Endurantism, Fixity, and Fatalism

Science, Religion & Culture Article Special Issue: Author Meets Critics: John Martin Fischer’s “Our Fate: Essays on God and Free Will” Endurantism, Fixity, and Fatalism Michael Almeida Department of Philosophy and Classics, University of Texas, San Antonio Editor | Gregg D. Caruso, Corning Community College, SUNY (USA). Correspondence | Michael Almeida, University of Texas, San Antonio; Email: michael.almeida@utsa.edu Citation | Almeida, M. (2017). Endurantism, fixity, and fatalism. Science, Religion and Culture, 4(2): 47-55. DOI | https://dx.doi.org/10.17582/journal.src/2017/4.2.47.55 Introduction Nelson Pike is often credited with offering the most compelling formulation of an ancient argument against human freedom.1 According to Pike, God’s foreknowledge that we will perform some action is inconsistent with our freedom to do otherwise. If freedom requires the freedom to do otherwise, then God’s foreknowledge undermines our freedom. Divine foreknowledge that we will perform some action makes alternative actions unavailable to us. Perhaps the most compelling set of counterarguments to Pike’s conclusion have been advanced by Ockhamists who distinguish hard facts and soft facts and argue that divine foreknowledge about what we will do is not a hard fact in the past. . . . Ockham distinguished between two kinds of past facts. One kind of past fact is genuinely and solely about the past; we might call this kind of fact a ‘hard’ fact about the past. Another kind of fact is not both genuinely and solely about the past; such a fact is a ‘soft’ fact about the past. The key claim of Ockham is that soft facts about the past do not carry the necessity that attends hard facts about the past.2 tion does not, on the Ockhamist account, make alternative actions unavailable to us. John Martin Fischer has argued that the best versions of Pike’s argument for theological fatalism depend on a principle of the fixity of the past. According to principles of fixity, the past—or significant parts of the past—are indeed counterfactually independent of anything we can do now. There are, most importantly, no true backtracking counterfactuals p > q such that p is a sentence describing something I can do now—say, mow the lawn at time t—and q is a sentence describing God’s foreknowledge that, say, I will not mow the lawn at time t. If God knows at 50 A.D. that I will not mow the lawn at t, then, according to Fischer, that knowledge is among the hard facts in the past. The hard facts in the past are all of those facts that are over and done with, and totally independent of anything I can now do. The principle of the fixity of the past is not implausible—indeed I’m prepared to concede, for the sake of discussion, the principle of the total fixity of the past. The principle of the total fixity of the past entails that there are no facts in the past that are not over and done with in the sense described. There are no facts about the past that are not causally independent of anything we can do now. It follows from total fixity God’s foreknowledge is not a hard fact that is coun- that the set of all past facts is just a set of hard facts.4 terfactually independent of anything we can do now.3 Total fixity of course entails that God’s foreknowlDivine foreknowledge that we will perform some ac- edge is among the hard facts. 2017 | Volume 4 | Special Issue 2 | Page 47 Science, Religion & Culture My aim in the paper is to show that the arguments for theological fatalism from God’s foreknowledge and total fixity—or any principle of fixity—are unsound. The problem with arguments from foreknowledge and fixity is the failure to directly address issues of persistence through times and worlds. But issues of persistence—the persistence of persons and objects— are central to the argument for theological fatalism. Nelson Pike’s original formulation of the argument for theological fatalism implicitly assumes a power entailment principle. Power entailment principles state that what an agent can do, or has the power to do, at any time is closed under entailment. If an agent can do A and A entails B then the agent can do B. Power entailment principles are, of course, controversial.6 Raising my hand entails that 2 + 2 = 4, for instance, but I cannot bring it about that 2 + 2 = 4. Raising The commonsense explanation of persistence is the my hand also entails that I exist, but I cannot bring endurantist account of persistence. Endurantism is the it about that I exist. There are, of course, a number of thesis that persons and objects persist through time other innovative and convincing counterexamples to by being wholly present at various temporal locations. power entailment. Fortunately, the argument for theAccording to endurantism I persist while driving to ological fatalism does not depend on formulating an work by being wholly present in the car at each mo- uncontroversial power entailment principle. ment of the drive. There is no temporal part of me that enters the car at home that is diverse from a tem- Fischer argues that better versions of Pike’s argument poral part of me that exits the car at work. Similarly, replace power entailment principles with fixity of the ordinary objects persist through time by being wholly past principles. The idea is to restrict what an agent present at each moment. The table you see today in can do at a time to actions that are counterfactually your living room is the table that was there yesterday. independent of the hard facts in the past. The initial It is not a temporal part or temporal stage of the table version of the fixity of the past principle is (FP). diverse from earlier temporal parts or stages of the FP. For any action Y, agent S, and time t, if it is true table. that if S were to do Y at t, some hard fact about the Endurantism explains how the proposition that I past relative to t would not have been a fact, then S mow the lawn at some time t and the proposition cannot at t do Y at t. that I—that is, the self-same or identical individual— might have failed to mow the lawn at t are consist- (FP) offers a necessary condition on what an agent ent with each other. Endurantism also affords a nice can do at a time t. The restriction in (FP) is a necesand natural explanation of how divine foreknowledge sary condition on what an agent can do in the relevant and the total fixity of the past are consistent with the sense of ‘can’.7 An agent S can do Y at t only if were S proposition that I mow the lawn at t and the proposi- to do Y at t then every hard fact about the past might tion that I might have—I could have or I had it in my still have been a fact.8 According to (FP), S can, in the power—not to mow the lawn at t. relevant sense, do Y at t even if in some of the closest worlds in which S does Y at t, some of the hard facts about the past are different. Fischer on Fixity According to the traditional conception of God, the properties of God include essential omniscience, omnipotence, perfectly goodness and necessary existence. These are the de re modal properties traditionally ascribed to a maximally great being. Essential omniscience entails, at least, knowledge of every true proposition. It might entail in addition knowledge of every indefinitely true proposition and, perhaps, de se knowledge.5 Arguments for theological fatalism assume that the traditional God is also essentially sempiternal. Sempiternity is omnitemporality, or existence at every temporal location. 2017 | Volume 4 | Special Issue 2 | Page 48 Fischer does not deny that there are true backtracking counterfactuals of the form were S to do Y at t it would be true that some hard fact in the past relative to t would not have been a fact. He allows, that is, that it might be true in all of the closest worlds in which S does Y at t that some hard fact in the past is not a fact. So, of course there is a sense in which S can do Y at t despite the fact that doing Y at t is not counterfactually independent of hard facts holding in the past. But S cannot in the relevant sense do Y at t. The relevant sense of ‘can’ is difficult to pin down precisely. In cases where performing Y is not counter- Science, Religion & Culture factually independent of the past hard facts, S can do Y at t in the same sense, say, that S can leap a tall building in a single bound. There are worlds in which S does so, but those worlds do not show that S can in the relevant sense leap a tall building. well-known support for the view that God’s beliefs are not counterfactually independent of what we can do now. But Fischer argues that these examples are all question begging. Consider, for instance, Alvin Plantinga’s well-received “carpenter ants” counterexample. Plantinga argues that there is an action an Fischer advances (FP) in support of a revised argu- agent can now perform which is such that if he were ment for theological fatalism. The argument of course to perform it, God would not have believed that a stipulates that God has the traditional divine attrib- colony of carpenter ants moved into Paul’s yard last utes including essential sempiternality. We suppose Saturday. that God exists and that S actually does Y at t2. It . . .if Paul were to mow his lawn this afternoon, seems evident that the divine attributes entail that (1) then the ants would not have moved in last Satand (2) are false. urday. But it is within Paul’s power to mow this afternoon. There is therefore an action he can per1. If S were to refrain from doing Y at t2, then God form such that if he were to perform it, then the would have held a false belief at t1. proposition [that the colony of carpenter ants 2. If S were to refrain from doing Y at t2, then God moved into Paul’s yard last Saturday] would have would not have existed at t1. been false.11 (1) is false in virtue of God’s essential omniscience and (2) is false in virtue of God’s necessary existence. It is impossible that God believes a false proposition Fischer’s objection is that the italicized sentence and it is impossible that God fails to exist. But con- above begs the very point at issue. While it is intuitively true that it is within Paul’s power to mow the sider (3). 3. If S were to refrain from doing Y at t2, then God lawn this afternoon—that is certainly the view that would have held a different belief from the one He commonsense invites us to believe—it is just this reactually held at t1, i.e., God would have believed ceived opinion concerning what we can do that the at t1 that S would refrain from doing Y at t2. argument for theological fatalism calls into question. Indeed theological fatalism calls into question many The backtracking counterfactual in (3) appears to be powers and abilities that we commonsensically betrue. Indeed, the truth of backtracking counterfactu- lieve we have. als like (3) have been taken as evidence that God’s beliefs are not counterfactually independent of what Fischer’s dialectical move in response to these examwe can do now.9 Nevertheless, Fischer argues—quite ples is fascinating in part because it leaves one wonunexpectedly—that (3) and (FP) entail instead that S dering what could possibly count as showing that God’s beliefs are not counterfactually independent of cannot refrain from doing Y at t2. what we can do now. It is not enough, according to Fischer, to discover an action A which is such that (i) … if (3) were true, then it would seem to follow A is a paradigmatically something S can do and (ii) in virtue of (FP) that S cannot refrain from doing 10 were S to perform A, then God would have held beY at t2. liefs distinct from the one’s he actually holds. Notice Here (FP) and (3) entail that S cannot refrain from that (i) and (ii) are exactly what David Lewis offers in doing Y at t2. We might have expected the fact that defense of the view that the time traveler can kill his S refrains from doing Y at t2 and (3) to entail that grandfather. (FP) is false. Roughly, the alternatives are the followGrandfather died in his bed in 1957, while Tim ing: (i) if S could refrain from Y at t2, then past foreknowledge wouldn’t be fixed, but it is, so S cannot do was a young boy. But when Tim has built his time machine and traveled to 1920, suddenly he that versus (ii) if past foreknowledge were fixed then S realizes that he is not too late after all. He buys couldn’t refrain from doing Y at t2, but he can, so past a rifle; he spends long hours in target practice; foreknowledge isn’t fixed. he shadows Grandfather to learn the route of his daily walk to the munitions works; he rents It is worth noting that Fischer acknowledges some 2017 | Volume 4 | Special Issue 2 | Page 49 Science, Religion & Culture a room along the route; and there he lurks, one winter day in 1921, rifle loaded, hate in his heart, as Grandfather walks closer, closer,. . . . Tim can kill Grandfather. He has what it takes. Conditions are perfect in every way: the best rifle money could buy, Grandfather an easy target only twenty yards away, not a breeze, door securely locked against intruders. Tim a good shot to begin with and now at the peak of training, and so on. What’s to stop him? The forces of logic will not stay his hand! No powerful chaperone stands by to defend the past from interference. . . In short, Tim is as much able to kill Grandfather as anyone ever is to kill anyone.12 The fixity of the past is an important metaphysical feature of the actual world, according to Fischer, that restricts what we can do now. The problem is that we just do not assess backtracking counterfactuals in ways that keep fixed all of the hard facts that fixity principles insist are fixed. According to principles of fixity, God’s past beliefs are among those important fixed facts. But when we assess the truth of backtracking counterfactuals—when we consider the closest worlds in which the antecedent of a backtracking counterfactual is true—we simply do not keep fixed God’s beliefs. We do not keep those facts fixed in the way that we standardly keep past facts fixed when assessing Emphasizing the time traveler’s intuitive ability to kill backtracking counterfactuals. The way we keep past his grandfather is central to Lewis’s argument that he facts fixed in assessing backtracking counterfactuals is can do it. Can we seriously deny that Tim can move to treat them as counterfactually independent of what his index finger and trigger the rifle? That cannot be we do now. To be fixed is to be independent. right. What more do we need to show that S can do Y at t and were S to do Y at t then some alleged hard Compare the assessment of backtracking counterfact about the past would not be a fact? factuals in the context of causal determinism. If it is causally determined that I not raise my hand, what A second fascinating consequence concerns the eval- would have happened had I raised it? uation of backtracking counterfactuals. Suppose God knows at 50 A.D. in world w that S will do A at t and . . . at least one of three things would have been suppose P is the totality of past facts in w up to time true. Contradictions would have been true tot. Is it true in w that S can do ~A at t? Fischer urges gether; or the historical proposition H would not that S cannot do ~A at t since doing ~A at t would have been true; or the law proposition L would require S to do something that is inconsistent with not have been true. Which? . . . Of our three alsome hard facts in P. Since the hard facts in P are ternatives, we may dismiss the first; for if I had fixed, it is false in w that S can do ~A at t. What would raised my hand, there would still have been no nicely explain why S cannot do ~A at t is that S’s dotrue contradictions. Likewise we may dismiss the ing so would bring about the impossible. It would second; for if I had raised my hand, the intrinsic state bring about the impossible given the unalterable facts of the world long ago would have been no different. about S’s powers or abilities and the unalterable facts That leaves the third alternative. If I had raised about the past. In order to do ~A at t, S would have my hand, the law proposition L would not have to alter some fixed historical fact, and no one can do been true.13 that. S would have to bring about a world in which S does something that is inconsistent with hard facts in If I had raised my hand then the historical facts that P. S can’t do that. constitute the intrinsic state of the world in the past would have remained unchanged. Those historiBut this is not what Fischer says or holds. He says cal facts are fixed and therefore counterfactually inthat we can entertain the possibility that S does ~A dependent of what we do now. And of course they at t. What we are entertaining—S’s doing ~A at t—is therefore do not affect what I can do now. not impossible after all. Were S to do ~A at t, according to Fischer, then some hard fact about the past rel- If God’s beliefs are among the fixed facts in the past, ative to t would not have been a fact. Nothing impos- we should expect them to be counterfactually indesible would have occurred had S done ~A at t. What pendent of what we do now. And, in fact, as we will then explains why S can’t do it? Presumably it’s the see, that is what we do find. We will also find that the principles of fixity that are supposed to explain why S counterfactual independence of those past facts does cannot do ~A at t. But how, exactly? not constrain what we can do now. 2017 | Volume 4 | Special Issue 2 | Page 50 Science, Religion & Culture at-to-t or that the clay stands in the cubical at relation to time t, we say that the clay is at t cubical or the clay According to endurantism, objects persist by hav- has at t the property of being cubical.15 ing multiple locations through time. The entire persisting object—not a mere temporal part or stage of Indexing the copula to times also permits us to say the object—is located at time t and then again lo- consistently that the clay is<at t> cubical and the clay cated at time t’.14 Of course, endurantists agree that is<at t’> spherical and, for that matter, the clay is<at objects can and do undergo change from one time to n> F, for any time n and property F whatsoever. And the next. An ounce of clay might be cubical at time again Leibniz’s laws entail that the clay must have all t and, at another time t’, the very same ounce of clay of those properties at every time at which it exists, might be spherical. The question is how such ordi- since the relation holding between the ‘objects’ existnary change could happen. How is this commonsense ing at different times is the relation of identity. view of persistence—the view that objects persist by being wholly present at various temporal locations— Now, consider what an omniscient God believes at consistent with the fact that objects undergo change. 50 A.D. about the properties of the clay. The clay is How could one and the same identical object be cubi- cubical at a certain time t and the clay is spherical at another time t’, there is no question about that. But cal and spherical? could God believe at 50 A.D. that the clay is cubical The simple solution seems to involve indexing proper- and that the very same clay is spherical? Certainly not, ties of the clay to times. On this analysis we find that or at least not consistently. Being cubical and being the piece of clay is spherical at t and cubical at t’. The spherical are intrinsic properties of the clay that are tension in exemplifying both cubicality and spher- inconsistent with each other; nothing can be both. If icality is quickly resolved. The clay does not in fact God believes at 50 A.D. that the clay is cubical, then exemplify either of those monadic properties. Instead the clay cannot be spherical as well. it stands in two different relations to time. The clay stands in the spherical at relation to one time t’ and But we obviously do not want God’s beliefs at 50 in the cubical at relation to another time t. We could A.D. to entail that the clay cannot change from one replace those relations with relational properties, if time to the next. And endurantists offer an analysis that seems more suitable. The clay would then have of persistence according to which God can believe at the relational properties bearing-spherical-at-to-t’ 50 A.D. that the clay is cubical and that the clay is and bearing-cubical-at-to-t. These are properties con- spherical. What God in fact believes is that the clay structed out of relations. And of course the clay stands is<at t> cubical and that the clay is<at t’> spherical. in both relations, or has both relational properties, at These beliefs are perfectly consistent. God’s beliefs at all times at which the clay exists. For any time t, it is 50 A.D. that the clay is<at t> cubical is perfectly contrue at t that the clay bears-cubical-at-to-t and bears- sistent with God’s belief at 50 A.D. that the clay is<at spherical-at-to-t’. The clay of course appears spherical t’> spherical. And of course it is perfectly possible that at t and appears cubical at t’, but the clay at t is in fact the clay is<at t> cubical and is<at t’> spherical. Given indiscernible from the clay at t’. There is no property the endurantist analysis, God’s beliefs do not preclude that the clay exemplifies at t that it does not exemplify the clay from changing over time and exemplifying at t’ and vice versa. Since the clay at t just is the clay at both of those properties. t’, on this simple analysis, it is a consequence of Leibniz’s law (of the indiscernibility of identicals) that it Endurantism and Fatalism must exemplify the same properties at t and t’. In discussion of Fischer’s argument for theological Endurantists do not in general endorse the simple fatalism, we noted that if God believes at t1 that S account of how objects persist. Better accounts reject actually does Y at t2, then the propositions in (1) and the proposal that the clay does not exemplify the mo- (2) are false. nadic properties of being cubical or being spherical. A better alternative is to index the copula to times, or 1. 1. If S were to refrain from doing Y at t2, then to ‘tense the copula’, rather than index properties to God would have held a false belief at t1. times. Instead of saying that the clay bears-cubical- 2. 2. If S were to refrain from doing Y at t2, then Endurantist Tensing 2017 | Volume 4 | Special Issue 2 | Page 51 Science, Religion & Culture God would not have existed at t1. Indeed, (1) and (2) are uncontroversially false. (1) is false because God is essentially omniscient and (2) is false because God exists necessarily. These are attributes of the traditional God stipulated in the argument. The central question is whether (3) is true. Recall that Fischer urges that (3) is indeed true, but that (3) together with (FP) entail that S cannot refrain from doing Y at t2. Temporally indexing the copula also permits God to believe at 50 A.D. that, over time, S does Y and that S refrains from doing Y. God cannot have both beliefs unless those ‘doings’ are somehow indexed to times. For instance, God might have the belief that S does<at t> Y and the belief that S refrains from doing<at t’> Y. The result otherwise—the result of indexing neither the ‘doings’ nor the properties to times—is hyperfatalism. If God believes at 50 A.D. that S does Y—rather than S does<at t> Y—then S cannot do anything inconsistent with doing Y at any other time in S’s existence! Of course, even fatalists reject hyperfatalism and instead index ‘doings’ to times. 3. 3. If S were to refrain from doing Y at t2, then God would have held a different belief from the one He actually held at t1, i.e., God would have But temporally indexing the copula does not preserve believed at t1 that S would refrain from doing Y consistency in cases of transworld change. It might be the case that the clay is cubical in one possible world at t2. w and the very same clay, at the same time, is spherical Were S to do Y at t2, it would be the case that a hard in another possible world w’. Similarly, it might be fact about the past relative to t would not have been a the case that S does Y in one possible world w and, fact. The hard fact that precludes S from failing to do at the same time, S refrains from doing Y in another possible world w’. Temporal indexing does nothing to Y at t2 is God’s belief at t1 that S does Y at t2. rescue consistency in cases of transworld change for But (FP) and God’s beliefs at t1—granting that those persons or objects. The analysis in (6) and (7) is as far beliefs are indeed hard facts in the past—do not in as temporal indexing can go towards consistency. fact preclude S from refraining from doing Y at t2. Since Fischer grants that (3) is true, we know that 6. S does<at t> Y there’s a possible world w’ in which S refrains from 7. S refrains from doing<at t> Y doing Y at t. And since God is essentially omniscient, God knows that there is a world in which S refrains It is obvious on inspection that the propositions in (6) and (7) are inconsistent. S cannot both do<at t> Y and from doing Y at t. refrain from doing<at t> Y. But the endurantist will ask how it is possible that one and the same identical person, S, has the property There seem to be two options available to the endurof refraining from doing Y at t in one world and the antist. We could simply deny that possible worlds property of doing Y at t in another world. The prob- overlap with respect to individuals. S exists, we can lem of course is perfectly analogous to the problem of stipulate, in one possible world w and his counterpart temporal persistence. The problem of temporal per- S’ exists in another possible world w’, and S ≠ S’. Since sistence was how the clay could have the property of there is no transworld identity we have no problem being cubical at one time and the property of being with the consistency of (6) and (7). Those proposispherical at another time. The simple solution recall tions become on proper analysis (6’) and (7’). was to analyze the monadic properties as implicitly relational properties. The better solution was to pre- 6’. S does<at t> Y serve the monadic properties and instead index the 7’. S’ refrains from doing<at t> Y having of properties to specific times. Indexing the copula to times makes it possible for an object to have It’s obvious on inspection that (6’) and (7’) are conat t the property F and to have at t’ the property ~F. sistent. S is diverse from S’, so both propositions could Temporally indexing the copula preserves consistency. be true. But endurantists might instead argue that Temporally enduring objects can have both of those possible worlds do overlap with respect to individuals—that objects and persons in fact endure across properties. worlds—and index the exemplification of properties 2017 | Volume 4 | Special Issue 2 | Page 52 Science, Religion & Culture to both times and worlds. To preserve consistency in the exemplification of properties across times and worlds, endurantism analyzes (6) and (7) as (6’’) and (7’’). 6’’. S does<t, w> Y. 7’’. S refrains from doing<t, w’> Y. (6’’) and (7’’) are the most natural analyses for those who find endurantism—the commonsense account of persistence—an appealing position. Since w is diverse from w’, (6’’) and (7’’) are perfectly consistent. The analysis makes it possible for S to endure though change and worlds. Indeed it is true in every possible world in which S exists that (6’’) and (7’’) are true. world, just as S does<at t> Y and S refrains from doing<at t’> Y are true at all times. We should conclude that (3) is false, and that the argument for theological fatalism is unsound. If S refrains from doing Y at t, then God would have exactly the same beliefs he actually has. Indeed, there would be no change at all in God’s beliefs. This is exactly the sort of result we should expect if God’s beliefs are among the fixed facts of the past. If endurantism is true, then God’s beliefs are counterfactually independent of what we do now, just as all historical facts—all intrinsic states of the past—are counterfactually independent of what we do now. There are additional reasons to expect God’s beliefs The endurantist analysis that yields (6’’) and (7’’) has to be counterfactually independent of what we do the major advantage of showing why the fixity argu- now. The traditional God exemplifies essential immuments for theological fatalism are unsound. Consider tability and essentially aseity—God cannot undergo the propositions in (1), (2), and (3) above. If S were to change and is necessarily independent of creation—so refrain from doing Y at t2, then God would not have God’s beliefs cannot undergo change and they cannot a false belief and God would not cease to exist. (1) be affected by any actions we undertake. and (2) are false under the analysis in (6’’). The only Can S refrain from doing Y at t2 in the relevant sense proposition that concerns us, once again, is (3). of ‘can’? According to (FA), S can refrain from doing 3. If S were to refrain from doing Y at t2, then God Y at t2 only if were S to refrain from doing Y at t2 then every hard fact about the past might still have would have held a different belief from the one been a fact. And given the commonsense account He actually held at t1. of persistence in endurantism, that of course is true. Consider what God actually believes at time t1 in our Were S to refrain from doing Y at t2 then God would world w. God has the belief in (8), since it is true that have exactly the same set of beliefs he actually has. There is no reason to believe, then, that S cannot reS does<t2, w> Y, where w is the actual world. frain from doing Y at t2 in any sense of ‘can’. 8. God believes at t1 that S does<t2, w> Y. But consider the strongest principle of fixity, the prinNow let w’ be the closest world to the actual world ciple of the total fixity of the past. The principle of the in which S refrains from doing Y at t2. Were S to total fixity of the past entails that there are no facts refrain from doing Y at t, then w’ would be actual and in the past that are not over and done with. It follows it would be true that S refrains from doing<t2, w’> Y. from total fixity that every past fact is a hard fact. So, God also has the belief in (9). TF. For any action Y, agent S, and time t, if it is true that if S were to do Y at t, some fact about the past 9. God believes at t1 that S refrains from doing<t2, relative to t would not have been a fact, then S w’> Y. cannot at t do Y at t. The beliefs in (8) and (9) are consistent and God actually has both beliefs. The fact that God believes at If (TF) is true, then for any agent S and time t, the set t1 that S does<t2, w> Y does not entail that God does of accessible worlds for S at t is just the set of worlds not believe at t1 that S refrains from doing<t2, w’> Y. that overlap with respect to the actual past up to time Indeed, (8) and (9) are true in every possible world t. We can think of the set of accessible worlds for an in which God exists. This is because S does<t2, w> Y agent S as determining what S can do at a time t in and S refrains from doing<t2, w’> Y are true in every the relevant sense of ‘can’. The set of accessible worlds 2017 | Volume 4 | Special Issue 2 | Page 53 Science, Religion & Culture for S at t, then, are worlds that branch off the actual past. Every accessible world for S at t has the very same past and not simply the same hard facts. Every possible world for S at t must therefore be consistent with the actual past. If S were to refrain from doing Y at t, would the actual past be any different at all? Would any past facts be different from the way they actually are? The answer is no. Let possible worlds w and w’ branch off the actual past. Suppose that in w’ S refrains from doing Y at t2 and in w S does Y at t2. Suppose further that an omniscient God exists in the actual past of these possible worlds and believes infallibly at t1 every true proposition. Does it follow that human freedom is affected in any way? The answer is again no. What God believes at t1 is that S does<w, t2> Y and that S refrains from doing<w’, t2> Y. These beliefs are both true and nothing S does in any possible world is inconsistent with these beliefs.16 The endurantist account of individual persistence thereby explains what goes wrong in arguments from principles of fixity to theological fatalism. Acknowledgment Many thanks to Mark Bernstein for comments on an earlier draft of this paper. Endnotes [1] See Nelson Pike, ‘Divine Omniscience and Voluntary Action’, Philosophical Review 74 (1965) 27-46. [2] See John Martin Fischer, ‘Soft Facts and Harsh Realities: Reply to William Craig’, Religious Studies, Vol. 27, No. 4 (1991) 523-539. [3] On one reading of Pike, God’s foreknowledge is counterfactually independent of what we can now do. Conclusions If the set of actions we can perform is S = {X, Y}, then God’s foreknowledge F is counterfactually independThe endurantist analysis of persistence is inde- ent of what we can do just in case X ☐⟶ F and Y pendently motivated. So the endurantist argument ☐⟶ F. Of course there might also be some action Z against theological fatalism is not ad hoc. The index- that you perform in some possible world such that Z ing of ‘doings’ to times is already accepted in every ☐⟶ ~F. But Z is therefore not in the set S of actions formulation of the problem of theological fatalism. It you can perform. Ockhamists deny that God’s forecan hardly be complained that the endurantist solu- knowledge is counterfactually independent of what tion indexes doings to worlds as well as times, since we can do, so Ockhamists would not exclude Z from that is independently required for endurantist analy- S. ses for transworld individuals. [4] I assume there is no Russell paradox in stipulating We could, of course, reject the transworld identity of a set of all past facts. individuals in favor of worldbound individuals. This amounts to the view that no individuals exist in more [5] There are several problems here that I can’t adthan one possible world. But the argument for theo- dress. There is a problem with an omniscient being’s logical fatalism would fail again. God’s beliefs at t1, knowledge of indefinitely true propositions, proposias noted above, would then be that S does<at t2> Y tions that are true under some, but not all, admissible and S’ (some counterpart of S) refrains from doing<at precisifications. There are larger worries for indefit2> Y, where S ≠ S’. There is no inconsistency in those nitely definitely true propositions, and other highbeliefs, so the fact that S’ refrains from doing<at t2> Y er order vague propositions. There is also a familiar does not entail that God’s beliefs are not identical to concern with quantifying over all true propositions. The collection of all true propositions is too big to his actual beliefs. be a set. The domain of true propositions would have Endurantism aims to explain how an individual S to be a class or some other kind of collection. Norcan persist—genuinely persist as the self-same indi- man Kretzmann has argued that an omniscient being vidual—through change over time and across possi- could not have de se or first-person knowledge of inble worlds. The endurantist explanation of individual dividual agents—the knowledge that Jones has when, persistence also explains how an omniscient being can for instance, he knows that he’s in the hospital (even have consistent beliefs about the very same individual if he fails to know that Jones is in the hospital). See through change over times and across possible worlds. Norman Kretzmann, ‘Omniscience and Immutabili2017 | Volume 4 | Special Issue 2 | Page 54 Science, Religion & Culture ty’, Journal of Philosophy, 63(14) (1966): 409–421 and Patrick Grim, ‘Against Omniscience: The Case from Essential Indexicals’, Noûs, 19(2) (1985) 151–180. See also John Hawthorne, ‘Vagueness and the Mind of God’, Philosophical Studies (2005) 1-25. [6] Phil Quinn, ‘Plantinga on Foreknowledge and Freedom’. In J. E. Tomberlin & P. van Inwagen (eds.) Profiles: Alvin Plantinga (Dordrechet-Holland: D. Reidel Publishing, 1985) 271–287. Thomas Talbott, ‘On Divine Foreknowledge and Bringing about the Past’, Philosophy and Phenomenological Research, 46, (1985) 455–469. [7] The exact sense of ‘can’ and ‘cannot’ in play in these discussions is both fascinating and extraordinarily elusive. I argue below that the right sense in which S cannot bring about p at t is in fact that p is false in every possible world accessible to S at t. [8] I make the standard assumption here that the negation of the would-counterfactual ~(A ◻︎⟶ ~B) is the might-counterfactual A♢⟶ B. [9] See Alvin Plantinga, ‘On Ockham’s Way Out’, Faith and Philosophy, 3, (1986): 235–69. John T. Saunders, ‘Of God and Freedom’, The Philosophical Review, 75, (1966): 219–225. [10] See John Martin Fischer, ‘Foreknowledge, Freedom, and the Fixity of the Past’, Philosophia [11] See Alvin Plantinga, ‘On Ockham’s Way Out’, Faith and Philosophy, 3, (1986) 254 ff. [12] See David Lewis, ‘The Paradoxes of Time Travel’ American Philosophical Quarterly (1976) 149 ff. Italics throughout are mine. [13] See David Lewis, ‘Are We Free to Break Laws?’ Theoria 3 (1981) p. 114. See also his ‘Counterfactual Dependence and Time’s Arrow’, Nous 13 (1979) 455476. My emphasis. [14] Some objects do not persist by enduring, of course, such as events and processes, baseball games, concerts and the like. [15] See Mark Johnston, Particulars and Persistence (Princeton: Princeton University Ph.D. dissertation 1984) and his ‘Is There a Problem about Persistence?’ 2017 | Volume 4 | Special Issue 2 | Page 55 Proceedings of the Aristotelian Society, supp. vol. 61 (1987) 107-135. Se also Frank Jackson, ‘Metaphysics by Possible Cases’ Monist 77 (1994) 93-110, David Lewis, ‘Tensing the Copula’, Mind 111 (2002) 1-14, and Sally Haslanger, ‘Endurance and Temporary Intrinsics’ Analysis 49 (1989) 119-25. [16] Some endurantists, of course, are presentists. What if we tried to formulate the argument for theological fatalism on the presentist assumption there exist no persons or objects that are not present. There exist, in particular, no past persons or objects and there obtain no past states of affairs or facts. The argument for presentist fatalism must argue that a state of affairs that obtained at t1 but does not obtain at t2—viz., God’s believing that S will do A at t2—is nonetheless at t2 a hard fact that constrains what S can do at t2. How could a state of affairs that does not obtain at t2 be a hard fact that constrains what S does at t2? How could a hard fact that does not obtain constrain what anyone can do at t2? Compare an analogue to presentism from actualism about possible worlds. Suppose it is true in some possible, non-actual world w that God believes that S does A in the actual world w’. Grant that God is infallible in w and that God’s belief in w entails that S does A in the actual world w’. Could a state of affairs that obtains in some possible, non-actual world w—but in fact does not obtain—constrain what S can do in the actual world w’? Isn’t it obvious that it could not? S’s behavior in the actual world w’ can only be constrained by facts or states of affairs that obtain in w’. Similarly, what S does at t can only be constrained by facts or states of affairs that actually obtain. What S does at t cannot be constrained by some fact that did obtain but presently doesn’t. Note that God does not believe at t2 that S does A at t2 until S does A at t2. But compare Michael C. Rea, Australasian Journal of Philosophy, Vol. 84, No. 4, (2006) 511– 524 and Michael Rea and Alicia Finch, ‘Presentism and Ockham’s Way Out’, Oxford Studies in Philosophy of Religion Vol. 1 (Oxford; Oxford University Press, 2008) 1-17.