Comments on a Recent Paper

Ian Spencer, a philosophy Ph.D. grad student at UC Davis, has commented on my paper (coauthored with Greg Boyd and Tom Belt) "Open Theism, Omniscience, and the Nature of the Future", which recently came out in Faith and Philosophy (Fall 2006 issue). In this post I'd like to address some of his concerns. I'll take short quotes from him and comment on them in turn.

One of things they do is to argue that the future's causal openness ... is incompatible with the denial of semantic openness for associated future-tensed sentences (a sentence is semantically open if it is neither determinately true nor determinately false).

This is inaccurate in two respects. First, our concern is not with future-tensed sentences but with propositions about the future. Sentences and propositions are not the same things. Sentences are physical things (typically strings of words or sounds) that can express a proposition. Propositions are semantically complete units of meaning that can have a truth value. Second, we say that a proposition about the future is 'semantically open' [I now prefer to say 'alethically open'] iff neither the 'will' nor the 'will not' variants of it are true and the corresponding 'might and might not' proposition is true. Thus, a proposition about a state of affairs X obtaining at future time t is semantically open iff neither nor is true (they may both be false) and is true.

They argue for this incompatibilism by arguing that 'will' in normal cases has 'causative force' - when we utter such future-tensed sentences we are indicating that there is some (high, perhaps) causal probability that what we are saying is going to occur. And supposedly that shows that if the future is causally open then such sentences cannot be semantically closed. ... Why think the causative force must show up in the semantics? After all, there's a very important distinction between saying and indicating - when I say that p, I am also indicating that I believe that p, but 'p' in my mouth doesn't have anything about me in its semantics. So 'It will be the case that p' may very well indicate something causal without that showing up in the semantics at all.

Why think that predictive propositions about the future have 'causal force' as part of their semantic content? We argue for this in some detail in the paper, but the gist of it is that to make a genuine prediction (as opposed to a mere guess) about the future is to make an assertion about the future. An assertion must express a sincere belief that the future to going to turn out a certain way rather than otherwise. And, we argue, such a belief must involve the conviction that the predicted event is at least probable in light of what is the case at the (putative) time the prediction is made. In other words, to genuinely predict that something 'will' happen is at least to say that it 'will probably' happen. That's what we mean by saying that predictions have 'causal force'.

Ian seems willing to concede that predictive uses of 'will' may indicate that the predicted event is believed to be probable, but he denies that this need amount to saying that the predicted event is probable. True enough, but the significance of this observation hinges upon an ambiguity in the notion of 'saying'. It is correct that for me to say (i.e., utter the sentence) "p is true" indicates that I believe that p, but does not necessarily say (i.e., assert the proposition) . But for that matter, it need not even amount to saying (asserting the proposition) that p. After all, I might be lying or joking when I utter "p is true". Now, the focus in my paper is on what it means to assert a proposition about the future. To do so in a way that could communicate to others, I have to utter a sentence like "X will obtain". But I could also utter that same sentence without having any intent to assert . Again, I might be lying or joking. So the issue of what an apparently predictive utterance like "X will obtain" cannot be settled by looking at the linguistic meaning of the words. We have to interpret the words, and apart from context, the basic principle that we have to work with is the principle of charity.

The principle of charity says that a person S’s utterances ought to be interpreted, provided it is textually and contextually plausible to do so, in such a way that we take them to reflect a coherent set of beliefs that is reasonable in light of the experiences and evidence available to S. In other words, before concluding that S is saying something stupid or talking nonsense, we should give S the benefit of the doubt and try to interpret S’s utterances in such a way that whatever claims we impute to S seem like they would be assertible for her. Thus, if S seems to make a claim that is not assertible for S in the context, then we should either (i) impute to S a different claim, one that does seem assertible for her in the context, (ii) impute to S additional beliefs so that the claim does become assertible for her, or (iii) construe S’s utterance as something other than a claim.

Given the principle of charity, any genuine predictive utterance must express the speaker's belief that the predicted event is probable with respect to state of the world at the (putative) time the prediction is made. If Sally says before a roulette wheel is spun, "The ball will land on 20", that is either because she really believes that it is likely to do (in which case she has made a genuine prediction, one that has a high degree of causal force), or it is not a genuine prediction about roulette wheel. Perhaps she is really making a different claim, like the autobiographical "I hope the ball lands on 20". Or perhaps she is not making a claim at all but engaging in an illocutionary speech act (e.g., placing a bet on 20).

In fact, their whole argument seems to trade on a confusion between evidence or conditions of rational assertibility on the one hand and truth conditions or semantics on the other.

This is a misunderstanding of our position. We do not base meaning and truth conditions on the speaker's evidence but on the speaker's beliefs, irrespective of their evidential grounding. In other words, when we speak of assertibility, what we have in mind is not epistemic or 'warranted' assertibility but merely psychological assertibility - the platitude that to assert p you have to believe p.

But let's say 'will' does function in the way the authors suggest. This tells us nothing about tenseless sentences that don't use 'will'. So you can still have sentences about the future with determinate truth values so long as you don't use 'will'. Or if that's not kosher, we could decide to use 'will' stripped of its causal significance and so still have sentences about the future with determinate truth values even in the face of causal openness.

These comments puzzle me. Tenseless propositions cannot be about the future. 'Future' is an inherently tensed concept, so for a proposition to be about the future as such it must be tensed. Tenseless propositions can be construed as disjunctively tensed - either was, or is, or will be - but, again, that doesn't give you a claim about the future. Since predictions are essentially about the future, we can't use these propositions to make predictions. As for the idea of stripping 'will' of any causal force, that is what Ockhamists do. For them, to say that X will happen just means that X does happen at some time subsequent to the (putative) time when the prediction is made. But if the argument of our paper is correct, then to strip the causal force of 'will' is to strip the proposition of its predictive character. This suggestion therefore fails to engage with our argument.