This is part six 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 6 by Katherin Rogers, “Foreknowledge, Freedom, and Vicious Circles: Anselm vs. Open Theism” (pp. 93–109). Rogers is a well-respected philosopher of religion and a huge fan of Anselm. Most of her philosophical work centers around explaining and defending Anselm’s positions, or what she takes those positions to be.
In what follows, I start with (1) some reflections on a previous exchange that I had with Rogers. I then dive into the current essay and discuss (2) her misunderstandings of open theism and of presentism, (3) her account of creaturely freedom (which I generally like), and (4) her “solution” to the freedom/foreknowledge problem (which fails, IMO).
1. Reflections on an earlier exchange with Rogers
I’ve interacted with Rogers before in print. My essay “Foreknowledge and Fatalism: Why Divine Timelessness Doesn’t Help” engages and critiques Rogers’ 2008 book Anselm on Freedom (Oxford). In her reply to my essay, she admits that she didn’t really understand my criticisms and so she concludes that I misunderstood her and that I begged the question against her position. Now, I’ll grant that my argument was not as clear as it could have been. In my recent Open Theism book (Cambridge, 2024), I significantly expand and improve the argument. But her response pretty much ignores the first two-thirds of my essay in which I develop the core argument. She focuses, instead, on the last third of the essay in which I apply the lessons from first two-thirds and leverage them against divine timelessness. So, her charge that I beg the question against divine timelessness and “isotemporalism”—her preferred term for a linear block ontology of time—is, as I see it, off base. I did not assume that those positions were false. I argued against them. By essentially ignoring the constructive part of my argument, Rogers convinced herself that I had merely assumed what I was supposed to be proving.
In any case, Rogers’ essay in the Arbour book, which was written after her exchange with me, doesn’t interact with my work at all, despite the fact that it would have provided a recent and relevant foil for her essay.
2. Rogers’ misunderstandings of open theism and of presentism
One of the reasons Rogers is gung-ho in championing isotemporalism is that she doesn’t see any viable alternative that preserves a robust libertarian view of creaturely freedom. She agrees with most open theists that God’s knowledge of future contingents must be grounded in reality, specifically in actual creaturely choices (p. 93). This rules out both the simple foreknowledge view and Molinism. According to the former, future events don’t yet exist and so aren’t yet available to ground any truths about the future. According to the latter, God’s knowledge of future contingents is derivable from God’s middle knowledge and creature decree, independently of actual creaturely choices. With those views out of the running, the metaphysical options for libertarian free-will theists reduce to open theism and some kind of timeless knowledge view. Regarding the latter, the main candidate positions are those of Boethius and Anselm. The problem with Boethius’ version is that he affirmed ontological presentism, and so cannot ground God’s timeless knowledge of the future in actual creaturely choices. So that leaves Anselm and open theism. I’ll say more about Rogers’ Anselmian position in section (4). For now, I want to look at why she rejects open theism.
In her essay Rogers seems to have two main reasons for rejecting open theism: (a) it diminishes God and, thus, conflicts with Anselmian perfect being theology (PBT), and (b) it entails presentism, which she finds completely implausible as an ontology of time. As we will see, these two reasons are closely connected in Rogers’ mind. Roughly stated, she rejects open theism in large part because she thinks it entails presentism, something she utterly rejects.
Regarding (a), she says the following of open theism:
- “open theism offers a God who is quite diminutive—He doesn’t even know the future!” (p. 93)
- “[Anselmian isotemporalism] allows for a much more adequate picture of the divinity than the open theist position where God lives from barely existing moment to barely existing moment.” (p. 96)
- “[God] would have absolute power over the course of things if He had not chosen to create agents with libertarian freedom, but He would have far, far less power if He had created libertarian free agents, but did not know the future and could act only upon the present instant.” (p. 100)
- “It is, perhaps, easier to wrap our minds around the open theist’s God, who is much more like us than is Anselm’s God. But any easily conceivable God is not likely to fit the description of ‘that than which no greater can be deceived.'” (p. 107)
The first of these claims, that the God of open theism “doesn’t even know the future,” is completely false. I’ve replied to it many times in published papers and in blog posts. Quite simply put, the best versions of open theism affirm unequivocally that God knows everything there is to know. He has perfectly accurate knowledge of all of reality. The issue that divides open theists from non-open theists like Rogers is not whether God knows the future, but what “the future” refers to. Does it, as Rogers supposes, refer to a unique actual future? Or does it, as open theists propose, refer to a branching array of possible futures because “the future” is objectively open-ended?
The last of these claims, that a more easily conceivable God is more likely to violate PBT, is silly. In the first place, open theists like myself have argued for years that open theism is perfectly compatible with PBT once we jettison certain dubious assumptions about what divine “greatness” must look like. In the second place, while PBT entails that finite beings like us should not be able to fully grasp the divine nature—a conclusion every serious open theist I know of would endorse—any model of God must strive to render God to some significant degree conceivable by us. A model that renders God completely inconceivable by us is useless as a model.
The second and third claims both reject open theism because of its alleged ties to presentism. If open theism is true, then (Rogers supposes) presentism is true, and if presentism is true, then (Rogers supposes) God is diminished because He “lives from barely existing moment to barely existing moment” and “can only act upon the present moment.” Both of these suppositions are wrong.
First, open theism does not entail presentism, the view that there are no past or future realities and thus that reality is coextensive with the temporal present. While it is probably the case that most open theists are presentists because that is the most popular and plausible dynamic model of time, there’s nothing in open theism as such that requires a presentist ontology. One could instead hold to a growing block model (the past and present are equally real, but the future does not yet exist) or to a branching block ontology (the past, present, and all causally possible futures are equally real).
Second, Rogers badly misunderstands presentism. Consider the following quotes:
- “The most common alternative to isotemporalism, and the view open theists defend, is presentism: all that exists is the present moment. And the present moment is simply the extensionless point at which the non-existent future becomes the non-existent past.” (p. 96)
- “Could the thought be that presentism captures change better because it takes account of the flow of events from the future through the present to the past? But it doesn’t. On presentism the future is non-existent, so no events are there to flow from it. The past is non-existent, so no events flow into it. The present is where two non-existents meet.” (p. 99)
- “On presentism there is no continuity between one’s past and one’s present, since one’s past is absolutely non-existent.” (p. 99)
- “Presentism would have it that we exist at the unextended point at which the non-existent future becomes the non-existent past.” (p. 99)
The way Rogers describes presentism suggests that she bases her understanding of it on a moving spotlight model according to which the present is the point at which future events “flow into” or “become” the past and then she conceptually lops off all past and future events to arrive at a present that is merely an “extensionless point” where “two non-existents meet.” So conceived, it’s easy to see why she thinks presentism makes no sense, metaphysically or theologically. If time involves a flow of events, then presentism seemingly gets rid of anything that can flow—it’s a flow of nothing to nothing. And it seems to imply that a temporally situated God lives precariously on a knife-edge of nothingness “from barely existing moment to barely existing moment.”
But this is not the right way to conceptualize presentism. The presentist does not start from a conception of time as spread out over past, present, and future and then strip it down to just the present. The presentist, rather, starts from the idea—seemingly attested by our everyday lived experience—that reality as a whole is dynamic. This doesn’t mean that every bit of reality is changing—some things are necessary and so stay the same—but that the whole is changing. That whole does not take the past and future to be utter nothingness. After all, it is presently the case that there was a past and presently the case that there will be a future. Rather, the present whole preserves the past in its “memory” and anticipates the future through the causal powers, propensities, and intentions of fully real concrete beings. In short, presentism does not view the present as a vanishingly thin boundary suspended between two nothings (pace Rogers) but as the metaphysically thick whole of reality, a reality that is actively becoming other than it currently is.
The role of God on a presentist metaphysics is that of a being who perfectly knows the whole of reality and is actively involved in guiding the becoming of that whole. This is not a “diminutive” or diminished conception of God. The main difference between an open theistic God on presentism and Rogers’ atemporal Anselmian God concerns not whether God is omnipotent, omniscient, omnibenevolent, etc., but how those attributes function on a fundamentally dynamic vs. a fundamentally static ontology.
3. Rogers’ libertarian account of creaturely freedom
I generally like Rogers’ account of creaturely freedom. It’s close, in fact, to the view that I hold. She calls it parsimonious agent causation (p. 94). According to this view, free creatures are ultimate or per se causes of their own free choices. This is an anti-determinist condition which means that, when it comes to free choices, the causal buck stops with the creaturely agent. There is no explanatorily prior causal chain necessitating which option the creature chooses. Moreover, like myself, Rogers seems to be a restrictivist about libertarian freedom. That is, free choices can only be made under certain conditions, which do not always obtain. For Rogers creaturely free choices only occur in a “torn condition” (TC), where we are “struggling to pursue two [or more incompatible] … motives” (p. 94). Rogers affirms agent causation in that “it is up to the created agent, and only to that agent, to opt for A over B or B over A” (p. 95). But her account is also “parsimonious” in that the choice is a “thin event” that “does not constitute some new entity added to the sum of what there is in the universe” and “does not introduce some new species of causality” (p. 95). Rather, it is simply that the agent’s desire for B, say, becomes stronger than the conflicting desire for A to the point that it eventually excludes the desire for A.
Now, I generally like this account, but I have reservations about the “parsimonious” part of it. I think it’s too parsimonious. The way Rogers describes an agent’s choosing, or “per-willing” as she calls it (p. 94), makes it sound like it’s just something that happens within the agent when one competing desire comes to dominate the other desires: “all that S does is continue desiring that desire [B] to the point where the desire for A is no longer viable” (p. 95, emphasis mine). This makes it sound like the agent’s (per-)will is merely a passive arena where competing desires duke it out until one of them wins. But then it seems to follow that the agent, qua agent, is not an ultimate per se cause. What I would say, in contrast to Rogers, is that in making a free choice an agent actively resolves the torn condition by prioritizing his reasons for, say, B over his reasons for A. That is, the agent settles which set of reasons or desires comes out on top.
In addition, the main reason she stresses that free choices are “thin events” is because she affirms, with Anselm, that God directly causes “all that exists” (p. 95). If creaturely choices had positive ontological status, then it would follow that God directly causes all creaturely choices, which would prevent creatures from being per se causes. Now, I’m fine with the idea that choices are thin events if we understand that to mean that they are basic events, i.e., events that are not constituted by other, smaller events. WWII, for example, is a “thick” event. It is constituted by millions of smaller events, including the Japanese attack on Pearl Harbor, the Allied ‘D-Day’ assault on Normandy, the bombings of Hiroshima and Nagasaki, etc. My problem is with the idea that God “causes all that exists, but He does not cause all that happens” (p. 95). Given Rogers’ isotemporalist ontology, God eternally causes every moment of time, including the before and after states of every free choice. But then it’s not clear how God can avoid causing “all that happens.” If God eternally causes the after state of my free choices, then how can I be free while in the before state to have chosen otherwise? To her credit, Rogers recognizes that there is a prima facie problem here: “[H]ow can a created agent choose with aseity in the universe of classical theism in which everything with any sort of ontological status at all is caused to exist and maintained in existence from moment to moment by God?” (p. 94) In the next section I argue that she has no good answer to this question.
4. Rogers “solution” to the freedom/foreknowledge problem
I put “solution” in scare-quotes because I don’t think Rogers’ proposal comes anywhere close to being successful. But, judge for yourself. Here’s her fullest statement, keeping the familiar distinction between primary (God-to-creation) and secondary (creation-to-creation) causation in view:
[H]ere is the picture regarding S choosing B at t2. From the perspective of secondary causation we can say that S’s desiring both A and B produces the torn condition, and then S desires B to the point of ousting the desire for A, and that constitutes the choice. S is a secondary cause. He does not bring anything into being ex nihilo. And here is the picture regarding S choosing B at t2 from the perspective of primary causation. First, at t1, God, as primary cause, is causing S in TC [a torn condition] with respect to A and B. Then, at t2, God, as primary cause, is causing the continuation of the desire for B, but God knows that S chooses B at t2, not because S exercises some sort of causal force or power over God. All the of the actual causal force is from the side of God. It is the desires for A and for B that provide the motive power for S’s choosing A or choosing B. S, if you will, “channels” that power by per-willing B. But God is not the passive recipient of a piece of information sent by S. Rather God knows what S does because God knows what He, God, is doing. (pp. 96–97, emphasis mine)
There are a lot of problems with this account.
First, note that I’ve underscored twice the phrase “from the perspective of.” This is a key indicator that her “solution,” rather than reconciling creaturely freedom and God’s exhaustively definite knowledge of the future, actually dodges the problem by engaging in what I like to call “perspective toggling.” Whenever creaturely freedom is in question, we toggle to the secondary causal perspective, look horizontally along the time series, and conceive of the future as causally open—it is really possible for S at t1 to choose A and it is really possible for S at t1 to choose B. From this perspective the future looks like it branches into multiple causally possible futures. But, whenever God’s exhaustively definite knowledge of the future is in question, we toggle to the primary causal perspective and imagine God looking vertically down upon (and causing) the entire temporal series so that He can “see” exactly how things play out. From this perspective, there is no branching, just a single, linear temporal series. In contrast with this perspective toggling approach, a true reconciliation would consider both creaturely freedom and God’s exhaustively definite knowledge of the future from a single, fully objective (i.e., God’s eye) perspective. The problem with Rogers’ approach is that her God’s eye perspective is the primary causal one dictated by her isotemporalist ontology. As already noted, from that perspective creaturely freedom completely drops out of the picture because there is no branching.
Second, Rogers claims that S “does not bring anything into being ex nihilo.” This has to be false if creaturely freedom is a se, as she herself insists. What a se creaturely freedom brings into being is information. Specifically, when S freely choses between A and B, S changes the explanatorily prior (before the choice) information state from one in which it is an open question what S shall choose to the explanatorily posterior (after the choice) information state in which it is a settled fact that S has chosen A [B]. As she herself insists a few pages earlier, “truth about an a se choice, and hence knowledge about an a se choice, can be grounded in (or dependent upon) only the actual choice itself” (p. 93, emphasis added). The reason truth and knowledge about an a se choice can be grounded “only” in the actual choice is because that event creates the information that the relevant instances of truth and knowledge require.
Third, despite what I’ve just argued about how a se choices create information, in the excerpted quote Rogers explicitly denies this. She insists that “God is not the passive recipient of a piece of information sent by S.” But this must be wrong if S’s choice is truly a se. If the truth that S chooses B at t2 and God’s knowledge of that truth are grounded in the actual choice itself, and if S’s choice creates the relevant information, then God is a passive recipient of information from S. This need not be understood as S’s exercising “some sort of causal force or power over God,” but it does entail that God is receptive toward creation. God does not know what S has chosen because God knows merely “what He, God, is doing.” Rather, God knows what S has chosen because of something S did.
In sum, Rogers’ positive “solution” to the freedom / foreknowledge problem is internally inconsistent on several levels. The reason (I surmise) why she doesn’t notice the inconsistency is because she engages in systematic perspective toggling, switching stances whenever the inconsistency begins to come into view. It’s a form of self-deception. Perspectivally relativizing two views prevents explicit contradiction between them at the expense of preventing any synthetic harmonization of them into a single, coherent whole. The self-deception consists in thinking that one has positively harmonized (by uniting into a synthetic whole) what one has merely negatively harmonized (by preventing explicit contradiction through perspectival relativization).
5. Concluding thoughts
Summing up, Rogers’ criticisms of open theism and of presentism seriously miss the mark. From a libertarian perspective, her analysis of human freedom is mostly good, but is too “parsimonious.” Finally, her positive response to the freedom / foreknowledge problem is multiply incoherent but masked by her perspective toggling. There’s more to Rogers’ essay than I have covered. She responds to various open theistic objections and defends her view against a vicious circularity charge (both unsuccessfully, in my opinion). In the course of her defense she constructs a rather elaborate scenario (pp. 103ff.) involving a time-traveling alien pseudo-Jesus named “TJ.” I won’t delve into this part of her essay. As a general rule of thumb, if your best way of making sense of your metaphysical position requires recherche time-travel scenarios, then your view is probably wrong or, at the very least, “sus.”