In philosophical theology, simple foreknowledge (SF) is the view that
- Minimal monotheism is true, i.e., there is one, necessarily existing, personal, omniscient, etc. God who freely creates ex nihilo.
- God is temporally everlasting and therefore has a past, present, and future.
- The future is causally open, i.e., there are multiple causally possible futures, and therefore determinism is false.
- The future is alethically settled, i.e., there is (and always has been) a determinate fact of the matter as to which one of the many causally possible futures is going to happen—a unique and complete true story of the future, if you will.
- The future is epistemically settled for God, i.e., God infallibly knows (and always has known) the unique complete true story of the future.
- God’s knowledge of the contingent future is receptive in nature. (God does not have middle knowledge.)
In this post I’m going to present a refutation of SF. I’m not going to take issue with 1, 2, 3, or 6. Rather, I’m going to argue that 3 and 4 are in conflict with each other. A causally open future must also be alethically open. And if the future is alethically open, then it cannot be epistemically settled either. So 3 entails the denials of 4 and 5.
SF and preventable futurism
To say that the future is causally open is to say that there is more than one causally possible future, which implies that there are future contingents and that determinism and fatalism are false. In one of my more recent papers, “Foreknowledge and Fatalism: Why Divine Timeless Doesn’t Help” (see esp. pp. 254–260), I argue that there are only two viable strategies for resisting fatalism, by which I mean simply the denial of future contingency.
The first anti-fatalist strategy and the one I endorse, is open futurism, which holds that there is no unique actual future, no complete linear extension of the actual past and present that either is or is going to be “the” future. According to the open futurist, the future is better conceived of as an open-ended branching array of possibilities. For example, suppose that I am about to make a libertarianly free choice between chocolate and vanilla ice cream. The open futurist would say that prior to my choice there is no fact of the matter about which flavor I “will” or “will not” choose. Determinate information about my actual choice doesn’t exist until I make the choice.
The second anti-fatalist strategy is preventable futurism according to which there is a unique actual future, but it is preventable because it is explanatorily (and ontologically) dependent on the actual occurrences of future contingent events. In terms of the ice cream example, let’s suppose I opt for chocolate. The preventable futurist would say that my choice brings about it’s always having been true that I will choose chocolate (on that occasion). In other words, our free choices somehow cast an information shadow backwards in time.
Now, it should be clear that SF endorses preventable futurism. It has to if it wants to reconcile its commitment to a causally open future (3) with its commitment to an alethically and epistemically settled future (4 and 5). Moreover, its receptive model of God’s knowledge (6) when it comes to creaturely future contingents leads directly to preventable futurism. God knows what He knows about our future free choices because in the future we make those free choices.
SF and informational incompatibility
We now have the tools in place to set up a fundamental problem for SF. Consider again the ice cream scenario. Let’s say that right now, at time T0, I make a free choice between chocolate and vanilla ice cream. And let’s say again that I opt for chocolate. According to SF, God knew at all times prior to T0, that I would choose chocolate at T0. So He knew at time T–1000 that I would choose chocolate at T0. Now, for God to know that at T–1000, determinate information about my future free choice has to be available at T–1000. So we have
- Determinate information about future contingents is available prior to the occurrence of those events.
By “determinate” information I mean information to the effect that a future contingent event will happen or will not happen, as opposed to information that it may or may not happen or that it will probably happen. The latter sort of information leaves the future occurrence of the event unresolved. If some event “may or may not” or “will probably” happen, it’s still an open question whether in due course it actually does happen.
So far so good, but now here’s the problem. By preventable futurism, past truths about future contingents are ontologically dependent upon the actual occurrences of those future contingent events. This means that until such an event occurs, and the future contingency is resolved one way or the other, there is no determinate information about it. There can’t be, not if that information is ontologically dependent on the actual occurrence of the event in question. Absent the event, determinate information about it simply doesn’t exist. And if it doesn’t exist (period), then it can’t exist at previous times either. So we have
- Determinate information about future contingents is not available prior to the occurrence of those events.
The problem is now evident: (a) and (b) contradict each other. Past information about future contingents can’t both be determinate and non-determinate at the same time. And yet it has to be if SF is true. It has to be determinate in order for there to be a unique and complete true story of “the” future for God to know. And it has to be non-determinate or indeterminate in order to be grounded in the actual occurrences of future contingent events.
In short, the SF view is internally contradictory and is therefore false. Causal openness (3) cannot be reconciled with alethic and epistemic settledness (4 and 5) apart from preventable futurism, which SF endorses. But the combination of causal openness and preventable futurism entails both (a) and (b), which is a contradiction.
Timelessness to the rescue?
I pursue this question in the paper of mine that I reference above, but it’s worth raising at this point. For one might plausibly wonder if what creates the problem for SF is the temporal gap between when God foreknows future contingents and when the foreknown events occur. What if we collapse that gap by assuming that God timelessly surveys a static block creation? Would that make determinate information about future contingent events always (or timelessly) available to a God who, from the standpoint of eternity, can just “see” how it all plays out?
I don’t think this proposal works (for various reasons), but that topic will have to wait for a subsequent blog post. 🙂
Hey Alan, I appreciate this. But I think some versions of SF assume the B=theory of time and that God is timeless and that might exclude point 2. However, I cannot think of a reference for that at this moment.
Thanks, James. I prefer to distinguish SF from the timeless “view from eternity” idea. I don’t think either approach works. The reasons are slightly different in each case, but both have a hard time accommodating preventable futurism. They generally try to substitute mere counterfactual dependence (e.g., If I had willed otherwise, then God would have known otherwise) for the explanatory/ontological dependence relation. But it’s quite easy to show that counterfactual dependence isn’t enough. One just has to observe that even a theistic determinist would affirm counterfactual dependency: If I had willed otherwise, then God would have known otherwise *because God would have efficaciously decreed otherwise*.
Pingback: A Refutation of Simple Foreknowledge Revisited – Open Future