On the idea that all future tensed contingents are false

IF 0.4 2区 哲学 0 PHILOSOPHY
Anthony Bigg, Kristie Miller
{"title":"On the idea that all future tensed contingents are false","authors":"Anthony Bigg,&nbsp;Kristie Miller","doi":"10.1111/phib.12335","DOIUrl":null,"url":null,"abstract":"<p>In ‘The Open Future’ (2021), Patrick Todd argues that the future is open and that, as a consequence, all future contingents are false (as opposed to the more common view that they are neither true nor false). Very roughly, this latter claim is motivated by the idea that (a) presentism is true, and so future (and indeed past) things1 do not exist, and (b) if future things do not exist, then the only thing that could ground there being future tensed facts, and hence make those future tensed claims true, is the present and the laws of nature. But (c) the present and the laws of nature cannot ground there being future tensed facts because they do not necessitate there being any such fact. Hence, future tensed claims are all false. Todd then goes on to present a semantics for his version of the open future in which all future contingents are false. In what follows, we take up two strands of Todd's view. First, we begin, in Section 2, by outlining Todd's argument that future contingents are all false. We suggest that the considerations that Todd adduces to this conclusion do not support this being so. Then, in Section 3, we consider the semantics that Todd offers and argue that it yields implausible consequences.</p><p>Let us begin with a recap of why Todd thinks that future contingents are all false. We start with some assumptions about our world. In particular, we begin with the assumption that (a) presentism is true and (b) causal indeterminism is true. Thus, we assume that future states of affairs do not exist and that the present state of the world, in conjunction with the laws of nature, does not entail what will be the case at any future time.2 Next, Todd articulates a core intuition that grounds much of what comes later in the book. The idea is that the present and the laws <i>produce</i> the future, and given this, they are what ground there being future tensed facts, if such facts there be (Todd, <span>2021</span>: 18–19). Thus, if the laws are indeterministic, then there can be no future tensed facts. This is because the existence of any such facts would be intolerably arbitrary, as these facts would not be necessitated by what is supposed to ground them.3 And the existence of facts that are not strictly required by their would-be explanans, Todd calls ‘mysterious and bizarre’ (ibid. 19). Hence, since there are no such facts, the future is open. Moreover, since there are no such facts, we should conclude that future tensed contingents are false. What it would be for there to be true future tensed contingents is for there to be future tensed facts, and there are none of these. It is the fact that the present and the laws <i>can</i> bring about a sea battle tomorrow that, according to Todd, entails that the proposition &lt;<i>It will be the case that there is no sea battle tomorrow</i>&gt; is false. Mutatis mutandis, the present and the laws, jointly, likewise have the power to bring it about that there is <i>no</i> sea battle tomorrow. And what is true of sea battles is true of all future contingents; hence, these are all false.</p><p>To reinforce this intuition, Todd draws a parallel between fiction and the future. Todd argues that it would be objectionably arbitrary to suppose that there are certain facts about, say, Harry Potter (and his ilk) if those facts are not grounded in facts about the Harry Potter fiction. For instance, the fiction remains silent on whether Harry's juice someday contains pulp.4 Given that the grounds for that fact do not exist, it would be peculiar to suppose that there is nevertheless some fact about the presence or absence of pulp. We are supposed to have the same intuition with regard to future tensed facts. It would be objectionably arbitrary to suppose there to be such facts if the ground of those facts does not obtain, and the ground does not obtain because the only ground there could be (conditional on presentism being true) are the laws and the present.</p><p>We can locate our key disagreement with Todd by considering an argument that appears analogous to the one Todd offers. We will present that argument and then note that it is surely not sound. We will then argue that reflecting on why it is not sound also sheds light on what has gone wrong with Todd's argument.</p><p>Suppose that there is, as of tomorrow, a sea battle. We know that said sea battle would be a product of (what is today) the present and the laws. No doubt the sea battle would be an <i>indeterministic</i> product of (what is today) the present and the laws; that is, the explanation of why the sea battle is occurring will make reference to the (now) present and the laws, but the fact <i>that</i> an explanatory connection holds between them will itself be a brute fact.5 But the (now) present and the laws are consistent not only with bringing it about that a sea battle occurs tomorrow, they are <i>also</i> consistent with bringing it about that there is <i>no</i> sea battle tomorrow. Thus, the existence, tomorrow, of the fact of there being a sea battle—a fact partly explained by the (now) present and the laws; a present and laws that might also have produced <i>no</i> sea battle—would be intolerably arbitrary. Hence, the (now) present and the laws cannot feature in an explanation tomorrow of a sea battle, on pain of arbitrariness. And since the would-be sea battle would be produced (if produced at all) by the (now) present and the laws, there could be no fact, tomorrow, of there being a sea battle. Moreover, what is good for there being a sea battle tomorrow is good for there being <i>no</i> sea battle tomorrow. So not only will there be no fact, tomorrow, of there being a sea battle, there will be no fact, tomorrow, of there being <i>no</i> sea battle. In essence, since the present and the laws can produce a sea battle, <i>and</i> an absence of a sea battle, the presence, tomorrow, of a sea battle fact should strike us as ‘mysterious and bizarre’.</p><p>Clearly, something has gone wrong here. <i>If</i> this second argument were sound, there could not be, tomorrow, a sea battle <i>or</i> the absence of a sea battle. But of course, if time passes from today to tomorrow, we will indeed wake up either to a sea battle or to no sea battle. So, if this argument is sound, time itself could not pass.6</p><p>Here is what went wrong. This argument has made a move from the <i>failure of necessitation</i> of tomorrow's sea battle to the <i>failure of an explanatory connection</i> between tomorrow's sea battle and the (now) present and the laws (since, given the failure of necessitation, the existence of an explanatory connection would be intolerably arbitrary). But, since the arguer recognises that <i>were there</i> a sea battle, it <i>would be</i> explained by the (now) present and laws (albeit indeterministically), the argument concludes that there can be no sea battle tomorrow.</p><p>But this move is illegitimate <i>precisely because</i>, when the would-be fact in question would be produced <i>indeterministically</i>, if produced at all, the failure of necessitation of the fact by its would-be explanans generates no reason to think the fact cannot obtain, or that were it to obtain, there would exist intolerable arbitrariness. After all, it is built into indeterminism that some facts will be obtained without any (necessitating) explanation. Thus, our hypothetical argument against the occurrence of a sea battle tomorrow is a bad argument.</p><p>Todd's argument is, in essence, the same argument as that just presented above, except that the argument above talks about tomorrow's sea battle, while Todd's talks about the future tense fact that there <i>will be</i> a sea battle tomorrow.</p><p>The above argument fails because it falsely relies on the idea that it is intolerably arbitrary to explain the sea battle by the (now) present and the laws. Todd makes a similar assumption. He maintains that the future tense fact that there <i>will be</i> a sea battle (or that there <i>will be no</i> sea battle) would have to be explained by the present and the laws, and that that explanation, if it actually obtained, would be indeterministic (non-necessitating),7 and that, therefore, there would exist intolerable arbitrariness if the present and the laws explained one rather than the other of the aforementioned future tense facts. But the same reasons exist to reject this argument as exist to reject the argument against the occurrence of a sea battle tomorrow. Namely, when explanations <i>would be</i> indeterministic, there is no intolerable arbitrariness in one obtaining, but not the other.</p><p>Moreover, there is no room to retreat by saying that indeterminism implies that indeterministic outcomes <i>have no explanation at all</i>. Consider what this would look like in the context of our analogous argument. Suppose we shed the assumption that the sea battle would be a product of the (now) present and the laws and say that if it occurs, it would have <i>no explanation at all</i>—not even one partly in terms of the (now) present and the laws. Now suppose we argue as follows: Were there to be a sea battle tomorrow, it would have no explanation at all. But a fact that had no explanation at all would be intolerably arbitrary. Hence, there will not be a sea battle tomorrow. Moreover, what is good for there being a sea battle tomorrow is good for there being <i>no</i> sea battle tomorrow. So not only will there be no fact, tomorrow, of there being a sea battle, there will be no fact, tomorrow, of there being <i>no</i> sea battle.</p><p>This new argument is even worse. If it succeeds, it shows that indeterminism per se leads to intolerable arbitrariness. Thus, Todd cannot amend his argument in an analogous manner, lest he fall foul of the same problem. That is, he cannot maintain that the fact that there <i>will be</i> a sea battle tomorrow has no explanation at all <i>and</i> that facts with no explanation are intolerably arbitrary, since this combination of views would undermine causal indeterminism.</p><p>What, then, is the alleged parallel between fiction and the future? Well, for simplicity, let us suppose that the fiction is the product of authorial intentions. Given this, it is plausible that it is an essential property of authorial intentions that they necessitate their effects. Thus, it is an essential property of Rowling's creative intentions that there be a fact that Harry Potter's name is ‘Harry’. Thus, the juice-pulp facts are necessitated by the authorial intention facts in such a way that the laws of nature and the present facts do not necessitate any future facts. Thus, the parallel between fiction and future tensed facts does not hold: where one plausibly involves the necessity of the relevant facts (in the case of fiction), the other clearly does not (in the case of future facts). And the only reason given to suppose that the present and the laws do not ground there being future tensed facts is just that they fail to do any necessitating of future facts.</p><p>Suppose, however, that authorial intentions produced some of their effects <i>indeterministically</i>, that is, suppose one thought, say, that Rowling's actual creative intentions, though they happen to have brought about the fact that Harry's name is ‘Harry’, are such that the <i>very same intention</i> might have, by chance, brought it about that Harry's name is ‘John’, or even that he had no name at all.</p><p>If that were true, could we argue that there cannot be a fact about Harry's juice's pulp content on the grounds that Rowling's actual intention fails to <i>necessitate</i> any such fact? Surely not. Once you allow that authorial intentions can generate certain effects indeterministically, this claim is no longer plausible. For if it could be that Rowling's authorial intentions indeterministically ground there being such a fact (even if, as a matter of fact, they do not). But our intuitions about these cases are the relevant ones to consider here. They are the cases that parallel the one about future facts. After all, Todd denies that it is an essential property of the present and the laws that they produce their effects deterministically (after all, he thinks they produce them indeterministically). If, however, we consider the correct parallel between fiction and the future, namely one in which necessitation obtains in <i>neither</i> case, our intuitions about the fiction case no longer provide us with any motivation to think that there are no future tensed facts.</p><p>So, we think, there are reasons to be suspicious of Todd's argument to the conclusion that all future tensed contingents are false. Moreover, in what follows, we argue that the semantics that accompany this claim, in particular the semantics for the WILL operator, are implausible. So, if this is the best semantics for the view that future tensed contingents are all false, then this is another reason to reject that claim.</p><p>In what follows, we first set out Todd's semantics for WILL before moving on to argue that this semantics is implausible.</p><p>Here is Todd's proposal in a nutshell. A <i>causally available future</i> is any complete, possible way that our timeline could unfold from the present moment, which is consistent with the facts about the present, the past, and the laws of nature. According to Todd's proposed semantics for the WILL operator, a sentence of the form WILL(p) is true iff p obtains at some point in time on every causally available future and is false otherwise. Similarly, a sentence of the form WILL<sub>n-units</sub>(p) is true iff p obtains in exactly n-units of time in every causally available future and is false otherwise.</p><p>If causal determinism is true, then there is only one causally available future, and hence it will trivially be the case that WILL(p) is true (if p obtains in the single causally available future) or WILL(p) is false (if p fails to obtain in the single causally available future).</p><p>If, however, causal indeterminism is true, then there are <i>multiple</i> causally available futures. Suppose that Freddie has libertarian free will and that it is up to him whether or not he raises his paw in the next 10 min.8 Then, there are causally available futures in which Freddie raises his paw and causally available futures in which Freddie does not raise his paw. As such, given Todd's semantics, the sentences WILL (Freddie raises his paw) and WILL (Freddie does not raise his paw) are <i>both</i> false. They are false because it is not the case in <i>all</i> futures either that Freddie raises his paw or that he does not raise his paw.9</p><p>Consider the following metaphysical possibility: First, causal indeterminism is true. Secondly, not only is the future not determined, but whether or not time marches forwards—whether or not time will pass from the present moment to the next—is <i>itself</i> an indeterministic matter.10 That is, the world faces the possibility that time will continue past this moment and the possibility that it will not. Call this possibility <i>temporal passage indeterminism</i>. Temporal passage indeterminism seems possible.11 If time passes, then it can either be the case that at any present moment, there <i>will</i> exist another moment after that, or it can be the case that at some present moment, there will exist <i>no</i> further moment. That, surely, is beyond doubt: time can be finite or infinite. If so, then in some possible world, it can surely be an indeterminate matter whether or not the present moment will be the last.12 Thus, temporal passage indeterminism is possible. Assume, then, that temporal passage indeterminism is true. How is the truth of a sentence like WILL(p) to be evaluated on Todd's semantics?</p><p>Of course, we know the answer to this question—check whether p obtains on <i>every</i> causally available future. But, given that indeterminism holds with respect to the passage of time itself, it is not determined whether there will even <i>be</i> a future after the present moment. How are we to make sense of this?</p><p>One possibility is that among the causally available futures, there is a <i>null</i> future. A sort of ‘empty’ future with no content, which serves to represent the possibility that time ends with the present moment.13 In what follows, we present Todd with a dilemma. Either there is nothing which plays the role of a null future, or there is, and in either case we are led to contradiction.</p><p>Suppose, first, there is no null future. Then, when evaluating the sentence WILL(p), we look only at those causally available futures on which there exist moments of time subsequent to the present. No matter what proposition is selected for <i>p</i>, on all such futures, <i>p or ~p</i> will be true, and so WILL(p or ~p) will come out true. Letting the present be denoted by ‘t’, on all such futures, WAS(t is present) will be true, so WILL(WAS(t is present)) will also come out as true. But <i>both</i> of these sentences—WILL(p or ~p) and WILL(WAS(t is present))—entail that the present moment is <i>not</i> the last moment of time. This is because there are truths about how the world <i>will</i> go. Moreover, there are truths about the present moment being past relative to some future time. But if it were currently the last moment of time, then there would be nothing <i>after</i> the present. As such, it would not be the case that it <i>will</i> be the case that p or ~p, and it would not be the case that it <i>will</i> be the case that it was once today. Therefore, since there are certain facts that are inconsistent with it presently being the last moment of time, it is not the last moment of time.</p><p>This means, however, that the present and the laws are sufficient to entail that there will be a moment of time after the present. But, <i>ex hypothesi</i>, temporal passage indeterminism is true, and hence the present and the laws are <i>not</i> sufficient to entail that there will be a moment of time after the present. This is a contradiction. Hence, we cannot suppose, on Todd's semantics, that the causally available futures include only those futures on which there exist times after the present.</p><p>This leads to the second horn of the dilemma. On this horn, among the causally available futures, there is something akin to a null future, which encodes the possibility that the present moment is the last moment of time.</p><p>The existence of the null future among the causally available futures will ensure that <i>p or ~p</i> and <i>WAS(t is present)</i> are not true on <i>all</i> of the causally available futures, and hence that WILL(p or ~p) and WILL(WAS(t is present)) are not now true. This avoids the objectionable results obtained above, but it raises a new set of concerns.</p><p>Indeed, for <i>any</i> proposition <i>p</i>, the sentence WILL(p) will be false. But this seems to entail that it <i>is</i> the last moment of time. To see this, notice that <i>what it is</i> for it to be the <i>first</i> moment of time is for there to be no past-tense facts. That is, for it not to have been the case that p, for any p. that is, for all p, WAS(p) is false. But the state of its being the last moment of time and the state of its being the first moment of time differ only in the relevant tenses involved. Whilst being the first moment in time concerns the absence of any past-tense facts, being the last moment of time concerns the absence of any <i>future</i> tense facts. Since, given the presence of the null future among the causally available futures, all future tense sentences (even those involving tautologies like <i>p or ~p</i>) come out false, it will be the case that there are no future tense facts <i>at all</i>. Hence, it will be the last moment of time.</p><p>This means that the present and the laws entail that there will <i>not</i> be a moment of time after the present. But, <i>ex hypothesi</i>, temporal passage indeterminism is true, and hence the present and the laws do <i>not</i> entail that there will not be a moment of time after the present (nor that there will be). This is a contradiction. Hence, we cannot suppose, on Todd's semantics, that the causally available futures include anything like a null future.</p><p>Moreover, one cannot escape the result of contradiction on the second horn by denying the equivalence between its being the last moment of time and there being no future tense facts, and this for two reasons. Firstly, such a denial would introduce an intolerable asymmetry between the metaphysics of its being the first moment of time and its being the last moment of time. But more importantly, it seems that the equivalence is a conceptual truth. To see this, consider that the denier of this equivalence needs to deny this conditional: If there are no future tense facts, then it is the last moment of time. But that conditional is equivalent to the conditional (by contraposition) that if it is <i>not</i> the last moment of time, then there <i>are</i> future tense facts. But this latter conditional (and hence the former) is unimpeachable. Suppose that it is <i>not</i> the last moment of time. Then there <i>will</i> be some <i>other</i> time. And moreover, if there <i>will</i> be some other time, and if we know a priori that tautologies are always true at all times, then it <i>will</i> be the case that p or ~p is true at some other time. In other words, WILL(p or ~p). Similarly, if it is not the last moment of time, then the current moment of time will once <i>have been</i> (otherwise, the passage of time would change history). In other words, if it is not the last moment of time, then WILL(WAS(t is present)) is true. But these results just go to show that if it is <i>not</i> the last moment of time, then there are at least <i>some</i> future tense facts. But this is equivalent to the claim that if there are <i>no</i> future tense facts, then it <i>is</i> the last moment of time. Thus, one cannot avoid the contradiction of the second horn by denying that conditional.</p><p>It is important to note at this point that the contradictions we derived above ensue only given Todd's semantics and are not inherent in temporal passage indeterminism itself. This is because, according to Todd's semantics, unlike Ockhamism, the truth value of all WILL statements supervenes on the distribution of causally available futures, which in turn supervene on the present and the laws.14 Thus, the present and the laws <i>necessitate</i> the WILL truths, and it was precisely this necessitation that led to contradiction. On Ockhamism, the very same present and laws, and hence the very same set of causally available futures (on indeterminism at least), are consistent with different truth values for the WILL statements.</p><p>Of course, Todd might argue that since his semantics is correct and since it results in contradictions in worlds where temporal passage indeterminism is true, then temporal passage indeterminism must not be possible after all. But, we think, Todd will not want to say this, and nor should he. He should not say this because this would be to derive robust metaphysical implications from a theory of semantics. Todd has argued extensively (and convincingly) that a theory of semantics should not have metaphysical implications. And indeed, we agree that it would be surprising if we could draw sweeping metaphysical conclusions from purely semantic facts.</p><p>So, we argue, there is strong reason to think that Todd's semantics for WILL is false. In turn, if this is the best semantics for a view of the open future on which all future contingents are false, it seems to us that we should deny that all future contingents are false.</p><p>Let us briefly take a look at the bigger picture, which we think generates our concerns with Todd's model of the open future. The core error, as we see it, is that this model makes indeterminism with respect to the future <i>inconsistent</i> with the facts <i>about</i> the future being brute. Prima facie, this is very odd. One might naturally think that it is precisely <i>because</i> the present and the laws do not determine the future that the facts about the future <i>must</i> be brute. After all, the present and the laws do not determine whether or not Freddie raises his paw. So plausibly, the fact that Freddie <i>does</i> raise his paw (say) is brute.15 Todd, by contrast, thinks that, <i>independent</i> of the truth of causal indeterminism, the facts about the future are decisively settled by the present and the laws (there can be no duplicates of our world, alike in regards to the present and the laws, which have different future tense facts). Whilst this thought fits naturally with a deterministic outlook on the world, it seems to us to sit somewhat uncomfortably with a view of the world as fundamentally chancy. Therefore, in our judgement, the best hope for a viable model of the open future is to be found either in Ockhamism or Supervaluationist Indeterminism.</p><p>There is much to like about Todd's book. In the end, though, we were unconvinced by the picture he presented, in which the future is open and all future contingents are false. We've argued, instead, that there are (at least) two good reasons to reject the idea that all future contingents are false.</p>","PeriodicalId":45646,"journal":{"name":"Analytic Philosophy","volume":"66 2","pages":"209-216"},"PeriodicalIF":0.4000,"publicationDate":"2024-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/phib.12335","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Analytic Philosophy","FirstCategoryId":"1085","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/phib.12335","RegionNum":2,"RegionCategory":"哲学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"0","JCRName":"PHILOSOPHY","Score":null,"Total":0}
引用次数: 0

Abstract

In ‘The Open Future’ (2021), Patrick Todd argues that the future is open and that, as a consequence, all future contingents are false (as opposed to the more common view that they are neither true nor false). Very roughly, this latter claim is motivated by the idea that (a) presentism is true, and so future (and indeed past) things1 do not exist, and (b) if future things do not exist, then the only thing that could ground there being future tensed facts, and hence make those future tensed claims true, is the present and the laws of nature. But (c) the present and the laws of nature cannot ground there being future tensed facts because they do not necessitate there being any such fact. Hence, future tensed claims are all false. Todd then goes on to present a semantics for his version of the open future in which all future contingents are false. In what follows, we take up two strands of Todd's view. First, we begin, in Section 2, by outlining Todd's argument that future contingents are all false. We suggest that the considerations that Todd adduces to this conclusion do not support this being so. Then, in Section 3, we consider the semantics that Todd offers and argue that it yields implausible consequences.

Let us begin with a recap of why Todd thinks that future contingents are all false. We start with some assumptions about our world. In particular, we begin with the assumption that (a) presentism is true and (b) causal indeterminism is true. Thus, we assume that future states of affairs do not exist and that the present state of the world, in conjunction with the laws of nature, does not entail what will be the case at any future time.2 Next, Todd articulates a core intuition that grounds much of what comes later in the book. The idea is that the present and the laws produce the future, and given this, they are what ground there being future tensed facts, if such facts there be (Todd, 2021: 18–19). Thus, if the laws are indeterministic, then there can be no future tensed facts. This is because the existence of any such facts would be intolerably arbitrary, as these facts would not be necessitated by what is supposed to ground them.3 And the existence of facts that are not strictly required by their would-be explanans, Todd calls ‘mysterious and bizarre’ (ibid. 19). Hence, since there are no such facts, the future is open. Moreover, since there are no such facts, we should conclude that future tensed contingents are false. What it would be for there to be true future tensed contingents is for there to be future tensed facts, and there are none of these. It is the fact that the present and the laws can bring about a sea battle tomorrow that, according to Todd, entails that the proposition <It will be the case that there is no sea battle tomorrow> is false. Mutatis mutandis, the present and the laws, jointly, likewise have the power to bring it about that there is no sea battle tomorrow. And what is true of sea battles is true of all future contingents; hence, these are all false.

To reinforce this intuition, Todd draws a parallel between fiction and the future. Todd argues that it would be objectionably arbitrary to suppose that there are certain facts about, say, Harry Potter (and his ilk) if those facts are not grounded in facts about the Harry Potter fiction. For instance, the fiction remains silent on whether Harry's juice someday contains pulp.4 Given that the grounds for that fact do not exist, it would be peculiar to suppose that there is nevertheless some fact about the presence or absence of pulp. We are supposed to have the same intuition with regard to future tensed facts. It would be objectionably arbitrary to suppose there to be such facts if the ground of those facts does not obtain, and the ground does not obtain because the only ground there could be (conditional on presentism being true) are the laws and the present.

We can locate our key disagreement with Todd by considering an argument that appears analogous to the one Todd offers. We will present that argument and then note that it is surely not sound. We will then argue that reflecting on why it is not sound also sheds light on what has gone wrong with Todd's argument.

Suppose that there is, as of tomorrow, a sea battle. We know that said sea battle would be a product of (what is today) the present and the laws. No doubt the sea battle would be an indeterministic product of (what is today) the present and the laws; that is, the explanation of why the sea battle is occurring will make reference to the (now) present and the laws, but the fact that an explanatory connection holds between them will itself be a brute fact.5 But the (now) present and the laws are consistent not only with bringing it about that a sea battle occurs tomorrow, they are also consistent with bringing it about that there is no sea battle tomorrow. Thus, the existence, tomorrow, of the fact of there being a sea battle—a fact partly explained by the (now) present and the laws; a present and laws that might also have produced no sea battle—would be intolerably arbitrary. Hence, the (now) present and the laws cannot feature in an explanation tomorrow of a sea battle, on pain of arbitrariness. And since the would-be sea battle would be produced (if produced at all) by the (now) present and the laws, there could be no fact, tomorrow, of there being a sea battle. Moreover, what is good for there being a sea battle tomorrow is good for there being no sea battle tomorrow. So not only will there be no fact, tomorrow, of there being a sea battle, there will be no fact, tomorrow, of there being no sea battle. In essence, since the present and the laws can produce a sea battle, and an absence of a sea battle, the presence, tomorrow, of a sea battle fact should strike us as ‘mysterious and bizarre’.

Clearly, something has gone wrong here. If this second argument were sound, there could not be, tomorrow, a sea battle or the absence of a sea battle. But of course, if time passes from today to tomorrow, we will indeed wake up either to a sea battle or to no sea battle. So, if this argument is sound, time itself could not pass.6

Here is what went wrong. This argument has made a move from the failure of necessitation of tomorrow's sea battle to the failure of an explanatory connection between tomorrow's sea battle and the (now) present and the laws (since, given the failure of necessitation, the existence of an explanatory connection would be intolerably arbitrary). But, since the arguer recognises that were there a sea battle, it would be explained by the (now) present and laws (albeit indeterministically), the argument concludes that there can be no sea battle tomorrow.

But this move is illegitimate precisely because, when the would-be fact in question would be produced indeterministically, if produced at all, the failure of necessitation of the fact by its would-be explanans generates no reason to think the fact cannot obtain, or that were it to obtain, there would exist intolerable arbitrariness. After all, it is built into indeterminism that some facts will be obtained without any (necessitating) explanation. Thus, our hypothetical argument against the occurrence of a sea battle tomorrow is a bad argument.

Todd's argument is, in essence, the same argument as that just presented above, except that the argument above talks about tomorrow's sea battle, while Todd's talks about the future tense fact that there will be a sea battle tomorrow.

The above argument fails because it falsely relies on the idea that it is intolerably arbitrary to explain the sea battle by the (now) present and the laws. Todd makes a similar assumption. He maintains that the future tense fact that there will be a sea battle (or that there will be no sea battle) would have to be explained by the present and the laws, and that that explanation, if it actually obtained, would be indeterministic (non-necessitating),7 and that, therefore, there would exist intolerable arbitrariness if the present and the laws explained one rather than the other of the aforementioned future tense facts. But the same reasons exist to reject this argument as exist to reject the argument against the occurrence of a sea battle tomorrow. Namely, when explanations would be indeterministic, there is no intolerable arbitrariness in one obtaining, but not the other.

Moreover, there is no room to retreat by saying that indeterminism implies that indeterministic outcomes have no explanation at all. Consider what this would look like in the context of our analogous argument. Suppose we shed the assumption that the sea battle would be a product of the (now) present and the laws and say that if it occurs, it would have no explanation at all—not even one partly in terms of the (now) present and the laws. Now suppose we argue as follows: Were there to be a sea battle tomorrow, it would have no explanation at all. But a fact that had no explanation at all would be intolerably arbitrary. Hence, there will not be a sea battle tomorrow. Moreover, what is good for there being a sea battle tomorrow is good for there being no sea battle tomorrow. So not only will there be no fact, tomorrow, of there being a sea battle, there will be no fact, tomorrow, of there being no sea battle.

This new argument is even worse. If it succeeds, it shows that indeterminism per se leads to intolerable arbitrariness. Thus, Todd cannot amend his argument in an analogous manner, lest he fall foul of the same problem. That is, he cannot maintain that the fact that there will be a sea battle tomorrow has no explanation at all and that facts with no explanation are intolerably arbitrary, since this combination of views would undermine causal indeterminism.

What, then, is the alleged parallel between fiction and the future? Well, for simplicity, let us suppose that the fiction is the product of authorial intentions. Given this, it is plausible that it is an essential property of authorial intentions that they necessitate their effects. Thus, it is an essential property of Rowling's creative intentions that there be a fact that Harry Potter's name is ‘Harry’. Thus, the juice-pulp facts are necessitated by the authorial intention facts in such a way that the laws of nature and the present facts do not necessitate any future facts. Thus, the parallel between fiction and future tensed facts does not hold: where one plausibly involves the necessity of the relevant facts (in the case of fiction), the other clearly does not (in the case of future facts). And the only reason given to suppose that the present and the laws do not ground there being future tensed facts is just that they fail to do any necessitating of future facts.

Suppose, however, that authorial intentions produced some of their effects indeterministically, that is, suppose one thought, say, that Rowling's actual creative intentions, though they happen to have brought about the fact that Harry's name is ‘Harry’, are such that the very same intention might have, by chance, brought it about that Harry's name is ‘John’, or even that he had no name at all.

If that were true, could we argue that there cannot be a fact about Harry's juice's pulp content on the grounds that Rowling's actual intention fails to necessitate any such fact? Surely not. Once you allow that authorial intentions can generate certain effects indeterministically, this claim is no longer plausible. For if it could be that Rowling's authorial intentions indeterministically ground there being such a fact (even if, as a matter of fact, they do not). But our intuitions about these cases are the relevant ones to consider here. They are the cases that parallel the one about future facts. After all, Todd denies that it is an essential property of the present and the laws that they produce their effects deterministically (after all, he thinks they produce them indeterministically). If, however, we consider the correct parallel between fiction and the future, namely one in which necessitation obtains in neither case, our intuitions about the fiction case no longer provide us with any motivation to think that there are no future tensed facts.

So, we think, there are reasons to be suspicious of Todd's argument to the conclusion that all future tensed contingents are false. Moreover, in what follows, we argue that the semantics that accompany this claim, in particular the semantics for the WILL operator, are implausible. So, if this is the best semantics for the view that future tensed contingents are all false, then this is another reason to reject that claim.

In what follows, we first set out Todd's semantics for WILL before moving on to argue that this semantics is implausible.

Here is Todd's proposal in a nutshell. A causally available future is any complete, possible way that our timeline could unfold from the present moment, which is consistent with the facts about the present, the past, and the laws of nature. According to Todd's proposed semantics for the WILL operator, a sentence of the form WILL(p) is true iff p obtains at some point in time on every causally available future and is false otherwise. Similarly, a sentence of the form WILLn-units(p) is true iff p obtains in exactly n-units of time in every causally available future and is false otherwise.

If causal determinism is true, then there is only one causally available future, and hence it will trivially be the case that WILL(p) is true (if p obtains in the single causally available future) or WILL(p) is false (if p fails to obtain in the single causally available future).

If, however, causal indeterminism is true, then there are multiple causally available futures. Suppose that Freddie has libertarian free will and that it is up to him whether or not he raises his paw in the next 10 min.8 Then, there are causally available futures in which Freddie raises his paw and causally available futures in which Freddie does not raise his paw. As such, given Todd's semantics, the sentences WILL (Freddie raises his paw) and WILL (Freddie does not raise his paw) are both false. They are false because it is not the case in all futures either that Freddie raises his paw or that he does not raise his paw.9

Consider the following metaphysical possibility: First, causal indeterminism is true. Secondly, not only is the future not determined, but whether or not time marches forwards—whether or not time will pass from the present moment to the next—is itself an indeterministic matter.10 That is, the world faces the possibility that time will continue past this moment and the possibility that it will not. Call this possibility temporal passage indeterminism. Temporal passage indeterminism seems possible.11 If time passes, then it can either be the case that at any present moment, there will exist another moment after that, or it can be the case that at some present moment, there will exist no further moment. That, surely, is beyond doubt: time can be finite or infinite. If so, then in some possible world, it can surely be an indeterminate matter whether or not the present moment will be the last.12 Thus, temporal passage indeterminism is possible. Assume, then, that temporal passage indeterminism is true. How is the truth of a sentence like WILL(p) to be evaluated on Todd's semantics?

Of course, we know the answer to this question—check whether p obtains on every causally available future. But, given that indeterminism holds with respect to the passage of time itself, it is not determined whether there will even be a future after the present moment. How are we to make sense of this?

One possibility is that among the causally available futures, there is a null future. A sort of ‘empty’ future with no content, which serves to represent the possibility that time ends with the present moment.13 In what follows, we present Todd with a dilemma. Either there is nothing which plays the role of a null future, or there is, and in either case we are led to contradiction.

Suppose, first, there is no null future. Then, when evaluating the sentence WILL(p), we look only at those causally available futures on which there exist moments of time subsequent to the present. No matter what proposition is selected for p, on all such futures, p or ~p will be true, and so WILL(p or ~p) will come out true. Letting the present be denoted by ‘t’, on all such futures, WAS(t is present) will be true, so WILL(WAS(t is present)) will also come out as true. But both of these sentences—WILL(p or ~p) and WILL(WAS(t is present))—entail that the present moment is not the last moment of time. This is because there are truths about how the world will go. Moreover, there are truths about the present moment being past relative to some future time. But if it were currently the last moment of time, then there would be nothing after the present. As such, it would not be the case that it will be the case that p or ~p, and it would not be the case that it will be the case that it was once today. Therefore, since there are certain facts that are inconsistent with it presently being the last moment of time, it is not the last moment of time.

This means, however, that the present and the laws are sufficient to entail that there will be a moment of time after the present. But, ex hypothesi, temporal passage indeterminism is true, and hence the present and the laws are not sufficient to entail that there will be a moment of time after the present. This is a contradiction. Hence, we cannot suppose, on Todd's semantics, that the causally available futures include only those futures on which there exist times after the present.

This leads to the second horn of the dilemma. On this horn, among the causally available futures, there is something akin to a null future, which encodes the possibility that the present moment is the last moment of time.

The existence of the null future among the causally available futures will ensure that p or ~p and WAS(t is present) are not true on all of the causally available futures, and hence that WILL(p or ~p) and WILL(WAS(t is present)) are not now true. This avoids the objectionable results obtained above, but it raises a new set of concerns.

Indeed, for any proposition p, the sentence WILL(p) will be false. But this seems to entail that it is the last moment of time. To see this, notice that what it is for it to be the first moment of time is for there to be no past-tense facts. That is, for it not to have been the case that p, for any p. that is, for all p, WAS(p) is false. But the state of its being the last moment of time and the state of its being the first moment of time differ only in the relevant tenses involved. Whilst being the first moment in time concerns the absence of any past-tense facts, being the last moment of time concerns the absence of any future tense facts. Since, given the presence of the null future among the causally available futures, all future tense sentences (even those involving tautologies like p or ~p) come out false, it will be the case that there are no future tense facts at all. Hence, it will be the last moment of time.

This means that the present and the laws entail that there will not be a moment of time after the present. But, ex hypothesi, temporal passage indeterminism is true, and hence the present and the laws do not entail that there will not be a moment of time after the present (nor that there will be). This is a contradiction. Hence, we cannot suppose, on Todd's semantics, that the causally available futures include anything like a null future.

Moreover, one cannot escape the result of contradiction on the second horn by denying the equivalence between its being the last moment of time and there being no future tense facts, and this for two reasons. Firstly, such a denial would introduce an intolerable asymmetry between the metaphysics of its being the first moment of time and its being the last moment of time. But more importantly, it seems that the equivalence is a conceptual truth. To see this, consider that the denier of this equivalence needs to deny this conditional: If there are no future tense facts, then it is the last moment of time. But that conditional is equivalent to the conditional (by contraposition) that if it is not the last moment of time, then there are future tense facts. But this latter conditional (and hence the former) is unimpeachable. Suppose that it is not the last moment of time. Then there will be some other time. And moreover, if there will be some other time, and if we know a priori that tautologies are always true at all times, then it will be the case that p or ~p is true at some other time. In other words, WILL(p or ~p). Similarly, if it is not the last moment of time, then the current moment of time will once have been (otherwise, the passage of time would change history). In other words, if it is not the last moment of time, then WILL(WAS(t is present)) is true. But these results just go to show that if it is not the last moment of time, then there are at least some future tense facts. But this is equivalent to the claim that if there are no future tense facts, then it is the last moment of time. Thus, one cannot avoid the contradiction of the second horn by denying that conditional.

It is important to note at this point that the contradictions we derived above ensue only given Todd's semantics and are not inherent in temporal passage indeterminism itself. This is because, according to Todd's semantics, unlike Ockhamism, the truth value of all WILL statements supervenes on the distribution of causally available futures, which in turn supervene on the present and the laws.14 Thus, the present and the laws necessitate the WILL truths, and it was precisely this necessitation that led to contradiction. On Ockhamism, the very same present and laws, and hence the very same set of causally available futures (on indeterminism at least), are consistent with different truth values for the WILL statements.

Of course, Todd might argue that since his semantics is correct and since it results in contradictions in worlds where temporal passage indeterminism is true, then temporal passage indeterminism must not be possible after all. But, we think, Todd will not want to say this, and nor should he. He should not say this because this would be to derive robust metaphysical implications from a theory of semantics. Todd has argued extensively (and convincingly) that a theory of semantics should not have metaphysical implications. And indeed, we agree that it would be surprising if we could draw sweeping metaphysical conclusions from purely semantic facts.

So, we argue, there is strong reason to think that Todd's semantics for WILL is false. In turn, if this is the best semantics for a view of the open future on which all future contingents are false, it seems to us that we should deny that all future contingents are false.

Let us briefly take a look at the bigger picture, which we think generates our concerns with Todd's model of the open future. The core error, as we see it, is that this model makes indeterminism with respect to the future inconsistent with the facts about the future being brute. Prima facie, this is very odd. One might naturally think that it is precisely because the present and the laws do not determine the future that the facts about the future must be brute. After all, the present and the laws do not determine whether or not Freddie raises his paw. So plausibly, the fact that Freddie does raise his paw (say) is brute.15 Todd, by contrast, thinks that, independent of the truth of causal indeterminism, the facts about the future are decisively settled by the present and the laws (there can be no duplicates of our world, alike in regards to the present and the laws, which have different future tense facts). Whilst this thought fits naturally with a deterministic outlook on the world, it seems to us to sit somewhat uncomfortably with a view of the world as fundamentally chancy. Therefore, in our judgement, the best hope for a viable model of the open future is to be found either in Ockhamism or Supervaluationist Indeterminism.

There is much to like about Todd's book. In the end, though, we were unconvinced by the picture he presented, in which the future is open and all future contingents are false. We've argued, instead, that there are (at least) two good reasons to reject the idea that all future contingents are false.

关于所有未来张紧式或然条件都是假的这一观点
1 引言 在《开放的未来》(2021)中,帕特里克-托德认为,未来是开放的,因此,所有未来的偶然性都是虚假的(而不是更常见的既非真也非假的观点)。粗略地说,后一种主张的动机是:(a) 现在论是真的,因此未来(以及过去)的事物1 并不存在;(b) 如果未来的事物并不存在,那么唯一能够作为未来张式事实的基础,从而使这些未来张式主张成为真的,就是现在和自然法则。但是,(c) 现在和自然法则不能作为未来张式事实存在的基础,因为它们并不必然存在这样的事实。因此,未来时态的主张都是错误的。托德接着为他的开放未来版本提出了一种语义学,在这种语义学中,所有的未来偶然性都是虚假的。接下来,我们将讨论托德观点的两个方面。首先,我们在第 2 节中概述了托德关于未来或然都是假的论证。我们认为,托德为这一结论所引用的考虑因素并不支持这一观点。然后,在第 3 节中,我们考虑了托德提出的语义,并论证了它所产生的难以置信的后果。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Analytic Philosophy
Analytic Philosophy PHILOSOPHY-
CiteScore
1.10
自引率
0.00%
发文量
34
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信