This is part three of eleven in a series responding to the essays in Ben Arbour’s edited volume, Philosophical Essays against Open Theism (Routledge, 2019).
In this post I tackle chapter 3 by Ben Arbour, “A Few Worries About the Systematic Metaphysics of Open Future Open Theism.” Unlike the previous chapter by Visser, this is a long and substantive essay (23 pages).
The essay focuses on open future open theism (OFOT), according to which there simply is no such thing as a unique actual future. Rather, “the future” refers to a branching array of causally possible futures. God thus knows the future as open because the future is open (by God’s own design).
Arbour distinguishes between two different versions of OFOT, depending on whether the version affirms bivalence for PCFCs (propositions concerning future contingents) or not. Bivalence is the thesis that all propositions are either true or, if not true, then false. To deny bivalence is to hold that some propositions either lack any truth-value whatsoever or have a truth-value indeterminate between truth and falsity. Arbour calls the bivalentist version of OFOT “Rhodanian” after yours truly. He calls the non-bivalentist version “Tuggyism” after Dale Tuggy. While he focuses primarily on Rhodanian OFOT (because “that is the version gaining the most traction among open theists,” p. 45), Arbour aims to refute OFOT in general. How so? By arguing that OFOT is driven primarily by the metaphysics of time in a way that is “incompatible with contemporary systematic metaphysics,” particularly the metaphysics of modality (p. 46). He closes by posing some Biblical reflections that, he thinks, rule OFOT out of bounds for Christians even if the metaphysical worries can somehow be resolved.
Before diving in to the meat of the essay, I should note that Arbour incorrectly claims in his introduction that OFOT requires “denying that God possesses any foreknowledge of future contingents whatsoever” (p. 45). This is false because Arbour fails to distinguish between two different kinds of propositions about future contingents (PCFCs). Some such propositions present the future as determinate in some respect. To say, for example, that some event “will” happen or “will not” happen conveys that the future is settled in that respect. Other PCFCs, however, present the future as indeterminate in some respect. To say, for example, that some event “might and might not” occur or “will probably” occur is to say that the occurrence of the event is not yet settled but could still turn out either way. Arbour’s charge that OFOT requires denying any foreknowledge of future contingents whatsoever depends on considering only the first (determinate) class of PCFCs. OFOT is, however, fully compatible with exhaustive divine foreknowledge of indeterminate PCFCs. That is, God knows exhaustively which events are causally contingent and how probable they are.
1. OFOT, Presentism, and Bivalence (pp. 46–49)
In this section Arbour discusses the open futurist semantics I develop in my early papers (2006–2011). In these papers, I regularly discuss Arthur Prior’s distinction between a “Peircean” semantics according to which the future-tense marker “will” carries determinative causal force and an “Ockhamist” semantics according to which “will” carries no causal force at all. I argue that, while neither maps perfectly onto ordinary language usage, the Peircean semantics is overall closer to ordinary language when we restrict attention to predictive assertions (as opposed to other kinds of speech act). While Arbour does contrast both types of semantics, he omits the important qualification I added in the previous sentence. And so he’s not engaging with ordinary language at all, but only with two idealized tense-logical systems.
Arbour also, quite oddly, identifies Peircean talk of what “might and might not” happen with “counterfactual language” (p. 47). This is flatly incorrect because to say, for example, “A sea battle might (and might not) occur tomorrow” is to speak categorically, not counterfactually or, more precisely, in subjunctive conditional terms. Nevertheless, Arbour proceeds to recast the categorical will/might square of opposition in terms of David Lewis’s subjunctive conditional would/might square of opposition (p. 48). This is a sloppy conflation on his part. He also mislabels the bottom row of the square as a “contrary” relation when it should say “subcontrary.” Fortunately, this section is mostly background information that doesn’t materially affect the main argument of his paper.
Before moving on, note 12 on p. 48 deserves mention. After observing correctly that, on a Peircean semantics truth-values can change over time, Arbour notes that “any view which denies the omnitemporality of truth faces an uphill battle” (p. 64). I’m sorry, but no. To suppose that truth is omnitemporal is to refuse to “take tense seriously” and to adopt a thoroughly static or B-theoretical view of reality. There are certainly arguments to be offered in favor of such a worldview, but only sheer metaphysical prejudice (or ignorance) could lead one to suppose that merely to question such a worldview is to face an “uphill battle.” Had Arbour been more careful, he would have restricted his claim to the omnitemporality of tenseless truth. That claim is widely entrenched among philosophers, with open futurists being the major exception.
2. OFOT, Possible Worlds, and the Metaphysics of Modality (pp. 49–52)
This section and the next are the heart of Arbour’s essay. He wants to argue that OFOT lacks (and badly needs) a developed metaphysics of modality, that the rudimentary modal metaphysics it does offer is woefully inadequate, and that, even if we grant OFOT’s metaphysics, “possible worlds Ockhamism” provides an undercutting defeater for OFOT’s claim that future contingency is incompatible with God’s having exhaustive definite foreknowledge. Every point of this argument is mistaken.
Let me begin by noting and rebutting several false assertions that Arbour makes:
- “[A] maximally consistent state of affairs [i.e., a possible world] necessarily includes a complete and total world history” (p. 49).
- No. This is question-begging. As philosopher Amy Seymour shows in her dissertation, a bivalentist version of OFOT can coherently take maximally consistent states of affairs to be tensed possible worlds.
- “[A]ccording to OFOT, there is no such thing as the actual world.” (p. 49)
- Yes and no. If we assume that possible worlds must include a complete world history, then OFOT denies that there is any such thing as an “actual” world, i.e., a unique possible world which is such that it obtains. However, if we drop that assumption (see point 1 immediately above), then OFOT can affirm that there is an “actual” world but no static actual world. In other words, which world is actual changes from moment-to-moment because reality is fundamentally dynamic and open-ended.
- “[O]n OFOT, it is impossible to determine which possible world picks out the actual world.” (pp. 49–50)
- No. Once we drop the gratuitous assumption that possible worlds must include a complete history we can say that whichever possible world corresponds to present reality is the actual world.
- OFOT “renders modal actualism … false” (p. 50).
- No. Modal actualism is the claim that there are no non-actual possibilities. The intuition behind this claim is simply that what is actual = what is real = what exists. If that’s right, then all possibility is included in actuality because we’re talking about real (actual) possibilities. A non-actual possibility is no possibility at all; it’s an impossibility. OFOT can only conflict with modal actualism if one assumes (again) that possible worlds must include a complete history. If that were so then, since OFOT says that there are multiple possible futures, it would follow that some of those futures are possible and yet non-actual. But (again) no OFOT proponent should concede that assumption.
- OFOT entails “the untoward conclusion that God doesn’t know which world he has created” (p. 50).
- No. Arbour is simply assuming that, in creating, God actualizes a complete, determinate history. This is question-begging against OFOT, which proposes, rather, that God knowingly actualizes the initial actual world, one with a built-in open-ended potential for developing in various different ways. As time unfolds and those potencies are actualized through divine and creaturely efforts, God continues to know exactly which world is actual at every stage.
In sum, Arbour consistently gets the modal implications of OFOT wrong because he takes for granted a conceptual framework (that every possible world must include a complete history) that OFOT categorically rejects. It’s no wonder that Arbour pejoratively describes OFOT’s modal perspective as “non-standard,” “deviant,” “aberrant,” “untoward,” and the like (p. 50). He doesn’t get it because he’s trying to cram a square peg into a round hole, so to speak. At one point he does seem to catch on, however: “we are constantly moving from occupying one world to a different world” (p. 52). That’s basically right, but then he immediately rejects this idea on the grounds that it “highly complicates” the metaphysics of persistence, the ontology of time, and modal metaphysics. Even if the complication charge were true, however, it’s not a good reason to reject a dynamic approach to temporal and modal metaphysics. The added complications are warranted by greater accuracy to reality as it is. If reality truly is dynamic and open-ended, then an OFOT approach to time and modality is exactly what we need. The dynamic approach is a feature, not a bug.
3. Is Possible Worlds Ockhamism an Undercutting Defeater for OFOT? (pp. 52–58)
Arbour next tries to show that OFOT’s commitment to the idea that creatures have real “say” about how reality unfolds sows the seeds for OFOT’s own rejection, for it allows us to solve the “dilemma of freedom and foreknowledge” (DFF) by way of “possible worlds Ockhamism” (PWO).
By PWO, Arbour means the dependence of which possible world (i.e., which complete history) is actual on creaturely free choices. By making the choices we do, we in some way bring it about that this world is actual rather than any of the other possible worlds that, had we made other choices, we could have rendered actual instead. PWO differs from standard Ockhamism because the later focuses on temporality not modality. That is, standard Ockahmism focuses on the temporal relation between God’s past foreknowledge and future creaturely free choices: If we had made different free choices from those we did make, then God would have foreknown differently. By abstracting from temporality, PWO works whether God is temporally situated or not. Had we acted otherwise, which world is actual would have been otherwise, and God would have known otherwise.
As Arbour understands DFF, it is based on the “fixity of the past.” But if, as on OFOT, “humans possess actual causal power to determine which possible world obtains,” and since (allegedly) “there is no such thing as an actual world on OFOT,” Arbour thinks it follows that “there is no good reason to deny possible world Ockhamism as a solution to the DFF, even if we affirm the … fixity of the past” (p. 55, emphasis his).
Arbour imagines OFOT proponents objecting that there are good reasons to reject PWO because it “fails to take seriously the fundamentality of time” (p. 55). In response, Arbour protests that this makes time “more fundamental to systematic metaphysics than modality” (p. 55). Not only is this supposedly against the “majority opinion” of systematic metaphysicians, but it “requires the addition of time as a fifth category of systematic metaphysics alongside properties, causation, laws of nature, and modality” (p. 55). In short, Arbour accuses OFOT of undermotivated metaphysical revisionism.
What should we make of this argument? Not much. First, as explained above, it is not the case that “there is no such thing as an actual world on OFOT”—rather, actuality is dynamic on OFOT: which world is actual changes from moment to moment because reality as a whole changes from moment to moment. Second, while I won’t argue the point in detail here (but see this post for some relevant argumentation and section 3.2 of my book for the fine details), there is a good reason to reject PWO. Namely, it’s metaphysically incoherent. Here’s the short argument: According to PWO there is both (a) a complete, true story of the future and (b) future contingency. Given (b), the contents of the story are still up-in-the-air; the story is still being written. Given (a), however, the story is already written; all of its content is finalized. This is incoherent because the content of a story cannot both be finalized and still up-in-the-air. A story can’t both be already finished and not yet finished. Third, what’s the problem with taking time to be a fundamental aspect of reality? Surely not that doing so goes against “majority opinion,” as if that even matters in philosophy. In addition, at least two of the supposedly fundamental categories of systematic metaphysics that Arbour lists, causation and laws of nature, are arguably fundamentally temporal. Causation is standardly understood in terms of the bringing about of events, and is widely thought to require that causes be temporally prior to their effects. Likewise, laws of nature describe how natural systems change over time.
But Arbour’s not done yet! He says that “OFOT must choose between two horns of a dilemma” (p. 56). The dilemma is not well articulated by Arbour. From what I can gather it seems to be roughly this: However the OFOT proponent develops his theory of modality, either (A) the DFF fails, or (B) God cannot entirely know the present. Frankly, I’m scratching my head to figure out what Arbour’s reasoning is behind this so-called “dilemma,” because neither horn seems remotely plausibly to me as an OFOT proponent. His argument for (A) seems to conflate logical necessity with causal necessity. Thus, he argues that “the type of necessity that distributes from ‘Necessarily, (If God knows x, then x)’ to the corresponding conditional ‘If God necessarily knows x, then necessarily x‘ gives us a false antecedent” (p. 56). I presume the false antecedent here is the idea that God necessarily knows x, where x refers to the occurrence of a causally contingent event, like someone’s freely refraining from drinking orange juice. But this is irrelevant because it’s not how arguments for the incompatibility of EDF and future contingency actually work. They don’t concern logical necessity but causal necessity or, if you will, now-unpreventability. If God unpreventably knows that x, then x cannot be causally contingent, because if x were causally contingent then events could transpire otherwise than God unpreventably knows, which would amount to preventing the unpreventable. In short, I don’t think Arbour understands the DFF very well. As for (B), Arbour contends that to circumvent possible worlds Ockhamism while preserving OFOT, proponents must adopt a modal metaphysics that prevents God from knowing “accidentally necessary truths about the present that involve free will” (p. 58). So far as I can see, however, Arbour gives no argument for this claim. Moreover, the claim is implausible on its face. Propositions about the present that correspond to the way present reality actually is are true and can be known by God simply by inspecting present reality.
4. The Modality of OFOT’s Metaphysics of Modality (pp. 58–60)
Next, Arbour contends that OFOT is problematic because it depends on a temporal ontology like presentism that is metaphysically contingent and, therefore, the “limitations” on divine knowledge that OFOT affirms are incompatible with perfect being theology (PBT), which requires that God’s omniscience be non-contingent.
This is a bad objection against OFOT on several levels. First, why should we grant that presentism is metaphysically contingent? I’m a presentist, and I believe it’s metaphysically necessary. I don’t think it could possibly be the case that non-present realities exist alongside present realities. Arbour provides no argument to the contrary other than to cite one author who (he claims) affirms presentism as metaphysically contingent. Um, okay. Second, it’s wrong to say that God’s knowledge is “limited” on OFOT if one means to imply that there are truths that God doesn’t know, for OFOT affirms that God is unqualifiedly omniscient. God knows all there is to know. Third, even if presentism is metaphysically contingent and has the consequence that God cannot know determinate (will/will not) PCFCs, this doesn’t conflict with PBT at all. All it means is that God has chosen to create a world with a temporal ontology that does not support certain kinds of information about the future. If God had chosen to create a block universe instead, then that information would have been available but then (arguably) would no longer have concerned future contingencies (because from God’s standpoint all of creation history would be objectively present). Either way, it seems there would be no true determinate (will/will not) PCFCs for God to know no matter which temporal ontology God had decided to instantiate. More fundamentally, though, it’s logic, not temporal ontology, that rules out infallible knowledge of how future contingents will turn out. If the information is infallibly knowable, then it is non-preventable. If it depends on a future contingency, then it is preventable. What is non-preventable cannot also be preventable, for that would be a logical contradiction. Therefore, such knowledge is logically impossible.
5. A Theological Defeater for OFOT? (pp. 60–62)
Arbour closes his essay with a biblical argument against open theism in general, not just OFOT:
- 2 Peter 3:12 tells us that the time of Christ’s return is a future contingent (b/c we can “speed” or “hasten” it).
- Matt. 24:36 tells us that the Father knows the time of Christ’s return.
- The Bible is accurate with regard to such matters (i.e., the time of Christ’s return is a future contingent and the Father knows when it will happen).
- Therefore, it is possible for God to know determinate (will/will not) PCFCs. (From 1–3)
This is a clever argument. If Arbour’s interpretation of both biblical passages is correct, and if the Bible is accurate with respect to what those passages are teaching, then it seems we have a clear biblical counterexample against open theism and any position that contends that God cannot know determinate (will/will not) PCFCs.
In response, I’d like to suggest that Arbour has over-interpreted Matt. 24:36. In the first place, Jesus is speaking as the incarnate Christ. As such, he willingly operates under human limitations (Phil. 2:5–9), submitting to the authority of the Father in all things and doing only what the Father authorizes him to do (John 5:19). Moreover, it is not for us [humans] “to know the times or dates the Father has set by his own authority” (Acts 1:7). In other words, it’s the Father’s prerogative, and no one else’s, to decide when to initiate Christ’s second coming. Thus, speaking under the limitations of the incarnation, when Jesus says that only the Father knows the time of Christ’s return, he’s not saying that the Father knows the definite time of Christ’s return. Rather, he’s saying that if anyone knows, the Father does (because it’s up to the Father to make that call). So understood, Matt. 24:36 meshes perfectly with 2 Peter 3:12. The latter verse tells us that the time of Christ’s return has not yet been decided. The Father hasn’t yet made the call.
Concluding Thoughts
Overall, I’m not much impressed with Arbour’s essay contribution to his own edited volume. His arguments fail to strike home mainly because he insists on taking for granted anti-open futurist ways of thinking about modality and metaphysics simply because they are more popular and/or more familiar to him.