Stress-testing the
mind of Christ

Where a recovering ex-atheist rams the Bible into other worldviews to see what breaks (note: Scripture cannot be broken)

Constructive criticism of The Unseen Realm #4: predestination and foreknowledge

In which I offer a friendly critique of some elements of Michael Heiser’s “Unseen Realm”—in this instance, his comments in chapter 9 on how God foreknows without predestining.

Continued from part 3, on perfection and freedom

Before I begin my critique, I want to make something clear: In this section of Unseen Realm, chapter 9, Mike is advancing an argument against God’s predestination, and as a Reformed Christian I obviously disagree with him. However, I am not critiquing his argument because I think it is unsound—i.e., because I disagree with one or more of the premises. If that were the only issue, I could simply say, “I disagree with Mike on this point.” Rather, my critique is motivated by the fact that his argument is invalid for a number of reasons.

My concern, as with my previous post, is not that Mike is teaching something I disagree with; it is that he commits logical fallacies in arguing for his position. Because of this, informed readers will begin to doubt his competence—and this has a negative halo effect on everything he says, including his later material that I’d want to endorse.

A real-world example

This is not just a theoretical concern. At the Bible study I attend for our church—Trinity Reformed Baptist—I suggested we work through Unseen Realm. For a while we did; but people got so disenchanted by the discussion in chapters 8–9 that we ended up dropping it before we reached the content that actually matters. The general feeling was that the book was too much of a mixed bag to be worth it. Most of the group’s members are not highly educated readers, but they have been taught well enough to realize that something was fishy with what Mike was saying. This is probably the most common scenario you’d find with readers in Reformed churches: they’re accustomed to reading discerningly, so they know there’s something wrong, but they’re not philosophically able enough to clearly pinpoint the problems.

But this is the worst possible outcome, because it is exactly the lack of clarity which will cause maximum doubt about Mike as a teacher in general, and prompt people to abandon the book.

It is relatively easy for someone like me to separate Mike’s discussion on freedom and foreknowledge from his discussion on linguistics and comparative mythology: I have the knowledge required to parse the arguments and to recognize that one deals with philosophical theology—where Mike lacks competence—and the other deals with something else entirely—where Mike has a great deal of competence. But typical Reformed readers are not as equipped for this task, and so the distinction is blurry to them. Their uncertainty about the one creates an overarching doubt that extends to the other.

The reason I add this lengthy preface is not merely to justify my critique, but because in the official Lexham Faithlife forums where Mike discusses Unseen Realm, one of his defenders states that the whole purpose of these chapters is to “disarm” Calvinists in order for Mike to be able to advance his main thesis. So I think it’s really important to have a Calvinist “on the ground” relay his own experience:

This chapter achieves the opposite of its stated goal.

I can’t see the need for such a disarming in the first place; Calvinists like me will have no problem with Mike’s central arguments; Calvinists of the dogmatically confessional variety will reject them and his argument about predestination anyway. So this entire part of the book seems completely counterproductive to Mike’s aim, as I understand it.

The critique proper

I almost feel like I should have made that a separate post, but never mind—here goes. I see four main fallacies in Mike’s argument against predestination:

1. False dichotomy

Mike begins with a rehearsal of the previous chapter, which acts as a bridge to set up his comments about predestination. The general issue is one of theodicy—why does God not do away with evil and suffering on earth? Mike’s answer is that he can’t—because that would require elimination of all his imagers (loc. 1104; all emphases are original).

In other words, this is a greater-good defense:

God’s decision was a loving one. Understanding that requires only a consideration of the two alternatives: (1) not having life at all, and (2) being a mindless robot, capable only of obeying commands and responding to programming.

If our decisions were all coerced, how authentic would those “decisions” actually be? If love is coerced or programmed, is it really love? … It isn’t. For a decision to be real, it must be made against an alternative that could be chosen. Ibid

Although I’ve dealt with the issue of freedom in part 3, Mike’s very specific concern to preserve the nature of love bears comment, because this is an objection I hear a lot from uninformed freewill theists. The problem is that the very nature of God himself refutes this understanding of love. Here’s a basic version of the argument:

  1. God exists necessarily (standard Christian doctrine)
  2. God is love (1 John 4:8)
  3. Therefore, love exists necessarily

To rephrase the conclusion slightly, the Father cannot fail to love the Son, the Son cannot fail to love the Father, and so on. If they could, love would not exist necessarily. This seems like an obvious enough truth, but it directly implies the falsity of Mike’s claim that love is not really love unless it can “be made against an alternative that could be chosen.” If, in the paradigm nature of love, there is no alternate possibility; if the “most real” kind of love does not admit of other live options; if love himself is not “free” as Mike defines freedom…then Mike’s definition of freedom is just wrong.

Notice that I am not claiming the Father does not choose to love the Son. The most robust, complete kind of love certainly is volitional; it is something that requires an act of will. The point is not that God does not choose, but that God cannot choose otherwise.

So Mike’s whole case falls apart right here, because he sets up a false dichotomy between libertarian freedom on the one hand, requiring other live options, and coercion or programming on the other, where no choice is made at all. But in fact both of these are wrong. Choice does not require other live options, yet the alternative is not coercion or programming; it is simply choice without other live options.

There are two sides to this false dichotomy’s dirty penny; two major errors which are going to leave a very poor impression with the Calvinists this chapter is supposedly trying to disarm:

  1. Mike begs the question by assuming that “choice” means libertarian choice—the exact point under dispute;
  2. Mike burns a strawman of the Calvinist view by equating compatibilist choice with “coercion” and “programming”—a category error that I can assure you Calvinists get very tired of correcting.

2. Non-sequitur or hasty generalization

Depending on how you want to cash out the problems here, you could name a couple of different fallacies. We could even call it a false dichotomy, but then I would have two headings with that name, and that would be Bad Writing.

The overall claim is that God cannot do away with evil and suffering on earth, because that would require the elimination of all his imagers. In other words, the only way to eliminate evil is to eliminate all free agents who can commit evil. But even if we grant Mike’s tendentious definition of freedom, this is just an assertion in lieu of an argument. It simply doesn’t follow without further reasoning. But what would that reasoning look like? Consider:

Does evil exist necessarily?

In order for Mike’s claim here to find ultimate traction, we would have to say that there is no possible world—no other state of affairs God could have brought about instead of this one—in which all free agents are always virtuous. But that is a truly staggering claim—directly equivalent to saying that evil exists necessarily given free will. In other words, if free agents exist, evil cannot fail to exist, and so God could not have created a world of free agents without also creating a world of evil. What possible reason could there be for believing that?

If you’re not familiar with modal logic, let me expand a little: possible worlds describe different ways that things could be; different possible states of affairs. If there is no possible world in which X exists, then X is metaphysically impossible: it necessarily cannot exist. Alternatively (and this is the easier way to understand it), if Y exists in every possible world, then Y is metaphysically necessary; it cannot fail to exist in this world or any other. A good example is God: under the Christian view, there is no possible world in which God does not exist, because God exists necessarily. But by the same token, if there is no possible world in which free agents exist but evil does not, then evil is necessitated by free will—which contradicts Mike’s very idea of freedom as the ability to choose live options, some of which are good. (Btw, it also won’t do to try cashing this out statistically, because that misses the point: even if there are few possible worlds in which free agents never sin, just as there are few possible worlds in which I toss heads a thousand times without ever getting tails, those worlds are nonetheless possible, and God could therefore choose to instantiate them, instead of the one he did.)

The alternative is to go the open theist’s route and claim that God cannot know in advance how free agents will act; in other words, that he has no knowledge of what Molinists call the counterfactuals of creaturely freedom. But that option isn’t open to Mike, as we’ll see under fallacy 3, below.

Much evil versus all evil

Aside from this, Mike’s statement is also broad in a rather weasly way—hence my suggestion that this is a hasty generalization. Even if God cannot eliminate all evil given free will, it is manifestly false that he cannot eliminate much evil. There are a staggering number of awful, horrific sins that God could prevent without interfering with free will at all.

For instance, why doesn’t God cause all the rounds that end up killing people to misfire? Why doesn’t he cause thick fog to blanket an area where refugees are being hunted by rape gangs? Why didn’t he make the brakes fail on Ted Bundy’s car, so he hit a pole and lost the use of his legs? And consider all the evil that happens by accident. Why can’t God nudge people to remind them to close the gate to the pool, so their toddler doesn’t wander in and drown? Why can’t he nudge them to remember to lock the gun safe, so their kids don’t play with the pistols and blow their own heads open? How is directly putting those thoughts into their heads notably different from another person reminding them verbally? And even if it does override free will, wouldn’t these people prefer he did that if it means saving their children from tragic, preventable deaths that wreck the lives of their families? Wouldn’t they freely choose that if they could?

Ultimately, this is just the standard failing of the free will theodicy. Mike wants to say that creating free agents was the loving thing to do—but he doesn’t mention (or doesn’t realize) that what he gives with one hand, he takes away with the other. God loves people enough to give them free will; but not enough to stop all manner of preventable evils that result from that free will.

Now, I don’t want to be unduly hard on Mike here, because obviously Unseen Realm is not a book about theodicy. But that is kind of my point. These chapters don’t add anything to the main purpose of Unseen Realm. Mike’s main thesis for the book is completely independent of his argument here. Whether you accept the one is completely irrelevant to whether you accept the other. So why include content that only turns off a sizeable reader-base, and makes the book harder to follow for everyone?

3. Category error

Let me step back to remind you that I like Mike and really want Unseen Realm to succeed. That is the context in which I say that this next fallacy is the most awkwardly egregious and embarrassingly inept of the lot, okay?

Citing 1 Samuel 23:10-13, Mike argues:

David asks the Lord two questions: (1) will his nemesis Saul come to Keilah and threaten the city on account of David’s presence? And (2) will the people of Keilah turn him over to Saul to avoid Saul’s wrath? Again, God answers both questions affirmatively: “He will come down,” and “They will deliver you.”

Neither of these events that God foresaw ever actually happened … This passage clearly establishes that divine foreknowledge does not necessitate divine predestination.

The theological point can be put this way:

That which never happens can be foreknown by God, but it is not predestined, since it never happened. Loc. 1173–1192

I quote at length because the bungle here is so remarkable that any informed reader would suspect I was misrepresenting Mike’s case. Unfortunately, as you can see, he really does try to get around a careful treatment of the genuine philosophical issues with predestination and freedom through a mere sleight of hand: by pretending that God’s counterfactual knowledge is the same as his foreknowledge.

This is simply not the case, and you won’t find any informed theologian who says otherwise (feel free to try; I’ll wait). Counterfactual knowledge is God’s knowledge of what would happen under some other conditions than actually occur in this world. His foreknowledge, by contrast, is his knowledge of what will happen.

These are two distinct types of knowledge. One is of possible worlds, and one is of the actual world; conflating them is therefore an outrageous category error.

Now, it’s true that under Calvinism God’s counterfactual knowledge is the same kind of knowledge as his foreknowledge: both are parts of what is called his “free knowledge,” which is his knowledge of contingent truths. Under Calvinism, God knows what he would cause to happen given the conditions of any possible world; and he knows what he will cause to happen given the conditions of the actual world. But clearly Mike is not making this kind of conflation; he is not simply using foreknowledge as a mistaken synonym for free knowledge. He is not a Calvinist.

The only alternative he can turn to is Molinism, where God’s knowledge of the counterfactuals of creaturely freedom is part of his so-called middle knowledge (falling between his free knowledge and his natural knowledge of necessary truths). But Molinism affirms what Mike denies: that the future is settled. The whole point of Molinism is that God is able to know what free creatures would do in any possible world, and he then chooses which possible world he wishes to instantiate, so that free creatures will do those things. But this makes the future settled, even though Molinists argue (wrongly) that it does not make our actions necessitated.

In case you’re wondering why I don’t canvass classical Arminianism here, it is because there is no place for counterfactuals of creaturely freedom there. Classical Arminianism affirms the doctrine of simple foreknowledge, which says that God only knows what will happen in the world; it tries to get God off the hook for sin by denying that he foreknew the fall until he had already decided to create the world. So there are no counterfactuals of creaturely freedom here for God to know in the first place—in contradiction to 1 Samuel 23.

Thus, when Mike says that, God foreknew a possibility—but this foreknowledge did not mandate that the possibility was actually predestined to happen, he is simply talking nonsense. God does not foreknow possibilities; he simply knows them. Foreknowledge refers specifically to future events in the actual world. God knew what would happen in the possible world that David was asking about. But that world was not the actual world, and God was not relaying any foreknowledge to David.

It’s hard to track what Mike is actually angling at here, because of his conflation of foreknowledge and counterfactual knowledge; but what it sounds like he is saying is that the future is not settled. That seems to be what he means by “predestined.” In other words, he seems to be trying to take a piece of Molinism—middle knowledge—and a piece of open theism—an open future—and mashing these incompatible theologies together to create a fiendish hybrid that will do everything he wants: preserve God’s omniscience, and human freedom.

But this is not a project for Unseen Realm. It’s not really a project for anything, because any theologian conversant with these issues would know that bird won’t fly. But it’s definitely not a project for a single chapter of a book about a completely different topic; and all Mike really achieves here is that he leaves most Christians as unknowingly confused as he is about a topic whose philosophical complexities he doesn’t understand or appreciate.

4. Self-contradiction

To explain a little further, let me turn to the final logical fallacy Mike commits: Ultimately, the view he is advocating is incoherent because it contradicts itself. Mike wants to affirm the following two inconsistent statements:

  1. What free agents would do in any possible world is settled (because otherwise God would not know counterfactuals like those in 1 Samuel 23)
  2. What free agents will do in the actual world is not settled (because otherwise the future would be predestined)

The problem should be obvious: since the actual world is a possible world, what free agents do in it must be settled according to statement (i). But that contradicts statement (ii). You can’t eat your cake and still have it too. Either God knows what free agents will do in a given situation, or he doesn’t. If he does, he knows what free agents will do in the actual world, and the future is settled. If he doesn’t, then he couldn’t know that Saul would come to Keilah, or that its people would hand David over to him.


To round out an already exhausting review, I want to briefly deal with the implications Mike draws out of chapter 9:

There is no biblical reason to argue that God predestined the fall, though he foreknew it. There is no biblical reason to assert that God predestined all the evil events throughout human history simply because he foreknew them … God does not need the rape of a child to happen so that good may come. His foreknowledge didn’t require the holocaust as part of a plan that would give us the kingdom on earth. God does not need evil as a means to accomplish anything. Loc. 1208–1216

Let me summarize the issues here, because they really summarize the whole problem with this chapter and the one before:

  1. What this amounts to saying is that there is no good reason for the rape of children. There was no good reason for the holocaust. God didn’t intend for those things to happen; they weren’t necessary parts of his plan. They were just tragic side-effects of free will. I don’t see how that is supposed to be a compelling alternative to a broadly Calvinistic view, in which God works all things for good—even something as appalling as the holocaust.
  2. Mike fails to anticipate obvious counterexamples to his overreaching claims. God doesn’t need evil to accomplish anything? Really? Come on, get serious—surely Mike agrees that God actually did need evil in order to accomplish his plan of redemption. He might disagree that glorifying Jesus through redemption was the original plan for creation—though that leaves God originally creating the world for a lesser reason!—but it is a simple matter of logical necessity that there must first be sinners if God is to redeem them. By the same token, if God wishes to manifest his perfections (aka glorify himself), and one of his perfections is wrath, he cannot do that without creating sinners to drink from that cup. Does Mike not think that God’s wrath is a perfection worth manifesting?
  3. Finally, Mike simply asserts his whole confused view in the teeth of the exegetical evidence that has brought people like me to Calvinistic convictions. He doesn’t interact with any of that evidence; he doesn’t even acknowledge it; rather, he simply takes a couple of passages, runs them through his very obviously modern, philosophical filter, and confidently asserts that, An ancient Israelite would have embraced this parsing of foreknowledge, predestination, sovereignty, and free will (loc. 1208). If I didn’t have good reason to think Mike was trying to disarm Calvinists here, I would think he was trying to infuriate them by rehashing all the trite old tropes we’re already familiar with: tropes which cannot be exegeted from Scripture, and which we’re confident can be refuted with exegesis from Scripture—as well as by simple, careful thinking like that I’ve tried to demonstrate here.

If this argument in chapter 9 is supposed to disarm Calvinists, then either Mike doesn’t understand Calvinism, or he just needs to go back to the drawing board and enlist the help of an actual Calvinist to write something that won’t piss us off. I’ve spoken with Doug van Dorn about this; he is a Reformed Baptist who works closely with Mike and I am certain he would be willing to help. As would I. But as it stands, there’s nothing compelling or surprising here to disarm Calvinists.


  1. G

    I have also read Unseen Realm and Reformed and also struggled with those aspects of the book. Thank you for writing this! It raised a lot of questions for me about authority of the believer, our role as God’s agents. Can you help sort this out? Quote from Unseen Realm, “But Yahweh’s decisions in the original Eden meant that he would not overturn human (or divine) freedom in his imagers. Yahweh had chosen to accomplish his ends through imagers loyal to him against imagers who weren’t. This commitment to humanity, his original imagers on earth, is one often-missed reason why, when humanity (Israel) failed to restore God’s rule” (p. 216) –Can we miss God’s potential for our lives? As one author wrote, “God is committed to his promises, but we can miss God’s potential for us.” Can we fall on plan B? Can we miss healing if we don’t pray hard enough, and miss bringing in the kingdom in a neighborhood (obviously not the entire kingdom) if we don’t cooperate with God and witness and work and share the gospel? How much does God limit himself to his imagers? What is our role in building the kingdom and his role? Can the kingdom not “break in” into a neighborhood today because we aren’t doing our role?

  2. Dominic Bnonn Tennant

    How much does God limit himself to his imagers?

    A great deal. God could raise up believers from the rocks if he wanted to. But he has chosen to raise them up through our preaching the word.

    What is our role in building the kingdom and his role?

    Our role is preaching the word. His role is changing hearts. This is the great comfort which Mike would actually deprive us of: in his view, God is not responsible for whether someone believes the gospel or not. It’s all on us. That’s incredibly daunting—and incredibly disheartening if we don’t see fruit. On the Reformed view, we simply do our part, and wait to see if God does his. If he doesn’t, it doesn’t necessarily reflect a failure on our part. As long as we preach the word faithfully, we’re doing everything we can. If people don’t respond, it’s because God did not draw them; not because we weren’t persuasive enough.

    Can the kingdom not “break in” into a neighborhood today because we aren’t doing our role?

    Of course. God works through us. But equally, if we do our role, the kingdom might still not break into our neighborhood, if God didn’t choose those people for it.

  I don’t post ill-considered articles and I don’t sponsor ill-considered comments. Take a moment to review what you’ve written…