Generic open theism

What is the bare minimum that one must believe in order to qualify as an ‘open theist’? The purpose of the question is to distinguish the primary or core commitments of open theism from other, secondary commitments. To define a generic open theism we ought to draw the circle as widely as possible without losing sight of the distinctives of open theism. In this vein, there seem to me to be four primary commitments, to which I'll add a few corollaries and qualifications.

The first obvious requirement for being an open theist is that one be a theist, as opposed to, say, a polytheist, pantheist, or process theist. Regarding the latter, open theists have been quite insistent that, while their position lies somewhere between the classical theism of high mediaeval orthodoxy and process theism, they mean to stay squarely on the classical side of that divide with respect to creation ex nihilo and the power of God unilaterally to intervene in the created order as He desires.Footnote ³ Indeed, open theists have generally seen their view as a relatively conservative modification or correction of the classical tradition – not a wholesale rejection of it – for the purpose of resolving what they see as otherwise irresolvable Biblical and philosophical tensions within that tradition.Footnote ⁴ As a result, I think it fair to nearly all open theists, and certainly to its chief proponents, to describe them as committed to a robust perfect-being theology in which God is conceived as a necessary being,Footnote ⁵ who essentially possesses a maximal set of compossible great-making properties, including maximal power, knowledge, and goodness.

Insofar as there are differences between open theists and non-open theists on this point, it is with respect to what that maximal property set consists in, and not with respect to whether God exemplifies such a property set. And it is important not to overstate the differences here, for it is quite clear that classical theists today are far less unified than they were in the days of Anselm and Aquinas on whether doctrines such as divine simplicity, impassibility, and timelessness ought to be included among God's great-making properties. Furthermore, as we shall see, each of the core commitments of open theism also has a large number of adherents among non-open theists. Accordingly, unless we are prepared to restrict the label ‘classical theism’ in such a way that few scholars apart from doctrinaire Thomists would qualify, it is somewhat tendentious to oppose ‘open theism’ to ‘classical theism’ and probably better to say that open theism is a species of classical theism. In sum, then, we have as our first core commitment of open theism:

1. (1) Theism:  God exists necessarily and possesses a maximal set of compossible great-making properties, including maximal power, knowledge, and goodness. He created the world ex nihilo Footnote ⁶ and can unilaterally intervene in it as He pleases.

The other core requirements for being an open theist are what put the ‘open’ in open theism. What is meant here is the openness of the future,Footnote ⁷ and that in two distinct senses, one broadly causal; the other, epistemic. We can define these as follows:

• Causal openness:  The future is causally open at time t with respect to state of affairs X and future time t* if and only if, given all that exists as of time t, it is really possible both that X obtains at t* and that X does not obtain at t*. (In other words, whether X obtains at t* or not is, as of t, a future contingent.)

• Epistemic openness:  The future is epistemically open for person S at time t with respect to some conceivable future state of affairs X if and only if for some future time t* neither ‘X will obtain at t*’ nor ‘X will not obtain at t*’ is known by S at t.

In terms of these definitions, we can state the second and third open theist commitments:

1. (2) Future contingency:  The future is, as of now and in some respects, causally open, i.e. there are future contingents.

2. (3) Divine epistemic openness:  The future is, as of now and in some respects, epistemically open for God.

Of these, future contingency is logically prior to divine epistemic openness. To see why, we must remember that the open theist wants to say that God has maximal knowledge, as per the commitment to theism. Consequently, if it is possible for God to know something, He must know it. So if the future is causally open with respect to whether X obtains at t*, and if it is possible for God to know either ‘X will obtain at t*’ or ‘X will not obtain at t*’, then He knows one of those, and the future is not epistemically open for Him in that respect. The reason, then, why open theists accept divine epistemic openness is that they believe that it is impossible for the future to be both causally open in some respect at time t and epistemically settled for God in that same respect at t. In other words, the future is epistemically open for God only because and only to the extent that it is causally open. We have, then, a fourth open theist commitment:

1. (4) EC incompatibility:  It is impossible that the future be epistemically settled for God in any respect in which it is causally open.

I take (1)–(4) to be the defining characteristics of generic open theism. Actually, since (3) follows from (2) and (4), we could define generic open theism solely with reference to (1), (2), and (4). It is helpful to make (3) explicit, however, since that has been a chief focus of debate. Indeed, the central dialectic consists in whether to accept both (2) and (4), and consequently (3), or whether to reject (3) and with it either (2) or (4). In these terms, open theism falls squarely between two competing traditions within classical theism: the theological determinism of the late Augustine, Luther, Calvin, and Edwards; and the ‘free-will theism’ of Ockham, Molina, and Arminius. Thus, open theists side with other free-will theists over against theological determinists by affirming future contingency. But they also side with most theological determinists over against other free-will theists by affirming EC incompatibility.Footnote ⁸

Before moving on to identify three specific variants of open theism, a few additional comments are in order. First, with respect to future contingency, it should be noted that the type of contingency that open theists chiefly have in mind is creaturely libertarian freedom. Nevertheless, there is some reason for not packing an explicit commitment to creaturely libertarian freedom into the definition of generic open theism. For it seems that one could hold to a version of open theism by accepting theism, EC incompatibility, and future contingency due to quantum indeterminacy, irrespective of whether any creatures have libertarian freedom. In order to allow room for that possibility, I think it best to define generic open theism in terms of a general commitment to future contingency rather than in terms of a more specific commitment to libertarian freedom.Footnote ⁹

Second, two important corollaries of divine epistemic openness are divine temporality and divine passibility. For the open theist, the future is epistemically settled for God in all and only those respects in which it is causally settled. This means that the content of God's foreknowledge changes over time, as matters that are future and contingent either cease to be future or cease to be contingent. Consequently, the open theist must allow that God can and does change in His epistemic states, which, of course, implies divine temporality. We should not say, however – at least not without further argument – that open theists are committed to essential divine temporality. For it is not immediately clear why open theists could not adopt Craig's suggestion that God is atemporal sans creation and temporal since creation.Footnote ¹⁰

As for divine possibility, it is helpful to refer to Creel's fourfold distinction between passibility in nature, in will, in knowledge, and in feeling.Footnote ¹¹ As theists, open theists will not admit that God is passible in nature. God's fundamental attributes and character are essential to Him and cannot change. Open theists are, however, clearly committed to divine passibility with respect to God's knowledge, for how God's epistemic states change over time depends on what free choices His creatures make. As for passibility in will and in feeling, these would seem to be optional for open theists. Some will point to Biblical descriptions of God ‘changing’ His mind, ‘repenting’, ‘feeling angry’, and so forth, as evidence that God is passible in both of those respects as well.Footnote ¹² Others, like Creel, accept passibility in knowledge, but reject passibility in either will or feeling.Footnote ¹³

Three versions of open theism

Within the broad framework of generic open theism, there are three important variants. To distinguish between them, we need to define a third sense in which the future may be thought of as ‘open’:

• Alethic openness:  The future is alethically open at time t with respect to some conceivable state of affairs X, and future time t* if and only if neither ‘X will obtain at t*’ nor ‘X will not obtain at t*’ is true at t.Footnote ¹⁴

Insofar as the future is not alethically open, I will say that it is alethically settled.

The relevance of this notion for open theism depends on the following principle:

1. (5) AC incompatibility:  It is impossible that the future be alethically settled in any respect in which it is causally open.

Any open theist who accepts (5) is thereby committed to saying that, if X's obtaining at t* is, as of time t, a future contingent, then it is not true at t either that ‘X will obtain at t*’ or that ‘X will not obtain at t*’. There are two ways in which one might arrive at this conclusion. First, one might hold that pairs of corresponding ‘will’ and ‘will-not’ propositions about future contingents are neither true nor false. Second, one might hold that all such propositions are false. The first option preserves consistency by denying bivalence. The second preserves consistency by holding that pairs of such propositions are contraries, not contradictories.

With these distinctions we can now formulate the three main variants of open theism:

• Limited foreknowledge:  The future is alethically settled but nevertheless epistemically open for God because true ‘will’ and ‘will-not’ propositions about future contingents are in principle unknowable, even for God.

• Non-bivalentist omniscience:  The future is alethically open and therefore epistemically open for God because ‘will’ and ‘will-not’ propositions about future contingents are neither true nor false.

• Bivalentist omniscience:  The future is alethically open and therefore epistemically open for God because ‘will’ and ‘will-not’ propositions about future contingents are both false.Footnote ¹⁵

The limited-foreknowledge version of open theism rejects AC incompatibility and accepts that the future is alethically settled. It follows that some ‘will’ and ‘will-not’ propositions about future contingents are determinately true. However, given EC incompatibility it also follows that no-one, not even God, can know these truths. Hence, on this view, God has maximal knowledge in the sense of knowing as much as any being can possibly know, but He is not omniscient in the sense of knowing all and only truths.Footnote ¹⁶ By virtue of creating a world populated by libertarian free agents, God willingly accepts a limitation on His foreknowledge. Among prominent open theists, William HaskerFootnote ¹⁷ has espoused this position.

The primary challenge facing the limited-foreknowledge view is to articulate in a principled way why true ‘will’ and ‘will-not’ propositions about future contingents cannot be known.Footnote ¹⁸ Unlike proponents of the other two versions of open theism, a defender of limited foreknowledge cannot say that these propositions are unknowable because they lack a truth-value or because they are uniformly false. Moreover, it seems that the primary reason for thinking that true ‘will’ and ‘will-not’ propositions about future contingents are unknowable is also a reason for thinking that such propositions are not true. Thus, as John Sanders puts it, ‘God's knowledge is coextensive with reality … [and] the future actions of free creatures are not yet reality, and so there is nothing to know.’Footnote ¹⁹ But reality doesn't just limit what can be known. On a correspondence theory of truth, reality also determines what is true. ‘Truth supervenes on being’, as the slogan goes. It appears, then, that in order to maintain the distinction between the knowable and the true that limited foreknowledge requires, its defenders must reject the correspondence theory of truth and deny that God's knowledge is coextensive with reality.Footnote ²⁰ Because that consequence is not particularly attractive, it is unsurprising that many open theists prefer one of the other two versions.Footnote ²¹

Those versions of open theism that affirm God's omniscience (in the sense of knowing all and only truths) are distinguished from the limited-foreknowledge position by the AC incompatibility thesis. The significance of this thesis may be gathered by reflecting on why advocates of the bivalentist and non-bivalentist omniscience positions say that there are pairs of corresponding ‘will’ and ‘will-not’ propositions of which neither is true. If their reason for saying this were that the future does not exist, the same rationale would yield the implausible result that no propositions about the future are true, and, if presentistsFootnote ²² are right, the utterly bizarre result that no propositions about the past are true either.Footnote ²³ Fortunately, this is not their rationale.Footnote ²⁴ Instead, what underlies these positions is a (usually tacit) commitment to what Prior has dubbed the ‘Peircean’ system of tense logic and opposed to the more common ‘Ockhamist’ system.Footnote ²⁵

Ockhamists hold that truths about the future (past) have their grounding in the future (past). In contrast, Peirceans hold that the grounding for all truths lies in the present. Thus, whereas Ockhamists say that ‘There will be a sea battle tomorrow’ is true if and only if tomorrow there is a sea battle, Peirceans say that ‘There will be a sea battle tomorrow’ is true if and only if sufficient conditions obtain today for a sea battle tomorrow.Footnote ²⁶ AC incompatibility is an immediate consequence of the Peircean semantics. Hence, it is because sufficient conditions for the realization of future contingents do not yet obtain, that omniscience-affirming open theists believe that neither ‘will’ nor ‘will-not’ propositions about future contingents are true.

Turning now to the non-bivalentist-omniscience version of open theism, this view accepts both AC incompatibility and the common view that pairs of corresponding ‘will’ and ‘will-not’ propositions are contradictories. Given those commitments, the only way to retain the contingency thesis is to give up bivalence. Thus, ‘will’ and ‘will-not’ propositions about future contingents are said to be neither true nor false. In contrast with the limited-foreknowledge view, this position does affirm that God knows all truths, including all truths about the future. As matters stand, there just aren't any ‘will’ or ‘will-not’ truths about future contingents for God not to know. Consequently, God, on this view, does not have limited foreknowledge. Rather, He knows the future exhaustively. The phrase ‘the future’, however, does not pick out a unique sequence of events, but instead refers to a branching array of causal possibilities stemming out from the present. Among prominent open theists, J. R. Lucas has espoused a sophisticated form of this position.Footnote ²⁷

The primary challenge facing the non-bivalentist-omniscience position is to motivate and defend the denial of bivalence and the attendant departure from standard logic. Denying bivalence for ‘will’ and ‘will-not’ propositions about future contingents has a long history in philosophy, but for many giving up standard logic is an unacceptably high price to pay. For example, since this position accepts that pairs of ‘will’ and ‘will-not’ propositions are contradictories, it seems to lead to counterintuitive truth-values for propositions like ‘Either X will happen or X won't’, and ‘X both will and will-not happen’.Footnote ²⁸

One solution is to adopt Thomason's ‘supervaluationist’ semantics according to which ‘X will happen’ is true if and only if X occurs on all causally possible branches of the future, and ‘X will not happen’ is true if and only if X does not occur on any causally possible branch of the future.Footnote ²⁹ This solves the above problems, but at the cost of denying that ‘and’ and ‘or’ are truth-functional. Furthermore, Thomason's approach admits the existence of a tertium quid between ‘will’ and ‘will not’, namely, the case in which X happens on some but not all causally possible branches of the future. In so doing, the supervaluationist approach undermines the assumption that ‘will’ and ‘will not’ are contradictories, which in turn undermines the motivation for denying bivalence for future contingents. This takes us to the third version of open theism.

The bivalentist-omniscience version of open theism accepts AC incompatibility but retains bivalence by holding that pairs of ‘will’ and ‘will-not’ propositions are not contradictories but contraries. Thus, for future contingent X, both ‘X will obtain’ and ‘X will not obtain’ are false and, instead, ‘X might and might not obtain’ is true. Unlike the limited-foreknowledge position, an open theist who holds to bivalentist omniscience can affirm unrestricted divine foreknowledge. Unlike the non-bivalentist-omniscience position, no departure from standard logic is required. For those reasons, the bivalentist-omniscience position enjoys a natural advantage over both of its chief rivals. Among prominent open theists, Greg Boyd espouses this position.Footnote ³⁰

The primary challenge facing the bivalentist-omniscience position is to motivate and defend what may initially seem to be the implausible claim that pairs of ‘will’ and ‘will-not’ propositions are not contradictories, but contraries. One obstacle here is to account for the common linguistic practice of retroactively describing statements about future contingents as ‘true’ simply because matters later turned out as the statement said they would.Footnote ³¹ A negative response to this objection requires advocates of this version of open theism to reject the assumption that common usage is an accurate guide to meaning in such cases. A positive response requires giving reasons why we should accept the claim that ‘will’ and ‘will not’ are contraries. Recent work drawing on insights from Arthur Prior suggests that there is a positive case to be made,Footnote ³² but as yet these arguments have not received detailed critical scrutiny. If this position can be sustained, then the advantages of bivalentist omniscience over its competitors make it the natural default position for open theists. If, however, it cannot be sustained then limited foreknowledge and/or non-bivalentist omniscience may serve as fall-back positions, assuming, of course, that the major difficulties facing them can be satisfactorily defused.