Powers, persistence, and the problem of temporary intrinsics

Abstract

David Lewis uses the problem of temporary intrinsics to motivate a perdurantist account of persistence in which four-dimensional individuals consist of temporal parts. Other philosophers use his argument to conclude that apparently persisting individuals are collections of temporal stages. In this paper, I investigate whether this argument is as effective in an ontology in which properties are causal powers and thus how seriously the problem should be taken. I go back to first principles to examine the ways in which individuals can change within an ontology of powers and then consider whether any of these ways are compatible with Lewis’s problem. I conclude that if powers are intrinsic, they are not temporary; and if they are temporary, they are not fully intrinsic. However, the situation with respect to changes in which powers are manifesting is not so clear cut, and so I explore how different conceptions of manifestation affect whether the problem of temporary intrinsics applies and what the powers theorist may say about them.

1 Introduction

David Lewis’s problem of temporary intrinsics is regarded as a significant result by many contemporary philosophers interested in problems of persistence (1986, 202–4). For instance, the rod is straight at time t₀ and then the rod is bent at time t₁. How can one individual have intrinsic properties which are inconsistent with each other? At the very least, Leibniz’s Law of the Indiscernibility of Identicals is threatened, since at t₀ the rod has a property which the same rod lacks at t₁ (and vice versa). By some, including Lewis himself, this is taken to motivate perdurance theories or stage theories of persistence, where the conflicting properties are instantiated by distinct temporal parts of four-dimensional space time ‘worms’, or by temporal stages which together compose the persisting object (Lewis (1983, 1986); Sider (1996, 2000, 2001); Hawley (2001)). On the other hand, endurantists have countered these proposals with attempts to resolve the conflict and to explain how the intrinsic properties of an individual entity—a continuant, say—may indeed conflict with each other in this way (Lowe (1987, 1988); Haslanger (1989)).

There is an interesting line of enquiry which considers whether powers theories are compatible with these solutions: are powers theorists bound to be perdurantists or endurantists as a consequence of the ontology of properties which they favour? But in this paper, I will take a step back from that debate and examine the extent to which powers theorists should be concerned about the problem of temporary intrinsics at all. In the first two sections, I will sketch out some ways of understanding powers and their manifestations which are relevant to their role in persistence and change. In Sect. 4, I will investigate from first principles what we might say about the persistence and change of powerful individuals had the problem of temporary intrinsics never been raised and, perhaps more importantly, had the ontological paradigm within which philosophers try to understand persistence and change not been focused for so long on understanding the world in terms of categorical properties rather than powerful ones. In Sect. 5, I will consider whether the powers theorist should worry about the problem of temporary intrinsics: first, from the point of view of changing powers; and second, when considering changes in which power is manifesting. I will conclude that the problem of temporary intrinsics is not a genuine problem for the persistence of powerful individuals given the different kinds of change they can undergo: if powers are temporary, they are not intrinsic, and if they are intrinsic, they are not temporary. The incompatibility which is associated with changes in intrinsic categorical properties does not arise in the case of powers. However, it is less clear whether the problem of temporary intrinsics can be avoided when we consider the manifesting of an individual’s powers, since the plausibility of different resolutions depends in part upon what one thinks is involved when powers manifest. Nevertheless, the powers theorist should be able to argue successfully that changes to which power is manifesting do not present problems for the persistence of an individual: either because manifestations can avoid the problem of temporary intrinsics, or because an alternative endurantist solution can be used. Alternatively, one might accept that changes in the manifestation of powers require separate treatment and should be understood in terms of perdurance, resulting in a dualistic account of the persistence of powerful individuals. In Sect. 6, I will consider some potential limitations and objections to my account.

The broader motivation behind this project is a suspicion that, while powers theorists have forcefully rejected some of the philosophical solutions given by the supporters of categorical or qualitative properties, they have often unreflectively accepted the philosophical problems which categorical property theorists raise. It is, I will contend, these very problems which should be brought into question as they may no longer be apt when raised in an alternative ontology.¹ Only once the powers theorist is satisfied that there is a genuine problem of temporary intrinsics in an ontology of powers should she set about resolving it.

2 Preliminaries

By ‘power’ I intend to mean a neo-Aristotelian, irreducibly dispositional, property which has its causal role as a matter of necessity; that is, ceteris paribus, such a property responds to the same causes and produces the same effects in every possible situation in which it is instantiated. (Causal power, when necessary to a property in this way, is irreducible and to be taken as a primitive notion. On the other hand, if there are categorical properties which have causal powers, they do so contingently; in the categorical case the causal power may be open to further analysis in terms of other entities or processes which contingently fix what the property does.) The necessity associated with which outcome or outcomes a specific power can bring about need not extend to governing the production of its effect or manifestation and I will leave open the possibility that a power P in the right causal context–that is, in the company of the powers with which P usually manifests–may either manifest according to nomological or metaphysical² necessity, or else may manifest as a matter of dispositional modality which is stronger than contingency but weaker than nomological necessity (see Mumford and Anjum 2011a, 2011b, 2018). It will be important to my argument that powers produce powers, but nothing will turn on the modal strength with which this occurs.

The entities which I am calling ‘powers’ are also known under a variety of other names: as dispositions, potentialities, capacities, or dispositional properties. I will ignore any subtle ontological differences between these options and treat them as terminological variants of each other. I will treat all properties as being powerful and not reducible to a categorical ones (although the category under discussion will include powerful qualities which, as their name suggests, are qualitative as well as being powerful), and I will not require that statements about powers are open to analysis in terms of counterfactuals. I will not assume a one category ontology and so will allow that powers are not the only category of entity in play. It is not my aim in this paper to defend the coherence of properties being powers, or pan-dispositionalism more generally, and so there are some familiar objections which I will neglect (on the assumption that the powers theorist has a plausible response).

There are some significant metaphysical variations between powers theories but, for the sake of generality and to avoid tangential debates which may not be easily resolved, I will conduct the discussion in as neutral a way as possible with respect to the different accounts of powers on offer.³ First, it should not (I think) matter to the role which powers play in change and persistence that one might characterise powers as universals—either Aristotelian or Platonic—or as tropes, although in the case of powers being Platonic transcendent universals, we are most probably interested in the role of their instances in particulars, rather than the abstract powers themselves.

Second, I will avoid the question of whether the manifestations of powers are themselves powers (Mumford & Anjum, 2011a, 2011b, 5), or whether they are events (Carnap, 1928; Choi & Fara, 2021; Ryle, 1949), processes (Ellis, 2001; Handfield, 2010), or states of affairs (Williams, 2020, following Armstrong’s conception of a state of affairs (1997)), although I will consider some of the implications of these different views in Sect. 6.⁴ In the interim, as long as in all these cases more powers are produced in some way or another as a result of the manifestation of powers, we can ignore the details of the exact mechanism by which this occurs for the sake of simplicity. Furthermore, even without performing a full evaluation of the cases, we can be (almost) assured that powers do produce powers because, on the pan-dispositionalist view of properties, powers are essential to causation. It would be very odd universe indeed if the manifestation of powers did not sometimes generate more powers since, as we will see, their manifestation can lead to powers ceasing to exist; were powers not produced by powers, the overall stock of powers in the universe would be decreasing and eventually causation might cease. Of course, odd universes are not logically impossible ones, but the position that powers produce powers is also borne out by the nature of the different categories which are candidates to be the manifestations of powers, since these all involve powers as constituents. For instance, one can think of states of affairs as being structured, complex entities such as an object having a power at a time (Williams, 2020); while processes can be thought of as protracted, structured events (Handfield, 2010; Williams, 2020, 75) or as essentially changing, dynamic entities which also instantiate powers (Lowe, 2006, 170; Dupré, 2012; Meincke, 2020). Events are either similar to states of affairs (Kim, 1973, 1976; Lombard, 1986), or else to processes.⁵

In light of these considerations, I will maintain generality and charitably assume that each of these metaphysical variations has the resources to be prima facie as good a candidate for the ontological basis of powers and their manifestations as the others. The examination of independent arguments for one metaphysical variation rather than another can be postponed for another occasion.

Nothing I have to say here is intended to establish that powers exist or that a specific version of powers theory is true. Nor am I arguing that a powers theory is the best way, or the only way, to account for persistence and change (or any other phenomenon for that matter). These arguments—if I ever find good ways in which to make them—can wait for another time, as I am interested here in the consequences of the truth of a powers theory, specifically what kind of account of persistence it can offer and whether this needs to take the problem of temporary intrinsics into account. A corollary of this position is that nothing I have to say here is intended to establish the primacy of powers theories as opposed to other accounts of the fundamental ontology of the world such as those involving processes, non-powerful trope theories, or objects having categorical properties. Although a successful account of persistence and change in terms of powers will count towards the explanatory utility of powers theory as an account of the ontology of nature (or perhaps even of everything there is) which will thereby strengthen the justification for holding it, this would not without further argument rule out the explanatory utility of other theories and leaves the possibility open that some version of pluralism is true.

3 Powerful individuals

Let us take a look at individuals and the powers which they have, before considering the ways in which such a powerful individual might persist or change. Consider a concrete individual b at a time t₀. It has powers. It may even, depending upon one’s ontological inclinations, consist of those powers and be nothing more than a bundle of powerful tropes or instances of universals. Alternatively, instead of thinking about b as an individual object or event, one might prefer to think of it an individual process which instantiates powers. For the purposes of this discussion, the type of entity (if any) which instantiates powers, or which powerful tropes combine to produce, is not important. So I will stick to talk of individuals for now as neutrally as I can.

It seems likely that some of the powers which b has will be manifesting—or else b would have no actual features were the ontology to be one of pure powers rather than powerful qualities—and some will remain potential.⁶ The results of these actually manifesting powers may be more powers, or one might think that the manifestations are entities of a different sort such as states of affairs or processes. If b is an ordinary physical object like a cow, a melon, or a grain of salt, it will have mass, colour, shape, extension, perhaps also smell, genetic or chemical structure, the powers to reproduce, to ruminate, to lactate, to produce methane from plants, to decay, and even, in the case of the cow, a bovine analogue of some psychological and emotional dispositions too. If the individual in question is a microphysical one such as an electron, a photon, or a quark, the properties it has–here thought of as powers, either manifesting or not–might include charge, mass and spin. (Some philosophers maintain that the only legitimate powers are those of the latter kind, the sparse, fundamental ones instantiated by microphysical entities (for example, Heil (2003, 2012); Williams (2019)). Those who think this can simply restrict the discussion to those powers which they do consider to be real and disregard the rest, reformulating my examples as required. Nothing which I say here is going to require the objective existence of the powers of ordinary middle-sized objects, or indeed any powers instantiated by ‘individuals’ which are not fundamental.)

3.1 The manifestation of powers

For simplicity (and because I think that it is the most plausible account available), I will presume that for the most part powers manifest when they are instantiated in proximity to other relevant powers, and that these powers interact or mutually manifest. The presence of other, specific powers is necessary for a power P to manifest but it may not be sufficient for it because other powers present might inhibit or mask P’s manifesting or causal contribution (and will perhaps similarly affect other powers present). The ambiguity of the terms ‘manifesting’ and ‘manifestation’ may cause confusion in this discussion: one can both talk about a power’s manifesting—its acting to bring about its effect whether or not that effect is produced—and about the manifestation as being that effect (which may also involve the manifesting of other powers). Furthermore, Molnar (2003, 194–98) and Mumford (2009, especially 103–4) prefer to distinguish between the manifestation of a power (when this is considered to be its individual contribution to the effect) and the effect itself (which may be the result of a complex interaction with other powers). I will not adopt their usage of ‘manifestation’ although I am sympathetic to the underlying concerns which motivate it. Whatever terminology is adopted, it is important that an account of powers should: (1) distinguish the contributions of different powers to an overall effect from each other and from that effect; (2) capture the stability of contribution that one power makes towards different effects brought about in the presence of different sets of powers; and (3) allow that powers may be manifesting when there is no effect (because their respective causal contributions cancel out).

There are two ways of thinking about the nature of a power and its manifesting, according to whether the presence of other powers with which it mutually manifests brings about an alteration in the power itself. We could call these ways the two-phase and one-phase conceptions of powers. On the two-phase conception, one might think that a power P is dormant until the presence of other properties triggers it to become active (in a reciprocal manner given the mutual manifestation, so that the presence of P prompts other powers to manifest too). Alternatively, on the one-phase conception, one might regard powers as always being active; the presence of other powers permits a power P to interact such that an effect is produced, but the onset of this mutual manifestation brings about no intrinsic alteration to the state of P itself (although the power may, in some cases, cease to exist having manifested). There is no epistemically detectable difference between the one-phase and two-phase conception of powers, since we can only detect powers which are manifesting, and so the choice between them will have to be based on a priori considerations. By way of an analogy with the two conceptions, one could imagine a locked windowless room with a bright main light. In one scenario, analogous to the one-phase conception, the light perpetually remains on although no-one can see it (and it can have no effects outside the room) unless the door is opened; in the other, analogous to the two-phase conception, the light is off when the door is closed but the opening of the door triggers the light to come on. One cannot tell which scenario is true without interfering with the system.

On the one-phase conception, a power is always making its customary causal contribution, even in the absence of other powers with which it might bring about an effect. Such powers are ‘ready to go’ but they do not produce a manifestation if the circumstances—that is, the other powers in their environment—are not right. Powers may appear to have different internal phases on this view but in fact they are always active, even when that activity does does not have any external effect on the rest of the world. Apparent dormancy in a power on the one-phase view is either the result of a lack of mutual manifestation partners, or else the power is interacting with other powers but is doing so in equilibrium and so, from certain perspectives, no change can be detected. One could conceive of powers on the one-phase conception as being rather like processes in the sense that whatever it is that these different powers have the respective powers to do, they (intrinsically speaking) are always doing it, whether or not there are manifestation partners present. (It should be noted however that considering one-phase powers as being, or being similar to, processes requires a different conception of processes to that which allows processes to instantiate powers. I will not examine the relationship between these ways of understanding powers and processes here.)

It could be seen as an advantage of the one-phase conception of powers that there is no internal difference in ‘activity’ between a power which is not manifesting and one which is except for the presence of other powers, and so complicated questions about triggers (and the regresses associated with them), or how a power moves between distinct phases of dormancy and activity, do not arise. This may also allay the worry that the move from a power’s being dormant to being active is itself a change, in the power and in the individual, and might be considered an intrinsic change, thereby presenting a new difficulty for the attempt to explain away the problem of temporary intrinsics, or more broadly the problem of change, in an ontology of powers.⁷

However, if the move from dormancy to activity is brought about by mutual manifestation with other powers, it is not obvious that one need regard this change in the internal phase of a two-phase power as being intrinsic change, and thus we might dispel the worry that phase-change presents an additional problem of change. Even if this approach is ultimately unsuccessful as a comprehensive response to this problem, because mutual manifestation turns out not to be the only way in which a power becomes active, this may not count against the two-phase conception of powers, since the two-phase conception may better explain a range of ways in which powers manifest.

For instance, one might think that there are powers which are able to manifest spontaneously without interacting with other powers. If such powers exist, they both provide a counterexample to the account of mutual manifestation just given and threaten the coherence of the one-phase conception of powers. On the first count, one could accommodate spontaneously manifesting powers alongside those which mutually manifest and accept that powers produce manifestations in two different ways: either spontaneously or by interacting with other powers. There seems little lost except parsimony if the univocal account of manifestation is abandoned. Nevertheless, the supporter of the one-phase conception might still worry about the consistency of spontaneously manifesting powers with his account, since spontaneously manifesting powers seem to have two states—dormant and manifesting—and so the two-phase conception seems to be required. To deal with this problem, one could either adopt the two-phase conception for all powers, accept that such powers are of a different type to one-phase powers, or argue that their dormancy and manifestation should be treated as one continuous state or process, perhaps in the manner of a periodic cycle. I will not dwell on the details, but if required, I will consider spontaneously manifesting powers separately.⁸

There is additional support for the two-phase conception if one believes that the modal force with which powers bring about their effects is weaker than necessity (such as Mumford and Anjum’s dispositional modality (2011; 2018)). On this account, the presence of the requisite mutual manifestation partners is not sufficient for the effect to occur; a power cannot simply already be active, ready to go in the presence of its mutual manifestation partners and so a move between phases of some kind does seem to be required.

On the basis of this discussion, the question of which conception of powers is preferable remains undecided, although the advantages of the two-phase conception only really seem to come into play when one considers spontaneously manifesting powers and those which act with dispositional modality, and the existence of both these classes of powers is open to question. So, one might be able to formulate a coherent theory with one-phase powers after all. But, although adopting the one-phase conception is prima facie a neater solution, the supporter of the two-phase conception might still be able to explain away the problems with their view. The primary difficulty for the two-phase conception of powers in the context of alleviating the problem of temporary intrinsics is that the move from dormancy to activity could itself be regarded as a change.⁹ As we saw above, the supporters of the two-phase conception could argue that the change of phase is not a change which is wholly intrinsic to the power if the active phase is brought about by mutual manifestation with other powers. Alternatively, the supporters of this conception could argue that because it is part of the intrinsic nature of the power to move from being dormant to being active and perhaps back to being dormant, these descriptions do not mark out real change at all. This response will be especially plausible if one considers change to involve a change in properties over time, but even without that characterisation one might argue that an individual power’s being dormant and active are different aspects of a unified entity and cannot be entirely abstracted from each other as separate stages.

3.2 Higher order powers

At t₀, b has multiple powers which determine what b could do: some are manifesting, and others are not but they have the potential to do so and would thereby produce more powers. Does this exhaust the powers which individual b has? If b has powers which are not manifesting and which have the power to produce more powers, at least some of which would be instantiated by b, one might think that the powers which b could produce should count in some sense as powers which b has at t₀ as well. Then, in turn, these produced powers could produce powers of their own, which then go on to produce further powers, and so on. In short, one might think that b has higher order or iterated powers in addition to (and in virtue of) the ones which are manifesting or have an immediate potential to do so (Borghini & Williams, 2008, 30–1; Vetter, 2015, 135–9). For example, I have the power to speak Hungarian fluently (although I currently speak no more than three words of Hungarian) on the basis of powers which I have and other powers in the environment, and thus I have the power to become a Hungarian translator, or even an acclaimed writer of Hungarian novels. Having the power to become an acclaimed writer of Hungarian novels is a higher order power that I have (at least partially) in virtue of prosaic first order ones such as having the power to use language and the power to write.

Higher order powers are useful, especially when we are trying to understand the possibilities associated with an individual, to find truthmakers for modal claims in terms of powers, and (I will claim in the next section) to determine the causal behaviour of an individual. But the set of higher order powers associated with this latter task for a specific individual is not coextensive with the set associated with the former two and should not be confused with it. If higher order powers are regarded as being the truthmakers of modal claims, it quickly becomes clear that the possibilities associated with a specific individual cannot be accounted for solely by the powers or higher-order powers which are instantiated by that individual. To continue our theme from the example above: although the claim that I could be a Hungarian speaker is made true by my having the power to speak Hungarian, which is a higher order power partially determined by powers which I have, the counterfactual claim that I could have been born Hungarian is not. This latter claim requires a change in the history of my ancestors (and perhaps much more besides), and so I do not have a higher order power to make this true even though the possibility is true of me. The fact that the higher order powers which are relevant to the truth of modal claims need not be determined by powers which are local to the individuals mentioned in those claims is all the more obvious the more far-fetched the possibility we consider. For instance, Daisy the cow could have been an acclaimed novelist but there is probably not much about Daisy herself which makes this true; rather, we have to look far back into the evolutionary history of cows and much more besides to find the powers which could have produced the powers (via as many iterations as required) to make this true. While the dispositional actualist about modality has no difficulty in accounting for the existence of powers, and thereby higher order powers, to make such counterfactuals true, one should be wary of attributing the relevant truthmaking powers to the individuals named in them.

Given that we are interested in the future of b and what it may do, the higher order powers determined by the powers of far-flung individuals distinct from b which act as truthmakers for modal claims about b are not those at issue.¹⁰ Rather, we can restrict our focus to the first order powers instantiated by b at a time and the higher order powers which are wholly or partially grounded by b’s first order powers. Thus, there is no ontological cost in including higher order powers alongside first order ones, since the former are derivative entities and are not additional constitutive parts of the individual. Only first order powers have immediate causal potential to change an individual or else to maintain its stability, since these could immediately manifest given the right conditions without the production of further intermediate powers. However, the higher order powers which those first order powers determine have a bearing on the future states and behaviour of that individual.

4 The progression of powers

Let us turn now to a discussion in general terms of what might happen to our individual b, assuming here that b is not a special type of individual which consists solely of an individual power. What happens to b when (as we could loosely and colloquially say) time passes? Let us–for the sake of the example–talk about b in common-sense terms as a persisting particular and withhold judgment about whether it would be better to treat the temporally extended b as being divided into temporal parts or as consisting in a collection of temporal stages. (The philosophical risks of such naive, loose talk are not lost on me: I not only risk the ire of both endurantists and perdurantists, but may also provoke followers of McTaggart, presentists and eternalists too. However, I hope that it is still possible to consider how a powerful individual could behave without getting caught between the competing commitments of these different theories. We can untangle the relationship between these theories and persisting particulars on another occasion.)

Given individual b and the powers it instantiates at t₀, there initially seem to be two options about how b can evolve: either b changes or b does not change over time. On considering the matter more carefully, however, there are some subtle ways in which the powers b has may contribute to b’s changing or not changing, and changing is not coextensive with the presence of causation in the sense of there being powers which are interacting or manifesting, leaving us with the following options:

1) There is no causation and no change, either (i) (on the two-phase view) because all b’s powers are dormant or (ii) (on the one-phase view) because none of b’s powers are located in the vicinity of any powers with which they will mutually manifest and they are not spontaneously manifesting powers.¹¹

2) There is causation, but no change. At least some of b’s powers are manifesting in virtue of the presence of powers with which they mutually manifest, but these interactions do not produce any change (except within a closed system or loop); the interaction between the powers is in equilibrium and produces no new powers nor other entities outside the system. Such an equilibrium situation could occur: (i) internally to b, that is, solely in virtue of interactions between b’s powers; or (ii) between some of b’s powers and powers instantiated by entities external to b. For example, the contents of a bottle of a fizzy (carbonated) drink which is still sealed are in internal equilibrium. At a specific moment in time, the bottle has CO₂ in the space between the drink and the bottle lid and CO₂ dissolved in the liquid. As time passes, the gaseous CO₂ dissolves and the CO₂ in the liquid is released and becomes gaseous, but while the bottle is sealed there is no sign of this equilibrium reaction. For an example of (ii): when the bottle is placed on a table, the powers associated with its mass move it downwards due to the Earth’s gravitational field; however the table also pushes upwards and the bottle stays where it is.

One might argue that the difference between equilibrium and non-equilibrium cases (that is, between causation which involves change and that which does not) is always a matter of scale or the boundaries of whichever individual or system we are interested in, such that there is always change involved in causation from some perspective. For instance, a small portion of the liquid within the sealed bottle of fizzy drink (a 1 cm high disc, say, 4 cm from the bottom of the bottle) would not be in equilibrium: on that scale, the liquid is undergoing changes. Williams (2019, 151–2) argues that not only are there such equilibrium cases of causation which lacks change but also that in some cases this is not a matter of perspective. I am sympathetic to his view here but I won’t add my own defence. If it turns out that, given fixed boundaries, causation always involves change, then the behaviour of individual b will be like one of the options in (3) below.

3) There is causation and change. Here we have four options:

• (i). b’s powers manifest spontaneously or in virtue of other powers which b has, thereby changing the constitution of b. Let us call this ‘strictly internal change’.¹²¹³

• (ii). The causation involves the manifestation of b’s powers and powers external to b, and brings about changes in both internal and external powers.

• (iii). The causation involves the manifestation of b’s powers and powers external to b, and brings about changes in b’s powers while the external powers remain in equilibrium.

• (iv). The causation involves the manifestation of b’s powers and powers external to b, and brings about changes in external powers while b’s internal powers remain in equilibrium.

Let us summarise the options so far: Cases (1) and (2) both amount to b’s persisting, albeit for different reasons. In (1), nothing happens because the conditions are not right for any of b’s powers to manifest, whereas in (2) the instances of b’s powers are manifesting although the overall inventory of b’s powers does not change. b persists because it is stable, not because there are no causal processes going on. Case (3iv) is similar to (2) from b’s point of view, since the powers of b are in equilibrium even though those in the environment are not. This only leaves cases (3i), (3ii) and (3iii) in which b changes, either (3i) in virtue of the powers which b alone has, or (3iii) because its internal powers were made to manifest by those in the environment (the latter remaining in equilibrium). The third, mixed case (3ii) involves change on the part of both b and its environment in virtue of a causal interaction between the two. Finally, if there are powers which manifest in different ways on different occasions—spontaneously, or mutually manifesting with other powers either internal or external to the individual—the change involved will be of the specific type for that particular manifestation.

5 Powers and temporary intrinsics

Let us now return to the problem of temporary intrinsics to see what Lewis’s problem looks like in an ontology of powers. When we consider the problem in terms of categorical properties, Lewis’s problem is that, at different times, an apparently persisting individual rod r has different, incompatible, intrinsic properties, being straight S_c and being bent B_c. (Furthermore, since being bent is one way of not being straight, one might think that this incompatibility amounts to inconsistency.)

To put this more formally, Lewis points out that there is an inconsistency in the following claims:

• At time t₀: r₀ is S_c.

• At time t₁: r₁ is B_c.

• Identity: r₀ = r₁.

• Incompatibility: being S_c precludes being B_c.

• Leibniz’s Law: For any individuals r, r’ and any property P, if r = r’ and Pr, then Pr’.

(Identity), (Incompatibility) and Leibniz’s Law of the Indiscernibility of Identicals cannot hold together and there is a choice of which one to reject. Lewis recommends that we reject (Identity) which holds that r₀ is numerically identical to r₁, in favour of the claim that r₀ and r₁ are temporal parts of the four-dimensional individual r*. This amounts to replacing an account of persistence in terms of endurance with one involving perdurance, since endurance requires an individual to be wholly present at each point of its existence and thereby requires (Identity). (There are, as noted at the beginning of this paper, other ways in which one might make sense of a rejection or revision of (Identity) with perdurance or stage theory.) Alternatively, one could reject (Leibniz’s Law) or maintain it in a restricted form (such as maintaining that it holds at a time rather than over time). In all but a minority of cases, (Incompatibility) has remained unchallenged.¹⁴

This is the story for categorical properties but I will argue that when an individual has powers, there is a clear sense in which (Incompatibility) does not hold and so the inconsistency outlined above does not arise. Thus, Lewis’s worries about endurance, the inconsistency of properties through change, and Leibniz’s Law can be forestalled in a plausible way. Initially, however, the change in our powerful individual b looks similar to that of individual r above which instantiates categorical properties:

• At time t₀: b₀ is straight S.

• At time t₁: b₁ is bent B.

• Identity: b₁ = b₀.

where, in a pan-dispositionalist ontology, being straight S and being bent B are both themselves powerful properties.¹⁵ Since manifesting being straight is incompatible with manifesting being bent, and we aim to preserve Leibniz’s Law, it seems that a similar inconsistency between basic principles has arisen as arose in the case of categorical properties.

This section aims to challenge this conclusion; a challenge which will involve two stages. First: I will argue that what matters to the nature of an individual (both at a time and over time) is which powers that individual has, not which of those powers are manifesting. Second, I will argue that once we consider an individual’s powers and do not focus solely on its manifesting powers, there is an important disanalogy between the cases of powers and categorical properties which allows the powers theorist to avoid (Incompatibility).

5.1 Individuals, powers and manifesting powers

The intrinsic nature of a powerful individual does not solely consist in how it actually is, but also includes how it could be: that is, it is determined by all its (first-order) powers and not only those which are manifesting. (Recall, as well as this, that since the individual’s higher order powers are determined by the first order ones, we get these additions to the ontology for free.) By ‘intrinsic nature’ in this context, I simply mean what the individual is like intrinsically and not what its essence is (if it has one) and thereby I take the nature to include features such as shape which are typically presumed to change. We might want to think of there being a core set of powers determining the essence of the individual, or we may want to avoid that ontological commitment and think of the powers which the individual has in their entirety, but I will leave that matter undecided for now. The relative ontological importance of all the powers of an individual, rather than just the subset of those which are manifesting to make it the way that it is, arises in part because the existence and survival of an individual is characterised by modal facts about it: the fact that an individual could or could not be certain ways is more fundamental to its identity at a time and its persistence as an individual than just the way that it is (although the ways it could be include the manifesting powers which determine the way that it is). Such modal facts are determined by the powers which the individual instantiates, not just which powers are manifesting (which is a slightly peripheral and contingent matter, since it often depends in part on the other powers and the environment of the individual). It matters to the nature of an individual sugar cube that it is soluble, brittle, sweet and can be stacked with other cubes, even though these powers may never manifest. A bilingual speaker of Spanish and English is importantly intrinsically different from a monolingual English speaker because of the powers which she has to speak these languages even when both speakers are speaking English (again, the powers matter more to the nature of the individual than their manifesting at one time or another). Furthermore, the fact that powers and not the just manifesting powers matter to the nature of an individual is reinforced by the suggestion in Sect. 2 that there could be an individual in which no powers were manifesting: if manifesting powers were the basis of an individual’s nature, then such potential individuals could not exist.¹⁶

A second reason why the ontological importance of manifestations might be called into question is because there is much more manifesting of powers going on in an individual than that associated with the powers which determine its actual features. In particular, it is likely that not every manifesting power will result in change in the individual (in the sense defined in Sect. 4) because some manifestations may result in causal equilibria such as those described in (2i) where the manifesting of some powers is cancelled out by the manifesting of others, leaving no resultant effect. Even those manifesting powers which do result in internal change–which we could think of as overtly manifesting powers which determine the actual composition of the individual at that time, such as being bent or being straight–are only such that they (and not other manifesting powers) play this role in the actual nature of the individual because of the other powers with which they are instantiated. If we focus in isolation on an individual’s overtly manifesting powers–and alterations in these–as being the most relevant factor in intrinsic change, we are missing the causal background of other powers (manifesting or not) which makes them so.

A third, related reason for not privileging manifesting powers (whether overtly manifesting or not) with greater ontological importance than those which are not manifesting comes from the discussion of what it is to be a manifesting power in Sect. 3.1. Recall that on the one-phase conception of powers, all powers are perpetually active and their manifesting simply comes about by mutual interaction when they are instantiated in proximity to certain other powers. These mutual manifestation partners may be powers which the individual instantiates, or they may be external to the individual (as outlined in Sect. 4) but, although their presence may bring about a manifestation or effect, they do nothing to change the internal nature of the power itself which is always active whether or not such mutual manifestation partners are present. There is no internal difference between powers which are manifesting and those which are not.

It is less obvious how the two-phase conception of powers is consistent with the claim that manifesting powers cannot be clearly distinguished from those which are not manifesting, since two-phase powers can be dormant or active and so there is something prima facie different about manifesting powers. Nevertheless, the context in which the power is instantiated remains important to whether it manifests on this conception, since powers are triggered into activity by mutual manifestation partners (in all but the case of spontaneously manifesting powers) and so which powers are manifesting is due to factors which are external to the individual powers themselves. Furthermore, as discussed in Sect. 3.1, if the two phase conception is to avoid an internal problem of change for powers, it is important that the dormant and active phases of the power be regarded as integral aspects of the same entity. If the account of two phase powers can avoid the criticism that the change of phase involves intrinsic change, then the difference between manifesting and non-manifesting powers is lessened and (except in the case of spontaneously manifesting powers) is a difference which is determined by the other powers of the individual and the environment.

If the preceding points hold, then neither the distinction between overtly manifesting powers and manifesting ones, nor the distinction between powers which are manifesting and those which are not manifesting mark important differences in the powers themselves. Whether a power belongs to one of these subclasses or another can only be understood contextually, relative to the other powers which the individual has. For this reason, I will primarily focus of the individual’s powers, not just its manifesting powers, in the discussion of change in the next section. I will, however, return to consider manifesting powers in Sect. 5.3.

5.2 Powers and the problem of temporary intrinsics

Once we focus on the potential which an individual has, rather than the potential that it is actualising, as being the key factor which determines its nature–that is, we consider the powers of an individual and do not privilege its manifesting powers–some key differences emerge between powerful individuals and those instantiating categorical properties with respect to the way individuals can change and the problem of temporary intrinsics. These differences arise from the modal strength of the relationships between a power and those around it, preceding it and after it, in virtue of which power it is.¹⁷ As we saw in Sect. 4, future states of a powerful individual are either caused internally entirely by the individual’s own powers, or brought about via the manifesting of a combination of its powers and powers instantiated by other entities. (In some senses, this point is so obvious that it could have been made without the lengthy preamble of the previous section, although I suspect that to someone most comfortable with an ontology of categorical properties, this point is not quite so transparent.) In the case of powers, rather than categorical properties, we are entitled to ask about the individual rod: how come rod b became bent? A power does not arrive spontaneously or get instantiated by an individual independently of its causal past: the change of b from being straight to being bent was either brought about by b’s internal powers, by powers external to b, or by a mixture of the two, according the changes described in (3) in the previous section.

First I will distinguish the production of powers by those internal to an individual as in (3i) and consider two cases of apparent intrinsic change: reversible and irreversible. In the former case, if being bent was the result of b’s internal powers as in (3i), then B must have causal precursors instantiated by b: the bent shape was brought about by (to simplify) a power to be bent B_p and this power must have been (at least) a higher order power which b had at the time t₀ when it was still straight. In fact, if we presume that b went directly from being straight to being bent, B_p would have been instantiated by b as a first order power while b was straight (S having been brought about by the power to be straight, S_p). Had b been spherical immediately prior to being straight, b would have instantiated B_p and S_p simultaneously as first order powers: the power to be bent B_p and the power to be straight S_p are not inconsistent with each other even though they cannot manifest simultaneously. In general, B_p and ¬B_p (the power to be B and the power to be not B, respectively) are not inconsistent with each other because of the modality intrinsic to powers and can be instantiated simultaneously by the same individual. (It is also worth noting that there need not be negative powers as such: having ¬B_p is more likely to consist in having a power to do something which cannot be manifested alongside B.)

Does b at t₁ have different intrinsic powers from b at t₀? In the case of a piece of iron which is bent and then straight, there is a good argument for saying that which powers b has does not alter, but this is perhaps due to the fact that the change from being straight to being bent is reversible, and so the conclusions drawn here may not extend to all cases of strictly internal change. There are two reasons one might think this. First, one might argue that the power to be bent B_p is retained by b on its manifestation B when the iron becomes bent; that is, an object still has the power to be bent while it is actually bent. (There is a parallel here between powers and modality: statements about what is possible are not falsified, nor do modal properties concerning possibility cease to be instantiated, by the possibility in question being actualised.) The process of bending is reversible in the case of iron and so the alternative to accepting the retention of powers on manifestation would be a complicated pattern involving the disappearance and reappearance of powers: S_p reappears at t₁ when B_p manifests to give B, and then B_p reappears at t₂ when S_p manifests to give S, and then S_p reappears at t₃ when B_p manifests to give B, and so on. While the loss of powers is in keeping with the way in which an individual can change intrinsically (as will be discussed when irreversible change is considered below), the spontaneous, uncaused instantiation of powers is not, since powers require causal precursors. But if the respective reappearances of S_p and B_p are brought about intrinsically (as required in the current example), then b retains S_p and B_p as powers of some order or other throughout their respective manifestations; manifesting can alter the order of a power but, if the change is a reversible one, it does not alter whether that power is instantiated. Alternatively, and more contentiously, one might suggest that S_p and B_p are themselves caused by a single (determinable, multi-track) power: the power to be a certain shape (say).¹⁸ In neither case do b’s powers actually change as b goes from being straight to being bent and back again. Since b’s powers do not change, then Leibniz’s Law is not violated either: b at t₀ has the same powers as b at t₁. There is no problem of temporary intrinsics for reversible internal change.

However, as mentioned above, I am not convinced that this reasoning about the powers involved in reversible changes generalises to all strictly internal changes. The claim that the powers of the individual do not change as they manifest is quite plausible in the case of reversible change, but what happens in cases where a power’s manifesting brings about irreversible internal change? Can the powers theorist still avoid the problem of temporary intrinsics? Again, let us stick to causation and change which comes about strictly in virtue of the individual’s own powers: individual d is P at t₀ and Q at t₁. If t₀ and t₁ are contiguous, then d possesses the power to Q, Q_p, at t₀ which it then loses at t₁ when Q_p manifests to produce Q.¹⁹ Do we have a case here where the intrinsic powers of d have changed? On the face of it: yes. d at t₁ might have fewer powers than d at t₀; it has, at least, lost Q_p, the power to Q.²⁰ But, on the other hand, all the powers which d has at t₁ are essentially caused by powers d had previously, including Q_p. Furthermore, d at t₀ had all the powers to produce the powers which d has at t₁, although (if t₀ and t₁ are temporally separated) the orders of powers at t₀ and t₁ might be different. To describe the example in a way in which an endurantist could understand: at any moment of d’s existence up to a time t, d’s future powers are wholly present (the future here being relative to t) but the powers d instantiated in the past–prior to t–are not always all still instantiated. However, at any time in d’s history, d’s past manifested powers have caused its present constitution (that is, the powers which it has) at that time and so the lost powers have left traces in the form of powers which are their effects. Eventually, the direct effects may be lost too, but just as we can think of there being higher order powers for a certain outcome, we could also think of there being effects of effects, and effects of the effects of effects, and so on; if we permit ordering, the traces of a power in an individual are not eliminated when its direct effects are lost. d at t₁ might have fewer intrinsic powers than d at t₀, but it does not have different ones (and certainly not inconsistent ones), so there seems no strong reason to think that we should distinguish d into temporal parts.

One could also explain the case of irreversible change in another way by postulating that powerful individuals have essences which consist in a core set of powers: it is these which the individual must retain and to which Leibniz’s Law applies. Thus, the loss of a non-essential power by irreversible change has no consequence for the identity of the individual through time, while the loss of an essential power coincides with the destruction of the individual. Leibniz’s Law does not fail on the account which accepts that powerful individuals have essences.²¹ In terms of the claims with which we began this section, (Incompatibility) does not hold and so (Identity) and (Leibniz’s Law) can be retained.

Let us consider whether the problem of temporary intrinsics applies to the other cases of powerful causation in which the powers of the individual (let us say individual e in this case) interact with the powers of e’s environment to produce change in e together with change in the environment, or change in e and equilibrium in the environment; that is, cases (3ii) and (3iii) from Sect. 4. There are some similarities with the previous example of d’s causal behaviour in (3i) because whichever powers e has, they all have causal precursors; but in this case, those causal precursors are not bound to be powers instantiated by e itself. Because of this, e can gain powers which it did not possess as higher order intrinsic powers at earlier times in its history as well as (in some cases) losing powers as powers manifest. In this scenario, it genuinely seems to be the case that e could have different powers at t₀ and t₁: say e is F at t₀ and e is G at t₁, where the power to G was not one of the intrinsic powers (of any order) which e had at t₀.²²

It seems, with the kinds of change in (3ii) and (3iii), that we might have discovered genuine examples of changing intrinsic properties in the powers case and so Lewis’s problem would remain. However, before drawing this conclusion, we should pause to consider what counts as an intrinsic power and whether G is an intrinsic power of e at all. After all, we are looking for a problem of temporary intrinsics.

The assessment of this situation is complicated by the fact that there are different accounts of the distinction between intrinsic and extrinsic properties. For simplicity, let us borrow one of Lewis’s own accounts of intrinsicality, as formulated with Rae Langton, to give us an indication about whether causation of type (3ii) or (3iii) would permit changes in intrinsic properties by Lewis’s own lights (1998; 2001).

A necessary condition of Langton and Lewis’s characterisation of intrinsicality is that a property must be independent of loneliness and accompaniment such that it can be instantiated by a particular whether or not anything else exists. We can summarize this as follows:

Independence:	Property P is independent of loneliness and accompaniment if and only if
	(i) It is possible for a lonely thing to have P23
	(ii) It is possible for a lonely thing to lack P
	(iii) It is possible for an accompanied (non-lonely) thing to have P
	(iv) It is possible for an accompanied thing to lack P

To make this into a sufficient condition, an independent property must also be neither disjunctive nor the negation of a disjunction. Care must be taken to distinguish genuine disjunctive properties from spurious ones,²⁴ for which Langton and Lewis rely upon Lewis’s primitive notion of there being degrees of property naturalness: an apparently disjunctive property does not count as genuinely disjunctive if its disjuncts are less natural than the disjunction as a whole; or, conversely, a genuinely disjunctive property is one such that its disjuncts are more natural than it is.

It would be easy to claim that powers are never intrinsic properties in Langton and Lewis’s sense: because of the essential causal relationships between powers, one might argue, a power cannot exist in isolation and can only be instantiated in a network of causally connected powers.²⁵ Thus, the criterion of independence is never satisfied. This would be an easy way out of the problem of temporary intrinsics for the powers theorist (which would remove the need for much of this paper) but I think that this view is mistaken and is based on a misunderstanding of what powers are. First, a power has its causal potential or powerful nature whether or not it is instantiated alongside other powers with which it can mutually manifest, this nature is intrinsic to the power even though it might be defined or described in terms of the effects which it has. Given this, it seems possible for a lonely individual to have powers which would interact with powers in the environment were there an environment. Were the entire universe but for the mug of tea on my desk to be annihilated, the mug would still retain the power to break if dropped, and the tea would still retain the power to scald skin, or to spill, or to go mouldy if left for a month or two. These outcomes would not happen because there are no powers in the environment in virtue of which mutual manifestation could occur, but the causal potential which constitutes the respective powers would still exist. Second, even if one rejected the first reason, Langton and Lewis’s criterion would still allow for a powerful individual which underwent strictly internal causation to have intrinsic powers, since such an individual could be either lonely or accompanied. Thus, the easy strategy of denying that any powers are intrinsic is not plausible. The powers which individual e has are intrinsic powers and so a change in these powers would (one would think) present us with a genuine problem of temporary intrinsics for powers.

However, although it is plausible to think of e’s powers at any particular time as being intrinsic to e, this is not true of the causation which is required to change e from having F at t₀ to having G at t₁. By hypothesis, changes in e of kinds (3ii) and (3iii) require the manifestation of powers in the environment as well as those instantiated by e, and so the change is not intrinsic to e (it is, to use the usual terminology, partially extrinsic).²⁶ Thus the move from e having F to e having G cannot be considered as part of the problem of temporary intrinsics since for e’s powers to change in this way, we need to consider their essential causal relationship with powers instantiated by individuals other than e.²⁷ The same is not true in Lewis’s original case which dealt with categorical properties—properties which have their causal roles contingently—because there is no necessity of causal origin with categorical properties: we can just discount the contribution of the environment in changing e’s intrinsic properties as being a contingent fact, rather than it being a necessary feature of the change in properties. The point here is not that G could not have been brought about by any other powers. The instantiation of a specific power might be brought about by many causes.²⁸ However, there is some cause (the power to G) for the instantiation of G which is either extrinsic, partially extrinsic or intrinsic to e. If it is the latter, then it has been dealt with by the earlier discussion, if it is extrinsic or partially so, then one should question whether this evolution in e’s powers is a case to which the problem of temporary intrinsics is applicable.

To summarise the findings of this section, when the powers of an individual change according to strictly internal causation (that is, by mutually manifesting with the individual’s other powers or self-manifesting (if that is plausible)), then the powers of the individual do not change if the change is reversible. There is no change in temporary intrinsics. If the change is irreversible, then no powers are gained although some may be lost. When the change in an individual involves powers external to it, then it can gain and lose powers, but there is no way to frame the gains so that the problem of temporary intrinsics applies. Although it makes sense to consider the powers of an individual as being intrinsic at a time, we cannot continue to ignore their essential relationship to powers external to the individual when it undergoes change of types (3ii) or (3iii).²⁹ There is no good reason to conclude that an individual should be distinguished into temporal parts or stages for any kind of powerful change it might undergo.

5.3 Manifesting powers (again)

Having argued in 5.1 that all an individual’s powers are important for making it what it is, rather than just its manifesting powers, I have now also argued that there is no problem of temporary intrinsics for powerful individuals considered in this way. Nevertheless, even if an individual’s having the power to be one way does not rule out its actually being another, one might still object that there is intrinsic change involved as one power’s manifesting gives way to the manifesting of another. Furthermore, such a change in manifesting powers–such as the change from an individual’s being straight to being bent–seems to revive (Incompatibility) once more and to bring it into conflict with (Identity) and Leibniz’s Law; that is, considering the manifesting powers of an individual seems to have reintroduced the problem of temporary intrinsics in another form. In this section, I will sketch three responses which are increasingly sympathetic to this objection.

There is an intuitive pull to concerns like this that the problem of temporary intrinsics has been relocated rather than resolved.³⁰ One may not find the argument of Sect. 5.1 convincing³¹ and even if one does, we are not accustomed to thinking that the ways an individual actually presents in terms of its overtly manifesting powers are the result of the instantiation and (in some cases) manifestation of other powers which are not overtly manifesting because their effects are cancelled out. In ordinary life, overtly manifesting powers are those we take notice of³² and so it requires something of a change of perspective to treat them as a temporarily ‘visible’ aspects of permanently instantiated powers. (I am here only considering the case of strictly internal change, since there would be no renewed problem if changes in which powers were manifesting in an individual turned out to be extrinsic.) The proposed solution to the problem of temporary intrinsics given in the previous section relies on a distinction between how persisting and changing individuals appear to us and the ontological reality of interacting powers.

The first, hard-nosed response to the objection invokes this distinction between appearance and reality, and on the account of manifesting powers given thus far, and insists that the problem of temporary intrinsics is about which powers an individual has and how these can change, while manifesting or not is something a power does. An individual’s having distinct, seemingly incompatible, manifesting powers at different times is only apparent variation in that individual, in the sense that it is an artefact of our perspective, and is not a genuine intrinsic change. Although what we are inclined to predicate of an individual varies according to which powers are overtly manifesting, giving the illusion that the properties of the individual vary, that need not indicate a genuine ontological variation in the individual.³³ There will be further discussion of the distinction between genuine and apparent change in Sect. 6, so I shall not pre-empt that here, since this response is unlikely to gain acceptance by someone who is already convinced by the power of the objection, although it may help to assuage the doubts of those already sympathetic to the arguments of this paper.

A second, more conciliatory response takes the objection seriously as a problem of change and provides an account of how seemingly incompatible manifesting powers can be instantiated by the same individual at different times. This might either either help to explain away the problem and serve to strengthen the claim outlined in the first response that variation in manifestations is not genuine intrinsic change in an individual, or admit that there is a specific difficulty associated with manifesting powers and try to alleviate that. The latter admission will lead us to the third response which accepts that manifesting powers must be treated differently from the powers of an individual considered more generally, resulting in a bilateral approach to the problem of temporary intrinsics depending upon whether we are concerned with change in the powers of an individual or in its manifesting powers.

Let us return to consider our rod, this time in terms of the manifestations of powers being straight S_m and being bent B_m:

At time t₀: r₀ is S_m.

At time t₁: r₁ is B_m.

Identity: r₀ = r₁.

Incompatibility: being S_m precludes being B_m.

This problem looks very much like the original problem of temporary intrinsics for categorical properties and so could be countered with one of the many suggested ways in which an endurantist might seek to resolve the problem, including relativising the manifested powers, or their manifesting, to times. In the context of our example, the result of such relativisation is that either r₀ is S_m at t₀ and r₁ is B_m at t₁ (Lewis, 1986, 204) or r₀ being S_m manifests at t₀ and r₁ being B_m manifests at t₁ (Lowe, 1988, 75; Haslanger, 1989). Lewis challenges suggestions such as these in the case of categorical properties, primarily on the grounds that the resulting temporally relativised properties are extrinsic (while the shapes of being straight and being bent are not), and to some extent the success of this kind of an approach to the problem of changing manifesting powers stands or falls with the success of counterarguments to Lewis’s complaint, based on the case of categorical properties. (See, for instance, Haslanger, 1989.³⁴) Were these responses to be successful, the problem of temporary intrinsics for manifesting powers would be avoided.

However, one could also point to some differences in the case of overtly manifesting powers compared to that of temporary intrinsic categorical properties. First, one might argue that Leibniz’s Law is not applicable in the case of manifesting powers which are for the most part fleeting and transient ways an individual’s powers are overtly active, and so should be excluded from the catalogue of properties which an enduring individual maintains throughout its lifespan. In a similar way, we do not usually require active verbs and phrases involving them to be covered by Leibniz’s law, even though these can be predicated of an individual. If this point holds, the problem of temporary intrinsics does not apply. Second, in the case of manifesting powers, one could argue against Lewis that whether or not a power is manifesting, and especially whether or not it is overtly manifesting as in our current example, is an extrinsic feature of the power anyway. Recalling the points made in 3.1 and 5.1, it is only in the case of spontaneously manifesting powers that powers manifest independently of other powers, otherwise they only do so in the context of their coinstantiation (and mutual manifestation) with other powers, even though in the case of internal change, this does not make the power which is manifesting extrinsic relative to the individual which instantiates it. If the manifesting of the power is already extrinsic, then either the problem of temporary intrinsics does not arise as a specific problem for manifesting powers (and we were mistaken in thinking that it was), or the Lewis-inspired charge that temporal relativisation makes the manifesting powers extrinsic is not as obviously damaging. (The latter option is added because one might think that the problem of temporary intrinsics remains for manifesting powers despite the manifesting of a power being an extrinsic feature of the power if the powers upon which the manifestation depends are also intrinsic to the individual.)

One might still have concerns about these responses–either with or without the temporal relativisation–and think that changes in manifesting powers still present a prima facie case of incompatibility which makes it hard to see how being straight and being bent can manifest in the same individual. Furthermore, one might suggest that if powers are manifesting powers relative to the other powers instantiated by an individual (as I have argued), then manifesting powers are best thought of as the kinds of entity which are possessed by an individual at a time (or which exists for an extended period of time); that is, crucially, that manifesting powers are powers which are possessed by the temporal parts of the individual and not by an enduring individual simpliciter. On this way of characterising the ontology, we can understand powers as belonging to an enduring individual, but we must understand the manifesting of powers as being associated with a perduring individual. Both the endurantists and the perdurantists turn out to be right about persistence, depending upon which aspect of the ontology we are focussing upon. The persistence of powerful individuals would, on this way of understanding powers and their manifestations, be similar to the mixed cases of perdurance and endurance which Lewis allows for in his original argument for perdurance (1986, 202).

6 Conclusions and concerns

The preceding discussion about powerful particulars and their persistence strongly suggests that the problem of temporary intrinsics does not apply to powerful individuals and so it does not require a solution; consequently, it should not motivate the powers theorist to accept perdurance rather than endurance. The extent to which this resolves the problem of temporary intrinsics entirely is dependent upon what one thinks about manifesting powers and whether the problem of temporary intrinsics also applies to them. Having mapped out some prospective ways of approaching this latter problem but ultimately having left it undecided, I will move the discussion away from manifesting powers in this section and consider the fate of persisting powerful individuals once again: one might still be suspicious of claims to have resolved a long-running philosophical problem, even if only for a restricted category of entities. One important question is how well key endurantist tenets fare given my account of powerful causation and change, a second is the recurring problem of whether my solution of the problem of temporary intrinsics for powers has merely shifted the problem elsewhere.³⁵ I will discuss these in turn.

To frame the former concern, it is useful to draw on a dilemma presented by Johanna Seibt to a class of endurantist attempts to avoid the problem of temporary intrinsics (1997, 155–7)³⁶: in these, she argues, either Leibniz’s Law fails and so persistence could not consist in identity, or there is no genuine account of change. Does a similar criticism apply to the account of powerful individuals which has been given?

First, how does Leibniz’s Law fare on this account? As we have seen, Leibniz’s Law still applies to powers in reversible cases of strictly internal change (3i) since individuals retain their powers over time, and also in the case of irreversible change if we restrict the application of Leibniz’s Law to essential powers.³⁷ In the case of non-essential powers (which will be all an individual’s powers if one rejects essences), Leibniz’s Law works in one temporal direction only because an individual can lose powers as it changes: any power which c has at t₁, c also had at any earlier time t₀, but not vice versa. This seems a plausible compromise for non-essential powers in conjunction with strictly enduring essential powers, but perhaps this is even enough to sustain a form of non-essentialist endurantism: the powers which c has at any time are always identical with the powers it had earlier; at any time, the enduring individual existed in its entirety at earlier times in its history, it is just that there may be less of it at later times than earlier ones. Individuals gradually lose powers until (at some point) they fade away.

As for making a choice between these options, one might argue that conceiving endurance to be a doctrine which requires the identity of all the powers of an individual over time—to ensure that the individual and all its powers is wholly present at all times—is too strong a demand even for the endurantist. The ‘whole’ which we worry about being wholly present is more plausibly a core of essential powers, with the scope of Leibniz’s Law restricted to these (such that when an essential power is lost, the individual ceases to exist).³⁸ Since there is no incompatibility or conflict due to changes or loss of non-essential powers, the problem of temporary intrinsics does not arise. However, despite these advantages, there are familiar problems with essences, and especially with the view that essences are possessed universally by every individual. Although essentialism seems to be the most plausible option to sustain endurantism, I will withhold judgment on it here.³⁹

What about the other horn of Seibt’s dilemma: if, as in the case of reversible, strictly internal change, the powers of an individual are neither gained nor lost, is this an account of change at all? One could accept this charge and insist that persisting powerful individuals do not really change intrinsically—after all, there is still plenty of extrinsic change to be had—but this response seems extreme and rather counter-intuitive. Fortunately, Seibt’s worry can be allayed in another way because there are two perspectives in play on the same entities: while identity over time is a matter of which powers an individual has, there are other ways in which an individual alters over time, including which powers happen to be manifesting (and also what the orders of the individual’s powers are), rather than powers being gained or lost.⁴⁰ Thus some strictly internal change can occur in the latter sense although the individual retains its powers. For example, there is some sort of change when the bilingual English and Spanish speaker speaks Spanish rather than English, but it is not a change in her powers and not one which affects her identity over time. Likewise, a jazz band persists through the performance of a piece without a change in powers (presuming no change in members mid-piece) regardless of which member of the band is taking a turn to do a solo, and yet from another perspective there is a difference between a solo on saxophone or on drums. Differences in which powers are overtly manifesting are interesting and notable, not least because they are our epistemic window onto which powers an individual has, but they do not have a bearing on the persistence of an individual.

One might be concerned about a sleight of hand here, or even two. First, according to the account of strictly internal change in (3i), while the powers of an individual do not change, the orders of those powers do change: why is change in order not a case of the individual’s powers changing too? Initially, it may seem as though there is a choice about how we should identify and individuate powers—either by their causes and effects, the order or both—and that I have chosen the former over the others, a choice which may matter to the success of what I have to say about the individual’s powers not changing in causation like (3i). For instance, if b’s third-order power to Q becomes a second-order power to Q. We might say that b has the same power to Q but that its order has changed, or we might treat the third-order power to Q and the second-order power to Q as distinct powers since, after all, the third- and second-order powers may be determined by b’s first-order powers in different ways. However, the choice to identify powers of different orders for the purposes of understanding change can be justified, since altering the order of a power does not change what the individual has the power to do or to be, just the likelihood that it will do so and the number of causal processes involved. Furthermore, changes in the way higher order powers are determined by first order powers (which result in changes in order) do not change what the individual can do either. If we are interested in powers for what their manifesting would mean for the future of the individual, then distinguishing between powers of different orders with the same outcome is not a good way to individuate powers.

The second, related concern is that although the preceding argument might have alleviated the problem of temporary intrinsics for powers—and so can account for apparent incompatibility between properties of persisting individuals—there are other ways in which an individual can alter and these too could present a problem of change. According to this assessment, the problem of change has been shifted along the philosophical road a bit, but not solved. One example of this is whether my account deals with the changes in which of an individual’s powers is manifesting, a problem which was discussed in the previous section with both optimistic and pessimistic outcomes having been explored. At worst, while powerful individuals are not susceptible to the problem of temporary intrinsics, the manifesting powers may require an independent account of persistence which does acknowledge the problem. Another issue might be the worry–somewhat neglected in this simplified discussion of manifestation–that the manifestation of powers may bring about novel categories of entities associated with those powers (or with the powers which they can cause) and perhaps also the individual concerned, such as states of affairs or processes. This is an interesting question: for instance, in addition to the causal progression of its powers, does an individual b also change because the power B_p’s manifesting produces a novel state of affairs which is partially composed of B,⁴¹ or brings about a process involving B? Prima facie, this presents quite a compelling problem but we should be suspicious of accepting such metaphysical shifts as genuine changes because this brings with it the danger of multiplying changes beyond our current understanding of them and classifying phenomena as changes which do not have ontological significance for the future of the individual involved or of the natural world more broadly. First, not every phenomenon which can be described as a change should be treated as one. There are Cambridge changes, such as Xantippe becoming widowed (which are ontologically dependent on changes in other individuals, in this case on Socrates dying) (Geach, 1969, 71–2), and gruesome changes which we intuitively take to mark no change at all. (For instance, let ‘grue’ mean ‘green if observed before time t and otherwise blue’ and ‘bleen’ mean ‘blue if observed before time t and otherwise green’. If it is a cloudless, sunny day and time t is the time you began reading this sentence, then the bleen sky has just changed to a grue one. See Davidson, 1995, 271.) It seems that the world would be overpopulated with changes if every way in which an individual can satisfy a predicate at one time t₀ and then not satisfy it at another time t₁ is treated as a genuine change. This is one reason why the changes in examples such as these are not generally regarded as being genuine and thus are not thought to be the kind that are at issue when we are concerned with the philosophical problem of change,⁴² while a second reason is that they do not appear to influence the future causal trajectory of the individual which undergoes them (in the case of Xantippe, her widowhood will only affect her once she herself undergoes some change as a result of it). The crucial question here is whether an individual’s undergoing changes in ontological categories or structures—if this does happen on the manifestation of its powers—should be treated as genuine changes. If our criterion for significance is the future causal trajectory of the individual, then one could argue that it is the powers the individual has which matter rather than whether they are partially constituting a process or a state of affairs, or whether the individual is undergoing a process or partially constituting a state of affairs. However, it is impossible to do justice to the metaphysical and meta-ontological subtleties of this question in the space I have here, which include which category changes there are, how the members of the category are identified and individuated (and thus whether these are different from each other and from the kinds of change already covered), what the ontological status of these categories is, whether these are relevant to the future of the individual and if they too can avoid a problem of change. Unfortunately, a full evaluation of these issues is beyond the scope of this paper but they do not affect the conclusions drawn about powers. In view of this, what am licensed to conclude here concerns the problem of temporary intrinsics when these are powers or their manifestations (where, as I have argued, the former but not the latter are significant to the persistence of the individual). I will leave discussion of the significance of changes in other ontological categories for future consideration.

Finally, one might also be concerned about how much change turns out to be extrinsic on the account presented here. Much of the loss and all the gain of powers comes about via interactions with powers in the environment which, although it avoids the problem of temporary intrinsics, might damage our understanding of what individuals are. My view is that the powers theorist could and should embrace this consequence, regarding individuals for the most part as being closely causally embedded in their environment (with their status as separate entities perhaps being regarded as a matter of degree).

I suggest that these concerns about my account should not be thought of as compelling: the powers theorist has a way of understanding change and persistence in terms of powers so that the problem of temporary intrinsics does not apply. Nevertheless, there is still some work to be done to clarify exactly how the problem should be understood when applied to manifesting powers. I will save that exploration for further research.⁴³