This is part five 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 5 by Paul Helm, “The ‘Openness’ in Compatibilism” (pp. 80–92). Helm is a well-respected philosopher of religion and a long-time staunch defender of theistic determinism, the idea that God is the ultimate sufficient cause of everything that ever happens.
In this essay, Helm argues that open theism is poorly motivated, incurs unacceptable theological costs, and that “compatibilism” (by which Helm means theistic determinism—on p. 85 he refers to “the determinism of compatibilism”) can provide more than enough openness to “satisfy all who desire their theology to be ‘open'” (p. 80).
In this essay review, I argue that Helm’s case against open theism has the same “flimsiness” (p. 80) that he thinks the case for open theism does. I also argue that his case for “compatibilism” is quite flimsy as well and, moreover, ignores the serious theological costs of theistic determinism.
Helm on why open theism is poorly motivated (pp. 80–83)
In this section Helm discusses one of my earliest papers, “The Philosophical Case for Open Theism.” In that 2007 paper, I present four very brief arguments in favor of the “future contingency thesis,” the claim that there are future contingents, i.e., causally indeterministic future event-types. These reasons are (1) that in deliberation it seems to us that our options are open-ended and that it is generally up to us (and nothing else) to determine which option comes to pass; (2) that quantum mechanics suggests that there is genuine causal indeterminism in nature; (3) that commonly and deeply-held intuitions about moral responsibility suggest that it traces back to what Robert Kane calls “self-forming actions,” which are inherently indeterministic; and (4) that the problem of evil poses a significantly harder challenge for theistic determinism than it does for libertarian free-will theism. Helm considers each argument in turn.
With respect to (1), the “argument from deliberation,” Helm correctly notes that I only claim that this argument offers prima facie evidence of future contingency, but he immediately challenges even that modest claim: “it is prima facie evidence of future contingency only if we conceive of our situation as one in which we possess unconditional power to choose A or to choose B” (p. 81, emphasis mine). Helm apparently thinks this conception is wrong and that, instead, “our experience is of the kind of being able to do A if we want to or B if we want to, or neither if that is what we want” (p. 81, emphasis mine).
In response to Helm, I think he’s doing very poor phenomenology here. When I actively deliberate between two incompatible options—with Helm, let’s call them A and B—I’m not entertaining competing conditionals (should I do A if I want to or do B if I want to?) but categoricals (should I do A or do B?). Moreover, I don’t “conceive” of my deliberation as a situation in which I have “unconditional power to choose A or to choose B.” That way of framing things abstracts from the concrete deliberative struggle and re-presents it at a meta-level in a purely conceptual manner. In contrast, when I deliberate, rather than “conceiving” of my deliberation as a “situation” of some sort, I perceive my own divided agency and I feel my own power to choose between A and B. Nor do I feel my choice as something that happens to me when one set of reasons becomes clearly dominant. Rather, I feel myself making that set of reasons dominant. (If my reasons for one option or another had been clearly dominant all along, then there wouldn’t have been anything to deliberate about.) Now, this feeling might be an illusion. Perhaps, unknown to me, there are subconscious psychological forces that will eventually cause me to settle on A over against B. And perhaps, when that happens, I will feel as if I played a decisive role in settling on A when it was really a foregone conclusion all along. That’s why this argument is only prima facie evidence of indeterministic freedom. But it’s evidence, nonetheless, and evidence that Helm casually dismisses because he does not approach deliberation with phenomenological sensitivity.
With respect to (3), the “argument from intuition,” Helm claims that my argument is “question-begging” because “compatibilists” can make sense of self-forming actions too. [Note, when Helm speaks of “compatibilism,” he really means soft determinism. Compatibilism proper is silent on the question of whether determinism is true. Helm is not silent on that point. He is an outspoken determinist.] On the soft determinist account, moral responsibility depends on whether the causal chains determining one’s choices and actions run through one’s psyche in the right sort of way, such that the proximate cause of one’s moral choices is, in some relevant sense, the agent himself. For comparison, a compatibilist who is not a determinist would typically say that moral responsibility depends on whether one’s choices and actions emerge from one’s psyche in the right sort of way. Here, only the proximate relation of choices to psyche is relevant. Either way, only proximate causes and relations matter. In contrast, the moral responsibility argument for libertarianism is based on the intuition that the ultimate, not merely proximate, cause of one’s moral choices must be the agent himself. This intuition conflicts with determinism because, if determinism were true, then the agent could at most be a proximate cause of his choices—whether he chooses moral good or moral evil on some occasion would ultimately be decided by determining causal factors laid down independently of the agent. A common libertarian worry is that, if this were true, then everyone would have a perfect moral alibi/excuse—the causal system made me do it!—absolving them of any real moral responsibility.
Now, this seems like a compelling argument to me, and not just to me. Most of the early Church Fathers of Christianity pressed exactly that sort of argument in favor of libertarian creaturely freedom. But Helm claims this is question-begging. Is it? Well, begging the question is a dialectical fallacy that occurs when one’s argument assumes, without independent support, a premise that would in all likelihood not be accepted by the audience the argument is aimed at. The relevant question, then, is what audience is my paper most plausibly aimed at on a charitable reading? If my argument were aimed at refuting hardened determinists like Helm, I would need to have started from premises they accept and show that those premises lead to conclusions that they would have difficulty accepting. But that’s not how my argument is presented. It’s presented as part of a cumulative case for causal indeterminism / future contingency and not as a refutation of determinism. My appeal to the prima facie evidence of deliberation, for example, could not plausibly be construed as a refutation. As such, my argument is most plausibly aimed at those who are neutral / on the fence with respect to determinism. I’m suggesting, for example, that if they reflect on their own moral intuitions, they will see that we need some degree of ultimate and not merely proximate responsibility. Some (like Helm) will insist that they don’t have such intuitions, so my argument surely won’t convince them, but it’s not question-begging if my argument wasn’t aimed at such people in the first place.
With respect to (2), the argument from quantum-mechanics, my contention was simply that, according to the most “widely accepted” interpretation of quantum mechanics, there is causal indeterminism at the micro-physical level. This is obviously not a “refute determinism” type of argument. It’s part of a cumulative case for indeterminism. Helm responds in two ways. First, he suggests that quantum indeterminism is merely apparent. From the all-determining God’s eye perspective, there is no real indeterminism. Second, he says that quantum indeterminism is irrelevant to creaturely libertarian freedom unless there’s a way to scale such indeterminism up from the micro-level to the macro-level (p. 82). I grant both of these points. With respect to the first, however, I didn’t present this argument as an empirical proof, but merely as evidence pointing toward indeterminism that determinists, for their part, need to explain away. With respect to the second, Helm’s point is irrelevant because I’m merely arguing here for indeterminism / future contingency, not creaturely libertarian freedom as such.
Helm closes his brief discussion of (2) with the puzzling line that “the argument has little to do with logically indeterminate futures” (p. 82, emphasis mine). Indeed, he speaks of “logical indeterminacy” rather frequently in this essay in connection with open theism. Frankly, however, I don’t understand what Helm means by this phrase. Nor do I understand its relevance. And he doesn’t explain it. Arguments (1)–(4) have to do with causal indeterminism, not logical indeterminism (whatever that could mean). And they have to do with indeterminism, not indeterminacy. The only plausible sense I can give to the phrase “logical indeterminacy” is that of logical undecidability—as Kurt Gödel proved, in any axiomatic system that includes basic arithmetic, one can formulate theorems that cannot be proven true or false in that system. But that sort of indeterminacy is, as far as I can tell, irrelevant in the context of Helm’s paper.
Finally, with respect to (4), the argument that the problem of evil is significantly worse for theistic determinism than for free-will theism (and esp. open theism), Helm’s response completely misses the point. Basically, he says that, according to theistic determinism, God has an “overall good” plan for creation, which plan “may” include notorious evils (p. 82). In contrast, he thinks open theists “say little about God’s plans” (p. 83), but because God can (on open theism) prevent all evils, God must have some sort of reason for not doing so. Hence, theistic determinism and open theism are in the same boat when it comes to the problem of evil.
What Helm misses is that there’s a significant difference between God’s merely permitting (and thus not preventing) evil E to occur and God’s specifically intending and ultimately causing E to occur. This difference is particularly salient with respect to moral evils or sins. Open theists can and typically do hold that no moral evils are specifically intended or caused by God. All moral evils are objectively tragic. The world would have been overall better had they not occurred. Their actual occurrence isn’t necessary for any “greater good.” It is only the possibility of their occurrence that is necessary—because it is good, all things considered, that creatures have moral freedom, and thus the ability to misuse that freedom by sinning. The theistic determinist can’t say that. He is committed to the highly counterintuitive claim that a perfectly good God specifically intends and ultimately causes every single instance of moral evil that ever occurs. This makes God, quite literally, the author of sin. Moreover, it generates an incoherence between God’s decretive “sovereign will” and His prescriptive “moral will” with God commanding us not to sin (as in the Ten Commandments) while at the same time setting up a causal system that ensures we commit the very sins God tells us not to do. Finally, theistic determinists are committed to defending the highly implausible claim that every single evil that ever occurs is somehow necessary for the “greater good” God is aiming at. But, really??? Every single evil??? It seems much more plausible to me that whatever those greater goods might be, if God causally micro-manages all the details of history (as He does on theistic determinism), then He could easily have achieved all of those greater goods, or something comparably wonderful, without nearly as many (as as severe) evils as we actually witness. Thus, even on theistic determinism, it seems a priori highly plausible that most actual evils, including all or nearly all moral evils, should be considered gratuitous and thus not necessary for any greater good. This being so, theistic determinism calls the goodness of God seriously into question because it strongly suggests that God has specifically caused a great many evils for no good reason. Open theism, in contrast, does not entail that any specific evil is necessary for a greater good. All it entails is that the possibility of the types of evil that we witness is necessary for a greater good. That claim, too, may seem implausible to some, but it’s not nearly as implausible as the idea that every actual evil is necessary for a greater good. For these reasons, Helm’s claim that theistic determinism and open theism are in the same boat with respect to the problem of evil strikes me as morally obtuse.
Summing up this section, Helm thinks that my reasons for affirming future contingency are “distinctly weak” (p. 83) and “flimsy” (p. 80). For my part, I think his dismissals of my reasons are distinctly weak. It seems to me that his ideological blinders prevent him from honestly considering how theistic determinism looks from the outside.
Helm on the costs of libertarian freedom (pp. 83–84)
As Helm sees it, free-will theists “are prepared to go to extraordinary … lengths” to defend creaturely libertarian freedom (p. 83). He proffers Molinism as an prime example of this claim: It’s apparatus of middle knowledge is widely thought by many philosophers and theologians (including myself) to be extremely problematic. Helm then shifts to open theism, which, he says, is “even more” problematic than Molinism. How so? Helm mentions two costs, though they are really the same point:
- On open theism God’s “omniscience has to be radically rewritten, in its scope if not its definition.” At the very least, the principle of bivalence must be rejected (p. 83).
- On open theism “God either … cannot know all the truth-values of all propositions, or [has] foregone such knowledge. … That is, God … forebears to know some future facts, and he cannot know, that is, he cannot identify, what he has chosen not to know” (p. 84).
Ironically, the very papers of mine (this one and this one) that Helm himself cites in the first section of his paper reveal both of these charges to be straw men. In both of those papers I point out that open theism is compatible with God’s unqualified omniscience, that is, with God’s infallibly knowing all truths and doing so in a manner compatible with bivalence. Nothing about the fact or scope of God’s omniscience has to change. The only change required by open theism is that the content of God’s omniscience does not specify a unique actual future. And it’s not like there are facts specifying a unique actual future that God has chosen not to know. Rather, God has chosen to create an open-ended world in which no such facts exist.
So, Helm’s discussion of the “costs” of libertarian freedom is pathetic and lazy. He ought to have known better. Moreover, the way he presents things, he seems to think that theistic determinism is all benefits and no costs whereas free-will theism is all costs and no benefits. If he were more honest in his presentation he would balance things out with some acknowledgement of the costs of theistic determinism and/or the benefits of free-will theism. At the very least, there are prima facie problems with theistic determinism that are prima facie mitigated or eliminated by various forms of free-will theism.
Helm on why theistic soft determinism gives us all the “openness” we need (pp. 84–90)
The rest of Helm’s paper is mainly concerned with arguing that “compatibilism can furnish us with all the openness that we need” (p. 84). As noted above, however, by “compatibilism” Helm really means soft determinism. Hence, in the interest of clarity and accuracy, I will describe Helm’s position in those terms.
So, what sort of “openness” does soft determinism entail? According to Helm, it requires epistemic openness on the part of creaturely agents: “If the action is to be spontaneous, free from external coercion, as compatibilists affirm, then … [i]t is not until we have finally and irrevocably made up our minds that we ourselves can be said to know what we have decided to do” (pp. 84–85).
That deliberation must occur under conditions of epistemic openness concerning what is to be decided upon seems obvious: To already know what you’re going to decide is to already have decided and thus to no longer be deliberating about it. But when Helm describes soft determinist decision-making as “spontaneous” and “free from external coercion,” he’s dissembling a bit. Given determinism, the outcome of every creaturely choice is objectively a foregone conclusion from the beginning of creation. In that sense, it’s not spontaneous, but forced. And it’s not free from external coercion unless we restrict our focus to the proximate causes of a choice. By considering only proximate causes, those running through the creaturely psyche in the right sort of way, a soft determinist can say that creaturely agents “freely” choose when they start out in a state of epistemic uncertainty about what to do and then, without any subjective sense of having being coerced, later arrive at a settled intention to act absent that uncertainty. [Note: Different versions of soft determinism give different accounts of what counts as “the right sort of way” for the causal chain to run through a creature’s psyche. Typically, these accounts involve things like reasons responsiveness (i.e., being moved in accordance with one’s reasons) and absence of higher-order volitional conflict.]
In any case, Helm imagines open theists objecting that this is a sham sort of freedom because, if God has decreed a creature’s choices in advance, then the creature is, effectively, God’s “puppet.” The creature may “feel” free, but his action really isn’t. In stating this objection, however, Helm straw-mans the objector in two ways. First, he supposes that the objector holds that the creature merely “seems to possess epistemic openness” but really doesn’t because his pending decisions are in fact being manipulated. And, second, he supposes that the objector holds that the creature’s decisions merely “have the appearance” of being “reason-based, unconstrained action[s]” (p. 86). But this completely distorts the libertarian’s objection. The objector’s claim is not, as Helm seems to think, that on determinism creaturely beliefs and desires become mere “epiphenomena” and thus play no causal role in producing creaturely decisions and actions (p. 87). The libertarian objector can (and should) grant that soft determinism is compatible with genuine (and not mere “seeming”) epistemic openness and reasons responsiveness. The objector’s claim, rather, is that it is the determined creature’s freedom that is merely apparent and not genuine. That’s the point of the puppet analogy. It’s not that creatures, like actual puppets, lack genuine and/or causally effective mental states.
Helm goes on to talk about primary and secondary causation and reiterates his point that soft determinism is compatible with a robust and normal seeming mental life. Adding theistic determinism into the mix doesn’t change that basic point. I agree with him here. Indeed, it’s a corollary of my argument that deliberation is merely prima facie evidence of libertarian freedom. Of course, Helm isn’t willing to concede even that much, but my point is that even if from the inside we have good reason for thinking we are libertarian agents, that’s not a proof that we are, for we could feel indeterministically free at the conscious level while still being determined at the subconscious level.
Helm wraps up his paper with “the Compatibilist Challenge.” It’s a challenge for libertarians to give “an account of a libertarian action or choice, either at the explanatory or at the phenomenological level, that would not also be … consistent with compatibilism [i.e., soft determinism]” (p. 90). Given what I’ve just said in the previous paragraph, I don’t think this challenge could succeed at the phenomenological level, but I’m not sure why Helm thinks it can’t succeed at the explanatory level. For one thing, a libertarian can appeal to explanatory parsimony to argue that the phenomenological fact that we seem to possess libertarian freedom when deliberating is best explained by the hypothesis that we do at least sometimes have that sort of freedom than by the hypothesis that hidden, subconscious factors always determine our choices. For another thing, a libertarian can offer an objective, agent causal account of free will that is incompatible with soft determinism simply by building into the account a requirement that the agent be an ultimate and not merely proximate cause of his free choices. Helm would, of course, question the facticity of such an account, but merely posing such an account does meet the stated terms of his challenge.
Finally, in considering whether the deliberative epistemic openness of soft determinism gives us, as Helm claims, “all the openness we need,” I am strongly inclined to say “no” for reasons along the lines of the “puppet” analogy discussed above, moral responsibility, and the problem of evil.