitive piece of ideology in their accounts,³⁸ and perhaps a primitive, unanalysable notion of tensed constitution is what this amounts to. The crucial point to realise is that tensed constitution, whether or not it can be understood in terms of temporally relative constitution, is inconsistent with ABSOLUTISM—that much, I think, is made sufficiently clear by the argument from atemporality.³⁹