Lecture Openness Theology: A response to Gregory Boyd
By Paul Helm
This paper was read at a Colloquium organised by the Association of Confessing Evangelicals, Colorado Springs, Colorado, in June 2000
This paper is a longer version of my response to Gregory Boyd’s advocacy of ‘Openness’ theology. Both Professor Boyd’s paper and the shorter response are to be found in Divine Foreknowledge: Four Views ed. James K. Beilby and Paul R. Eddy (Downers Grove, InterVarsity Press 2001. )
This paper was read at a Colloquium organised by the Association of Confessing Evangelicals, Colorado Springs, Colorado, in June 2000. Page references in the text are to Professor Boyd’s published paper.
One cannot but admire the skill and the attention to detail with which Professor Boyd presents his case for a partially open creation. In responding to what he has to say, I wish to make two preliminary points, and then focus on what I shall argue is the main issue that Professor Boyd’s paper, and indeed that the whole controversy between the proponents of classical theism and the proponents of open theism raises.
The first preliminary point concerns the accuracy of Boyd’s opening statements on foreknowledge. Boyd says that ‘The classical view of divine foreknowledge affirms that the content of reality, and therefore the content of God’s infallible foreknowledge, is exhaustively settled…Possibilities are known by God only as what might have been, never as what might be’ (p. 13)
In the interests of both philosophical and historical accuracy, I suggest that some further discrimination is needed. On what might be called the standard Arminian view of divine foreknowledge, or on any view that thinks that perfect divine foreknowledge and the libertarian freedom of the creature are consistent, the divine foreknowledge is knowledge of what choices men and women with such freedom will make. Thus the possibilities that might have been that are known by God are possibilities that are excluded by the choices of human beings. The outcomes of free actions are logically prior to God’s foreknowledge of them. Were the choices to be different, as (prior to the effective choices in question) they could have been, then what God would have foreknown would have been correspondingly different. So on this view although Boyd is correct to say that given that God foreknows a given choice that choice is settled, had the choice been different God’s foreknowledge would have been different. The truth conditions of such foreknowledge crucially include what people will in fact choose to do.
So I am not as sure as Professor Boyd is that the debate between open theism and classical theism is a debate about the content of reality that God perfectly knows. For both the openness view and Arminianism share the view that men and women have libertarian powers such that what is might not have been. Where they differ is over the question of whether God foreknows the outcomes of such libertarian choices before they occur: the Arminian (and all who hold that divine foreknowledge and human libertarian freedom is consistent, such as Molinists and Ockhamists) hold that God does foreknow such choices, openness theism denies this. So the difference of view, at least as far as Arminianism and openness theism are concerned does seem to be over the scope of God’s cognitive powers. Arminianism holds that God knows matters which on the openness view he does not. They seem to have a pretty similar view of reality, a view that includes human beings with libertarian powers. Professor Boyd seems to recognise the coincidence between Arminianism and openness theism towards the end of his paper. (p.43)
Each of these views differs, of course, from that which holds that the truth conditions of God’s foreknowledge are settled by God’s decree, including what God willingly permits. Of this view it is certainly true that (in Boyd’s words) ‘from all eternity every detail of what shall transpire in the creation has been settled’ (p.12), settled by God’s eternal decree.
Perhaps a word should also be said here about another of Professor Boyd’s opening claims, that ‘the sovereign Creator settles whatever he wants to settle ahead of time…And he leaves open whatever he wants to leave open’ (p.14) It is not quite like this on the openness view, is it? For these words suggest that whether or not the future is open, and the extent to which it is open, is a matter of pure sovereignty, of what the sovereign Creator wants. Had to wanted to ‘close’ the future, or to close it more than it is already closed, then he could have, just like that. But he only could have done so, on the openness view (as I understand it), if he had actualised creatures with very different powers of agency than the creatures which he has in fact actualised. For having actualised creatures with libertarian powers then the Creator could not settle things ahead of time, even though he may have wanted to. And despite Professor Boyd’s confident affirmations, it is not obvious (to me at least) that God could ensure the fulfilment of his ends without ensuring the fulfilment of the means to those ends. (p.15) 
Calvin on Accommodation 
The second preliminary point concerns Professor Boyd’s discussion of the objection to the openness view from what he calls anthropomorphicism. Boyd thinks that Calvin’s appeal to accommodation to justify treating many passages in Scripture about the actions of God as anthropomorphic is weak and self-refuting. (p.38). In particular he believes that there is a flaw in Calvin’s argument in that Calvin appears to appeal both to God’s accommodating of himself and at the same time to know what the non-accommodated truth of the matter is.
My reason for looking at what Calvin says is not simply to attempt to put the historical record straight. Since a classical theist must adopt something like Calvin’s view that when Scripture depicts God as changing in using such language of himself God is accommodating himself to his creatures, Boyd’s charge that Calvin’s view is flawed is of some importance.
If knowing what the accommodated language literally means is a flaw in the idea of divine accommodation, then it’s one that is to be found in the very idea of anthropomorphism, or in the recognition of any trope. For anyone who recognises anthropomorphism in language about God for what it is (including Professor Boyd) recognises both the force of the anthropomorphic language, and the literal truth that that language depicts in non-literal fashion. Identifying some utterance as anthropomorphic requires that one has some idea of what the non anthropomorphic reality is. Thus we all know that God does not have eyes, and that the scriptural language about God’s eyes is anthropomorphic, depicting God’s immediate knowledge of human circumstances (say) in a vivid and forceful way. The interesting thing about tropic language, however, is that it works best when our attention is wholly taken up with it.
Let us take a closer look at what Calvin says. Writing of those scriptural passages where God is said to repent, Calvin says:
What, therefore, does the word ‘repentance’ mean? Surely its meaning is like that of all other modes of speaking that describe God to us in human terms. For because our weakness does not attain to his exalted state, the description of him that is given to us must be accommodated to our capacity so that we may understand it. Now the mode of accommodation is for him to represent himself to us not as he is in himself, but as he seems to us… So we ought not to understand anything else under the word ‘repentance’ than change of action…Therefore since every change among men is a correction of what displeases them, but that correction arises out of repentance, then by the word “repentance” is meant the fact that God changes with respect to his actions. Meanwhile neither God’s plan nor his will is reversed, nor his volition altered; but what he had from eternity foreseen, approved and decreed, he pursues in uninterrupted tenor, however sudden the variation may appear in men’s eyes. 
So God accommodates himself to human spatio-temporal conditions, by the use of sensory, figurative, anthropomorphic language about himself. The movement is from God to mankind; such accommodation is an act of divine condescension and grace. And the very fact that we know certain expressions to be divine accommodations implies that it is possible to think of God in non-accommodated ways, even though Calvin stresses that when we think of God in such ways, God ‘as he is in himself’, we cannot fully grasp what we think of.
Perhaps one way to understand what Calvin is saying is to think in terms of the idea of two standpoints, the divine and the human. Christian thought about God and the world must always reckon with such standpoints, that of the transcendent creator, and that of the time-bound and space-bound creature. When Calvin (together with the long tradition of speaking in this way that is to be found in Christian and Jewish thought) invokes the idea of accommodation, it is to cater for the fact that God condescends to speak in terms which creatures with a time-bound and a space-bound standpoint need if they are to respond appropriately.
For at the heart of the idea of divine accommodation as Calvin uses it is a logical point; that it is a logically necessary condition of dialogue involving the creaturely standpoint that such creatures should act and react in time. If dialogue is to be real and not make-believe, then God cannot represent himself as being only immutable, for then dialogue, real dialogue, would be impossible.
The contrast, between the human standpoint and the divine, is vividly brought out by something else that Calvin says in the Institutes. Calvin is here contrasting the existential human challenge of evil with the place of evil in God’s purposes.
But we must so cherish moderation that we do not try to make God render account to us, but so reverence his secret judgements as to consider his will the truly just cause of all things. When dense clouds darken the sky, and a violent tempest arises, because a gloomy mist is cast over our eyes, thunder strikes our ears and all our senses are benumbed with fright, everything seems to us to be confused and mixed up; but all the while a constant quiet and serenity ever remain in heaven. So we must infer that, while the disturbances in the world deprive us of judgement, God out of the pure light of his justice and wisdom tempers and directs these very movements in the best-conceived order to a right end. 
Suppose that God wishes to convey some truth about himself to his creatures existing in space and time; in particular, some truth about his relation to them, seeking to elicit a response from them. Suppose that he wishes to test his people in some way. Then it would seem that he is bound to represent the matter to his creatures in the only terms which they are going to be able to comprehend, in terms of action in space and time, and not to disclose his purposes all at once. It is not simply that God chooses, for the sake of pedagogic effectiveness, to accommodate himself to mankind, but that he must accommodate himself in these ways if he is faithfully to represent his relations to mankind. The ‘must’ here is an ontological or metaphysical ‘must’. Given the metaphysical distinction between changeless creator and the changing creature there is nothing else for it than that God represent himself to them in such ways. So the language about God in Scripture which is at odds with his immutability is the language of divine accommodation.
Boyd is correct to draw attention to the theme of God testing his people. (p. 31f.) But a classical theist such as Calvin interprets the theme of testing differently. A person could not be tested if God were to disclose to him the outcome of the test before it occurred. And so when the Lord says, after the completion of the test ‘Now I know…’ (e.g. Gen.22.12) it is not that the Lord has learned something, but these are simply words of acknowledgement and assurance to the one who has undergone the test that he has come through the fire unscathed. Given that, in order to initiate the test, the Lord had to hold back what he knew, the words confirm in a public way God’s recognition of the success of the one tested.
Despite all this one may still be left with the nagging feeling that given the idea of divine accommodation the divine action and reaction in the world which the Scriptures recount is only shadow action, and that the real action, or inaction, takes place from the divine standpoint. We may ask, does God really change his mind, or not?
To try to answer this question let us briefly consider Calvin’s treatment of the story of Hezekiah’s sickness and recovery (Isaiah 38), a case also discussed by Professor Boyd (p. 34). The Lord tells Hezekiah that he will die and then, in answer to a prayer from Hezekiah about this, promises recovery. Here is a case not simply of God apparently changing his mind, but of him apparently changing his mind having implied that he will not change his mind. What are we to make of this? Can this be made consistent with what the Bible elsewhere says about God’s immutability and steadfastness? If God is steadfast, how could God’s first utterance, that Hezekiah will not recover, be sincere?
Here is part of Calvin’s discussion of the case
But it may be thought strange that God, having uttered a sentence, should soon afterwards be moved, as it were, by repentance to reverse it; for nothing is more at variance with his nature than a change of purpose. I reply, while death was threatened against Hezekiah, still God had not decreed it, but determined in this manner to put to the test the faith of Hezekiah. We must, therefore, suppose a condition to be implied in that threatening; for otherwise Hezekiah would not have altered, by repentance or prayer, the irreversible decree of God….and thus we must suppose an implied condition to have been understood….nor are we at liberty to infer from it that God used dissimulation by accommodating his discourse to the capacity and attainments of man…In order to prepare Hezekiah by a spiritual resemblance of death, and gradually form him to a new life, he keeps back a part of the discourse. 
Does Hezekiah have to believe what is false, that God is capable of changing his mind? No, there would be nothing to stop Hezekiah believing the following:
. And because Hezekiah does not know which possible outcome has been decreed, and (naturally enough) wishes to live longer, he prays accordingly. What Hezekiah’s story does require, however, is that he believes that what God says on one occasion is not necessarily a full account of what he has decreed. And maybe he sees that one rationale for this partial disclosure is that God wishes to bring about certain changes in the lives of his time and space bound creatures, perhaps (as in the case of Hezekiah himself), to put them to the test. So there is no need for Hezekiah to believe that God has changed, only that there may be a certain dislocation between what God has willed in total and what he says he has willed as the dialogue unfolds, and that there may be a sufficient reason for this dislocation.
So when Boyd asks, what truth is communicated by such language, and claims that there is no adequate answer to this question (p. 25), one response might be: in saying one thing and then doing another God is acting in the only way possible for him to act in order to elicit certain responses from his human dialogue partners.
Hezekiah may recognise that some language about God implies change even when God does not and indeed cannot change. Let us try to work out why and how it is that sometimes our language about God is necessarily changeful even if God is necessarily changeless. Suppose that at t1 God wills A, which X notices at t1 and that at t2 he wills not-A, which X notices at t2. Then, from the point of view of this cognizer of A and then not-A, there is a change, in precisely the same sense in which a human person may be said to have changed his mind if he declares he wills A at t1 and then wills that not-A at t2. That is, God is perceived to have willed what, were this to have been a human action, would constitute a real change in the agent. And this is because what has been willed by God is truly though perhaps rather loosely expressible in contrary or contradictory terms. It is loosely expressible in such ways because these ways omit the temporal indexing of God’s actions. One can temporally index God’s actions, just as one can human actions, and so account for change over time as the possession of different temporally ordered properties. By temporally indexing, one can see that willing A at t1 and willing not-A at t2 is not necessarily contradictory. Furthermore by coming to see that God might eternally will both A at t1 and not-A at t2 one can see that such actions, not in themselves contradictory, do not even require a real change in the will of God.
So that if God eternally decrees the recovery of Hezekiah at t2, having eternally decreed to announce his death at t1, then God has not changed (since his eternal will was to bring about the recovery of Hezekiah) and his announcement that Hezekiah will die and then that he will live can be expressed as component parts of one eternal will. Putting the point even more fully, what God in fact decreed was not the recovery of Hezekiah, but the recovery of Hezekiah upon request, the request being a component part of the entire decree and that for the eliciting of which the original sentence of death was announced. (Hence Calvin’s reference to an ‘implied condition’). God may nonetheless be thought of by Hezekiah as having changed; it is most natural to do this; because one rarely has the time or energy to think in the fashion that we have been laying out.
So another part of the explanation for the language of accommodation is the desire for economy in expression, as when we say ‘The sun has moved behind the clouds so it’ll be cooler in the garden now’. We might ask, in the style of Professor Boyd, what truth is communicated by the manifestly false claim that the sun has just moved behind the clouds. (p. 25) But the answer is obvious, isn’t it?
In summary, while the justification for using the language of accommodation is partly pedagogical, the Lord speaking to his people in vivid and unpedantic ways, at the centre of Calvin’s remarks about divine accommodation is not so much a pragmatic or pedagogic as a logical or metaphysical point. It is a logically necessary condition of dialogue between people, or between God and mankind, that the partners in the dialogue should appear to act and react in time. If dialogue with God is to be real dialogue, then God’s language about himself cannot be restricted to characterising himself as eternal and immutable, but he must accommodate himself to speak in ways which are characteristic of and essential to, persons in dialogue with each other.
The main issue
I come now to what I believe is the main issue that is raised by Professor Boyd’s paper. This is not about the meaning of this or that text of the Bible, or about some point of theological or philosophical clarification, but it is a question of method, of theological method if you like. To try to put what I believe is the main intellectual difference between openness theism and classical theism, whether Arminian or Calvinist, into perspective, I wish briefly to review two or three other controversies about fundamental matters that have plagued Christians and the Christian church over the years.
There is a long-standing difference between those - let us call them Anabaptist or Baptist Christians - who while stressing the unity of the Scriptures emphasise the difference(s) between the Old Testament and the New Testament and those (let us call them Presbyterians) who though recognising elements of difference, nevertheless place greater stress on elements of continuity between the two dispensations. This stress either on similarities or on differences has implications (as each see it) for ecclesiology, including of course the ordinance of baptism, but implications which I do not need to dwell on here.
The point is: these differences in the estimation of continuity and discontinuity between the testaments cannot be settled by a simple appeal to texts in the New Testament which (as Baptists think) underline the discontinuity between them. For these very same texts will be seen by Presbyterians in a different light, in terms of how they see the relation between the testaments in terms of continuity.
Or take, as a second example, the vexed question of the canonicity or otherwise of certain texts. Many people accept the sixty-six books of the Old and the New Testaments as forming the canon of scripture. But nowhere in these books are the sixty-six books listed and a statement made to the effect that the listed books, and only these, form the canon. Indeed, any such statement, were it to have been made, would have the form of an addendum or coda to the sixty-six books themselves. And this raises an interesting logical issue. A statement of the form ‘The following books, namely Genesis, Exodus…and only these are canonical’ would only have force, or ought only to have force, if this statement is itself regarded as being canonical. But this, as is easily seen, leads to a regress. For then a further statement, validating this addendum or coda, would be needed; and a further statement, validating this statement would be needed, and so on ad infinitum. And this logical feature of the very idea of a canon of writings presents formidable problems for anyone who wishes to try to argue for this or that as the Christian canon. For how are they to argue? Only in ways, it seems, which already assume an answer to the question of canonicity; only in ways that beg the question at issue.
Or take a third example. There are deep divisions between those who affirm and those who deny the proposition ‘The character of the Christian church in the Acts of the Apostles is normative for the character of the church in all subsequent eras’. But how is this difference to be settled? Not, presumably, by reference to the documents of the New Testament, which are all documents of the apostolic era, in fact if not by definition. So how?
Now I want to suggest that the difference between open theists and classical theists is one of this kind. One which, since it is a fundamental difference about the interpretation of whole swathes of what Scripture has to say about God, appealing to Scripture in the way that Professor Boyd does cannot in fact be used to settle it.
A-data and D-data
Scripture contains such statements as ‘There is no creature hidden from his sight, but all things are naked and open before the eyes of him with whom we have to do’ (Heb.4.13); ‘God works all things after the counsel of his own will’ (Eph. 2.11); ‘And we know that all things work together for good to those who love God, to those who are the called according to his purpose’ (Rom.8.28); ‘For of him and through him and to him are all things, to whom be glory for ever. Amen’ (Rom.11.36); ‘Lord, thou knowest all things, thou knowest that I love thee’ (John 21.17); ‘According to the working whereby he is able even to subdue all things unto himself’ (Phil.3.21); ‘For if are heart condemn us, God is greater than our heart, and knoweth all things’ (1 John 3.20). Professor Boyd is surely mistaken when he claims that no passage of Scripture asserts that God foreknows everything that is going to occur. (pp.15-16) Such texts are representative of a class of scriptural data which assert, on the face of things, God’s knowledge of all things past, present and future, and his sovereign control of all that he has created. Let us call these the ‘all things’ data, A-data for short.
There are, in addition, those texts which Professor Boyd considers with such expertise in the main part of his paper, texts in which the Lord interacts with his people in very human-like ways. Let us call these the dialogue data, D-data for short. Boyd’s strategy is to diminish the force of the A-data while bringing the D-data into centre stage. In other words, his method is to allow the D-data to control the A-data, and to interpret all A-data in the light of the D-data at least insofar as these data bear on the doctrine of God. The method of classical theism is to work in the other direction, to place the A-data at centre stage and to interpret the D-data in the light of the A-data.
Neither side is going to convince the other side by swapping texts, however expertly or fully such swapping is attempted. For this debate is about how entire categories of texts are to be interpreted. Just as debates about the canon, or about the charismatic gifts, or the present-day use of the Old Testament, are debates about the status of the Bible or of fundamental parts of it, not debates about its authority as such but debates about the range and nature of that authority, so the debate between classical theism and the openness position is about which texts are to be given hermeneutical priority in our endeavour to formulate a consistent biblical account of the doctrine of God as part of developing our overall theological position.
So while I admire and applaud the detail of Professor Boyd’s paper, and the ingenuity he shows in the interpretation of the various texts he examines, I want to suggest that regrettably such effort is, in the context of this particular debate, so much wasted breath. For the question is, which motif, the control and knowledge motif, or the creative flexibility motif, (p. 34) is to have priority. And the examination of however many individual texts cannot settle that question.
Boyd claims that on this question of the two sorts of data the openness position does not distinguish between literal and anthropomorphic language but ‘Generally speaking, it interprets the second motif the same as it does the first. It believes that both motifs can and should be interpreted as describing what God and creation really are like.’ (p.23)
But anyone can see that this is not the case, and that it cannot be the case, simply because the two sorts of data present prima facie contradictions. In this situation of apparent contradiction something must give, and in Boyd’s treatment of the A-data, something does give. To start with, Boyd’s treatment of the A-data is highly selective. There are certain obvious passages of Scripture which are A-data which he does not refer to. But in the second place, in order to avoid the contradiction between the A-data and the D-data when each are understood without qualification, he has to interpret the A-data somewhat elastically. Thus Boyd says that the scriptural motif which exalts God as the sovereign Lord of history ‘clearly entails that much of the future is settled ahead of time and is therefore known by God as such’ (p.15) And here it is evident that Boyd is not taking the A-data with the same literalness as the D-data. For the A-data do not assert that much of the future is settled ahead of time they assert that all of it is. For some reason, Boyd simply denies that there are such passages in Scripture. ‘Neither this (Isaiah 46.9-11) nor any other passage in Scripture says that God foreknows or declares everything that is going to occur” (p.16). So Boyd must be saying that the ‘all things’ texts, the A-data, have another meaning than the meaning they in fact have when interpreted literally.
Or take Boyd’s more detailed proposal regarding Peter’s statement about God’s foreknowledge of Christ’s death (p. 22). He says that in his view that ‘saying that someone carried out a predestined or foreknown wicked event is much different from saying that someone was predestined or foreknown to carry out a wicked event. Scripture affirms the former but not the latter’ (p. 22). So while Peter refers to foreknown actions, the Lord leaves open which individuals perform these. This conveys the distinctly odd idea that God foreknows wicked acts which are in search of owners! Whether this interpretation is plausible is one thing. (It seems to be at odds with other things that openness theism claims; for suppose that one after another, the likely candidates for being wicked crucifiers of Jesus reject that dubious honour. What then?) But what is clear, whatever may be its merits as an interpretation, is that Professor Boyd’s suggestion represents a considerable gloss on Acts 2.23. Peter does not say anything that approaches what Professor Boyd claims is implied by what he says.
These remarks are not to be taken as in any way implying a criticism of Professor Boyd’s handling of these texts, or of any other texts. The remarks are about method. Professor Boyd thinks he’s interpreting both sets of data evenly. But in fact he is giving priority to one set of data, the D-data, over the A-data. This is not clumsiness or ineptness on his part; rather, it is something that he cannot avoid doing if he is to avoid blatant self-contradiction. And the same goes for the proponents of classical theism who claim, as we have seen, that in certain texts God is accommodating himself to human understanding and to the conditions of human existence.
Theism and Philosophy
How are we to approach the difference in method between the classical theist and the openness theist? I am not at all sure. But I am sure that certain ways that are or have been offered of settling the differences won’t do. I’ll mention a couple of these.
One general suggestion for attempting to break the impasse that has been proposed is that in interpreting the Scripture one should interpret the less clear passages of scripture by the clearer. But of course this simply invites each of the parties to the dispute to beg the question in their favour, for each of them will claim that different passages are clear, and different passages are less clear.
Another favourite ploy, going back at least to Adolf Harnack, and currently employed with considerable gusto by the openness theists, is to blame philosophy, particularly Greek philosophy, for the crime of imposing alien forms of thought on the pure word of God. Professor Boyd himself appears to endorse this idea when he says ‘exegesis should always drive our philosophy, instead of the other way around’ (p.14) and that classical theism has developed ‘overly motivated by philosophical considerations’. (p.47) He says that classical theism is motivated by a philosophical presuppositions of what is worthy of God (p.35), the influence of ‘ancient Greek philosophers’ (p. 40). But of course (as he sees at the end of his paper) Professor Boyd cannot himself avoid philosophical preconceptions and preferences in constructing his openness position. Whether the philosophical preconceptions that he favours he adopts amounts to ‘sound philosophy’, (p.47) as he claims, is another question.
The impression that he conveys is that in the case of classical theism the philosophical influences were both unsound and also that they were imposed from above. The philosophy drives the exegesis. Whereas in the case of open theism the philosophy is sound and provides support for the openness position from below. But in fact whichever philosophy is sound, philosophical influences play exactly the same role in the openness position and in classical theism.
How so? It is a plausible principle that if p is a metaphysical position, then the denial of p, not-p, is also a metaphysical position. So if it is a metaphysical position that human persons are essentially immaterial then the denial of that position is likewise metaphysical. If it is a metaphysical position that God cannot change, then the denial of that position, the claim that God can change, is likewise metaphysical. And if I understand the position correctly Professor Boyd and the openness theists assume from the outset that God changes in a similar sort of way to that in which you and I change. Or at least they assume that it is possible for God to change in this way. They may say that they do nothing of the sort, that such a view of God is drawn by careful induction from relevant biblical data. But the biblical data, however carefully interpreted, will only yield this theological conclusion, the openness conclusion, if the protagonists of that position assume ab initio that such a view of God is cogent. Precisely the same is true with classical theism.
Why is this? It is because philosophy deals, inter alia, with what is possible. And what is possible, while it does not dictate what is true, determines what can be true. Thus if it is possible that God changes his mind, if this represents a possibility, then biblical texts which indicate such a change may be interpreted literally, as reporting such a change. But if it is impossible for God to change his mind, then biblical texts which report such changes must be interpreted non-literally.
Of course if as a matter of fact all the language about God in the Bible implied that God changes somewhat in the same fashion as we do, then it would be easier to assent to the metaphysical view that God changes. But even such uniform language about God would not entail that God really changes, for it is possible to believe that all such language about God in scripture is non-literal. Moses Maimonides seems to have held a view rather like that. But the debate between classical and openness theists is acute precisely because each appeal to the same data, and these data are not uniformly of one kind; some of them seem to favour the openness position, others seem to favour the classical position. Hence the importance that is played by the acknowledgement or denial of certain metaphysical possibilities not so much in indicating the correct way forward as in ruling out certain ways.
Whether these respective metaphysical assumptions about the nature of God ought to be thought of as being impositions on the biblical text or not, one thing is sure; that if classical theism can fairly be convicted of imposing alien ideas on the biblical text, then so can openness theism. And if the openness theists are to be exonerated from this nefarious activity, then so are the classical theists.
So, if we cannot appeal to the clarity of certain texts by which to interpret the less clear, and if we cannot dismiss one view as being corrupted by false philosophy, while the other view is purely biblical, how are we to proceed?
Theism and Christian supernaturalism: the analogy of faith
Earlier I mentioned that developing a doctrine of God from scripture is only part, though an important part, of developing an overall theology. And I believe that the best prospect of a genuine debate between openness theism and at least some proponents of classical theism is to widen the terms of the debate, not to restrict it to the doctrine of God, but to show what the implications of each view are for other areas of Christian theology; soteriology, for example.
I suggest that what is at the heart of the contrast between openness theism and classical theism is not only or principally a different understanding of the nature of God, or even a difference of interpretation of this or that passage of Scripture, but a profoundly different appreciation of the plight of mankind and the saving grace of God.
On the openness view however, in detail, divine saving grace is understood and defined, it is that action of God which is causally necessary for human salvation, but not causally sufficient. The human will can always frustrate the grace of God. Let us suppose, with all evangelical Protestants, that it is faith in Christ that is the instrumental cause of a person’s salvation. Then on the openness view God’s grace can only ever be causally necessary for the production of such faith, not causally sufficient. What, in addition, is causally necessary, is an autonomous free choice on the part of the would-be believer, a choice which, if exercised, could frustrate God’s loving purposes towards that individual. Divine grace and such a choice are then together causally sufficient for faith in Christ. Without such grace no human being would come into a right relationship with God, but by itself such grace is never sufficient. For even when such divine grace is exercised, human beings have the power to resist or thwart such grace from God. And given that mankind has a nature which is antipathetic to the rule of God, we might expect such power to be exercised in the rejection of the overtures of grace.
It is not clear whether libertarian freedom is a premise or a conclusion of the openness position. But whichever it is, there is no good reason to be committed to the openness position without being committed to libertarian freedom, with all that this implies about the nature of divine grace.
For the openness position, the plight of man cannot be so great, (for he possesses inalienable freedom to choose for or against Christ), and the power of God need not be so great either. There is no need of a covenant of sovereign grace to which God is immutably faithful, and of divine promises on which the sinner can utterly rely. In particular, there is no need for saving grace from God which is efficacious in renewing an alienated will. This is why, ultimately, the issue of openness in connection with the doctrine of God, isolated from other facets of the Christian faith, implies a superficial view of the plight of man and the power of God. On a more radical view, human need is such that men and women cannot want to want God, and his grace is needed not only to help them to want to want him, but to ensure that they do so. So the claim that such incompatibilistic freedom is consistent with divine omniscience, as Professor Boyd claims at the start of his paper, even if correct, does not get to what, for the Christian, should be the heart of the issue; for only the efficacious grace of God can ensure the salvation of a person in bondslavery to sin. And for God’s grace to be efficacious God needs to be able to work all things after the counsel of his own will.
Boyd claims that the origins of the idea of divine immutability go back to ancient Greek philosophers who sought to demythologise popular stories told about the gods. (p. 37, footnote 18). But perhaps he does not fully reflect on the significance of this fact. For the gods of classical mythology were subject to fate, to natural forces, and to much else. They were, in a real sense, part of the natural order. By contrast, classical theism is not so much a product of natural reason (p. 40) as an effort to distance the God and Father of Jesus Christ from the gods of Canaanite religion and of Greek mythology. For only classical theism provides a clear conceptual safeguard in the doctrine of God for that supernaturalism that is at the heart of the Christian faith. Only on this view can God escape being a prey to time and change, to frustration and disappointment, as the gods of Greek mythology were, and as the God of openness theism is. This is a recurring theme in Professor Boyd’s paper, the theme of God’s frustration with the human rejection of his influence (p.28f.). Is not openness theism in danger of remythologising the God of classical theism?
I believe that the same deep-seated motivations or intuitions were at work in the formulation of classical theism as were present in the development of the doctrines of the Trinity and the Incarnation, namely to safeguard the fully divine character of our salvation, its monergism. In carrying out their work the fathers, no doubt, used whatever appropriate conceptual tools were to hand. The terminology was, in any case, flexible and open to redefinition, as Professor Gerald Bray has recently emphasised.  What is more significant than their use of the philosophical terminology of the Greeks is that in putting it to use the fathers had the sure instinct for safeguarding the purity of biblical theism and the supernaturalism of our salvation which such theism makes possible.
Classical theism and Christian supernaturalism go together. Two sorts of queries might be raised about this. First, it might be pointed out that Arminians hold that classical theism is consistent with certain forms of synergism. But whether they are correct in this is most certainly debatable. And secondly it might be argued that a full Christian supernaturalism could be made consistent with the idea of a God who chops and changes in the exercise of creative flexibility. But while one could certainly attempt to develop an openness type theism in the interests of safeguarding Christian supernaturalism there would be little point in doing so. For to invoke such a hypothesis, the hypothesis of the ever-resourceful God of the sort portrayed by openness theism, in order to safeguard Christian supernaturalism would be an unnecessarily complex and pointless strategy. What could possibly be gained by it? The God of classical theism is the simpler hypothesis, the hypothesis which best safeguards the monergism of salvation by grace. ‘But our God is in the heavens: he hath done whatsoever he hath pleased’ (Ps.115.3).
In the main part of this paper I have tried to argue that what is at issue between classical theism and openness theism cannot be settled by a direct appeal to the Scriptural data about God. If there is to be real dialogue the terms of the debate must be widened from the doctrine of God narrowly considered to other areas of Christian theology, and I have tried to illustrate this from soteriology. Classical theists must hope that once openness theists see the full implications of postulating a God whose saving purposes can always be frustrated by the will of the creature they will have second thoughts.
London WC2R 2LS
 There are other claims which, if space allowed, one might take issue with. Thus Professor Boyd seems to confuse indeterminism with the absence of prediction. (pp.5-6) Failure to predict does not signal libertarian freedom, only (maybe) human cognitive limitations.
 This section is an adaptation of material which first appeared in ‘God in Dialogue’ in Interpreting the Bible (ed. A.N.S. Lane), Leicester, Apollos (1997)
 Institutes of the Christian Religion trans. F.L. Battles (London: S.C.M.Press, 1960) I.17.13.
 Institutes I.17.1
 Commentary on Isaiah (Calvin Translation Society, reprinted Grand Rapids, 1979) III.158
 Gerald Bray, The Personal God (Carlisle 1998) p. 50. He cites Christopher Stead, Divine Substance (Oxford 1977)