p44

4.2. Attempt 2: Temporal Rationalism Another attempt to independently motivate Ontological InertiaMonism begins with the observation that Future Tuesday Reset involves an entirely brute and inexplicable change in reality. Why did the world endure until next Tuesday instead of next Wednesday? And why did the world get replaced by ψnext Tuesday rather than some other possible world-individual? There don’t seem to be any good answers to these questions. It therefore seems that any world that violates Ontological InertiaMonism would involve completely brute and inexplicable changes in reality. In recent times, some philosophers have ventured to revive the Principle of Sufficient Reason (PSR), according to which no fact can be brute and inexplicable in this way.21 If the endurantist endorses the PSR, then they could motivate Ontological InertiaMonism on the grounds that any violation of Ontological InertiaMonism would violate the PSR. In fact, the endurantist need not endorse the PSR in full generality to derive Ontological InertiaMonism. They only need to endorse the following: Temporal Rationalism: Necessarily, if the world changes from one state to another, there is an explanation of why the world changed in that way.22 20 This analysis of the H2O-water case can be made more precise using the tools of two-dimensional semantics (e.g. see Chalmers 2008). 21 For defenses of different versions of the PSR, see Pruss (2006), Della Rocca (2010), Dasgupta (2016a), and Pruss and Koons (2021). 22 Note that Temporal Rationalism need not imply determinism, since one could explain why the world changed in the way it did by appealing to objective chances (see Pruss 2006: ch. 8). 16 Temporal Rationalism is a weakening of the PSR that allows for certain aspects of the world to be brute, such as the initial conditions of the universe, the laws of nature, or the conjunction of all contingent truths. Because of this, Temporal Rationalism avoids perhaps the biggest objection to the PSR, namely that the PSR implies that every truth is a necessary truth.23 Moreover, there are also independent motivations for Temporal Rationalism. For example, Tugby (2017) has formulated a principle analogous to Temporal Rationalism, which he calls the “Principle of Change,” which has the advantage of resolving an important objection to dispositional essentialist views regarding fundamental physical properties. Ultimately, whether one accepts Temporal Rationalism as an adequate motivation for Ontological InertiaMonism will depend on one’s prior metaphysical commitments. On the one hand, one might be attracted to a certain kind of link between conceivability and possibility, which seems to tell against Temporal Rationalism. After all, it seems conceivable for a world to change in inexplicable ways. Somewhat ironically, perhaps the best motivation for endorsing a link between conceivability and possibility is due to a thesis which Chalmers (2002) calls Modal Rationalism. Chalmers (1996: 136-38) motivates Modal Rationalism on the grounds that violations of a suitably idealized connection between conceivability and possibility would involve brute and inexplicable restrictions on the space of possible worlds. On the other hand, the best motivation for Temporal Rationalism also involves a rationalist rejection of the possibility of brute and inexplicable changes in the universe. We won’t take a stand between these two rationalist principles here, but we do think that Temporal Rationalism at least provides a better way forward for the endurantist than Essential Ontological InertiaMonism in their task of motivating Ontological InertiaMonism. That being said, we suspect that most endurantists would be loath to embrace a strong rationalist principle like Temporal Rationalism. 4.3. Attempt 3: Hume’s Dictum for Non-Humeans In our view, a more promising way forward for the endurantist involves endorsing a certain sort of non-Humean view that recognizes necessary connections in nature; in particular, that 23 See Bennett (1984) and Van Inwagen (2015) for defenses of this argument. 17 the present state of the world necessitates future states in accordance with the laws of nature.24 On the sort of views we have in mind, if the laws dictate that state A evolves into state B then the fact that A evolves into B will turn out to be metaphysically (rather than merely nomically) necessary. For example, any “Aristotelian” view of natural necessity—according to which fundamental physical properties have intrinsic causal powers or dispositions to behave in certain ways—will be an instance of such a view (see footnote 32 below for some proponents). However, proponents of other metaphysical accounts of laws of nature have also argued that the laws of nature describe metaphysically necessary truths (for example, Wilson (2013)). A common objection to such views is that they seem to violate Hume’s Dictum: Hume’s Dictum: Necessarily, there are no necessary connections between wholly distinct entities.25 Two entities are wholly distinct just in case they do not mereologically or spatiotemporally overlap. The caveat of being “wholly distinct” is important because even defenders of Hume’s Dictum will acknowledge that there can be necessary connections between an entity and itself, an entity and its parts, and so on. For example, the fact that x is 2 kg might necessitate that x is not 4 kg, the fact that x has a color might necessitate that x has a shape, the fact that x is 2kg and mereologically complex might necessitate that all of x’s proper parts have a mass of at most 2kg, and so on. However, necessary connections between wholly distinct entities have seemed much more problematic. How could one entity being a certain way make it impossible for another wholly distinct entity to be some other way? Again, one could motivate Hume’s Dictum by means of a conceivability argument. No matter what way some entity is, it seems 24 Even if the laws of nature are chancy, the present state of the world still necessitates some facts about the future. So long as there are some non-trivial features of the future that get assigned a maximal chance of 1 (e.g. black holes not spontaneously forming in the absence of matter), then there will be necessary connections between the present state of the universe and future states of the universe. 25 See Wilson (2010) for a critical assessment of the different kinds of motivations that have been offered for Hume’s Dictum, and various nuances in formulating the principle. See Segal (2014) for a recent defense of the principle. 18 conceivable for a second wholly distinct entity to have any collection of compossible intrinsic properties. Our suggestion is that the non-Humean endurantist can use Hume’s Dictum to their advantage. In particular, they can use Hume’s Dictum to motivate the following thesis: Non-Humean Ontological InertiaMonism: Necessarily, if three-dimensional Monism is true, then if non-Humeanism is true, then for all times t and t’, ψt = ψt’ Note that for the non-Humean, Non-Humean Ontological InertiaMonism is sufficient to bridge the gap between qualitative laws and full laws, because the non-Humean will think that the only nomically possible worlds are ones where the present state of the world necessitates future states in accordance with the laws of nature. The crucial point is that the monist non-Humean endurantist is able to satisfy Hume’s Dictum if and only if they endorse Non-Humean Ontological InertiaMonism. To argue for this claim, consider again the case of Future Tuesday Reset, where the world is replaced by a distinct world-individual next Tuesday. If this is to be a nomically possible world for the non-Humean, it must be that the world before next Tuesday bears a necessary connection (in accordance with the laws of nature) to the world after next Tuesday. Since the world before next Tuesday is wholly distinct from the world after next Tuesday, this is an explicit violation of Hume’s Dictum. More generally, if there are any two times t and t’ such that ψt is wholly distinct from ψt’, then if the laws of nature generate necessary connections these would be connections between wholly distinct entities, violating Hume’s Dictum. However, if for any two times t and t’, ψt = ψt’, then there would be no violations of Hume’s Dictum. The world would never bear any necessary connections to anything besides itself.26 26 Two qualifications about the biconditional in this paragraph. First, to fully argue for the left-to-right direction, we’d need to argue against some more exotic views, according to which ψt is distinct yet mereologically coincident with ψt’, rendering them not wholly distinct, and hence blocking the relevant violation of Hume’s Dictum. However, we don’t think there’s much to recommend this option, and so we ignore it in the main text. For a sympathetic exploration of something in the vicinity of this option, see Segal (2014). Second, to fully argue for the right-to-left direction, we must set aside views according to which the laws themselves count as an entity that 19 The intuitive idea behind this appeal to Hume’s Dictum is that the causal connection between the present and the future would be broken if the world simply popped out of existence and was replaced by an entirely new and wholly distinct world that popped into existence. How could the new world possibly “know” about what went on with some wholly distinct entity in the past? How could the present world manage to make necessary demands on some future time in which it doesn’t even exist? Although she does not appeal to Hume’s Dictum, Haslanger (1989) defends a similar conclusion. She summarizes her discussion as follows: [T]he past can be causally efficacious in the present only through things presently existing. Therefore, if nothing from the past [endures] to the present, the past can set no constraints on the present; the “causal message” cannot be communicated across the gap […] From this we can see that [endurantism] does provide us intelligibility in explanations of change. Natural explanations work by showing the systematic causal interconnections between things. Without [endurance], the causal story becomes unconnected; neither the past nor the future can get a hold on the present in a way that is causally efficacious. (21) A similar point can be read into Thomson’s (1983) famous complaint about perdurantism: I said this seems to me a crazy metaphysic. It seems to me that its full craziness only comes out when we take the spatial analogy seriously. The metaphysic yields that if I have had exactly one bit of chalk in my hand for the last hour, then there is something might witness a violation of Hume’s Dictum. However, just about every candidate view in the metaphysics of laws doesn’t construe laws as an entity, and hence would vindicate our reasoning. The only exception here would be a version of primitivism about laws, according to which laws are sui generis entities (suggested by some remarks in Maudlin (2007)). However, more orthodox understandings of primitivism take some nomic ideology as fundamental, such as a primitive ‘it’s a law that’ operator (analogously to how modalists view metaphysical necessity). These more orthodox versions of primitivism don’t construe laws as entities of some kind, and so pose no threat to our reasoning in this paragraph. 20 in my hand which is white, roughly cylindrical in shape, and dusty, something which also has a weight, something which is chalk, which was not in my hand three minutes ago, and indeed, such that no part of it was in my hand three minutes ago. As I hold the bit of chalk in my hand, new stuff, new chalk keeps constantly coming into existence ex nihilo. That strikes me as obviously false. (211) One way to account for the remark that new chalk keeps constantly coming into existence “ex nihilo” is by reference to a background intuition that there can’t be intelligible causal connections between the past pieces of chalk and the wholly distinct future pieces of chalk. The natural way to avoid the absurdity of new chalk continually popping in and out of existence is instead to countenance a single piece of chalk that endures across time. We see, then, that endurantists sympathetic to non-Humean necessary connections in nature might have entirely independent reasons to endorse Non-Humean Ontological InertiaMonism: by doing so, they can satisfy Hume’s Dictum, which many take to be independently plausible. It just also turns out that by endorsing Non-Humean Ontological InertiaMonism, the nonHumean endurantist also thereby bridges the gap between qualitative lawfulness and full lawfulness. Both of us take there to be strong independent arguments for non-Humeanism, and are sympathetic to Hume’s Dictum, and hence regard this route as the most promising one for endurantists to pursue in their task of motivating Ontological InertiaMonism. 4.4. Attempt 4: Distinctively Nomic Ontological Inertia So far we have been pursuing the strategy of supplementing the laws of nature with a principle that holds with metaphysical necessity, along the lines of Ontological InertiaMonism, in order to rule out cases like Future Tuesday Reset. However, full lawfulness can also be secured by a weaker view which renders cases like Future Tuesday Reset metaphysically possible yet nomically impossible. On this view, some feature of our laws of nature ensures that the world always endures through time. Notice that this strategy need not require abandoning the common view that laws of nature are qualitative. So far we have been working with the assumption that the laws of nature take the following qualitative form: if the fundamental things are qualitatively so-and-so at some 21 time, then there are fundamental things that are qualitatively so-and-so at such-and-such other times. These laws simply take no stand on whether the fundamental objects at one time continue to exist at other times, thereby leaving them vulnerable to threats to full lawfulness like Future Tuesday Reset. Yet there are other forms for qualitative laws to take that aren’t neutral on cross-time identity facts; we’ll call any such laws strong laws. For example, these laws might state that if some particle is qualitatively so-and-so at some time, then that very particle is qualitatively so-and-so at such-and-such other times. Or a monistic analogue: if the threedimensional world-object is qualitatively so-and-so at some time, then that very world-object is qualitatively so-and-so at such-and-such other times. Notice that laws of this form are still qualitative, yet they build in endurance, thereby ruling out cases like Future Tuesday Reset above: if objects pop out of existence then laws of this form are simply false. Even though strong laws can be qualitative, we want to flag that they would still be unfamiliar in various respects. For example, such laws would no longer be neutral on the metaphysics of persistence: if it were built into the laws of nature that any fundamental objects there are at a time endure for the rest of time, then the truth of endurantism (and the falsity of perdurantism) would be an immediate consequence of the laws of nature themselves. In addition, strong laws would be unlike ordinary scientific laws in that they govern features of the physical world that are in principle empirically undetectable. For example, no scientific experiment could possibly detect the difference between worlds like Future Tuesday Reset and No Reset. Nevertheless, we don’t find these unfamiliar features of strong laws to be at all worrying: in our view, it is unsurprising that different metaphysical views about persistence across time might lead to different formulations of the laws of nature, which dictate how the universe evolves as time goes on. We leave it to those who find these unfamiliar features problematic to make good on this worry. If they are right then that would only further support our pessimistic evaluation of the strong laws strategy. In our view, the reasons to be skeptical of strong laws emerge by examining different views about the metaphysics of laws of nature. Let us consider the main theories of laws in turn. 22 To start, consider the Dretske-Tooley-Armstrong (DTA) theory, according to which laws describe a second-order “necessitation” relation between properties.27 Crucially, the necessitation relation here is meant to hold between sparse universals or fundamental properties, such as mass or charge. A familiar complaint about this theory is that it limits the form that laws of nature can take. 28 As concerns the present strong laws strategy, let us grant that the theory can deliver laws to the effect that if some fundamental property is instantiated at a time then some other fundamental property is instantiated at such-and-such other times, even though there’s room here for debate.29 The trouble is that the strong law strategy also requires laws to pin down which particular things have the fundamental properties at the other times, via cross-time identity facts. And we don’t see a plausible means of including these additional facts within the strictures of the DTA theory. Doing so seems to require recognizing infinitely many non-qualitative haecceitistic properties as sparse universals (e.g. being ψt), which can then themselves stand in the second-order necessitation relation. However, such a view not only requires a very implausible account of sparse universals, but it also requires countenancing infinitely many non-qualitative laws of nature. We doubt anyone would want to embrace the resulting DTA-friendly vision. Let us now turn to Humean theories of lawhood, in particular the popular Ramsey-Lewis “best systems analysis” (BSA), according to which laws are just certain especially simple and informative summaries of the ways things are across all of space and time.30 Here there’s no obstacle to the endurantist’s strong laws being candidate laws, for the BSA imposes no inprinciple constraints on what form laws of nature can take. Our central worry is that, given this metaphysic of lawhood, the strong law strategy doesn’t in fact independently motivate the nomic analogue of Ontological InertiaMonism. And our reasons for thinking this are just the familiar complaints about the explanatory circularity of Humean theories.31 Once 27 See Dretske (1977), Tooley (1977), and Armstrong (1983) for classic developments of this view. 28 See, for example, Wilson (1987) and Maudlin (2007). 29 The issue turns on whether the DTA theorist can accommodate temporal-derivatives. 30 For an introductory survey of Humeanism about laws of nature, together with how the view might respond to various objections, see Bhogal (2020a). 31 For versions of this worry see Lange (2013, 2018), Shumener (2019), and Emery (forthcoming).