Philos Stud (2011) 155:23–35 DOI 10.1007/s11098-010-9564-6

There is a problem of change Michael J. Raven

Published online: 12 May 2010  Springer Science+Business Media B.V. 2010

Abstract Impostors are pseudo-problems masquerading as genuine problems. Impostors should be exposed. The problem of change appears genuine. But some, such as Hofweber (2009) and Rychter (2009), have recently denounced it as an impostor. They allege that it is mysterious how to answer the meta-problem of saying what problem it is: for even if any problem is genuinely about change per se, they argue, it is either empirical or trivially dissolved by conceptual analysis. There is indeed an impostor in our midst. But it is the meta-problem of change. I defend the appearance that the problem of change is a genuine metaphysical problem about change. This vindicates philosophers’ lasting interest in it. It also illuminates what makes a problem metaphysical, how metaphysics relates to other inquiries, and how best to respond to attempts to undermine metaphysical problems. Keywords

Change
Metaphysics
Time

1 The Puzzle of change Light a new candle. Let it burn out. It was straight. It is bent. So: Possibility. It is possible for a candle to change its shape from straight to bent. Possibility entails both: Persists. The candle itself persists through the change from straight to bent. Incompatibility. Straightness is incompatible with bentness. Without Persists, nothing changed. Without Incompatibility, nothing changed.

M. J. Raven (&) University of Victoria, Victoria, BC, Canada e-mail: [email protected]

123

24

M. J. Raven

One and the same candle persisted through the change from straight to bent. So Persists entails: Identity. The straight candle is the bent candle. But nothing could be straight and bent. So Incompatibility entails: Non-identity. The straight candle cannot be the bent candle. We have a puzzle: an apparently valid argument from the manifest possibility of the candle’s change to a contradiction.1 But the candle’s change isn’t special. Any alleged change would generate an analogous puzzle. So, generalizing, we get the Puzzle of Change (or just the Puzzle). One reaction to the Puzzle is that it shows that change is itself incoherent or incoherent with other entrenched beliefs. The philosophical problem is thus to cope with the impossibility of change, despite its apparent manifest possibility. But let us ignore this overreaction. Grant that we know change is possible; the Puzzle can’t show otherwise. Instead, the problem is to explain how change is possible given an argument for its impossibility.2 Those who dismiss this problem usually claim either to have a resolution or else that none is possible. Neither reaction rejects the appearance that the problem of change is a genuine metaphysical problem about change. Instead, they merely say either that it is solved or else that it cannot be solved. But no-problemers, such as Hofweber (2009) and Rychter (2009), dismiss the problem by saying there is no genuine problem to resolve. They reject the appearance that the problem of change is a genuine metaphysical problem about change. They allege that the problem of change is merely an impostor on the grounds that it is mysterious how to answer the meta-problem of saying what problem it is. To denounce the problem of change as an impostor merely on the grounds that it is hard to say what problem it is obviously is a non sequitur. So the burden is upon no-problemers to defend their allegation. Hofweber (2009, pp. 289–290) distinguishes two methods of defense. The ‘‘topdown’’ method assumes a meta-metaphysics which says generally what metaphysical problems are and employs it to determine whether there is a particular metaphysical problem of change. The ‘‘bottom-up’’ method, favored by Hofweber and also Rychter (2009), examines particular candidates for what the metaphysical problem of change might be.3 A more direct approach, however, is to see whether any problem is both metaphysical and about change. The Puzzle appears to be all three. So the burden of 1

My notion of puzzle is like Rychter’s (2009, pp. 9–10) and like Schiffer (2003)’s notion of paradox.

2

My presentation of the problem of change in terms of the Puzzle closely resembles that of Hinchliff (1996), Haslanger (2003), Kurtz (2006), Wasserman (2006), and many others.

3

Hofweber’s own method is inconsistent. At one point Hofweber (2009, p. 290) rejects the top-down method; but elsewhere Hofweber (2009, p. 287) uses the top-down method by assuming that a problem is metaphysical only if it modestly respects science’s authority and then rejecting a formulation of the problem of change because it does not satisfy this constraint. This methodological inconsistency is not atypical among no-problemers.

123

There is a problem of change

25

proof is on no-problemers to explain why this appearance is deceiving. And, indeed, the diverse arguments no-problemers give are best unified as objections to the appearance that anything is all three. That is, they argue that even if any problem is genuinely about change per se, it is either empirical or a pseudo-problem trivially dissolved by conceptual analysis. My goal is to argue that the Puzzle is as it appears: a genuine metaphysical problem about change. I defend this by refuting no-problemers’s objections to the contrary. There’s no good challenge to the appearance, so the appearance remains. I first respond to the objection that the Puzzle poses no genuine problem because a bit of conceptual analysis trivially dissolves it (Sect. 2). Next I respond to the objection that the Puzzle is merely an empirical problem (Sect. 3). Then I respond to the objection that no problem the Puzzle might pose is about change per se (Sect. 4). I conclude with a discussion of how the failures of no-problemers’s objections might generalize, thereby illuminating what makes a problem metaphysical, how metaphysics relates to other inquiries, and how best to respond to attempts to undermine metaphysical problems (Sect. 5).

2 Is the Puzzle a problem? No-problemers object that the Puzzle trivially dissolves with the help of some conceptual analysis. They conclude that it is not a genuine problem. They defend this on the basis of the legitimacy of reading Incompatibility in either of the following two ways4: Incompatibility-1a. Nothing can be straight and bent at the same time. Incompatibility-1b. Nothing can be straight and bent, not even at different times. No-problemers claim that there is no genuine problem on either reading. First, Incompatibility-1a is irrelevant because the Puzzle did not require the candle to be straight and bent at the same time. Second, Incompatibility-1b not only fails to imply Non-identity, it is also implausible to boot. Whence the problem? But the problem emerges once we realize that no-problemers overlook a competing pair of readings of Incompatibility. Consider: Incompatibility-2a. Nothing is now straight and now bent. Incompatibility-2b. Nothing now straight was, is, or will be bent. This second pair is similar to the first pair in that each temporally relativizes the straightness and bentness of the candle. But the second pair differs from the first pair by relativizing differently. The first pair relativizes tenselessly: the candle is straight at a time earlier than when it is bent. The second pair relativizes tensely: the candle was straight and is bent. (And among these two different ways of relativizing, there are yet further variations on precisely how the relativization is achieved.) As with the first pair of readings of Incompatibility, it seems there is no genuine problem for the second pair. First, Incompatibility-2a is irrelevant because the 4

Here I follow Hofweber’s (2009) presentation; but Rychter (2009) discusses these issues too.

123

26

M. J. Raven

Puzzle did not require the candle now to be straight and bent. Second, Incompatibility-2b not only fails to imply Non-identity, it is also implausible to boot. Whence the problem? But there is a problem. The problem is not with the general strategy of temporally relativizing; it is rather with there being two competing ways of achieving relativization. Each involves a competing view of reality. The second pair suggests a tensed view which asserts an objective distinction in reality between past, present, and future. The first pair suggests a tenseless view which does not suggest such an objective distinction.5 These views of reality are incompatible. It is a serious question which of them is correct. So it is a serious question which (if either) of the two temporal relativizations succeeds. What’s more, the tensed and tenseless views comport with (if not entail) different accounts of the nature of change.6 The tensed view comports with taking an object to change just in case it itself had some property it now lacks.7 The tenseless view comports with taking an object to change just in case one of its temporal parts has a property which another temporal part does not.8 By assuming the tenseless view without argument, no-problemers assume a controversial view of reality which supports a controversial view of change in the course of a controversial argument for there being no metaphysical problem of change. Far from dissolving the problem, this is at best an ex post facto justification of why there is no problem from a theoretical perspective which purports to have resolved it. Indeed, the controversy over the tenseless and tensed views makes the Puzzle a problem, much like the controversy over revising the traditional analysis of knowledge makes Gettier cases problems. Now, no-problemers might object that the Puzzle’s interaction with the controversy over tenseless and tensed views threatens its being a problem about change per se. I will discuss this objection in Sect. 4. But notice that it assumes that there is a problem (just not one about change per se), and so is useless as a defense of there being no problem. In sum, no-problemers’ alleged dissolution is, at best, just an attempted resolution. There is a problem said to be resolved. So there is a problem.

3 Is the Puzzle metaphysical? Even if the Puzzle poses a genuine problem, no-problemers object that it is not metaphysical.

5

Mellor (1998) and Sider (2001) defend the tenseless view, whereas Prior (2003) and Fine (2005) defend the tensed view.

6

See Haslanger (2003) for a detailed discussion of just how particular views about persistence are logically independent of, but nevertheless comport with, particular views about the tensed or tenseless nature of reality.

7

See Hinchliff (1996) for an influential example of this endurantist view.

8

See Sider (2001) for an influential example of this perdurantist view.

123

There is a problem of change

27

But this might seem indefensible in light of the previous section: for if resolving the Puzzle depends on sorting out other metaphysical views, then that might suffice to make the Puzzle metaphysical (innocence by association!). In reply, no-problemers might object that this merely shows that the Puzzle was never really about change per se (guilt by association!). But that’s the topic of the next section. For now, focus on whether the Puzzle is a metaphysical problem. No-problemers might argue that the Puzzle poses no metaphysical problem by arguing that there are no metaphysical problems. Were they to do so, they’d face the tough challenge of arguing against the appearance that there are metaphysical problems. But no-problemers tend to avoid this route. Instead, they tend to distinguish metaphysical and empirical problems and claim that the Puzzle is at best empirical. The metaphysical/empirical distinction is elusive. Some problems (e.g. math and logic) aren’t neatly categorized as either while others (e.g. Quinean ontology) might be both. And the distinction raises hard questions about how metaphysics relates to the empirical.9 (I return to some of these issues later.) However inchoate the metaphysical/empirical distinction is, it is revealing enough as a starting point. Granting that, it seems that the Puzzle is at least partly empirical since Possibility may have an empirical justification. But it is a mistake to take a problem’s being partly empirical to preclude its being metaphysical. For example, physics tells us much about time even if it leaves work for metaphysics. For another example, the problem of free will is metaphysical even if physics can illuminate it by telling us whether the physical world is deterministic or indeterministic. Still, the Puzzle wouldn’t be metaphysical if it were wholly empirical. No-problemers claim that there is a wholly empirical problem of how the candle changes. It belongs to material science. Material science solves it. That problem has a wholly empirical solution. So it is not a metaphysical problem.10 But this objection is not compelling for at least two reasons. One reason is that it simply declares that the Puzzle is avoided without giving any guidance as to how it is. Material science’s empirical explanation of how the candle changes by itself says nothing about which premise of the Puzzle is false or which step of the reasoning is invalid. A second reason the objection is not compelling is that it ignores that the Puzzle concerns the possibility tout court of change, whereas the wholly empirical problem concerns the nomological possibility of change. While empirical explanations may shed some light on the possibility tout court of change, it is doubtful that they can do so satisfactorily. This is because an empirical explanation of the nomological possibility of change ignores possible but nomologically impossible changes.

9

To demand that an answer to the meta-problem of change must answer these questions is to adopt a strong version of the top-down method rejected by Hofweber and Rychter. Indeed, everyone should reject this demand. We needn’t settle on a view of the whole of metaphysics prior to settling on a view of one of its component problems.

10

See Sect. 2.1 of Hofweber (2009).

123

28

M. J. Raven

So whatever solves the wholly empirical problem needn’t solve the Puzzle. They are different problems. Still, no-problemers might object that the empirical explanation of the candle’s actual change does explain the possibility of change. For what is actual is possible. So the possibility of change is trivially explained on the basis of any empirical explanation of any actual change.11 But this objection misconstrues the Puzzle. For we might know that change is possible without knowing how. This is not perversely to hold our knowledge that change is possible hostage to our ability to explain how it is. Not knowing how change is possible needn’t shake our confidence that it is, just as not knowing how 2 ? 2 = 4 or how dogs bark needn’t shake our confidence that 2 = 2 = 4 or that dogs bark. One reason we don’t trivially know how change is possible is that we don’t yet have a unified explanation of its possibility. Actual changes are disparate and could have been explained empirically in ways that don’t actually explain them. So an explanation of how change is possible should be unified in a way analogous to the functionalist’s explanation of why my firing c-fibers, Marvin the Martian’s inflating cavities, and my non-actual counterpart’s xyz-fibers are all pains: each realizes the pain role. But inferring the possibility of change from an empirical explanation of an actual change won’t provide a unified explanation of how change is possible. Unlike ‘velocity’ and ‘mass’, ‘change’ doesn’t belong to the scientist’s distinctive vocabulary. So an empirical explanation of an actual change could bear on all possible changes only by invoking some or other metaphysics of change. To illustrate, dyeing my hair from brown to blonde changes my hair color. Science explains why my hair actually changed color. But this explanation does not explain why my hair color’s change and the candle’s change are changes. Unifying both explanations as explanations of changes requires invoking a metaphysics of change which classifies both the hair-dyeing and the candle-melting as changes. For another illustration, the actual explanation of the candle’s change might not have held. A counterfactual explanation could have. Unifying both explanations as explanations of changes requires invoking a metaphysics of change which classifies both the actual and counterfactual events as changes.12 No-problemers might reject the need for unifying these explanations of actual and counterfactual changes as explanations of changes. But that is to raise what noproblemers would presumably reject: a new metaphysical problem of defending a ‘‘fragmented’’ metaphysics of change. In sum, the Puzzle resists being characterized as a wholly empirical problem because it inescapably involves the metaphysics of change.

11

Hofweber (2009, p. 287) agrees: ‘‘The empirical problem of saying why the candle got bent, or how it was possible that it changed its shape, even though no one touched it, has been solved.’’ 12 No-problemers might try to lessen the force of this second illustration by assuming that how things actually are is how they must be. But that is to make the argument for the Puzzle not being a metaphysical problem reliant upon a controversial metaphysical assumption.

123

There is a problem of change

29

4 Is the Puzzle about change? Even if the Puzzle poses a genuine metaphysical problem, no-problemers offer two objections to taking the problem to be about change per se. The first objection is directed toward those who think that the interest in the Puzzle derives from how it answers questions such as13: Question. Why is it possible for the candle to be straight and bent at different times, but not both at the same time?14 But according to no-problemers, the answer is easy: Answer. It is possible for the candle to be straight and bent at different times because the candle changed between them; and it is impossible for the candle to be straight and bent at the same time because the candle couldn’t have changed at that one time. The answer to Question invokes change and thus doesn’t explain it. So the Puzzle was not about change per se. Any inclination to think that it was implicitly begs the question by assuming that ‘‘there is something wrong with change’’, as Hofweber (2009, p. 301) puts it. But this argument is not cogent because it is circular. The following principle seems unobjectionable: Change. Necessarily, something changes if and only if it has incompatible properties at different times. Thus a legitimate restatement of Question is: Question*. Why is it possible for the candle to change from straight to bent, but impossible for it to be straight and bent? But Answer circularly answers Question*. Since Question* just restates Question, Answer therefore circularly answers Question.15 No-problemers cannot avoid this objection by rejecting Change. For that would raise the problem of coherently specifying change’s non-standard nature. And that, of course, is just to raise precisely what no-problemers reject: a metaphysical problem of change. Nor should no-problemers object that Change fails to support the substitution required to get Question* from Question. Change is close enough to a conceptual truth to warrant the substitution here, even if not elsewhere. The second objection no-problemers raise against the Puzzle being about change per se is that it rests on confusing numerical and qualitative identity. Hofweber (2009, p. 294) pursues this strategy when he writes: 13

Here I follow Hofweber (2009).

14

No-problemers overlook the tensed version of this question: ‘‘Why is it possible for the candle to have been straight and now bent, but not now both?’’ 15 It is no help to answer Question* with ‘‘Because the candle can be straight and bent at different times, but not both at the same time’’. For that circularly answers Question, but Question* just restates Question.

123

30

M. J. Raven

…[The problem of change] confuses numerical and qualitative sameness and difference. Change requires an object to be numerically the same and qualitatively different. It has to be one and the same object that first has one feature, and then a different one. But numerical sameness and qualitative difference are quite different, and thus there is no conceptual contradiction when something is both the same and different in the relevant sense, since it is the same in one sense of the word, numerical sameness, and different in another, qualitative difference. And Rychter (2009, p. 15) pursues this strategy as well. This strategy, however, is inconsistent on the standard assumption that numerical identity entails qualitative identity.16 The numerical sameness of the straight and bent candles precludes their qualitative difference. No-problemers cannot glibly deflect this inconsistency by saying it merely shows the need for a new sense in which the straight candle and the bent candle are numerically identical but qualitatively distinct. For it is controversial how to satisfy this need.17 Is it consistent for the straight candle to be the bent candle because the candle is only how it presently is (and never is how it was or will be)? Or is the straight candle the bent candle because a spacetime worm has them as temporal parts? Or are the straight and bent candles the same candle even though numerically and qualitatively distinct? The deflection fails until these questions are answered. But this last move might seem to give no-problemers’s their ‘‘ace in the hole’’. If the Puzzle’s resolution depends upon settling controversial issues in the metaphysics of identity, time, tense, and persistence, then surely the ‘‘real’’ metaphysical problems are about them and not about change per se. A fable will help show by analogy what no-problemers suggest. Imagine that skeptics once argued that we know far less than we thought. This was ‘‘the problem of knowledge’’. Philosophers later noticed that knowledge requires justified true belief and that the skeptical arguments most directly challenged whether our true beliefs were justified. They declared ‘‘the problem of knowledge’’ passe´: for it was ‘‘really’’ a problem about justification and not about knowledge per se. Analogously, no-problemers declare ‘‘the problem of change’’ passe´: for it is ‘‘really’’ a problem about identity, time, tense, and persistence and not about change per se. As Hofweber (2009, p. 312) puts it: The problem of change is not a goal for the philosophy of time to solve, it is a distraction from the real questions in the philosophy of time. We should thus give up trying to solve the problem of change, there is no such problem, and focus on the central questions in the philosophy of time instead. That’s the no-problemers’s alleged ‘‘ace in the hole’’.

16

Wasserman (2006) raises this same objection.

17

Indeed, the need is apparently rejected by standard views about persistence through change such as perdurantism and stage theory. By leaving these views off the table, no-problemers once again take without argument a controversial metaphysical stand in the course of an argument that there is no metaphysical problem of change.

123

There is a problem of change

31

But there are many problems with this alleged ‘‘ace in the hole’’. First, the alleged ‘‘ace in the hole’’ assumes a bizarre conception of how change relates to identity, time, tense, and persistence. We tend not to find it even prima facie puzzling how things could persist without change.18 We do tend to find it prima facie puzzling how things could persist through change. That’s partly why the problem of change raises ‘‘real’’ questions in addition to being intimately connected to other ‘‘real’’ questions no-problemers recognize. Second, the alleged ‘‘ace in the hole’’ invokes fallacious reasoning. The main inference moves from the premise that the problem of change involves problems about other things to the conclusion that the problem of change is really about those other things, not change. This is obviously a non sequitur. This point is worth emphasizing. Philosophical problems are notoriously hard to isolate. The problem of knowledge expands to cover truth, belief, and justification. This does not undermine the sense in which the problem of knowledge is about knowledge per se. But no-problemers incomprehensibly urge a different attitude toward change. Because the problem of change expands to cover identity, time, tense, and persistence, we are told that it is not ‘‘really’’ about change per se. This is just an unjustified and arbitrary difference in attitude; or, at best, a difference grounded in philosophical prejudices. This reveals a third problem. No-problemers claim to illuminate our dim grasp of ‘‘real’’ metaphysical problems (e.g. identity, time, tense, and persistence) by rejecting any problem about change. But this is no less baffling than claiming to illuminate our dim grasp of ‘‘real’’ epistemological problems (e.g. justification) by rejecting any problem about knowledge. However philosophical problems are individuated, it is mysterious what’s lost in recognizing a problem about change or what’s gained by refusing to do so. Indeed, it is beneficial to acknowledge a problem about change. For, as we’ve seen, certain responses to the Puzzle more naturally fit certain views about identity, time, tense, and persistence and, certain views about identity, time, tense, and persistence more naturally fit certain responses to the Puzzle. It is as if change is a center of gravity pulling the other topics into its orbit. So these topics should be assessed together. It is thus no less a methodological mistake to ignore the Puzzle when investigating the metaphysics of identity, time, tense, and persistence than it is to ignore the metaphysics of identity, time, tense, and persistence when investigating the Puzzle. Perhaps no-problemers might object that nothing but a pragmatic justification has been offered to take the Puzzle to be a problem about change. But why is this any objection at all? If imposing some order on a mess of other topics is achieved by construing the Puzzle to be a problem about change, then that seems reason enough to construe it in just that way. Indeed, this practice is familiar and widely adopted. For example, what unites instances of the sorites paradox (e.g. heaps, baldness) as members of a kind is that there is a common phenomenon which gives rise to them. The phenomenon is that there is a (possible) spectrum of cases with those on one end clearly satisfying a 18

Although some find it puzzling upon reflection; c.f. Descartes (1996) and Unger (2006, pp. 266–300).

123

32

M. J. Raven

vague predicate which is clearly not satisfied by those at the other end. In a word, the phenomenon is vagueness. The philosophical task is to illuminate what vagueness is and how it is not paradoxical (if it isn’t). This task encompasses a diverse range of topics apparently distant from the initial phenomenon. Semantic issues arise since some say vagueness is a kind of incompleteness of meaning.19 Psychological issues arise since some say vagueness is a peculiar kind of mental state.20 Epistemological issues arise since some say vagueness is a special kind of ignorance.21 Metaphysical issues arise since some say vagueness is a kind of indeterminacy in the world.22 But the interaction of these diverse topics with vagueness, far from showing that the sorites paradox isn’t about vagueness, reinforces that vagueness is (so to speak) a center of gravity pulling the other topics into its orbit. If it is legitimate to construe the sorites paradox as a problem about vagueness partly for this reason, it would be ad hoc to refuse to do the same for change. In sum, while I have not settled the hard question of what makes a problem about this per se rather than about that, it is clear that the answer implicit in the noproblemer’s objection is utterly bizarre. So it offers no serious challenge to taking the Puzzle to be about change.

5 Lessons for metaphysics Exposing impostors is vital work. But we must temper our zeal for it, lest it turn into a witch hunt. No-problemers denounce the problem of change as an impostor. But they misidentify the impostor. The impostor is the meta-problem of change. There is a problem of change. Its mysteries, far from making it an impostor, vindicate its status as a central problem of metaphysics. The interest in this issue goes beyond the problem of change and ‘‘leads to the question how metaphysics relates to other disciplines, and what is distinctive about it and its methods,’’ as Hofweber (2009, p. 286) puts it. One way to formulate this question is by raising Hofweber’s (2009) separation problem: how can metaphysics modestly respect science’s authority while being separate from it? This question raises further questions about the domain and method of metaphysics: (i) what is the subject matter of metaphysics? and (ii) what method should we use to answer questions about its subject matter? No-problemers hope that exposing the problem of change as an impostor will illuminate these difficult and perplexing questions. But, ironically, their failure illuminates these questions in a way diametrically opposed to the way they intended. This is clear once we recognize several principles implicit in their arguments:

19

C.f. Fine (1975) and Lewis (1986).

20

C.f. Schiffer (2003).

21

C.f. Williamson (1994).

22

C.f. Williams (2008).

123

There is a problem of change

33

(1a)

Metaphysics cannot modestly respect science’s authority while being separate from it if it tries to explain particular actual changes (e.g. the candle’s melting). (1b) Particular actual changes are not a subject matter of metaphysics if they are explained by empirical science. (1c) If a question putatively about the nature of change is answerable but not answerable by empirical methods, it is (at best) answerable by conceptual analysis or is not a genuine question about change. No-problemers implicitly reason as follows. By (1a), metaphysics has no business answering questions about particular actual changes. This is because, by (1b), science already answers these questions. To the extent that any genuine questions about change remain, they are, by (1c), trivially answerable by conceptual analysis. But it is far from clear what these principles can teach us about the nature of metaphysics. One reason is that (1a) and (1b) are irrelevant. As we’ve seen, metaphysics is not in the business of trying to explain particular actual changes. So (1a) and (1b) are vacuous, if true. Neither can be expected to shed significant light on what metaphysics is in the business of trying to do. So it can hardly be reasonable to ground substantive conclusions about the nature of metaphysics upon them. A second reason is that (1a)–(1c) jointly express a sort of ‘‘either-science-orconceptual-analysis’’ attitude which is not only undefended but also untrue to the phenomena. In contrast to (1a)–(1c), we saw that an adequate conception of the problem of change should: (2a)

Distinguish the tasks of explaining how change is possible from explaining that change is possible. (2b) Permit empirical science to explain particular actual changes. (2c) Permit empirical science to fail to explain other actual changes or non-actual possible changes. (2d) Permit change to interact with identity, time, tense, and persistence without collapsing any into the others. It is a hard question which (if any) conception of metaphysics (or metametaphysics) accommodates these desiderata. But any conception that could needs a subtler and more qualified view of how metaphysics relates to the sciences than (1a)–(1c) allow. To illustrate, such a subtler and more qualified view might approach change jointly from two angles: science illuminates how change is actual or nomologically possible by explaining particular actual changes while metaphysics proper illuminates it by investigating how change is possible tout court (even if nonactual or nomologically impossible). Metaphysics thus at once concerns what is contingent (e.g. particular actual changes) while also concerning what is merely possible tout court (e.g. possible but nomologically impossible changes). So, in a sense, metaphysics elusively outruns both conceptual analysis (by concerning the contingent) and empirical science (by concerning the possible tout court). That is the picture of metaphysics which emerges from reflecting on the problem of change. It is elusive. We do not yet know how to accommodate it. But it is hard to

123

34

M. J. Raven

see how no-problemers could do so even in principle, given their implicit adherence to (1a)–(1c). Nor is it reasonable to reject the need to do so, since that need is implicit in our grasp of the problem of change—a problem which cannot be dismissed as an impostor. These are some of the lessons we learn about the nature of metaphysics by reflecting on the problem of change and, specifically, on why it is not an impostor. But they are not the only lessons we learn. The three challenges posed to the problem of change seem to be instances of three general strategies for undermining any alleged metaphysical problem: (i) show that it is not genuinely metaphysical (often by showing it to be amenable to empirical methods); (ii) show that it is not a genuine problem (often by showing that conceptual analysis trivially dissolves it); and (iii) show that it is not genuinely about what it purports to be about (often by showing that it collapses into other issues, which are often alleged to be less obscure, more tractable, or independently motivated). None of these challenges against the problem of change succeed. But can analogous challenges against other metaphysical problems be blocked in analogous ways? Surely, individual differences between various metaphysical problems make defending a general answer to this question difficult. I won’t attempt such a defense here. Nevertheless, my suspicion is that the ways in which these challenges failed to undermine the problem of change will generalize, mutatis mutandis, to analogous challenges against other metaphysical problems. At the very least, those wishing to defend a metaphysical problem against these challenges now have a recipe to follow for devising cogent replies. There is a problem of change; it is not an impostor. Seeing just why it is not an impostor illuminates the nature of metaphysics, and thereby offers much-needed guidance on how to respond to attempts to undermine its central questions. Acknowledgments Thanks to Margaret Cameron, Anna-Sara Malmgren, Colin Marshall, Ryan Robinson, Ted Sider, Jonathan Simon, Peter Unger, Audrey Yap, the University of Victoria students in my fall 2009 seminar ‘‘Time, Tense, Change, and Persistence’’, the audience of the 2010 Western Canadian Undergraduate Philosophy Conference held at the University of Victoria, and an anonymous referee for helpful comments. I am also indebted to Kit Fine and Mark Hinchliff for influencing the general picture of metaphysics sketched in Sect. 5, along with its particular application to the problem of change.

References Descartes, R. (1996). Meditations on first philosophy (J. Cottingham, Trans., J. Cottingham, Ed.). Cambridge: Cambridge University Press. Fine, K. (1975). Vagueness, truth, and logic. Synthese, 30, 265–300. Fine, K. (2005). ‘‘Tense and reality’’. In modality and tense: philosophical papers. Oxford: Clarendon Press. Haslanger, S. (2003). Persistence through time. In M. J. Loux & D. W. Zimmerman (Eds.), The Oxford handbook of metaphysics. Oxford: Oxford University Press. Hinchliff, M. (1996). The Puzzle of change. Philosophical Perspectives, 10, 119–136.

123

There is a problem of change

35

Hofweber, T. (2009). The meta-problem of change. Nous, 43(2), 286–314. Kurtz, R. M. (2006). Introduction to persistence: What’s the problem? In S. Haslanger & R. M. Kurtz (Eds.), Persistence: Contemporary readings. Cambridge, MA: Bradford Books. Lewis, D. (1986). On the plurality of worlds. Oxford: Blackwell Publishers. Mellor, D. H. (1998). Real time II. London: Routledge. Prior, A. N. (2003). Papers on time and tense (P. Hasle, P. Ohrstrom, T. Brauner, &J. Copeland, Eds.). Oxford: Oxford University Press. Rychter, P. (2009). There is no Puzzle about change. Dialectica, 63(1), 7–22. Schiffer, S. (2003). The things we mean. Oxford: Oxford University Press. Sider, T. (2001). Four-dimensionalism. Oxford: Oxford University Press. Unger, P. (2006). All the power in the world. Oxford: Oxford University Press. Wasserman, R. (2006). The problem of change. Philosophy Compass, 1(1), 48–57. Williams, J. R. G. (2008). Ontic vagueness and metaphysical indeterminacy. Philosophy Compass, 3, 763–788. Williamson, T. (1994). Vagueness. London: Routledge.

123