The Atlas of Reality
Page 74
Since Anti-Tensers hold that the differences between past, present, and future are merely perspectival, they would have no reason to treat the past or future differently in this respect. Some Tensers, such as Simple Tensers would agree. Modal Tensers and Presentists could take either position, since there could be primitive, simple facts about what will in fact happen (contingently) in the future. Thus, the truth-valuelessness of future contingent propositions is an option for Tensers but not for Anti-Tensers. If we have good reason to think that future contingent untensed propositions do lack a truth-value, this would provide another argument for Tensism.
20.3T Indeterminacy of Future Contingents. Some future contingent untensed propositions lack any truth-value.
20.3A Determinacy of Future Contingents. All future contingent untensed propositions are either true or false.
The main argument for the Indeterminacy of Future Contingents appeals to the existence of free will, in a metaphysically serious sense. Here is a standard version of the argument against what is called ‘logical fatalism’:
If future contingents are determinate, then any untensed proposition concerning what one will freely choose to do tomorrow has already been either true or false in ages past.
Any fact about ages past is beyond one's control, both now and in the future. (The fixity of the past)
Therefore, if future contingents are determinate, the truth or falsity of any untensed proposition concerning what one will freely choose to do tomorrow is beyond one's control, both now and in the future. (From 1 and 2)
The truth or falsity of any untensed proposition concerning what one will freely choose tomorrow logically entails that one will choose or not choose accordingly, and vice versa. (That is, the truth of p logically entails that p, and vice versa.)
If one cannot control whether it is the case that p, and p and q logically entail one another, then one cannot control whether it is the case that q. (The Transfer Principle)
Therefore, if future contingents are determinate, then one cannot control whether it is the case that one freely choose to take some definite course of action tomorrow. (From 3, 4, and 5)
One can control whether it is the case that one freely choose to take some definite course of action tomorrow. (Freedom entails self-control.)
Therefore, some future contingents are indeterminate. (Reductio from 6 and 7)
Anti-Tensers have the option of denying the reality of free will, in the strong sense involved in step 7. However, most Anti-Tensers will object already to step 2. They might deny that the past is already fixed. For example, some Anti-Tensers accept the possibility of time travel. Even if they admit that the past is fixed in some way, they might deny that it is fixed in all respects. In the Middle Ages, scholastic philosophers such as William of Ockham proposed a distinction between hard and soft facts about the past. The fact that the Battle of Hastings occurred in 1066 AD is a hard fact about the past: it is clearly beyond our powers of control. However, the fact that it is now true that a sea battle will take place tomorrow might well be a soft fact, one that is still under our control.
On Ockham's view, the future is genuinely open relative to the present, in the sense that multiple futures are now possible, even though only one of them is the actual future. In effect, Ockhamists can accept McCall's Falling-Branches picture of time, while holding that there is a unique stem that leads from the present through all future branch-points, a stem that represents the unique, actual future. Hence, for Ockhamists, branches don't fall off the tree of time: whether a branch B is possible at a time t depends simply on whether t is located before the root of B.
Tensers could respond by insisting that this distinction is forced and ad hoc. Either the past is fixed in toto or it is totally open. In addition, Tensers could argue that Anti-Tensers have no account of why the past is in any respect fixed or closed or beyond our control, while the Tensers can explain this fact in terms of the determinacy of the past, in contrast to the indeterminacy of the future. We will take this issue up again when we consider the problem of the direction of time and of causal control in Sections 27.1.1.2 and 28.5.
20.5.6 Circular time and time travel
Is time travel possible? If Tensism were true, then it would seem to be impossible. For example, if Presentism is true, then there is no future or past to travel into. If we adopt instead Simple Tensism, time travel still seems impossible, since it would involve some “spark” of actuality breaking away from the unique present and wandering off into a region of near-total non-actuality. What kind of time travel would that be? The realms of past or futurity we reach would be devoid of any reality. Finally, time travel in a world in which Modal Tensism is true seems equally absurd. Travel to the future would be travel to a realm lacking any closure or definition, while travel to the past would involve encasing oneself in the inexorable grip of fixed fate.
In the case of Anti-Tensism, it is not obvious that time travel is possible, but neither is there any obvious barrier to its possibility.
So, is time travel possible? Could one travel into the past or the future?
There are at least two reasons for thinking that it might be possible. First, it seems to be conceivable, as evidenced by the many time travel stories written in the last hundred years. Given Imagination as Guide to Possibility (PEpist 1), this is good reason to think time travel is possible. Of course, this reason is defeasible, so arguments for Tensism might count as evidence that defeats the evidence supplied by the conceivability of time travel. Second, there are solutions to the equations of Einstein's general theory of relativity discovered by mathematician Kurt Gödel that describe a world in which time is partly circular. This would mean that some objects could, by traveling far enough through the universe, arrive back at their destination before they departed. This would seem to involve real time travel, and it is reasonable to take our best scientific theories at face value:
PMeth 2 Scientific Realism. Other things being equal, adopt the theory that implies that our best scientific theories are straightforwardly true, as standardly represented.
However, general relativity by itself cannot establish the real possibility of time travel, since even assuming that it is true, it still might be incomplete as a description of the physical constraints of the world. Something outside the scope of general relativity might render such circular time impossible.
A well-known argument for the impossibility of time travel is the grandfather paradox.
If time travel were possible, it would be possible for someone, Mr. X Jr., to travel back in time and kill his own paternal grandfather before Mr. X's father, Mr. X Sr., was conceived.
It is impossible for Mr. X Jr. to exist if his paternal grandfather died before his father was conceived.
It is impossible for Mr. X Jr. to do anything, including travel back in time, if he doesn't exist.
Hence, it is impossible for Mr. X Jr. to travel back in time and kill his own paternal grandfather before Mr. X's father was conceived.
Hence, time travel is impossible.
The defender of time travel might reject step 2, with its implicit assumption that the genealogical origin of both Mr. X Jr. and of Mr. X Sr. are essential to their identities. Still, it seems we could change the story by enabling Mr. X to annihilate the earth, or even the entire cosmos, before his own conception. Thus, the most vulnerable premise is the very first one. Why think that just because time travel in general is possible, it must be possible to kill one's own grandfather as a toddler?
The response to this challenge is to resort again to patchwork principles like Finite Patchwork:
PMeta 5.1 Finite Patchwork. If an event or process of (intrinsic) type A is possible, as is an event or process of intrinsic type B, and if there is enough room in the history of the world to locate in it instances of both events (or processes) without overlap in time and space, then it is possible for events (or processes) of both types to occur together.
Let's suppose that it is possible for Mr. X Jr.
to travel into the past, so long as he doesn't disturb anything essential to his own existence. Call this process A: Mr. X Jr. travels into the past and passively observes events, without disturbing anything. It is surely possible for Mr. X Jr. to kill a toddler, so long as the toddler is not his own grandfather or other ancestor. Call this toddler-killing scenario process B. Finite Patchwork justifies us in assuming that there must then be a possible world in which processes A and B are spliced together, resulting in the grandfather paradox. The only solution seems to be to deny that any time travel is possible at all.
The cogency of this line of reasoning depends on the Intrinsicality of Powers (PMeta 2). If powers are intrinsic, as Powerists (4.4A.3) and Nomists (4.4A.2) believe, then Finite Patchwork (PMeta 5.1) is applicable to the toddler scenario, with the result that time travel is impossible. However, if powers are not intrinsic to their bearers, as Neo-Humeists (4.4T) contend, then Finite Patchwork does not apply, since we have not limited ourselves to intrinsic facts in describing the toddler-killing scenario (Lewis 1976). Whether one is able in a certain situation to kill a toddler depends on the pattern of qualitative facts that pervades the actual world.
The cogency of the argument also depends on the question of Transworld Identity (16.1T) versus Worldbound Individuals (16.1A). According to the Transworld Identity, in order for Mr. X Jr. to be able to kill Mr. X Sr.'s father as a toddler, there must be a world in which they both exist and in which the first kills the second. This seems to be impossible, since Mr. X Jr. cannot exist in any world in which Mr. X Sr.'s father has died as a toddler. However, if we adopt Worldbound Individuals, and in particular Counterpart Theory (16.1A.1), instead, all we need is for Mr. X Jr. to have a counterpart in a world who kills a counterpart of Mr. X Sr.'s father. This seems possible, since there doesn't seem to be any bar to Mr. X Jr.'s having a counterpart that isn't a grandson of Mr. X Sr.'s father's counterpart (in that world).
20.6 Conclusion
In this chapter, we have been examining arguments in favor of Tensism. These arguments fall into two main categories. First, there are those arguments, like the Thank Goodness argument, our experience of the flow of time, and the reality of intrinsic change, that appeal to how time appears to us in our ordinary, everyday experience. Metaphysicians must take such data seriously if they are not to embrace a global skepticism about the world of appearances. However, such considerations are in principle defeasible. Second, there are arguments that presuppose a particular conception of time and of the nature of power. Those who embrace Powerism and an Aristotelian views about time have several good arguments for embracing Tensism, namely, the threat of fatalism, the impossibility of time travel, and the present as an actual boundary between past and present. In the next chapter, we consider arguments against Tensism.
Note
1. Well, we don't suppose that they would have said it then in modern English. Better: ‘Imperium Romanum intactum est,’ or something like that.
21
Arguments for Anti-Tensism
In this chapter, we look at six arguments against Tensism (20.2T). They are, equivalently, arguments for Anti-Tensism (20.2A). The arguments are of three basic kinds: those that argue that Tensism is incoherent or mysterious, those that argue that it is in irresolvable conflict with modern science, and those that fault Tensism for its unexplainable or brute necessities. In Section 21.1, we consider the objection that Tensism cannot sensibly account for the rate of the flow of time. This is followed by the long Section 21.2, in which a variety of objections based on Truthmaker Theory (2.1T/2.1A.1T/2.1A.1A.1T) are lodged against various versions of Tensism. In Section 21.3, we look to science, especially Einstein's theory of relativity, as the basis for an argument against Tensism. Relativity theory seems to deny the possibility of a cosmic “now”, since it denies the existence of absolute, or observer-independent, simultaneity relations. The final three sections concern additional philosophical objections to Tensism. Section 21.4 considers epistemological problems for Tensism; Section 21.5 considers McTaggart's argument that Tensism is self-contradictory; and Section 21.6 explores the number of brute necessities for which Tensism can provide no simple explanation.
21.1 How Fast Does Time Flow?
Both C.D. Broad (1938) and J.C.C. Smart (1949) raised the following sort of argument against Tensism:
Tensism implies that time itself passes, that events and times continually change their position in the A series.
Any process of passage or change takes place at some sort of rate.
The pure passage of time is a process that cannot take place at any rate, since (i) if its rate is measured against some kind of meta-time, then Tensists face an infinite regress, and (ii) the trivial rate of one second per second is absurd.
Therefore, Tensism is false.
Some Tensers (for example, George Schlesinger 1980) are willing to embrace the reality of meta-times but deny that an infinite regress results. (Schlesinger proposes that there are two A series, each flowing at a rate to be measured against the other.) One could go further and embrace an infinite series of meta-times while denying that this infinite regress is vicious. This isn't a very popular line, however. Therefore, we ignore it here.
Ned Markosian (1993) seems to have the right Tensist response to this argument. Markosian distinguishes two possibilities. First, it may be that all talk of rates of processes is irreducibly relative. The most we can do is compare the progression of one process with that of another. Second, it may be coherent to talk about the absolute rate of a process, that is, its rate when its progression is compared with the passage of time itself.
On the first possibility, there is no difficulty in specifying the rate of the passage of time. Time passes, for example, at the rate of one year per 365.25 revolutions of the Earth. We can pick any process whose regular progression is constant when compared with the flow of time itself.
Markosian seems to suggest that the second possibility entails Tensism. If there were no pure passage of time, there could be no absolute rates or velocities, since absolute rates are simply the comparison of the progression of one process with the progression of time itself (p. 841). However, this isn't right. Anti-Tensers could insist that the time dimension has an intrinsic measure, so that it would make sense to measure the absolute rate of process by using equally spaced benchmarks along the B series.
Markosian divides the second possibility into two sub-cases: (a) it makes sense to determine the rate of time's passing by comparing its progress with itself or (b) it does not make sense to do so. If (a), we should embrace the trivial answer that time passes at the rate of one second per second. If (b), we should say that the rate of time's passage is undefined because all absolute rates are defined by reference to it. One might defend (b) by saying that we are not allowed to use the rate of time's passage to define its own rate, for the sort of reason that Wittgenstein proposes for saying that it makes no sense to ask for the length of the standard meter stick. If we do so, one might think we are committing what Gilbert Ryle called a “category mistake”, applying a description to something in a category of the wrong sort. Of the two, (a) seems more plausible. Wittgenstein is wrong about the standard meter stick—it seems to be straightforwardly true that it is one meter long!
So, what exactly is supposed to be wrong with saying that time passes at the trivial rate of one second per second? Robin Le Poidevin (2005) offers a suggestion:
If a process progresses at a rate of x, then it is metaphysically possible for it to progress at some rate y ≠ x. In fact, it is metaphysically possible for it to progress (regress) at the rate −x.
It is metaphysically impossible for time to progress at any rate other than one second per second, and in the standard direction (with times moving into the past).
Hence, time does not pass at the rate of one second per second.
Markosian might respond that step 1 derives its credibility from the first possibility above: it should be metaphysically possible fo
r time to progress at different rates, when those rates are defined by reference to the progression of other processes. We could imagine a world in which all physical and mental processes progressed at exactly twice their current rate. In such a world, time would pass at a rate exactly one-half as fast (as defined with reference to any of those other processes). Once we realize that we are considering the second possibility, though, it may not be so plausible to assume that time could progress at a different rate. When that rate is measured by comparing the passage of time to itself, it's not obvious that we can imagine it passing at a different rate.
The problem of the rate of pure passage is a problem for anyone who believes in absolute units of time, whether Tenser or Anti-Tenser. Anti-Tensers should concede that time passes in exactly the same sense that a child grows, by using an At-At Theory of change (compare 24.5A.1T). The child changes its height in the sense that we find, as we pass mentally along the time-axis, the child having progressively greater heights. In exactly the same way, as we pass mentally along the time-axis, we find events having progressively more past-ish B-times. If time has an intrinsic measure, then Anti-Tensers must also embrace the one second per second account of its rate. Tensers could avoid the problem of explaining how fast time flows by denying that time has an absolute, intrinsic measure. Thus, the problem seems to be orthogonal to the debate between Tensers and Anti-Tensers.
What about the possibility of time changing direction? This would seem to make sense only for those who thought that events had a position in the B series that was metaphysically independent of the A series. That is, if, like McTaggart, we thought that the B-times of events supervened on their A-times, then time flowing backwards would make no sense. The B-times of events necessarily correspond to the order in which they had become present.