| ID | REGIOSTAR-7 NOMENCLATURE | N | SHARE OF VL, 2021 | CHANGE IN VL AREA, 2011–21 (%) | ||
|---|---|---|---|---|---|---|
| REDUCTION IN VL | NEW VL | NET BALANCE | ||||
| MR1 | Regiopolis and large city | 3 | 2.99% | –20.17% | 9.49% | –10.69% |
| MR2 | Medium-sized cities and urban areas in metropolitan regions | 46 | 3.82% | –27.71% | 13.17% | –14.54% |
| MR3 | Small town areas and village areas in metropolitan regions | 12 | 4.46% | –20.70% | 27.22% | 6.52% |
| RR1 | Central city in rural regions | 5 | 3.99% | –26.26% | 12.32% | –13.93% |
| RR2 | Medium-sized cities and urban areas in rural regions | 27 | 4.18% | –21.55% | 17.84% | –3.71% |
| RR3 | Small town areas and village areas in rural regions | 26 | 4.40% | –18.57% | 26.17% | 7.60% |
Introduction
I don’t have any great-grandchildren. Nor do I have any great-grandparents. This is because my great-grandparents are no longer, and my great-grandchildren are (perhaps) yet to come. The philosophical view that affirms both of these is called presentism. Presentists say that only present things exist, nothing is at a temporal distance from anything else, the most inclusive quantifier ranges over only those things that are present, and everything that exists is simultaneous with everything else.1 Most precisely, presentism is the thesis that for all x, x is present.2
If every action that anyone performs is causally determined by what happened in the distant past, then nobody does anything freely. But people do, at least sometimes, act freely. The view that is the conjunction of these two theses is called libertarianism; it is the conjunction of the free will thesis and incompatibilism with respect to free will and determinism. Determinism is the thesis that the conjunction of the past and the laws of nature entails a unique future. The free will thesis is that sometimes agents act freely. Incompatibilism is the thesis that it is impossible that both determinism and the free-will thesis be true.
Bivalence says that every proposition is either true or false.3 Bivalence, then, is the thesis that: for all propositions p, p is true or p is false.
Finally, it is standardly thought that if God exists, God is omniscient, and if God is omniscient, then God has foreknowledge.4 That is, there are truths about the future – including truths about the free actions of free people – and God knows that they are true.
One often hears that libertarianism is incompatible with divine foreknowledge, but there have been persuasive responses to this charge. However, there is a further charge – that these responses rest on the existence of non-present objects. So this charge says that libertarianism, divine foreknowledge, and presentism are incompatible. This has been pressed recently by Philip Swenson (2016). Some go even further; they say that divine foreknowledge is not even required – the conjunction of libertarianism and presentism is incompatible with there being truths about the future actions of people, regardless of whether there’s a being that knows these truths. This has also been pressed by Michael Rea (2006).
In this paper, I respond to Swenson’s and Rea’s arguments. In each case, I develop a parody argument that seeks to show that actualism – the view that everything is actual – is inconsistent with the conjunction of (in the case of Rea) libertarianism and bivalence and the conjunction of (in the case of Swenson) libertarianism and divine foreknowledge. Seeing how these parody arguments using actualism go wrong helps us see how the arguments using presentism go wrong. I conclude that we have not yet been provided with a sound argument that presentism is inconsistent with the conjunction of libertarianism and bivalence or that presentism is inconsistent with the conjunction of libertarianism and divine foreknowledge.
Rea’s Argument
Rea’s argument assumes the truth of presentism and bivalence and concludes with the denial of libertarianism.
(M1) Presentism is true. (Assumption)
(M2) Sally will stand 1000 years hence was true at t*. (Assumption)5, 6
(M3) If presentism is true and if Sally will stand 1000 years hence was true at t*, then the truth of Sally will stand 1000 years hence at t* was not even partly grounded in the occurrence of any event involving Sally or in any exercising of her agent-causal power. (Premise)
(M4) Therefore: The truth of Sally will stand 1000 years hence at t* was not even partly grounded in the occurrence of any event involving Sally, or in any exercising of her agent-causal power. (From M1, M2, M3)
(M5) If the truth of a proposition p at a past time tn was not even partly grounded in the occurrence of any event involving S, or in the agent causal activity of S, then S has never had and will never have a choice about whether p was true at tn. (Premise)
(M6) Therefore: Sally has never had and will never have a choice about whether M2 is true. (From M2, M4, M5)
(M7) M2 entails that Sally stands now (at t, one thousand years later than t*). (Trivial)
(β3) If p and if x never had and never will have a choice about p, and if p entails q, then x never had and never will have a choice about q. (Premise)
(M8) Therefore: Sally has never had and will never have a choice about whether Sally will stand 1000 years hence is true. (From M2, M6, M7, β3)
It’s worth noting that this isn’t merely an argument that presentism, libertarianism and bivalence are inconsistent. It’s an argument that they, combined with other assumptions, are inconsistent. These assumptions are: that the world is at least 1000 years old, that there are true singular propositions about non-existent things, and β3. One could deny any or all of these additional assumptions in order to block the argument.7 I do not deny any of them, however, so I must come up with a different response.
One response to this argument is to deny M2 by saying that all future contingent propositions are false.8 On this view, WILL(P) and WILL(NOT(P)) can both be false, because ‘X will obtain at t’ and ‘X will not obtain at t’ (and their tense-neutral counterparts) are contraries, not contradictories’.9 So Sally will stand 1000 years hence was false at t*, and It is not the case that Sally will stand 100 years hence was also false at t*. While I think this view responds to the problem, I think that even if we grant that there are future contingent truths, the argument still fails. And I do think there are future contingent truths. So, I’ll grant this for the sake of argument.
One could also deny M5, invoking an ‘Okhamist’ strategy. The Okhamist strategy grants that the truth of Sally will stand 1000 years hence at t* was not even partly grounded in something involving Sally, but the truth of Sally will stand 1000 years hence was true at t* is, at t, at least partly grounded in something involving Sally; so Sally has a choice about it.10
My response is that the argument trades on an ambiguity in M2. Once this ambiguity is exposed, either the resulting reading of M7 is false, or the resulting readings of M7 and M8 do not entail the denial of libertarianism. The easiest way to see the ambiguity is with a parody argument intended to establish the inconsistency of actualism, bivalence, and libertarianism.
Parody of Rea’s argument
Presentism is often explained by analogy to actualism. Actualism is the view that everything is actual. So, much like presentism says that I don’t have any great-grandparents or great-grandchildren because if I did they would be in the past or future, actualism says that I don’t have any brothers or step-sisters; if I did, they would be merely possible. I could have had brothers and step-sisters, much like I did have great-grandparents and will have great-grandchildren — but I don’t.
In this section, I exploit the similarities between presentism and actualism to give a parody of Rea’s argument. First, a note on the analogy between presentism and actualism. The presentist I have in mind is an ersatz presentist’; that is, a presentist who thinks of times as abstract objects.11 Times are maximal, consistent conjunctions of present-tense propositions that are ordered by earlier-than and later-than relations, forming an ersatz B-series.12 The ersatz B-series is contingent; that is, the times that stand in the B-relations exist necessarily, but that they stand in the B-relations to the other times they do is contingently true.13 So, a proposition p is true at a time t just in the case that t has p as a conjunct or t entails p. And for any t* such that t* stands in a B-relation to t, it is contingent that t and t* stand in that B-relation.
The actualist I have in mind thinks of possible worlds in very much the same way that the presentist thinks of times.14 Possible worlds are abstract objects, and are maximal consistent conjunctions or classes of propositions. They differ from times in that they do not merely have present-tensed propositions as conjuncts. A proposition p is true at a time t when p accurately describes t, and a proposition p is true at a world W when p accurately describes W.15
There are relations between possible worlds. One such relation is accessibility; at least some worlds are accessible from others, and many think that every possible world is accessible from every other world.16 The accessibility relation is an equivalence relation. But, there are other relations that hold between worlds, and not all are equivalence relations. Some of them are nearness relations. Nearness relations fall under the more general category of ordering relations. The series of times also has ordering relations: ‘is earlier than’ and ‘is later than’. And this comes in varieties, like ‘is n years earlier than’, ‘is n minutes later than’, and so on. Rea’s argument uses these ordering relations – ‘is 1000 years hence’.
For the analogy, we must stipulate an ordering relation by which possible worlds can be ordered; call it a ‘zear’. Zears order worlds in the way that years order times. Every world is 0 zears from itself, and which world is actual determines which worlds in fact stand in the zears relation to which other worlds, and how many zears each world is from each other world (much like which time is present determines which times stand in the years relation to which other times, and how many years each time is from each other time). So, years are measures on the ordered series of times, and zears are measures on the ordered series of possible worlds. So the n zears from relation is the modal analog of the temporal n years from relation. But ‘is n years from’ is symmetric, and Rea’s argument uses an asymmetric relation – ‘is 1000 years hence’. We need an asymmetric ordering relation on worlds. So let’s stipulate that the ‘is n zears from’ relation is actually two relations, both of which are asymmetric – ‘is n zears ance’, and ‘is n zears zence’. The names of the directions don’t matter; what’s important is that they are each asymmetric. They could be determined, for example, by the number of peaches eaten, or by the total number of hairs on every animal.
Here, then, is a parody of Rea’s argument. Assume that Sally does not exist in W ever, at any time, and that α (the world that happens to be actual) is 1000 zears from W.
Parody:
(M1*) Actualism is true. (Assumption)
(M2*) Sally stands 1000 zears zence is true at W. (Assumption)
(M3*) If actualism is true and if Sally stands 1000 zears zence is true at W, then the truth of Sally stands 1000 zears zence at W is not even partly grounded in the occurrence of any event involving Sally or in any exercising of her agent-causal power. (Premise)
(M4*) Therefore: The truth of Sally stands 1000 zears zence at W is not even partly grounded in the occurrence of any event involving Sally, or in any exercising of her agent-causal power. (From M1*, M2*, M3*)
(M5*) If the truth of a proposition p at a possible world W is not even partly grounded in the occurrence of any event involving S, or in the agent causal activity of S, then S has no choice about whether p is true at W. (Premise)
(M6*) Therefore: Sally has no choice about whether M2* is true. (From M2*, M4*, M5*)
(M7*) M2* entails that Sally stands in the actual world (at α, one thousand zears zence from W). (Trivial)
(β3) If p and if x has no choice about p, and if p entails q, then x has no choice about q. (Premise)
(M8*) Therefore: Sally has no choice about whether Sally stands 1000 zears zence is true. (From M2*, M6*, M7*, β3)
Since the world that happens to be the actual world is 1000 zears from W, one might think that Parody shows that Sally has no choice about whether she stands. But clearly there is a problem; actualism, libertarianism and bivalence are consistent! And whatever the problem is with Parody, I’ll argue that there is an analogous problem with Rea’s argument.
Note also that the parody argument doesn’t rely on presentism or eternalism being true in W or in α. There is no time in W at which Sally exists. Since Sally never exists at any time in W, then according to actualism, in W she doesn’t exist at all. So, she can’t even partially ground the truth in W of Sally stands 1000 zears zence. So, if the parody argument is right, it is right completely independently of presentism and eternalism.17
Here’s the problem. Look again at M2*, and imagine a speaker in W contemplating the sentence ‘Sally stands 1000 zears zence.’ There is an ambiguity in the term ‘zence.’ When taken in one sense, an expression in which it appears designates rigidly (‘1000 zears from W’). In the other sense, an expression in which it appears designates non-rigidly (‘1000 zears from the actual world’). M2*, then, is logically equivalent to one of the following, but not both. First:
M2*Rigid: Sally stands 1000 zears from W is true at W
Or:
M2*Non-Rigid: Sally stands 1000 zears from the actual world is true at W18
M2*Rigid is necessarily true, and Sally doesn’t have a choice about its truth. Furthermore, since it’s necessarily true, we cannot get to M7*. All we can get is:
M7*Rigid: M2* entails that Sally stands at α
But M7*Rigid is also a necessary truth, and of course Sally doesn’t have a choice about it. But then the conclusion of the argument must also be a necessary truth.
M8*Rigid: Therefore: Sally has no choice about whether Sally stands 1000 zears zence is true.
If ‘1000 zears zence’ rigidly designates α, then this is equivalent to
M8*Rigid*: Therefore: Sally has no choice about whether Sally stands at α is true.
And this doesn’t undermine libertarianism. Sally doesn’t have a choice about whether she stands at α, but that doesn’t entail that she doesn’t have a choice about whether she stands at the actual world. She has a choice about whether α is the actual world. So, reading M2* as M2*Rigid does not show that actualism, bivalence, and libertarianism are inconsistent.
However, M2*Non-Rigid is not necessarily true. Presumably Sally does have a choice about its truth, since she has a choice about which world is actual. But if Parody is sound, she doesn’t. So, what happens to the rest of the argument? Well, it’s a contingent fact that M2*Non-Rigid is true, which is to say that there are worlds in which it’s not true. But then M7* is logically equivalent to the following:
M7*Non-Rigid: Sally stands 1000 zears from the actual world is true at W entails that Sally stands in the actual world.
But that’s false. After all, it might have been the case that the world 1000 zears from W (α, say) not be the actual world. So, there is a problem with Parody. On one reading, its conclusion doesn’t entail the inconsistency of actualism, bivalence, and libertarianism. On the other reading, it’s unsound in virtue of M7*Non-Rigid being false.
Response to Rea’s argument
Look again at M2, and imagine a speaker at t contemplating the sentence ‘Sally stands 1000 years hence.’ There is an ambiguity in the term ‘hence.’ When taken in one sense, an expression in which it appears designates rigidly (‘1000 years after t’). In the other sense, an expression in which it appears designates non-rigidly (‘1000 years after the present time’). M2, then, is logically equivalent to one of the following, but not both. First:
M2Rigid: Sally stands 1000 years after t* is true at t*
Or:
M2Non-Rigid: Sally stands 1000 years after the present time is true at t*19
M2Rigid picks out the present time t rigidly, and says that Sally stands at it; but this is eternally true. Since it’s eternally true, it cannot entail temporary truths; we cannot get to M7. All we can get is:
M7Rigid: M2 entails that Sally stands at t
But M7Rigid is also eternally true, and of course Sally doesn’t have a choice about it, since it’s true iff t includes Sally stands. But then the conclusion of the argument must also be eternally true.
M8Rigid: Therefore: Sally has no choice about whether Sally stands 1000 years hence is true.
But if ‘1000 years hence’ rigidly designates t, then this is equivalent to
M8Rigid*: Therefore: Sally has no choice about whether Sally stands at t is true.
And this doesn’t undermine libertarianism. Recall that the presentist thinks of times as maximal, consistent conjunctions of propositions that are ordered by earlier-than and later-than relations, forming an ersatz B-series.20 If p is true 1000 years after now is true at t, that means that p is true at t* and t is 1000 years earlier than t*. That, in turn, means that t* includes p and t is 1000 years earlier than t*. Of course, it is necessary that t* includes p, but it’s contingent that t is the present time, and contingent that t* is 1000 years later than t. So, reading M2 as M2Rigid does not show that presentism, bivalence, and libertarianism are inconsistent.
However, M2Non-Rigid is not eternally true. Sally does have a choice about which time is the present time, and so Sally does have a choice about the truth of M2Non-Rigid. But if Rea’s argument is sound, she doesn’t. So, what goes wrong? Well, it’s a temporary truth that M2Non-Rigid is true, which is to say that there are times at which it’s not true. But then M7 is logically equivalent to the following:
M7Non-Rigid: Sally stands 1000 years after the present time is true at t entails that Sally stands at the present time.
But that’s false. After all, it might have been the case that the time 1000 years after t (t*, say) not be the present time, since the B-relations between times hold contingently. Had Sally not stood at t, then t* would not have stood in the 1000-years-earlier-than relation to t, since t* has as a conjunct Sally stands 1000 years hence, and had Sally not stood 1000 years hence, that conjunct would have been false. This suffices for Sally to be free with respect to standing at t.
So, Rea’s argument is faulty. On one reading, its conclusion doesn’t entail the inconsistency of presentism, bivalence, and libertarianism. On the other reading, it’s unsound in virtue of M7Non-Rigid being false.
Swenson’s Argument
Unlike Rea (2006), the ultimate goal of Swenson (2016) is not to argue against presentism. Rather, Swenson’s goal is to provide a response to the following puzzle. A lot of people are incompatibilists about freedom and determinism but compatibilists about freedom and divine foreknowledge. Since it seems like both determinism and divine foreknowledge entail that we perform the actions that we in fact perform, how can we be unfree given determinism but free given divine foreknowledge? Swenson’s solution is that we can be free as long as our actions explain God’s foreknowledge – Swenson calls his solution ‘The Dependence Solution’. Our (future) actions explain divine foreknowledge (or at least, they explain why God knows certain things about the future involving our free actions), so that’s why we can still be free even if God has exhaustive foreknowledge. Our actions don’t explain determinism, so that’s why we can’t be free if determinism is true.
Swenson says, though, that his solution is unavailable to the presentist, because according to the presentist our future actions don’t exist. Here’s how Swenson puts the argument:
Does ‘Jones will sit at t’ depend on Jones’s act itself? If it does not, then ‘Jones will sit at t’ is explanatorily independent of Jones’s act. And, presumably, there are no worlds where ‘Jones will sit at t’ was true and Jones does not end up sitting at t. Thus, by the Fixity of the Independent Past, it appears that Jones cannot refrain from sitting at t. Alternatively, if ‘Jones will sit at t’ does depend on Jones’s sitting then the worry for presentism returns. Jones’s sitting is not around 1000 years ago to explain ‘Jones will sit at t.’ (666-667, emphasis mine)
Since Swenson thinks that the Dependence Solution is the only way to solve the puzzle, he thinks that presentists can’t make sense of our freedom given divine foreknowledge. So, he thinks, presentists have to be incompatibilists about freedom and divine foreknowledge. That is, presentists must think that if God has foreknowledge, nobody is free.
To put it another way: God’s belief must depend on Jones’s action in order for it to be a free action, and if it does, then Jones’s action must exist; if Jones’s action exists, then since it’s in the future, presentism is false.
Let’s try to turn this into an argument in premise-conclusion form. Let ‘tpast’ be (the proposition that accurately describes the state of the world at) the time 1000 years ago.21
At tpast, God believed that Jones sits 1000 years after now.
If at tpast, God believed that Jones sits 1000 years after now, then, if Jones sits 1000 years after tpast freely, then God’s belief at tpast depended on Jones sitting 1000 years after tpast.
If Jones sits 1000 years after tpast freely, then God’s belief at tpast depended on Jones sitting 1000 years after tpast.
If God’s belief at tpast depended on Jones’s sitting 1000 years after tpast, then Jones sitting 1000 years after tpast exists at tpast.
If Jones sitting 1000 years after tpast exists at tpast, then presentism is false.
So, if at tpast God believed that Jones sits 1000 years after now and Jones sits freely 1000 years after tpast, then presentism is false.22
Parody of Swenson’s argument
By now my strategy should be familiar: make an actualist analog, see where it goes wrong, and then use that as a guide for what to say about the original. So, here’s the actualist analog with descriptions of worlds, where again our world is the one and only world that’s in fact 1000 zears ance from W:
1*) At W, God believes that Jones sits 1000 zears zence from the actual world.
2*) If at W, God believes that Jones sits 1000 zears zence from the actual world, then, if Jones sits 1000 zears zence from W freely, then God’s belief at W depends on Jones’s sitting 1000 zears zence from W.
3*) If Jones sits 1000 zears zence from W freely, then God’s belief at W depends on Jones’s sitting 1000 zears zence from W.
4*) If God’s belief at W depends on Jones’s sitting 1000 zears zence from W, then Jones’s sitting 1000 zears zence from W exists at W.
5*) If Jones’s sitting 1000 zears zence from W exists at W, then actualism is false.
6*) So, if at W God believed that Jones sits 1000 zears zence from the actual world and Jones sits freely 1000 zears zence from W, then actualism is false.
Again, this clearly cannot be right; the beliefs God could have had do not make us actually unfree. Here’s what I think has gone wrong. Look at the consequent of 2*. ‘1000 zears zence from W’ does not rigidly designate our world, α.23 Rather, ‘1000 zears zence from W’ refers to whatever world is 1000 zears zence from W. But then God’s belief at W that Jones sits 1000 zears zence from the actual world doesn’t depend on Jones sitting in the world that is 1000 zears zence from W. Rather, it depends on Jones actually sitting. The zears relation depends on what world is actual, so if Jones hadn’t sat, then α (which represents Jones as sitting) wouldn’t have been actual, and so wouldn’t have stood in the 1000 zears zence from relation to W.
Jones could have not actually sat. In that case, a different world would have been actual, and so a different world would have been 1000 zears zence from the actual world and then God would have believed that Jones sits 1001 zears zence from the actual world, or however many zears zence from W a world is in which Jones sits. It’s up to Jones whether to sit, and his choice influences what God believes.
Before I get to the problem with the formalized version of Swenson’s original argument, I’ll reiterate the role that the parody arguments are playing. They are trying to show that if presentism and libertarianism aren’t consistent, then neither are actualism and libertarianism. But it seems quite obvious that actualism and libertarianism are consistent. So there must be a problem with the parody arguments. I haven’t proved that the same problem afflicts the originals as the parodies, but in the absence of reasons to think they have different problems, it’s reasonable for the presentist to insist that the objector (a) show what’s wrong with the parody arguments and (b) show why the presentist can’t say that the same thing is wrong with the originals. So, even if the response I’m about to give has a problem, we still know that the parody argument is unsound, and given the similarities in the arguments and the positions, we have good reason to think that the originals are unsound for the same reason.
Response to Swenson’s argument
Look at the consequent of 2. ‘1000 years after now’ does not rigidly designate our time, t; of course Jones cannot avoid sitting at t, since t essentially includes the proposition that Jones sits. But he can avoid sitting now, and had he done so t wouldn’t have been the present time. ‘1000 years after t’ non-rigidly designates whatever time is 1000 years after tpast. But then God’s belief at tpast that Jones sits 1000 years after now doesn’t depend on Jones sitting 1000 years after tpast. Rather, it depends on Jones sitting now. The years after now relation depends on what time is present, so if Jones hadn’t sat, then t (which represents Jones as sitting) wouldn’t have been the present time, and so wouldn’t have stood in the 1000 years after relation to tpast.
I’ll put it another way. Consider the following three statements. A) At W, God believes that Jones sits 1000 zears zence from the actual world. B) Whichever world is actual is 1000 zears zence from W. C) Jones sits at whatever world is actual. A and B entail C, so if neither A nor B depend on Jones, then neither does C. But intuitively if C doesn’t depend on Jones, then Jones isn’t free. It seems like A doesn’t depend on Jones. The actualist should say that B depends on Jones. The ‘zears zence’ relation depends on which world is actual, and which world is actual depends on Jones. Had Jones not sat, a different world would be actual, and so a different world would be 1000 zears zence from W.
The same goes for the following three statements. A*) At tpast, God believed that Jones sits 1000 years after the present time. B*) Whichever time is present is 1000 years after tpast. C*) Jones sits at whatever time is present. A* and B* entail C*, so if neither A* nor B* depend on Jones, then neither does C*. But intuitively if C* doesn’t depend on Jones, then Jones isn’t free. It seems like A doesn’t depend on Jones (though Okhamists would disagree). The presentist should say that B* depends on Jones. The ‘years after’ relation depends on which time is present, and which time is present depends on Jones. Had Jones not sat, a different time would be present, and so a different time would be 1000 years after tpast.
To sum up: Jones could have not presently sat. In that case, a different time would have been present, and so a different time would have been 1000 years after tpast and then God would not have believed that Jones sits 1000 years after now. Instead, God would have believe that Jones sits 1001 years after now, or however many years after tpast Jones ends up actually sitting. It’s up to Jones whether to sit, and his choice influences what God believed.
Conclusion
Michael Rea and Philip Swenson have offered arguments that presentism entails fatalism. Rea has sought to show that presentism is inconsistent with libertarianism and bivalence, and Swenson sought to show that presentism is inconsistent with libertarianism and divine foreknowledge. I have shown that both of these arguments fail. I have not shown that presentism does not entail fatalism. But I have shown that promising attempts to show that it does fail. At least for now, presentists can accept libertarianism, bivalence, and divine foreknowledge.
Notes
[1] Some presentists, like Cameron (2015: 6) restrict the scope of presentism to concrete objects. Others include everything (e.g. abstracta) in the scope. This paper will make use of abstracta, but won’t take a stand on whether they exist in time or lie outside the scope of presentism. To my knowledge, nobody has argued that accepting presentism requires that one deny the existence of abstracta.
[2] Those who are worried that this is trivial should see Crisp (2003). Merricks (1999) would add that for any x, all the properties x has are the properties x has at the present time.
[3] I will speak of propositions as the fundamental bearers of truth and falsity. If you disagree, feel free to substitute whatever you think are the fundamental bearers of truth and falsity; nothing should hang on this.
[4] Open theists deny that God has exhaustive foreknowledge, for a variety of reasons. See, e.g., Rhoda (2008).
[5] Here and throughout, I follow Rea by using italicized declarative sentences as names for propositions expressed by those sentences.
[6] Since Sally does not exist at t*, this assumption denies existentialism, the thesis that singular propositions existentially depend on the things they’re about. The existentialist has a response to the argument; deny M2.
[7] I’m not aware of anyone who denies that the world is at least 1000 years old. On denying that there are true propositions about non-existent things, see Plantinga (1983) and Adams (1981). On the prospects for denying β3, see, inter alia, Carlson (2000), Pruss (2013), and Waldrop (forthcoming).
[8] See Hartshorne (1939) and (1964: 476). Alan Rhoda and Patrick Todd also endorse the view; see Rhoda (2006) and (2008: 231), and Todd (2016).
[9] See, e.g., Rhoda (2011: 7).
[10] Moon (2008) offers this as a response to Rea, and Alicia Finch and Rea defend M5 from that kind of argument in their (2008).
[11] See Bourne (2006: chapter 2) and Crisp (2007).
[12] For more on this, see Crisp (2007).
[14] I’m thinking of the view in Plantinga (1974) and van Inwagen (1986), which Lewis (1986) calls ‘magical ersatzism’.
[15] On truth-at, see Adams (1981: §3.2ff).
[17] Again, see the truth-in versus truth-at distinction in Adams (1981).
[19] Perhaps these ways of phrasing aren’t enough to force the rigid and non-rigid readings of M2. But that’s not important. What’s important is that there are two ways to read M2; let’s just stipulate that M2Rigid picks out the present time rigidly and M2Non-Rigid does not.
[20] Not that abstract times are required for the response I’m suggesting. Even if the argument is formulated with tense-operators, one can still respond that M2 is ambiguous between the rigid and non-rigid reading I suggest, because the tense operators can be read rigidly or non-rigidly. And the presentist can insist that slice tense-operators (as opposed to the standard tense operators like WAS and WILL) are just disguised quantification over abstract times.
[21] Though the argument could be run using any time as long as the time at which God believed the proposition in question isn’t the present time.
Acknowledgements
Many thanks to Alex Arnold, Andrew Bailey, Tom Crisp, Mike Rea, Amy Seymour, Peter van Inwagen, and Dean Zimmerman for helpful discussion and comments. Special thanks to Trenton Merricks and Mike Rea for detailed written comments and Philip Swenson for extensive discussion.
Competing Interests
The author has no competing interests to declare.