Prominent Christian philosopher, apologist, and analytic theologian William Lane Craig is in the process of releasing his magnum opus, a 5-volume Systematic Philosophical Theology (Wiley Blackwell, 2025–2026). Volume 2a (released in 2025) focuses on God’s attributes and includes a 100-page chapter on omniscience. Now aged 76, Craig presumably intends this chapter to be the crowning summary of his nearly 40-year publication record on the topic of divine omniscience and foreknowledge.
In a 4-part blog series I will review and critique Craig’s omniscience chapter. The chapter reveals very clearly that Craig has not kept up with the literature on open theism and the foreknowledge problem. Indeed, of the 160 footnotes in the chapter, only 10(!) reference any post-2000 sources. Of those 10, only three(!) footnotes bear upon the foreknowledge problem and only one(!) citation is to an open theistic source (from 1994), and that only in the context of discussing how to define ‘omniscience’. The chapter is divided into four main sections—§1 Biblical data, §2 Omniscience, §3 Foreknowledge, and §4 Middle Knowledge—which will correspond to the four parts of this blog series.
Part 1. Biblical data concerning divine omniscience (pp. 185–202)
(a) God is omniscient and has exhaustive knowledge of the past and present
Craig begins by citing three Biblical verses to support God’s general omniscience: Isaiah 40:28; Psalm 147:5; Psalm 139:17–18. None of these explicitly teach omniscience, however. Better verses for this are Hebrews 4:13 and 1 John 3:20. He then cites additional verses to support God’s exhaustive knowledge of the present (Job 28:24; Jeremiah 17:9–10; 1 Chronicles 28:9; Hebrews 4:13), and God’s exhaustive knowledge of the past (Malachi 3:6; Psalm 56:8). Since I do not contest any of these claims, I will merely note them and move on.
(b) Is the future a LINE or a TREE?
Things start to get controversial when Craig comes to God’s exhaustive knowledge of the future. A basic problem here that Craig never mentions or considers is whether “the future” is best conceived as a unique and complete time-LINE or as a branching TREE of possibilities, despite the fact that open theists like myself and Dale Tuggy have explicitly singled out this issue since 2006. Without even noting it as an assumption, Craig simply assumes that “the future” refers to a unique timeline. This is an assumption that Patrick Todd has since 2016 dubbed the “unique actual future” (UAF) and that Nuel Belnap and Mitchell Green had earlier (1994) called the “thin red line.” By failing to highlight and defend his stance on this issue, Craig sets himself up for a question-begging discussion of divine foreknowledge in relation to open futurism.
(c) The importance of foreknowledge
In any case, Craig begins by emphasizing the importance of divine foreknowledge. God’s foreknowledge and providence, he contends, are “fundamental to the biblical conception of history and are a source of comfort and assurance to the believer in times of distress” and also “essential to the prophetic pattern that underlies the biblical scheme of history” (p. 189). As an open theist, I have no problem with these declarations unless it is assumed—as Craig tacitly does throughout—that God’s foreknowledge and providence meticulously single out, in the familiar words of the Westminster Confession, “whatsoever comes to pass” (i.e., a UAF). From an open theist perspective the importance of divine foreknowledge consists in the fact that God has a well-considered plan. He’s not just “winging it” with creation.
(d) Psalm 139
For an example of Craig’s unspoken assumption of a UAF at work consider his discussion of Psalm 139:1–6. He says that “just as God knows the thoughts humans have, so he foreknows the very thoughts they will have” on the grounds that God’s knowing the psalmists thoughts “from afar” (v. 2) connotes “temporal distance” (p. 190). Craig, therefore, thinks the psalm implies that God knows everything the psalmist ever thinks long before the latter entertains any thoughts at all, and presumably long before the psalmist comes to exist. But this is a pretty weak argument, as there’s no clear and compelling reason to take “from afar” in a temporal rather than in a spatial sense. Indeed, the psalm begins (v.1) with an assertion of God’s intimate acquaintance with the psalmist (“O Lord, you have searched me and known me”). This implies that the psalmist already exists and sets the context for the rest of the psalm. So God’s knowledge of the psalmist’s thoughts isn’t prior to his existence, much less long before then. Indeed, the dominant orientation of the psalm is spatial, not temporal (see vv. 5, 7–12).
(e) New Testament pro- words
Craig moves on to a discussion of New Testament pro– words: proginosko (foreknow); prognosis (foreknowledge); proorao (foresee); proorizo (foreordain); and promarturomai (foretell). He rightly notes that in context many of these words are used in ways that imply God’s advance knowledge of and planning for various types of events, such as Christ’s betrayal and crucifixion. But Craig is quick to overinterpret these texts in favor of his Molinism (for a precise explanation of Molinism, see §1 of this post). For example, he argues that if the foreordination of the crucifixion spoken of in Acts 4:28 is based on God’s “knowledge of what Herod, Pilate, … etc. would do, should Christ be sent” then this is an instance of the Molinist doctrine of middle knowledge (p. 192). But this does not even remotely follow, for middle knowledge is prevolitional, and there is no suggestion anywhere in the Bible that God has prevolitional knowledge of conditional future contingents. God’s knowing what Herod etc. would do comes, presumably, from God’s intimate post-volitional acquaintance with Herod etc.
It is also worth noting that the Greek prefix pro- can simply mean before some contextually specified reference point. It need not imply, unless the context makes it clear, that what God is said to have foreknown or foreordained was such from the very beginning of creation. For example, when Romans 8:29 says that “those whom He foreknew He predestined” the verse does not imply that there is some definite group of individuals foreknown and predestined from the very beginning of creation to be “conformed to the image of the Son.” In the larger context of the chapter, those God “foreknew” can plausibly be read as referring to those God knew beforehand, that is, people of faith who lived prior to when Paul is writing. This is exactly how the word is used in Romans 11:2. So understood, Paul is talking about past saints and saying that just as they are now predestined for glory so will we be if we join the company of the faithful.
(f) Judas’s betrayal and Peter’s denial
Regarding Jesus’s predictions of Judas’s betrayal and Peter’s denial, Craig asserts that “to suggest that these … predictions were not founded on foreknowledge, but were inferences based on the character of the disciples and the context of events soon to occur, is to evacuate the incidents of all theological significance whatsoever” (p. 192). This is quite an overstatement on Craig’s part. The theological significance of Jesus’s predictions on the proposed (open theistic) reading is exactly that of Psalm 139, namely, that Jesus (as God) is intimately acquainted with the thoughts and characters of Judas and Peter (and the larger context) to a degree that no mere human could have been. That the passages can be read in a way that doesn’t signify Craig’s theology hardly amounts to stripping them of “all theological significance.”
(g) Foreknowledge of future free acts?
Craig also contends that in the Bible God’s foreknowledge “seems to extend to future free acts, events which could not possibly be inferred from present causes” (p. 193), but he doesn’t reference any passages that clearly show this. His whole case is based on a weak inference from unspecified verses. Here it is:
We have seen examples [which ones?] throughout Scripture of God’s foreknowledge of such events [i.e., of undetermined future events], including even the thoughts which individuals will have [a likely allusion back to his overinterpretation of Psalm 139]. Therefore, in the biblical conception God’s foreknowledge comprises foreknowledge of future contingents. (p. 193)
It only takes a moment’s reflection to see that Craig’s “therefore” is woefully unsupported. I don’t see how any theistic determinist would be remotely fazed by his biblical case for future contingency. He could at least have cited passages like 1 Cor. 10:13 and Deut. 30:19. For a scholar of his caliber, Craig’s biblical case for God’s foreknowledge of how future contingencies will turn out is pathetically weak, indeed, almost non-existent.
Things get a little better when Craig considers God’s knowledge of conditional future contingents. His chief example here is the dialog between God and David regarding the men of Keilah (1 Samuel 23:6–10). But even granting for the sake of argument that this passage implies God’s knowledge of conditional future contingents (i.e., that if situation S were to obtain, then event E would occur and occur contingently), there is no proper way to infer from this that God has categorical knowledge of future contingents (i.e., that event E will occur contingently) any more than one can properly infer “Q” from “if P then Q.” But while the conditionality of this knowledge may be evident in the text, it’s not immediately obvious that this is an example of a conditional future contingent. Nothing in Craig’s exegesis—if one can even properly call it that—implies that the connection between antecedent and consequent is anything but necessary, and thus non-contingent. Nor is there any biblical reason to project this conditional knowledge back into the deep recesses of time as Craig’s Molinism requires. To all appearances whatever knowledge God does have of conditional future contingents is grounded in His intimate acquaintance with creation.
(h) What about open theism?
Craig next considers “two denials of God’s knowledge of future contingents.” The first (open theism) affirms future contingency but denies that God can know (categorically) how future contingencies will turn out. The second (theistic determinism) denies future contingency.
Regarding the open theist challenge, Craig doesn’t use the now-standard label “open theism” nor does he reference any open theistic literature in this section. Instead, he refers to open theists as “revisionist theologians.” This is shamefully petty on Craig’s part. He’s clearly trying to poison the well against open theists by calling them “revisionists.” And his not citing any open theists is poor scholarship and cowardly. Does he fear that readers might consult what open theists actually say and see that Craig’s criticisms are straw men? For example, he says that according to “revisionists” (i.e., open theists), God is “ignorant of vast stretches of forthcoming history” (p. 195). This is a straw man because one can only be ignorant of information that is there to be known. But according to most open theists there simply is no such thing as a unique actual future precisely because there are future contingents.
Craig does concede that some Biblical passages seem to suggest that from God’s perspective there is no unique actual future. He mentions Genesis 18:20–21, Jeremiah 26:3, Jeremiah 39:3, and Ezekiel 12:3. But Craig dismisses this biblical evidence for two reasons.
First, he says, “a consistent application of this way of interpreting the Bible leads to a defective concept of God” (p. 196) one that, like Mormonism, depicts God as essentially corporeal, localized, and ignorant even of some present realities. But this is a really bad objection. It has been responded to at length by prominent open theistic scholars like John Sanders and Greg Boyd. (It would have been nice if Craig would at least have cited them and interacted with their arguments.) Without going into details, the basic rejoinder is that we have principled reasons in context for judging corporeal descriptions of God as figurative and narrative uncertainty on God’s part as literal. After all, if you’re going to say that some biblical depiction of God is figurative, then you ought to be able to say what the literal point of the figure of speech is. Where God is the speaker, however, and is presented as uncertain, surprised, or using conditionals when God could easily have spoken confidently and categorically on the assumption that Craig’s view is correct, then it should at least deserve consideration whether that’s because the future is not a unique and complete timeline and the future is partly open-ended, even from God’s perspective.
Craig’s second reason for dismissing Biblical evidence for open theism is that it depends, he says, on a “naïve hermeneutic,” one that overlooks the fact that Biblical narratives are told from “the human perspective” and thus are “anthropomorphic” (p. 196). This seems a weird line to take, however, because the open theist-friendly passages that Craig previously referenced from Genesis, Jeremiah, and Ezekiel are all presented from God’s perspective, not from a human perspective. Thus, it is God who is presented as saying “perhaps” in Ezekiel 12 and as speaking in terms of what “may” happen in Jeremiah 26 and 39. Craig proposes that in these passages God is merely assuring his listeners that “it is not too late for them to change and avert disaster” (p. 197). But if that were the whole point, then why did God say as an aside to Jeremiah and not to the people targeted by the prophecy, “It may be they will listen, and every one turn from his evil way” (see Jeremiah 26:2–6).
In the end, for Craig, it all turns on the supposed “clear teaching of Scripture” that God foreknows “the future free acts of individuals” (p. 198)—nevermind the fact that Craig doesn’t supply any clear Biblical evidence that God has categorical knowledge of how bona fide future contingencies will resolve.
Finally, Craig considers how “revisionists” (i.e., open theists) try to explain biblical prophecies. They do so, he says, in one of three main ways: (1) as declarations by God of what He intends, (2) as inferences based on present conditions, and (3) as conditional statements of what will happen if something else does. Craig thinks these strategies are inadequate for four reasons. First, he thinks (3) either collapses into (1) and (2) or presupposes middle knowledge (p. 198). This is absurd, however, because as noted above middle knowledge is prevolitional whereas there is no reason to think any conditional prophecies are. Moreover, Craig himself proposes on the previous page that many prophecies have implicit ceteris paribus (i.e., unless) clauses. This is a type of conditional prophecy which, as conditional, obviously does not reduce either to (1) God’s categorical intentions or (2) what’s already categorically predictable. Second, he repeats his previous implausible assertion that (2) cannot explain Judas’s betrayal or Peter’s denial without stripping the predictions of “any theological significance” (p. 199). Third, he thinks the utility of option (1) is limited because many prophecies concern free human choices. An open theist can push back, however, and say that God’s having a settled intention to bring about X need not imply that it’s settled how God brings X about. Human free choices may be part of that how, with freedom-negating divine intervention only employed if necessary as a last resort. Fourth and finally, Craig says that none of (1–3) “comes to grips with … the didactic passages which explicitly teach that God foreknows the future” (p. 199). But, again, there are no didactic passages that clearly and explicitly teach that God has the kind of foreknowledge Craig supposes and open theists deny. All open theists believe that the future is partly settled and that God knows exactly how those aspects of the future are going to turn out. Nothing Craig has presented from Scripture even remotely implies that the future is wholly settled such that there is a unique future timeline, a unique actual future.
(i) What about theistic determinism?
Craig finally wraps up the “biblical data” section of his chapter by considering the challenge of theistic determinism. Why should we think that there is any future contingency? Why not conclude, as theistic determinists do, that God knows the unique actual future because God has causally determined that it comes to pass?
Craig’s main line of response to such questions is that if God ultimately causes all things, then this makes God the author of evil. He also notes (as open theists like to point out) that there are divine repentance passages like Genesis 6:6 and 1 Samuel 15:11, 35. Such passages suggest, he argues, that “in many cases what God does depends on how they [creatures] freely respond to his [God’s] initiatives” (p. 202).
What Craig doesn’t point out here is that his objections to theistic determinism apply to his own Molinist position as well. For Molinism God doesn’t deterministically cause everything that happens in creation, but He does specifically ordain everything. Nothing is “merely permitted” by God. In virtue of this there is a clear sense in which for Molinism God is the author of evil because, like the sole author of a novel, He deliberately chose to actualize a complete creation timeline containing all of the evils that ever occur. Furthermore, like theistic determinism, Molinism has no room for divine responsiveness to creation. For Molinism everything in creation is settled at the moment of God’s creative decree. There is no additional “input” from creation that God ever has to factor in.
My concluding thoughts on Craig’s biblical case
Overall, I am deeply disappointed by Craig’s biblical case for the position for that God has exhaustive foreknowledge of all creaturely future contingencies and how they will resolve. His biblical case is very weak and nowhere near cogent enough to falsify open theism. For a work that is supposed to be Craig’s magnum opus, the quality of his argumentation is no better than (and possibly worse than) Frank Turek-style pop apologetics. Craig’s overconfidence stems in no small part from his unexamined assumption that “the future” refers to a unique actual future, a singular timeline that obtains by God’s creative decree and, therefore, is known by God. His insistence on calling open theism “revisionism” and his failure to cite any of the relevant literature on it (in this section of the chapter at least) is scholarly malpractice.