This is the 9th and final installment of my series on Patrick Todd’s recently published book The Open Future: Why Future Contingents are All False (Oxford, 2021). (Previous installments: 1, 2, 3, 4, 5, 6, 7, 8.)
In this chapter Todd tackles the “assertion problem” for open futurism. He’s quick to clarify that it’s not a single problem but a “family of problems” (p. 181) which run something like this:
[W]e often assert claims about the future, and we seem rationally warranted in doing so—but the theory of the open future would imply that many such claims are untrue or false; but since … well, something or other, we have a problem for the doctrine of the open future. (p. 181, ellipsis in original)
I. Conversational norms
One way of fleshing out “something or other” in the above rough description of the problem is to appeal to commonsense norms of assertion, such as Grice’s conversational maxims:
- Quantity: Provide as much information as necessary but not too much information.
- Quality: Do not say what you believe to be false or that for which you lack adequate evidence.
- Relation: Be relevant.
- Manner: Avoid needlessly obscure or ambiguous expressions. Be orderly and brief.
The maxim of Quality enjoins us not to say things we believe to be false. We might broaden this to do not say things you believe are untrue. Since open futurists deny that there are any true “will” or “will not” propositions about future contingents—all such propositions are false or at least untrue—this maxim might pose a problem. Allegedly, when non-open futurists assert “will” and “will not” propositions about the future in general and future contingents in particular, they aren’t engaged in any systematic violation of the maxim of Quality, for they believe that such propositions can be true. But if open futurists regularly make “will” and “will not” claims about what they take to be future contingents, then it seems they are systematically violating the maxim. If so, wouldn’t that amount to a kind of hypocrisy—saying one thing (i.e., that no such propositions about future contingents are true) and yet acting contrariwise (i.e., talking as if some such propositions are true)?
In Ch.8 Todd develops two related “themes” in response to this sort of problem (p. 181):
- Communication: “[T]he open futurist can argue that although … we assert what is untrue or false, we nevertheless communicate what is true.”
- Replacement: The open futurist can offer “replacement talk … [in which] we replace our current talk (in which we assert future contingents) with nearby talk (in which we don’t) … without sacrificing anything of importance.”
In other words, the open futurist can maintain that we shouldn’t take what may seem to be open futurist assertions of untrue or false claims about future contingents at full face value. To put it another way, the open futurist can defend the practice of occasionally speaking as if “will” and “will not” propositions about future contingents are true by leveraging the other Gricean maxims (Quantity, Relation, and Manner) against the maxim of Quality. The interests of conversation are sometimes better served by avoiding needless exactitude with respect to truth values than by always talking in a strict open futurist manner (Todd’s communication strategy). When exactitude is required, however, the open futurist can translate ordinary talk about the future in a manner consistent with open futurism (Todd’s replacement strategy).
To develop these strategies, Todd compares open futurism to mereological nihilism, an eliminativist view according to which there are no composite objects. More moderately formulated, this is the idea that certain broad classes of things that we ordinarily think of as composite objects (e.g., chairs) do not exist as such. Chairs, such theorists will say, are nothing more than particles arranged chairwise and so talk of “chairs” is just convenient shorthand for talk about such chairwise arrangements. For ordinary communication purposes, mereological nihilists frequently engage in the shorthand talk (communication strategy), but when exactness is required they can replace “chair” talk with something more exact (replacement strategy).
II. Awkward conversations
The first variant of the assertion problem Todd examines stems from a 2012 paper by Christopher Hughes titled “Openness, Privilege, and Omniscience”. Hughes argues that open futurists must—if they consistently apply their theory—engage in various forms of conversational revisionism, such as (in Todd’s words) refraining “from asserting claims we otherwise would have felt perfectly comfortable asserting” or even correcting “others when they assert the relevant claims” (p. 182). Hughes illustrates with an imagined dialog between persons A, B, and C:
A: She’ll marry him.
B: That’s not true.
C (to B): But didn’t you say she could (still) marry him?
B (to C): I did say that, and I still think it’s true. But what A said is that she’ll marry him, and that’s not true. (Todd, p. 182)
Person B in this dialog is an open futurist. The conversation is not a philosophical conversation where exactitude might be required, and yet B corrects A and defends that correction against C by appealing to open futurism. Moreover, in a follow-up dialog, Hughes goes on to argue that open futurists in some sense should correct others in this fashion (cf. Todd, pp. 184–185).
Todd strongly disagrees with this assessment:
[T]here is something untoward about bringing to bear one’s obscure—even if perfectly justified—philosophical theory in an ordinary context[, especially one] in which the falsity of such a theory is taken for granted. … [Open futurism] is a philosopher’s theory. It is, I contend …, a good such theory—but it remains a philosopher’s theory. The problem with B‘s assertion, in this context, is that it reveals B to be a philosopher who has forgotten that he is not in the company of fellow philosophers discussing a theory, and who is … absurdly insistent on inflicting his philosophical theory on his unsuspecting peers. … Hughes seems to be suggesting that we open futurists (the handful of us there are) should … be prepared to correct people in ordinary life who say things that are inconsistent with the truth of our theory. … But no. We don’t. We have better things to do with our time … than to attempt to correct people in the way envisaged here by Hughes. And this is true even if our theory is true. Our theory, even if true, sadly, is just not that important. (pp. 184–185)
The main gist of this reply seems right to me. And Todd aptly supports his reply by drawing a comparison with mereological nihilism aka “ontological eliminativism” (pp. 185–186): Why should the eliminativist correct people in ordinary contexts about the existence of chairs and the like? It’s not worth it. It’s better just to play along with ordinary “chair” talk than to stir up unnecessary ideological controversy and confusion. Like open futurism, mereological nihilism is “not that important” such that eliminativists should normally pressure others to adopt their theory and talk accordingly in ordinary life.
Observant readers may notice that, in the preceding quote, I interpolated “especially one” in the first quoted sentence before “in which the falsity of such a theory is taken for granted”. I made the interpolation because Todd overstates the tension between open futurism and ordinary contexts. Open futurism is much less controversial in ordinary contexts than mereological nihilism is. The latter is highly counterintuitive. Indeed, almost everyone in almost any ordinary context would assume it to be false. Open futurism, however, isn’t particularly counterintuitive, but it is controversial and so can’t be safely taken for granted. In short, it is not that the falsity of open futurism is taken for granted in ordinary contexts, but rather that it’s truth is not taken for granted. And so to bring it up in most ordinary contexts would do nothing but generate needless confusion and controversy. Why open that can of worms if it’s not germane to the conversation? In this respect a better comparison than mereological nihilism is the adage that in polite company one shouldn’t bring up controversial matters of politics or religion. If someone does brings up such matters in a casual context, then most of the time you should just smile, nod, and move on.
Consider Hughes’ dialog again. When A says “She’ll marry him” what should B as an open futurist take A to be saying? In particular, why should B take A to mean anything incompatible with open futurism? Ordinary language, after all, is pervaded by vagueness and “loose use” because it aims to convey important information quickly and easily without unnecessary exactitude or nuance. So most plausibly all A means is “It’s very likely that, in the future, she marries him” rather than “It is now a definitively settled fact that, in the future, she marries him.” (Consider the maxim of Quality: Is A even in a position to know that the future marriage is a definitively settled fact?) So why should B get worked up over A‘s comment when it’s most likely not affirming anything an open futurist like B should disagree with? Understood in this way, B‘s response “That’s not true” is more naturally understood as challenging A‘s idea that the future marriage is likely than the claim that the future marriage is now a definitively settled fact. Viewed in this light, C‘s question to A is more out of place than B‘s initial response, for that B said she “could” marry him doesn’t at all conflict with the idea that it’s not likely to happen. In sum, I think both Hughes and Todd are reading too much into these kinds of dialogs by taking the surface grammar too much at face value. There’s simply no clear problem here for the open futurist in the first place.
III. Asserting that something “will” happen
Over the next several sections (pp. 188–196) Todd develops the idea that most ordinary language claims about the future are really about present intentions and/or tendencies. That’s the sort of information such claims are intended to communicate, and we can replace (strictly false) assertions about future contingents with assertions about those intentions/tendencies instead. I think that is broadly right, and I’ll give some examples in a bit, but first I want to raise an issue that Todd skips over: What does it mean to assert something? I raise the issue because it seems to me that Todd (and other philosophers) have a tendency to conflate assertion with utterance (of a statement).
An utterance is a spoken, written, or signed string of physical symbols (a sentence) that serves an expressive purpose, most commonly that of communicating information. When an utterance is intended both to convey information and the speaker’s endorsement of that information, then that information is asserted. In such cases the sentence used to convey the assertion is called a statement (a declarative sentence), and what is asserted is a proposition, a semantically complete unit of meaning, that is, one capable of bearing a truth value.
Now, return to Hughes’ imagined dialog above. A utters “She’ll marry him”, but what does A assert by that utterance? Todd and Hughes both seem to take the utterance at face value as asserting WILL<She marries him>. I think this is a mistake. I maintain that it’s not obvious what proposition A intends to convey and endorse by that utterance. A few paragraphs back I argued that probably all A is asserting is that it’s very likely (strong tendency) that she marries him in the future. Likewise, I don’t think it’s obvious what B is asserting in his response to A. That depends on what B takes A to be asserting. If B takes A to be merely asserting that there is a strong tendency toward a certain marriage then B‘s denial is most plausibly understood as asserting that it is not the case that there is such a tendency, i.e., that reality is not strongly tending toward that marriage.
If we distinguish what’s uttered from what’s asserted, said, or claimed, then we don’t need to make the distinction that Todd does between what’s asserted and what’s communicated (pp. 190ff). That someone utters a sentence that could (in some context) be used to assert that some future contingent will happen is not the same as asserting that a future contingent will happen. Most often, in ordinary contexts, such utterances are doing something different, such as making assertions about present intentions, tendencies, and the like—things that generally don’t imply that there is already a settled fact of the matter about what will happen.
Todd gives several examples (cf. pp. 190–191, including note 9 on p. 191):
- “His train will arrive at 4:30” = The train is scheduled to arrive at 4:30 and, as far as we know, that’s how the world is tending.
- “I will be at the conference next month” = I have a settled intention to be at the conference next month.
- “That will be the postman at the door” = That’s probably the postman who just knocked. I’ve been expecting him.
- “Oil will float on water” = Oil has a general tendency to float on water.
- “You will immediately tender your resignation” = My desire is that you resign immediately (and if you don’t, you will face negative repercussions).
I wrap up this section with a point that is seldom appreciated but that has been ably documented. In the book The Evolution of Grammar (1994), linguist Joan Bybee et al. explore tense, aspect, and modality across a sampling of 76 languages covering every major group of known languages. One thing they discover is that, in nearly every case, the language used to communicate the future tense stems from one of three main semantic sources: (a) volition/intention (e.g., “I will do X”), (b) tendency/motion (e.g., “it’s going to happen” or “it’s coming to pass”), and (c) obligation/need (e.g., “I have to X”) (see esp. Bybee et al., ch. 7). In other words, the future tense is fundamentally a projection from present conditions (intentions, tendencies, etc.) onto the space of future (causal, epistemic, optative, etc.) possibilities.
IV. The assertion problem for Ockhamists
Above I made the following claim:
Allegedly, when non-open futurists assert “will” and “will not” propositions about the future in general and future contingents in particular, they aren’t engaged in any systematic violation of the maxim of Quality, for they believe that such propositions can be true.
I now wish to challenge the claim that non-open futurists “aren’t engaged in any systematic violation of the maxim of Quality”. There is more to the maxim of Quality than the idea that one shouldn’t assert what one believes to be false (or untrue). The other half of the maxim is that one shouldn’t assert that for which one lacks adequate evidence. Open futurists have been charged by Hughes and others of violating the first half of the maxim, but Ockhamists can be charged with violating the second half. Here I’m drawing on Todd, pp. 196–201, but going beyond what he says in order to press Ockhamists a bit more strongly.
On p. 196, Todd summarizes his position as follows:
I contend that for any assertion that the theory of this book judges to be false, then either (a) that assertion, though false, communicated a truth regarding either (i) a plan or (ii) a worldly tendency, or (b) that assertion was inappropriate.
Regarding the appropriateness or inappropriateness of such assertions, Todd contrasts a case wherein a diehard fan of a pathetically bad sports team declares “My team will win the next game!” and a case wherein someone predicts “It will rain tomorrow” based on recent weather reports. The latter case seems appropriate as a justified projection based on evidence of the worldly tendencies, whereas the former seems like an inappropriate projection based on nothing more than wishful thinking (p. 196). Todd then asks us to consider whether an utterance like
- The world isn’t tending towards a Giants victory next year, but the Giants will win next year.
could ever be appropriate. Todd says no: (1) “is exactly the type of utterance the theory of the open future predicts to always be inappropriate” (p. 196). Someone asserting (1) would be making a claim that is not only not supported by worldly tendencies but that runs contrary to those tendencies. As such, a claim like (1) cannot be appropriately made in the absence of “crystal ball”-like access to the future, which we don’t have.
[I]t is definitive of the doctrine of the open future that there are no crystal balls—more particularly, that there are no truths for such crystal balls in principle to access … Further, even if there were such truths (viz., truths about what is going to happen, against the current tendencies), presumably even the Ockhamist agrees that we do not have access to such truths. Someone asserting [1] is thus always asserting something he could not possibly be in position to know” (p. 196).
This is undoubtedly right as far as it goes. “Will” and “will not” propositions about future events whose chances are believed to be < 0.5 are always unassertable. But what if the chances are believed to be really high but less than 1.0? We can argue that these claims are either unassertable or should not be taken at full face value. Consider (2):
- There is a 95% chance of rain tomorrow; therefore, it will rain tomorrow.
(2) seems like a reasonable thing to say, but there’s some flexibility with respect to how we interpret “it will rain tomorrow”. We can either take it at full face value as saying something like “It is now determinately the case that WILL<It rains tomorrow>” or we can taking as saying something more modest like “I am confidently expecting rain tomorrow”? The latter is clearly assertable, but the former, I contend, is not, for it conflicts with the second half of the maxim of Quality: Don’t assert that for which you lack adequate evidence. However strong we may think the world is tending toward rain tomorrow, if we suppose that tendency to be less than maximal, then any claim that the future is now determinate with respect to rain tomorrow exceeds our evidence, and so violates the maxim. All our evidence supports is that the world is strongly tending toward rain tomorrow. It does not support, at least not without additional metaphysical assumptions (like determinism, retro-closure, or will excluded middle), that rain tomorrow is now a determinate fact about future reality. In other words, unless we assume with the Ockhamist (and the determinist) that the future simply must be exhaustively determinate, evidence of strong but non-deterministic tendencies does not support Ockhamism but rather open futurism. This argument does not by itself show that Ockhamists (and determinists) cannot have good reason for making those additional assumptions, but it does show that the debate turns on those assumptions and not on questions of assertability per se.
In sum, if the Ockhamist tries to use the assertability problem as an independent argument against open futurism, then he is hoist on his own petard for his own view runs into prima facie assertability problems as well. The Ockhamist can of course avoid the assertability problem by not taking ordinary “will” and “will not” claims at full face value, but then so can open futurists. That is, both sides can recognize that most ordinary “will” and “will not” claims are shorthand for claims about present intentions, tendencies, expectations, etc. rather than claims that the future is now determinate with respect to this-or-that future event.