The Augustinian-Calvinist View
4
THE AUGUSTINIAN-CALVINIST VIEW
Paul Helm
have been invited to contribute to this volume an essay on divine foreknowledge and human freedom as viewed from the tradition of Augustine of Hippo and John Calvin. The issue, as classically stated, is whether divine omniscience, as far as it is concerned with the future, is logically consistent with human freedom, where this is understood in an incompatibilist sense.' The endless discussions of this issue have been motivated by the underlying assumption that a person can be held responsible only for those actions that are incompatibilistically free. So if divine omniscience is inconsistent with such freedom, then two unappetizing alternatives stare us in the face. Either we shall have to accept a modified, reduced account of divine omniscience or we shall have to accept that human agents are never responsible for what they do since they are never incompatibilistically free. Some have indeed taken the first of these alternatives;' I know of no one who has opted for the other.
On the view to be expounded in this chapter these alternatives do not confront us, for according to this view there is no need to claim that divine foreknowledge is logically consistent with human indeterministic freedom, and there is good reason for not doing so.3 I will present three broad arguments for the conclusion that a Christian theist may accept a different sense of "free," a compatibilist sense, and that indeed there are good reasons to think that Christianity requires such a view. The arguments are broad in the sense that it is not possible, in a piece such as this, to dot every i and to answer or anticipate every objection. Of these three arguments, the first has a more theological character to it than the others in that it depends crucially upon a particular view of God's grace. It offers an argument that should, other things being equal, motivate a Christian to accept a compatibilist account of human action. The second and third arguments are more strictly philosophical. One is an argument for the consistency of divine foreknowledge and human compatibilist freedom from the idea of divine perfection and the principle of simplicity; and the other is an argument for the logical inconsistency of divine foreknowl edge and incompatibilism.
These arguments are consistent with each other, and therefore they may, taken together, be said to present a cumulative case for the Augustinian view. But I do not claim that together they make the Augustinian view compelling; my only claim is that together they make a defensible and reasonable case for the view that divine foreknowledge is consistent with human freedom when this freedom is properly understood. If this case is accepted, then there is no need for, and no advantage in, either trying to provide an understanding of human freedom that is indeterministic or trying to argue that such freedom is logically consistent with divine foreknowledge.
Divine Foreknowledge
I shall begin by saying something about the meaning of "divine foreknowledge." It is possible to distinguish at least three different concepts of divine foreknowledge-at least three different senses in which God may be said to foreknow. One is the straight causal sense: God's knowledge is the cause of things, as Thomas Aquinas famously put it. On this view there would appear to be no distinction between what God causes and what he permits, because, on Aquinas's view, since God foreknows all events, he must cause them all. (Aquinas does in fact allow a sense of divine permission: since God cannot cause evil, since evil is not anything that is positive, God may be said to know it without causing it. We shall have more to say on divine permission later.) A second sense is that in which the foreknowledge of God is logically subsequent to his decree. On this view God's decree is all embracing, and his foreknowledge is simply his knowledge of what he has decreed before that decree takes effect in time. Thus while on this view divine foreknowledge and divine foreordination are necessarily coextensive, they differ in meaning. The third and weakest sense is a sense of divine foreknowledge that is logically prior to God's decree; it is in the light of what he foreknows that God decrees this or that; what he decrees is conditioned by what he foreknows.
My arguments entail, as a conclusion, an understanding of divine foreknowledge according to which at least some instances of divine foreknowledge, as far as they concern human free acts, are to be understood in one or other of the first two senses; these instances are inconsistent with the third sense of foreknowledge. None of the arguments requires one particular account of divine foreknowledge as a premise. Before setting out these arguments and defending them, since this chapter is offered as a contribution to a debate among Christians, I wish to say something further about what I take to be the Christian parameters within which this whole debate, and particularly the Augustinian contribution to it, is set forth.
The Fixed Points
In setting out to reflect, as Christians, upon the relation between God's knowledge and power on the one hand and human responsibility on the other, it is important that we understand what sort of task we face. Christians do not reflect on these concepts in a purely abstract way and then try, in a wholly a priori fashion, to establish the concepts' consistency. For we are constrained by the biblical witness insofar as it refers to and makes use of these ideas. Christians must hold this witness as a set of fixed points, and then we must reflect upon the cogency or coherence of the several parts of that witness. It is perfectly possible that at the end of these reflections we shall reluctantly have to draw the conclusion that we cannot at present see how these parts cohere, that we cannot demonstrate their consistency. Perhaps we shall have to be content with showing that the ideas are not inconsistent. (Think of similar difficulties that may be encountered in trying to understand the doctrine of the Trinity or the incarnation.) But it is not open to us to amend or modify that witness in any way in the interests of greater comprehensibility. Faith seeks understanding, but the understanding gained must not be at the expense of the faith.
What then are these fixed points? In brief, the scriptural worldview is one in which all things are created by God and ordered and governed by him; he numbers the hairs of our head, directs the fall of a sparrow and the flight of an arrow; he turns the hearts of humans as he wishes; like a potter, he has power over human clay. It is true that God is said to forget, to be surprised and to act and react toward his people in blessing and chastisement. These may appear to be the actions of someone with limited knowledge and power, but the scriptural language in such cases is usually recognized as metaphorical or symbolic, language "accommodated" to some human situation or need. I shall not at this point consider further the justification for regarding such language in this way nor the implications of such language for our understanding of God. I will try to say something about these matters in my rejoinders to the other views.
We may think, on purely a priori grounds, that the exact arrangement of molecules in a jar of sugar or of flecks of dust dancing in the sunlight is beneath God's notice. And it may seem that as far as God is concerned nothing turns on there being one arrangement of sugar molecules in the jar instead of another. But no scriptural statements about the nature and character of God suggest that his divine ordering and control of his creation are qualified or attenuated in even such marginal ways; nor, indeed, does any passage of Scripture say that they are not. Perhaps these matters are literally beneath his notice. But suppose that they are. Then Scripture appears to affirm that everything that is worth God's notice is governed by him. So God's universal and yet particular providence is one fixed point that our theorizing about divine knowledge and freedom should not attempt to overturn.'
On the other hand, Scripture teaches that although God ordains everything which comes to pass-even the evil actions and omissions of human beingsmen and women are nevertheless accountable to God for their actions and omissions, though they are not accountable for all of these, because a distinction may be drawn between matters that are outside a person's control and cancel responsibility and those matters that are under a person's control. Scripture also has a place for degrees of responsibility or accountability. A person may be more or less accountable for an action, as Scripture also teaches; responsibility may be partly canceled, for example, by partial ignorance.
We need, also, to make a distinction between two kinds of the responsibility. The first is the responsibility we each have before God: an accountability that depends upon God's knowledge of all circumstances, including the thoughts and intentions of our hearts, and that is exercised in accordance with God's loving, wise and righteous character. (I shall call this GR for "responsibility before God.") The second is the responsibility we have as judged by human agencies; by other human beings, such as parents and friends; and by the police and other law-enforcement agencies. (I shall call this HR for "responsibility toward humans.") Ascribing responsibility to oneself or to others is a human, social activity having serious limitations but an activity necessary for carrying out certain social functions and achieving certain ends. Ascribing responsibility under these conditions performs a different function than accountability before God does, a function that has as one of its primary aims the well-ordering of society.
It is important to stress that the basic biblical emphasis, certainly in the New Testament, is upon our accountability to God. Because the moral accountability of one person to another, or legal accountability to the state (HR), depends upon fallible human knowledge of the motivations and circumstances of ourselves and others; each can only be pale and imperfect reflections of GR, the accountability each of us has to God. Human justice and injustice are, when compared to divine justice, necessarily rough and ready.
So when Christians are exploring the relationship between divine knowledge and human responsibility, they might be doing one of two different things. First, they might be exploring the implications of divine knowledge for HR as it is exercised, say, in families or between friends and colleagues. This idea of moral responsibility is a familiar one and needs to be distinguished from legal responsibility, to which it is related. But both moral and legal responsibility need in turn to be distinguished from human responsibility before God.
Second, Christians might be exploring the implications of divine knowledge and GR. I take it that when philosophers and theologians, motivated in part by a desire to preserve human responsibility, have attempted to show that divine foreknowledge and human incompatibilism are consistent, it is GR that they have primarily in mind. A sense of responsibility that is not responsibility before God is not the sort of responsibility we ought to want to have.
Why may GR and HR differ? For the following reason: in order to hold others responsible for their actions, it is necessary that one know their state of mind and the circumstances of their actions (as we shall note later). Social and family imperatives may require the ascription of responsibility when full knowledge is unobtainable; perhaps full knowledge is never attainable by us. In such cases what we may have is an ascription of HR and not GR. In some particular case perhaps human and divine verdicts coincide, perhaps not. Thus certain kinds of HR may require strict liability, a responsibility for a particular action X that stands even though a person did not want X to happen and took all reasonable steps not to let X happen. But it is hard to see that there is any case of GR based upon strict liability in this sense.
In the case of GR it must be presumed that God himself has perfect knowledge of the state of mind and the circumstances of every agent and every action. Not only does he have knowledge of the facts, of course, but he also carries out his judgment in an immaculate fashion, not in the warped and biased way that human justice is often meted out. Part of the "righteous judgment of God" is, thus, that it is "according to truth," based upon all and only the relevant facts, and is founded upon perfect justice.
It would be perfectly possible to maintain the importance of GR while being skeptical about the basis on which judgments of HR are made. Indeed, when the apostle Paul urges his readers to "not pronounce judgment before the time" (1 Cor 4:5), he appears to be adopting such a skeptical position; this is not to say that HR bears no relation to GR but to say that the judgments of HR, because of our imperfect knowledge of ourselves and others, are only partial and provisional, open to revision at "the time." Perhaps this should be the general Christian attitude to HR.
Divine sovereignty and human accountability before God-these are two of the fixed points. When we are faced with problems about the consistency of these concepts, it is tempting to modify one or both of them. But we must make every effort to avoid such a course of action. Scripture holds them together, it even speaks of them in the same breath, and so must we. For if Scripture teaches them in this way, they must each be true and so together be consistent, even though it may be difficult for us to grasp this now.
Nowhere is this more noteworthy and striking than when Peter tells his listeners, "This man [Jesus], handed over to you according to the definite plan and foreknowledge of God, you crucified and killed by the hands of those outside the law" (Acts 2:23). Here a divine verdict is passed on the actions of the crucifiers. Whether or not the culpability of those responsible for the crucifixion was a case of HR, it was certainly a case of GR. From this text it is possible to draw the conclusion that there are at least some occasions when the action of a wicked person (in the sense of GR) is the result of the set purpose and foreknowledge of God, for one such occasion was the crucifixion of Christ, the focal event of the Christian faith.
It is tempting to say about the crucifixion of Christ that if the men who crucified Christ were wicked (in the GR sense), then they cannot have acted as a result of God's set purpose. It is equally tempting to say that if the crucifixion of Christ came about as a result of the set purpose of God, then the men who crucified him cannot have been wickedly responsible (in the GR sense) for what they did. But we must resist the temptation to reason either of these ways.
There is a third fixed point, one that perhaps we reflect upon less frequently: God's relation to the universe that he has created and that he sustains and directs is a relation without parallel. It is unique, incomparable, sui generis. This is not surprising, or it ought not to be; after all, there can only be one God, and if by "the universe" we mean the totality of whatever exists apart from God, there can only be one universe. We cannot have experience of more than one God nor of more than one universe.
Those critics of the "argument from design," such as Philo in David Hume's Dialogues Concerning Natural Religion, make an important point when they argue that we do not, in the nature of things, have experience of more than one world. Cleanthes claims that the universe reveals such marks of intelligent design as to make it likely that it has a designer, to which Philo replies:
Having found in so many other subjects much more familiar the imperfections and even contradictions of human reason, I never should expect any success from its feeble conjectures and in a subject so sublime and so remote from the sphere of our observations. When two species of objects have always been observed or conjoined together, I can infer, by custom, the existence of one wherever I see the existence of the other; and this I call an argument from experience. But how this argument can have place where the objects, as in the present case, are single, individual, without parallel or specific resemblance, may be difficult to explain. And will any man tell me with a serious countenance that an orderly universe must arise from some thought and art like the human because we have experience of it? To ascertain this reasoning it were requisite that we have experience of the origin of worlds; and it is not sufficient, surely, that we have seen ships and cities arise from human art and contrivance....
This argument reminds us of the general danger of modeling the divinehuman relation either on the relations between human beings or on relations between human beings and artifacts. For just as none of us has had experience of many universes but yet we recognize the singularity of this universe, so the relations between God and the universe are singular and cannot be closely modeled on the relationship between intramundane agents.
In our remarks about responsibility we have already seen a need to distinguish between the conditions under which a person may be held responsible for an action within a human society, such as the state or a family, and the conditions that obtain in the case of responsibility before God. When faced with the scriptural claim that the actions leading up to and including the crucifixion were in accordance with the determinate counsel and foreknowledge of God, it is tempting to infer that in the crucifixion God "manipulated" the crucifiers or that they were his "puppets." To say such a thing would be to apply to God concepts derived from (let us say) the behavior of a magician or an entertainer. And this is to forget that those factors which distinguish God from humanity are at least as great as those factors which they have in common. We must therefore resist the temptation to say that in all cases in which one person acts as a result of the set purpose of another intelligent agent, that person is thereby being tricked or made into a puppet.
Emphasizing the sui generis character of the relation between God and his creation is simply another way of saying that God's relations with his creation are incomprehensible. We cannot get our minds around them so as to fully understand them because they are unparalleled in human experience. The divine activities are not subsumable under laws, and the creative and interactive activities between human beings only provide a faint reflection of the profundity of the relation between the eternal Creator and his creation. But in saying that God's relations with his creation are incomprehensible, it needs to be borne in mind that comprehensibility is a matter of degree; for a state of affairs to be incomprehensible is not necessary for it to be fully or totally incomprehensible gibberish, but rather for it to be not fully comprehended.
We are mistaken in reasoning that because God's relations with his creation are sui generic-and that because, therefore, we cannot fully comprehend them-we must inevitably fall into paradox, self-contradiction or other kinds of nonsense whenever we attempt to think or speak about them clearly. In what follows we shall try to comprehend aspects of the relation of divine sovereignty and human responsibility while recognizing that we can never, in the nature of things, fully comprehend that relation.
Let us now turn to the arguments.
Argument 1: Foreknowledge, Freedom and Divine Grace
What motivates the Augustinian view at the most fundamental level, I shall argue, is a particular understanding of God's saving grace, an understanding that, this view holds, is the biblical view. I shall briefly elaborate this view of grace and try to show that it is reasonable to think it is consistent with compatibilism, the idea that human actions are free in a sense that is consistent with determiniSM.6 Basic to the difference of view between those who think that divine omniscience is consistent with human incompatibilism and those who think that it isn't, is not principally a different understanding of the nature of God or of human freedom, nor even a difference of interpretation of this or that passage of Scripture, but a profoundly different appreciation of the plight of humankind and the power of God.
One can take for granted on the incompatibilist view of human freedom that however divine saving grace is understood and defined, it is this action of God that is causally necessary, but never causally sufficient, for human salva tion. Let us suppose, with all evangelical Protestants, that faith in Christ is the instrumental cause of a person's salvation. Then on an incompatibilist view of freedom, God's grace can only ever be causally necessary, not causally sufficient, for the production of such faith. For otherwise faith in Christ would not be a free act. For on the incompatibilist view of freedom what must, in addition, be causally necessary for receiving God's grace is a free, incompatibilist choice on the part of the would-be Christian. Divine grace and such a choice are then together causally sufficient for faith in Christ, for the personal appropriation of Christ. Without such grace no human being would come into a right relationship with God. But on the incompatibilist view by itself, such grace is never sufficient. For even when such divine grace is exercised, given that humankind has a libertarian or incompatibilist will, such a will has the power to resist or frustrate such grace from God. And given that humankind has a nature that is antipathetic to the rule of God, we might expect such power to be exercised in the rejection of the overtures of grace. With such freedom, God's saving grace is always resistible, and so saving grace can never ensure its intended effect.
That is the first point. The second point is that according to Scripture such irresistible grace-the grace that alone is causally sufficient for faith and for a faithful appropriation of Christ-when it is received, is liberating. By receiving it a person is freed from his or her slavery to sin and granted spiritual freedom: the freedom to willingly and gladly serve the living and true God. This sense of "freedom" is, of course, rather different but nonetheless related to the sense of "freedom" that is at issue in the controversy between compatibilists and incompatibilists. It is freedom with a spiritual and moral dimension, freedom necessary to be remade as a person. So not only is the receiving of such grace inconsistent with a libertarian conception of freedom, but there is a further sense of freedom, a moral and spiritual sense, according to which one is free only when one has received such grace. Now if such grace alone ensures true human freedom, as the Augustinian maintains, and if the operation of such grace is inconsistent with incompatibilism, then such incompatibilism cannot itself be necessary for the maintenance of, or the achieving of, true human freedom.
How are we to understand such efficacious grace? Here is a typical expression of the way in which Augustine understands it:
Unless, therefore, we obtain not simply determination of will, which is freely turned in this direction and that, and has its place amongst those natural goods which a bad man may use badly; but also a good will, which has its place among those goods of which it is impossible to make a bad use-unless the impossibility is given to us from God, I know not how to defend what is said: "What hast thou that thou didst not receive?" For if we have from God a certain free will, which may still be either good or bad; but the good will comes from ourselves; then that which comes from ourselves is better than that which comes from Him. But inasmuch as it is the height of absurdity to say this, they ought to acknowledge that we attain from God even a good will.7
So if God's saving grace is efficacious in renewing the human will, then it is sufficient to attain the end for which it is given: the salvation of the individual to whom it is granted. It is reasonable to suppose that in bringing his saving grace to men and women, God has certain purposes, salvific purposes. And if these purposes are salvific and if grace is efficacious, then the provision of grace is sufficient for attaining that salvific end.
The incompatibilist holds that because the plight of humans is not so great (for they retain inalienable freedom to choose for or against Christ in an incompatibilist sense), 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 or of divine promises on which the sinner can utterly rely. More particularly, there is no need to receive efficacious saving grace from God. This is, ultimately, why the insistence upon incompatibilism in connection with the issue of divine foreknowledge and human freedom is so superficial. For the Augustinian account of the plight of humankind is such that people cannot want to want God, and his grace is needed not only to help them to want to want him, but also to ensure that they do so. So the claim that such incompatibilistic freedom is consistent with divine omniscience, even if it were convincing, would 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 slavery to sin.
What of the objection that some human partners do resist God's graceapparently successfully? The point about efficacious grace is, in a manner of speaking, a grammatical one. By definition, efficacious saving grace succeeds in the way indicated. Those who successfully resist the grace of God do so because in such cases that grace is not efficacious. For not all God's actions are efficacious in the sense that they bring the person on whom they act to true human freedom. Some of God's actions are resistible and are resisted. What does this imply? Not that there are no irresistible gracious divine energies but simply that those which are resisted are not among them.
Those discussing the idea of divine grace sometimes stress its personal character and, in particular, the view that because grace is personal it is therefore relational. This may be helpful, for grace is relational; it relates the divine persons and human persons. But saying this by itself does not help us forward. The question is, what sort of relationship is a personal relationship? Is it a relationship in which the human partner can always resist the divine partner or not?
The chief difficulty over stressing the idea of saving grace as being basically relational concerns the idea often associated with it: that parity between the partners and the ensuing reciprocity are necessary for any genuine personal relationship to be established. Of course it is possible to define a personal relationship in this way, in terms of the parity of those in the relationship, but the result is that whoever insists upon such a definition will simply misdescribe many genuine human relations in the process. For such a definition is surely unconvincing as a claim about ordinary human relations and unequally unconvincing about relations between human beings and God. You may have a relationship with another human being who cannot reciprocate: for instance, with a baby, or with someone of very low intelligence or with someone who is in a coma because she or he has fallen through the ice and is stuck beneath it. Someone in a coma has to be brought back to consciousness and, essentially, to life; but this process is itself a case of exercising a personal relationship. In a parallel way, all that is implied in divine rescuing and bringing to life is the establishing of a relationship, the key to which is not incompatibilistic human freedom but the unilateral establishing of a loving relationship that will not let go until it has secured reciprocal love by recreating it. According to Augustinianism, God in grace establishes a genuinely personal relationship with people who are incapable of establishing it for themselves. And God does this by an act of condescension and power-condescension because he is the infinite Creator and has no obligation to do what he does, and power because our plight is such that without that power the relationship could be neither established nor continued. This is a different sort of divine-human relationship than the one usually envisaged by those who claim that God's relations with humankind are personal relations. But isn't it the biblical sort?
I think we can now dismiss the theological argument for the consistency of divine foreknowledge and human incompatibilistic freedom. Divine efficacious saving grace is inconsistent with such freedom; and as (on the Augustinian view) such an account of divine grace is a fundamental feature of the Christian way of salvation, the Christian faith does not need to posit incompatibilism; in fact it needs to posit compatibilism.
Argument 2: Divine Perfection and Providence
My second argument, one that is more strictly philosophical, is derived from the ideas of divine omnipotence and omniscience. We may distinguish between the connotation and denotation (or between the sense and reference) of expressions such as "an omniscient being" or "an omnipotent being." For obvious reasons, let us chiefly consider omniscience, while not ignoring omnipotence altogether. And let us assume, fairly uncontroversially in present circumstances, that the denotation of "an omniscient being" is God alone.
The connotation of omniscience, "knowing all things" (Shorter Oxford Dictionary), may also seem uncontroversial. If God is omniscient, then he has unlimited knowledge. But (as is well known) matters become more difficult when we inquire more precisely into the connotation of omniscience; what exactly are the "all things" that omniscience knows? For example, is what God can know "limited" by powers possessed by beings other than God, particularly the power of incompatibilistic free choice? Can an omniscient God know what time it is now? Can he know what it is like to be a bat? Two people may each agree that God alone is omniscient but disagree on the connotation of the term. Thus while Aquinas believed that God knows all future free human actions, others, such as Richard Swinburne, believe that he does not. So there is disagreement among philosophers about the connotation of omniscience as this applies to God.
Omniscience might be defined as the knowledge of all truths. But omniscience is limited by what it is possible for the omniscient knower to know. Some have argued that it is not possible for God to know the entire future since there are parts of the future that are unknowable, and that it is not possible for him to know what it is like to be a bat, or what I am doing now, and so on.
So there is scope for disagreement about the precise connotation of omniscience. However, it seems a reasonable principle, in reflecting upon the concept of God in philosophical fashion, that the connotation of "omni-" term such as "omnipotent" and "omniscient'-should, when applied to God, be as wide in their connotation as possible. (Let us call this principle A.) Thus the term omniscient is more appropriately applied to God when it connotes knowledge of more types of actions and events than when it connotes knowledge of fewer types of actions and events. Further, the term is more appropriately applied to God when it connotes knowledge of more tokens of each type of action than knowledge of fewer such tokens. After all, the rationale for applying such "omni-" terms to God is to convey the idea of maximality; so it must be reasonable not to limit their application unnecessarily, otherwise the danger is that such terms, when applied to God, come to possess only rhetorical or hyperbolical value. The presumption must be, therefore, with respect to any type of event and to any token of that type, that an omniscient being has knowledge of them, knowing the truths that are the correct descriptions of all such actions, or that he is directly acquainted with all the relevant states of affairs. Of course omniscience may be said to extend further than knowledge over what is the case; it extends to what might possibly be or have been the case. An omniscient being knows not only all actualities but the contents of all possible worlds. This point will become relevant when we shall shortly consider Molinism.
I am claiming, then, that wherever possible one should interpret the term omniscience when applied to God as generously as possible, pushing its connotation as far as one can, unless there are overriding reasons not to do so. But it is at this very point that controversy begins. It seems clear to many that there is such an overriding reason in the case of human incompatibilistic freedom, for as we have seen in their view such freedom is necessary for human responsibility and (in some theological systems) necessary for the achieving of merit.
Given the common Christian view that God is omniscient and given principle A, it is a priori desirable that divine omniscience should extend to as many facets of creation as possible. It is undesirable that God knows about only some of the things that he has created, presiding over a creation in which some things happen that he does not know about before they occur. For an omniscient being who does not know the future course of free human actions is "less" omniscient than an omniscient being who knows the entire future.
Let us call the degree of knowledge that encompasses all events past, present and future "strong" omniscience and call anything less "weak" omniscience. My argument is that the ascription to God of strong omniscience, as part of ascribing perfection to God's nature, is most at home with compatibilist views of human freedom. It is best to think of a perfect God as one who is essentially strongly omniscient.
But is not essential strong omniscience inconsistent with other essential features of God's character, notably his goodness as expressed in his perfectly righteous and holy moral character? If God is essentially strongly omniscient, and so knows about all events before they occur, then surely he is implicated in the fact of evil and his essential goodness or righteousness is compromised.
At this point many argue that it is possible to preserve the divine righteousness only by postulating a class of actions-indeterministically free actions that God cannot actualize but only weakly actualize and that therefore he cannot govern in a positive sense, a sense required by universal and particular providence. Currently many who opt for this position-the compatibility of divine omniscience and human freedom-opt for Molinism, some version of the idea of middle knowledge.
If one accepts the distinction between God's free knowledge, natural knowledge and middle knowledge, then there are, in my view, two principal areas of concern with Molinism. The first is whether the idea of divine middle knowledge of free human actions is possible (a doubt raised by Robert Adams, among others); the second is how God performs the feat of actualizing possible worlds that include libertarian actions without thereby infringing such freedom. Let us suppose God knows that "if A were to be placed in circumstances C, he would freely do X." Then since God has no control over the truth of such counterfactuals, how exactly does God actualize this item of his middle knowledge, should he wish to? What exactly is it to actualize, even to "weakly" actualize, such a possible state of affairs?
The picture one gets from this is of God's knowing all possible worlds, including those that allegedly contain counterfactuals of freedom. These worlds have, so to speak, already run their courses, for they are complete, filled out in every detail. They are, we may suppose, exhaustively describable by a set of unconditional propositions. So the counterfactuals of freedom such worlds may initially be thought to contain are not really counterfactual; what is conditional is the possible world, not bits of it. It is as if God is in his video library, viewing all the videos. For reasons best known to himself, he is attracted to one particular video, the video with you and me in it, just as we are in fact. It is as if he says, by a decision fraught with enormous consequences for us all, "Let it be in fact as this video depicts it to be," knowing (by his omniscience) its most intimate detail. It is as if the videos are accounts of what already has happened in innumerable possible worlds. (How could it be otherwise, if God has full descriptions of them all?) God decides to actualize one such world-to run it, so to speak, or perhaps to rerun it. The problem is how, if each video is a complete description of a possible world, does thinking of God's relation to possible worlds in this way help us to preserve both providence and libertarian freedom? And how does it help with providence and evil? I conclude that Molinism is not a viable way of preserving divine righteousness in the face of evil but that God's righteousness can be preserved in another way.
Let us suppose what is not too difficult: that some human acts are evil. Could such evil actions as these be foreknown by an essentially righteous God? According to both Augustinians and Molinists, they could be. But I have offered a reason to reject Molinism. Yet does not invoking compatibilism (and not middle knowledge) as the reason why God knows all future free actions mean that God is the author of evil? No, the occurrence of evil is consistent with essential strong omniscience, even aside from the Molinist option, because while God could not positively govern evil acts he nevertheless can infallibly know of their occurrence. Although God is not the cause of evil actions, nevertheless evil actions have causes that God can know of and, if he is strongly omniscient, does know of. So God is essentially strongly omniscient with respect both to those acts that he brings about and to all other events the bringing about of which is inconsistent with his essential righteousness.
But how does God know of the causes of evil actions if he himself is not the cause of them? Augustine's answer is that God foreknows future evil by knowingly and willingly permitting particular evil actions. For God to permit some event to occur does not entail that he brings that event about, but it is consistent with his foreknowledge of such events. God does not and can not will evil actions, but he may nevertheless know that they will occur and be willing for them to occur. In permitting evil in this way God acts for the highest and holiest reasons even though the detail of such reasons may be at present hidden from us. But is not one who is willing for an evil action to occur the cause of that action, or at least an accessory, and so himself evil, despite his good intentions? Shortly I wish to present two alternative arguments for thinking not. But first, before we look at these arguments, it is necessary to get clearer on the meaning of willing permission.
The nature of such permission is well expressed by Augustine:
So for X willingly to permit an action A is at least this: for A to be the action of someone other than X; for X to foreknow the occurrence of A and to have been able to prevent A; and for A not to be against X's overall plan. So on this conception God foreknows everything and unconditionally governs everything, but he does not causally determine everything in the sense that he is the efficient cause of everything, though everything that happens has sets of efficient or deficient causes in a way consistent with compatibilist accounts of human actions. Nevertheless, nothing happens that God is unwilling should happen.
How can such willing permission be consistent with both compatibilism and the occurrence of evil actions of which God is not the author? Let us consider an illustration. Suppose that my young daughter is learning to ride a bicycle and that in order to help her retain her balance I hold onto the bicycle seat from behind. Her action on the bicycle and my steadying hand are together causally sufficient for her to maintain her course. But suppose that, in a moment of inadvertence, I take my hand from the seat, and as a result of this she crashes into the wall. There is a causally sufficient story that can be told of the crash in terms of her action together with my omission. In the case of evil actions God may be said, in a similar fashion, to withhold his steadying hand. He does not do so inadvertently but for ends that are entirely consistent with his character but most of which are presently hidden from us. Human nature being what it is, evil results. If one asks how human nature can come to have an inclination to evil in a world created by a wholly good God, then Augustinianism, in common with all other views that are faced with such a question, does not have a satisfactory answer.
For the incompatibilist has no satisfactory explanation of how it is that individuals created by a good God, with moral and spiritual characters of an appropriate kind, can nevertheless, by acts of their free will, choose evil. Invoking incompatibilism at this point does not explain the occurrence of evil in a world created by a good God any more than does the Augustinian account of God's willingly permitting evil in a world created very good. Accounting for the arrival of evil in a world created good by God is a problem faced by all Christians who take the idea of a fall seriously. In contrast to the incompatibilist, the Augustinian says that the removal of God's hand led to the encroachment of evil-to the operation of a causal force arising from a deficiency-that God is not and could not himself be the author of, though he willingly permits it. Such an account, though perfectly consistent, is not an explanation of how evil came to be desired by people created in God's image.
One might supplement this argument with an appeal to the principle of simplicity or economy of explanation. By appealing to simplicity in explanation, one is led to posit as few entities as possible. I believe that a Christian theism that accounts for the onset of evil in terms of an omniscient, omnipotent God and a compatibilist account of human action-even as it recognizes the unresolved problem of the entry of evil into a universe created good and its attractiveness to men and women made good-is preferable (on the grounds of simplicity) to a system of thought that grants God lesser omniscience and omnipotence and that invokes the mysterious faculty of incompatibilistic freedom while leaving the problems of evil's entry and its attractiveness unresolved. A compatibilist account of human action is simpler than an account that invokes incompatibilism because it extends the idea of causal explanations of events, which all recognize is fundamental to natural science, into the realm of human action. One should prefer a simpler hypothesis that provides an explanation equivalent to the explanation given by the more complex hypothesis even if the simpler hypothesis has the same problems as those found in the more complex hypothesis.10
One might also invoke simplicity in explanation as an argument against Molinism. If an account that requires God to have only two kinds of knowledge can explain the question of God, human freedom and evil as well as an account that requires God to have three kinds of knowledge, then, other things being equal, the simpler hypothesis is to be preferred.
God knows all future actions and positively governs all acts that are not evil.11 How exactly he does this raises other questions that we cannot address in detail (though I will make some brief comments later). He governs all other acts, evil acts, by permitting them, since he cannot positively govern them. However, if such permission is to be consistent with God's knowledge of the future, then it has to be a particular kind of permission: it has to be knowingly and willingly given, and it has to be permission of particular actions.12 The willing permission governs particular action tokens.
So a God who is essentially strongly omniscient positively governs all acts that occur except those which are evil, and he negatively governs evil acts by knowingly and willingly permitting them.
One may, therefore, make sense of the idea of divine permission in a way that is consistent with both upholding the divine righteousness and recognizing the existence of evil if one is prepared to maintain that there are types of actions which God can prevent but which he nevertheless cannot cause, even though he may be willing for them to occur and even though those actions have causes that are sufficient for their occurrence in the compatibilist sense. Then God can only control an evil action by knowingly and willingly permitting it, by deciding not to prevent it; and the evil action occurs because it is caused by the natures and circumstances of those who perpetrate it, not by God (because God cannot cause it), though God willingly permits it. (For though God cannot cause it, he can willingly permit it, and he does so, we might presume, as a necessary component part of some broader overall will.) As already noted, one is still left with the questions of exactly why God willingly permitted evil and exactly how evil comes about in a world created by an all-good God. But these questions, surely among the most fundamental of all theological questions, have to be faced by other accounts of God's relation to evil.
Objections to Argument 2
(i) Divine determinism? First, we will address the claim that an appeal to divine willing in the sense defined is a case of divine determinism. Here it is important to have in mind a point made earlier: that the relation between God and his universe is sui generis. This relation has a character that is basically incomprehensible, that our human models and analogies cannot fully capture. It is tempting, but I believe crude and misleading, to assimilate the working of such knowing and willing permission to intramundane models of causation, and particularly to general physical determinism. Such knowing and willing permission has this in common with determinism: that what is physically determined and what is willingly permitted will each, in virtue of the determinism and the willing permission, come to pass. However, to knowingly and willingly permit an action is not to cause that action; it is to provide a necessary but not sufficient causal condition for the action. Whereas physical determinism has a strong tendency to be reductionistic and has difficulty in finding a place for a range of objects having their own causal powers, the divine willing permission is most certainly not reductionist in this sense.
Hence it is a serious mistake to suppose that classical Christian theism claims that God monopolizes power. According to Augustine, "As He is the Creator of all natures, so is He of all powers: but not the giver of all wills; for wicked wills are not of Him, being against that nature which is of Him."15 God is the source of all creaturely power, but the powers of creatures, even when efficaciously empowered by God, are really their own and so are distinct from his. If God efficaciously empowers me to type this paper, the typing of this paper is still my action, not God's. The wicked men who crucified Jesus were the cause of his death, even though he was crucified by the determinate counsel and foreknowledge of God (Acts 2:23).
One way of expressing this difference might be as follows. While it seems clear that intramundane causation is transitive-given events A, B and C, if A causes B, and B causes C, then A causes C-there is no necessary transitivity in the case of any causal aspects or features of the divine knowing and willing 16 It is thus not necessarily the case that if God governs by knowingly and willingly permitting some event B, and B causes C, then God causes C; rather God may will by permitting that B causes C and so knowingly and willingly permit C. God's willing permission is thus not a straightforward case of intramundane causation, and those who seek to assimilate God's knowing and willing permission of evil to the actions of someone manipulating a puppet, or to hypnotism, or to brainwashing or programming, have not recognized the truly unique character of such permission.
Alternatively, one may allow that while God is the primary cause of all events that occur, even of all evil acts, he is not and cannot be the secondary cause of any evil act because he is not the secondary cause of any act. But this requirement as it stands is almost certainly too strong; it seems to have the deistic consequence that God cannot directly act in the world that he has created. One may modify this requirement to allow that God is the secondary cause of some acts, which is consistent with his being the secondary cause of morally indifferent acts and of morally good acts. The exact scope of what God causes and what he permits does not matter here, provided that we are clear that he cannot cause evil.
Those who hold that God knowingly and willingly governs whatever comes to pass may nevertheless make a distinction, within that overall government, between what God causes and what he permits. William Hasker says that the central idea of Calvinism is quite simple: "Everything that happens, with no exceptions, is efficaciously determined by God in accordance with his eternal decrees."" But we have seen, I hope, there is reason to doubt that the central idea of Calvinism need be as Hasker says it is, without surrendering God's universal and particular providence. To say that each particular action is providentially governed by God is not to say that everything is efficaciously determined by God, though it may be to say that everything that occurs is endorsed by God.
So there are ways of preserving the integrity of the divine righteousness in the case of human acts that are morally evil, namely the idea that God willingly permits particular evil actions, in the sense understood. It is possible that God knowingly governs whatever comes to pass, and it is plausible (if God is omnipotent and omniscient) to suppose that he does so. If, for any event E, E occurs, then God knowingly governs E either by knowingly bringing it about or being knowingly willing for it to occur. Whatever occurs, occurs because God knowingly governs it in this sense; whatever is true in virtue of what occurs is true because God so governs it. So saying that all events are knowingly governed by God, while it entails that all events are intended by God, is not equivalent to asserting that, for any event E, if E occurs, then God has caused it.
Just as many argue in developing a freewill defense that not even God can ensure that a free agent only does what is morally right, so the Augustinian argues that there is no possible world in which the righteous God can be the author of evil. He may, however, willingly permit evil; that is, he may actualize that possible world in which he foreknows that Jones will do a particular evil act.'8 This is an instance of particular permission-God permits particular acts-which is distinct from general permission, as when a teacher permits a class to write an essay on any topic they choose. And God may give such particular permission willingly, not because he is willing for the evil act to occur per se but because he ordains some wider good of which that act is a necessary part. God's willing permission of evil may in many cases be like the willingness of a parent to allow one of her children to undergo some extremely painful but necessary course of treatment, say the removal of a vital organ, to ensure the survival of another of her children, say by transplanting the removed organ into that child. So also God may willingly permit some particular evil action for some further good, though of course he does so without any of the feelings of psychological pressure or tension that accompany such human permittings.
So does it follow from such knowing and willing permission of evil that the universe is in every detail as God intends it to be?'9 This is an interesting question, but it is unclear as it stands. There is no reason to think that God intends the details of the universe separately; there is one divine will, which encompasses all events. It would be fallacious to suppose that the divine attitude is the same with respect to every detail of what God wills; for one thing, as we have already seen, God is the efficacious cause of some events and willingly permits others.
As Aquinas put it, "God, and nature, and indeed every causal agent, does what is best overall, but not what is best in every part, except when the part is regarded in its relationship to the whole. ,20 We may suppose that when God knowingly and willingly permits certain events he does so in furtherance of some wider consideration wholly consistent with his character with respect to which they are a logically necessary condition. And likewise some of those things which he causes are means to some further end. It is a fallacy to think that because some arrangement is wise, every detail of that arrangement, considered in isolation, is wise. It does not follow that every thread of my tartan tie is tartan.
(ii) Fatalism? A second objection is that this view is a species of fatalism. But it is certainly not logical fatalism, since the universe is the outcome not of some principle of logic alone but of the divine creative and providential will, which, we may assume, is logically contingent and either causal or willingly permissive in its operation.
(#f) God cannot respond A third objection is that on such a view God cannot be responsive to what occurs in the universe, and hence the position entails some version of deism.
Within the one creative and providential will of God it is possible to distinguish those aspects that are unconditional or unilateral from those that are conditional and bilateral. Unconditional aspects are of the form "Let X be," whereas conditional aspects are of the form "Given W, let X be" (where W is brought about by someone other than the one uttering the statement). An example of the first might be "Let the planet Earth be"; an example of the second "If A sins, let him be forgiven." However, we must interpret these conditional expressions in the light of the place of conditions in the overall will of God.
We find an element of conditionality about God's willing permission of such evil, since necessarily he is not the author of it. Nonetheless as Creator he upholds the perpetrator of the evil and knowingly and willingly permits the occurrence of the evil. So the way to understand such conditional aspects of God's overall willing is not as God's response to what he has merely foreseen will happen, but as his response to what he has both foreseen and been willing to permit: for instance, that A will sin. That is, God wills to permit the evil and wills the consequence. He wills evil by willing to permit it, willing it in such a way that he is not himself the author of the evil, which he could not be, while he may will what is not evil by being the author of it, by bringing it about. There is a crucial distinction between a willing of conditionals and a conditional will. God may infallibly know all truths, including all conditional truths, and he may know what his response to the antecedents of some of these conditional truths is. But it does not follow from this that his knowledge is conditional knowledge. God's knowledge that C will happen if A does B need not depend upon his first knowing the conditional "If A does B, then C will happen" and then deciding that because person A does B, God will bring about C.
Argument 3: The Incompatibility of Divine Foreknowledge and Human Freedom
So far I have argued for the Augustinian view that divine foreknowledge and human free action are consistent, understanding human freedom in a compatibilist sense, on the grounds that it is required by the Christian idea of efficacious grace and that it does better justice to the idea of divine omniscience than do some rival accounts. Its main rival, Molinism, has a fundamental difficulty. In presenting these arguments I acknowledge that all accounts of divine and human action face the common problem of accounting for how evil could first occur in a universe created good by a good God.
I turn now to my third argument for the view that human libertarian freedom is unnecessary for a Christian understanding of the relation between divine and human activity. As far as I am aware, although this is not an argument to be found in Augustine's own writings,21 the conclusion is consistent with a broadly Augustinian position.22 This is an argument to the conclusion that divine omniscience and human incompatibilist freedom are logically inconsistent. If this is a sound argument and if divine omniscience (in the strong sense in which we have been advocating it) is essential to God's nature, and therefore nonnegotiable for the Christian, then libertarian freedom is inconsistent with it. And if libertarian freedom is inconsistent with divine omniscience and if human sin occurs, then libertarian freedom cannot be necessary for an account of the Christian view of sin and salvation. Once I have offered this argument, I shall end my contribution by considering a couple of objections to it.
Here is the argument: Let us suppose that statement p, "Jones will freely eat a tuna sandwich tomorrow," is true and that "freely eat" means "freely eat in accordance with the tenets of human libertarian or incompatibilist freedom." Let us assume that it follows from this that if God were strongly omniscient in the sense that we have been advocating earlier, then God would foreknow that Jones will freely eat a tuna sandwich tomorrow.
Let us suppose, to be more precise, that God foreknew yesterday that Jones will freely eat a tuna sandwich tomorrow. (For the present we are waiving consideration of issues relating to God and time and for the moment are supposing that God is in time.) God's foreknowledge of what Jones will do tomorrow is something that is now past, for he foreknew yesterday what Jones will do. And if that foreknowledge is now past, then it is necessary. It is not logically necessary, since God need never have known about Jones since Jones himself need not have existed. But it is what is sometimes called accidentally or historically necessary. Given that God knew yesterday what Jones will do tomorrow, then God cannot not know. This is because, and on the assumption that, time is linear and what has happened cannot now not have happened. We shall return to this crucial point in a moment.
Let us turn our attention to the idea of knowledge. We may say this about the concept of knowledge, that it is necessarily the case that if B knows that q (whatever q is, and whoever B is), then q is true. If someone, anyone, knows that q, then q is true. This is a definitional point, setting out a part of the definition, or a necessary consequence of the definition, of knowledge. (As such it is open to anyone to reject this definition and to use one's own. But this is a widespread and common definition of knowledge and is certainly a definition that accords with basic theological accounts of God's knowledge.) Note that this claim-that if it is necessarily the case that if B knows that q, then q is true-must be distinguished from this claim: If B knows that q, then q is necessarily true. This latter claim certainly does not follow from the concept of knowledge that we are employing here. It does not follow that because I know I am typing at my word processor, then I am necessarily typing at my processor, for I could have chosen not to word-process but, say, to go for a swim instead. So we may say, "Necessarily, if B knows that q, then q1' (but not "If B knows that q, then necessarily q.")
So, returning to the argument, we have God's knowledge yesterday that Jones will eat a tuna sandwich tomorrow, knowledge that is accidentally or historically necessary, and we have it being the case that necessarily, what God knows is true. Now it is a well-known theorem in modal logic that if it is necessarily the case that s (necessarily the case now that God knew yesterday that Jones will eat the tuna sandwich tomorrow) and if it is necessarily the case that s entails t (as it is necessarily the case that God's knowledge yesterday that Jones will eat a tuna sandwich tomorrow entails that Jones will eat the sandwich), then it is necessarily the case that Jones will eat the tuna sandwich tomorrow.
This argument may seem slightly complicated, but the basic thought behind it is easy to grasp. The argument is this: if there is something in the past that entails something in the future and if what is past is necessary-accidentally or historically necessary-then what is entailed is similarly accidentally or historically necessary. The necessity of the past is transferred to or transmitted to whatever the past itself entails. In the case under consideration, what the past (God's knowledge yesterday) entails is that Jones will eat his sandwich tomorrow. And since the past is necessary, accidentally necessary, then what it entails (Jones's eating the tuna sandwich tomorrow) is similarly necessary. Therefore, given that God's foreknowledge of p is past and so necessary and given that necessarily if God knows that p then p is true, it follows that what is known, p, is itself necessary.
If this is so, then Jones's putative free act of choosing the sandwich must necessarily occur and therefore cannot be free. But if this is so, then divine omniscience is inconsistent with human incompatibilist freedom.
Arguments of this type have been subject to considerable philosophical scrutiny over the years, and objections have been raised to them. So I shall now consider three sorts of objection to this type of argument, objections frequently made but ones that seem to me to be unconvincing.
Objections to Argument 3
(i) Divine timeless eternity. One standard objection to this argument is based upon an appeal to divine timeless eternity:23 timelessly eternal God does not, strictly speaking, foreknow anything. His knowledge is never, strictly speaking, necessary in virtue of being past, since it is never past.
So on this view of God's relation to time, if God is timelessly strongly omniscient, then there timelessly exists in the divine mind a specification of all events, including all free actions at present future to us, but none of these items is foreknown by God. And this, it is said, rules out the argument from foreknowledge that we have just deployed.
However, let us suppose that God has already publicly inscripted (on a rock or piece of parchment in a cave in the Middle East) one specification in his mind of an action that is future to us now, namely, "Jones will freely eat a tuna sandwich tomorrow."24 (Think of the idea of biblical prophecy in this connection.) Let's call the inscription "A." It is composed of the proposition "Jones will freely eat a tuna sandwich x days after this inscription is made," and the inscription is suitably dated. It is certainly possible for there to be such an inscription, and since we do not know that such an inscription has not in fact been made, then for all we know "A" has been inscribed.
But now let us consider the character of such an inscription. It is the result of some divine action. And if we suppose it has occurred, then that inscription, if it has occurred, will now be past and therefore necessary, accidentally necessary-necessary in the sense that it is in the past and, so, over and done with. It cannot now be the case that the inscription has not been made if in fact it has been made. But now the dated inscription consists of a certain proposition "Jones will eat a tuna sandwich so many days after this inscription has been made." This inscription is made by an omniscient and infallible God (we are supposing), and the content of that inscription-the truth that it asserts, namely that Jones will eat a tuna sandwich at such and such a date-is also accidentally necessary. For the proposition expressed in the inscription cannot now be false. Hence it is necessary, accidentally necessary, that Jones eat the tuna sandwich on such and such a day, and hence Jones cannot be incompatibilistically free not to eat the sandwich on that day. So the cognitive content of the inscription cannot be that Jones incompatibilistically freely eats a tuna sandwich tomorrow. Therefore, Jones is not now free not to eat a tuna sandwich tomorrow. So if, for all we know, timeless omniscience is not consistent with human freedom, then we cannot know that timeless omniscience and human freedom are consistent. Therefore the appeal to timelessness is not sufficient to establish the consistency of divine omniscience and human freedom. This is my reply to the first objection.
(ii) Hard and soft facts. Another current objection to this argument is based upon the claim that some facts about the past are not wholly about the past (these are called "soft" facts). These facts carry implications for the future and so are not accidentally necessary; for in order for such facts to be true, certain claims about the future must also be true. Only facts that are wholly about the past are "hard" facts. And, it is claimed, a fact such as our fact about the past inscription "Jones will eat a tuna sandwich so many days after this inscription has been made" is not wholly about the past, since it is partly, at least, about the future, namely about the time when, according to the inscription, Jones will eat a tuna sandwich. Accordingly, whether or not it is true depends upon whether Jones will in fact eat the tuna sandwich.
One problem with the objection is that the distinction between hard and soft facts is itself not a very clear one, for it would seem that any fact entails some fact or other about the future.25 The fact that Caesar is now dead implies that Caesar will not be eating breakfast tomorrow. However, we may let this point pass and assume that there are soft facts. Is God's knowledge about Jones such a soft fact? There are at least two reasons to think not. In the first place, God's belief about Jones is one of innumerable beliefs that he holds infallibly, being part of his essential omniscience. So if any of his beliefs were falsified, then he would cease to be God; and so Jones, by his freedom, would have power to bring it about that God did not exist. And this seems to be implausible.
In the second place, we may suppose that God's believing that Jones will eat a tuna sandwich tomorrow involves his mind's being in a certain state akin to the states our minds are in when we believe that Jones will eat a tuna sandwich tomorrow.
Surely nothing can now bring it about that God did not believe that Jones will eat a tuna sandwich tomorrow (if in fact God did believe this). So although this belief of God's has certain features that may make it a soft fact (if indeed there are such facts)-chiefly the feature that it is a belief about the future-it has other features that suggest it is a hard fact. It may be, as John Martin Fischer puts it, a soft fact that is nevertheless fixed by the fixity of the past. It is a hard-core soft fact. To deprive God of such a belief, Jones would have to bring it about that some hard fact about the past was not such a fact, which he manifestly cannot do.26 The idea of God actually inscribing his belief is intended to make this point vivid.
Perhaps the objection from the hard-fact-soft-fact distinction is sound against divine foreknowledge, but it is not sound against divine forebelief, since (as we saw) for something to be a case of knowledge what is known has to be true. But this is not so in the case of a belief. A belief is still a belief even if it is false. But in the case of God it is necessarily the case that his beliefs are all true. So Jones cannot by any act bring it about that God did not have the belief that he had and that God's beliefs are (in virtue of his omniscience and infallibility) necessarily true.
(iii) Molinism again. It might be said that God could, by his middle knowledge, know what Jones would freely do when Jones was placed in certain circumstances and that God choose to instantiate a world in which Jones is placed in these circumstances. This would mean that God's foreknowledge of what Jones will do was a logical consequence of his knowing what Jones will freely do and, hence, that Jones's freedom is preserved. On this account what God's foreknowledge records is his knowledge of what Jones will freely do. But we have already found reason to reject Molinism. If this reason is cogent, or if others are, then this objection loses its force.
Conclusion
This concludes my exposition of the broadly Augustinian case for holding that it is reasonable to believe that divine foreknowledge and human freedom are consistent, where this freedom is understood in a compatibilist sense, in a sense that is consistent with causal determinism. The argument that I have presented is a cumulative one, consisting of three strands. Firstly, that only such an account of human freedom is logically consistent with divine efficacious grace; causal indeterminism therefore has serious adverse theological consequences. Secondly, that this account does full justice to divine omniscience and is simpler than its closest rivals. And finally, that there is reason to think that in any case divine foreknowledge and human incompatibilism are logically inconsistent. I realize that I have not answered all objections that may be made to these arguments any more than I have made all the serious objections to the rival views. But it is hoped that the dialogue with the various other contributors to this book will afford some opportunity to make good some, at least, of these deficiencies.
AN OPEN-THEISM RESPONSE
Gregory A. Boyd
shall begin with a significant point on which Paul Helm and I are in agreement, over and against William Lane Craig and David Hunt, and then I will discuss two points on which Helm and I disagree.
God's Infallible Beliefs and the Necessity of the Past
Craig spent a good portion of his essay attempting to establish the compatibility of libertarian freedom and exhaustively definite foreknowledge' by exposing the fallacy of the following argument:
(1) Necessarily, if God foreknows x, then x will happen.
(2) God foreknows x.
(3) Therefore, x will necessarily happen.
Craig is correct in concluding that this argument is fallacious, for it transfers the logical necessity of the knower (God) to the known W. What Helm establishes, however, is that exposing the fallacy of this argument doesn't go far in proving that libertarian freedom is compatible with exhaustively definite foreknowledge. For the "necessity" that renders Craig's (and Hunt's) position problematic is not a logical necessity, but an "accidental" or "historical" necessity. This is the kind of necessity possessed by past events. Though past events are logically contingent, they are historically necessary, for they cannot now be other than they are.
The problem with Craig and Hunt's position is that if God eternally foreknows future free acts, then future free acts are "accidentally" or "historically" necessary. Among all the unalterable facts of the past is the content of what God infallibly believes shall come to pass in the future. The definiteness of the future is part of the (now necessary) definiteness of the past. Hence, I agree with Helm that if God possesses exhaustively definite foreknowledge, we can exercise no more libertarian freedom with regard to our future acts than we can with regard to any past fact. Both are "accidentally necessary."
Of course, Helm and I differ sharply on how we apply this argument. Helm grants that God possesses exhaustively definite foreknowledge and so denies we have libertarian freedom. I rather affirm that we have libertarian freedom but deny that God possesses exhaustively definite foreknowledge (more on this below). But we agree that this argument exposes a logical difficulty in views such as Craig's and Hunt's, arguments that wish to affirm both exhaustively definite foreknowledge and libertarian freedom.
On the "Anthropomorphic" Nature of "Openness" Passages
The first disagreement I have with Helm's essay concerns his "metaphorical" interpretation of certain passages of Scripture. Helm begins his essay by arguing that "the scriptural worldview is one in which all things are created by God and ordered and governed by him.... He turns the hearts of humans as he wishes; like a potter, he has power over human clay" (p. 164). Helm acknowledges that some Scripture seems to stand in tension with this conclusion, for sometimes God is said to be "surprised and to act and react toward his people in blessing and chastisement" (p. 164). In such cases, Helm argues, we are to recognize that Scripture is using "metaphorical or symbolic language, language `accommodated' to some human situation or need" (p. 164).
The passages Helm has in mind are precisely the passages on which I build much of my case for a partly open future, and the way he explains them (viz., as metaphorical) is precisely the way most classical theists explain them. This hermeneutic is not unproblematic, however. I touched on this issue in my opening essay, but since this constitutes the central biblical objection to my position, it seems expedient to briefly expand on my response to this objection here.
First, as I mentioned in my opening essay, there is nothing in these texts themselves that suggest the authors intended their statements to be read anthropomorphically. Passages that say or depict God as changing his mind, as regretting decisions, as searching hearts "to know" what is in them, and so on, are written as straightforward reports-no different from (say) passages that say that God declares the end from the beginning (Is 46:10) or that God so loved the world that he sent his Son On 3:16). The two passages that are most frequently cited in support of the classical view that God cannot change his mind (Num 23:19; 1 Sam 15:29) do not suggest that God cannot recant or change his mind, only that in certain cases he will not do so (cf. Jer 15:6; Ezek 24:14; Zech 8:14).
Second, reinterpreting the passages that constitute the motif of future openness as "metaphorical" does not clarify these passages: it undermines them. We may grant that all God's revelatory language is "accommodated" to our human condition, and we may grant that God sometimes uses metaphors, symbols and anthropomorphisms to accomplish this task. But this nonliteral language accomplishes this task because it communicates something truthful about God. But what does saying that God regrets decisions he makes or that he changes his mind truthfully communicate about God if in fact God never regrets his decisions or changes his mind? What does God's reporting to us that he thinks of the future in terms of what might or might not happen (e.g., Ex 13:17; Jer 3:7, 19-20) truly reveal to us if in fact God never really thinks about the future in terms of what might or might not happen? If God in truth doesn't ever change his mind, regret decisions or think in subjunctive terms, then saying he does so doesn't reveal anything true about God at all. It only misrepresents him.
Third, some have argued that advocates of a partly open future are guilty of naive literalism, which they inconsistently apply to the biblical text.2 We wrongly interpret some but not all anthropomorphisms in a literal fashion. If we were consistent, it is argued, we would believe not only that God literally changes his mind, but also that God literally lacks present knowledge; literally has eyes, arms and a mouth; and literally travels to and from earth-for some passages speak of God in these ways.
If this argument is valid, it can be used against all evangelicals, for all evangelicals take some things in Scripture literally (e.g., Jesus is the Savior of all who believe) and other things nonliterally (e.g., God has eyes or wings). The issue is not whether some of Scripture is literal and some nonliteral: all agree on this. The issue is why many evangelicals assume that the entire openness motif of Scripture is in the latter category rather than the former. Open theists simply find no exegetical, philosophical or theological warrant for this assumption.
The Difference Between a "Condition" and a "Cause"
A second major disagreement I have with Helm's essay concerns his view of salvation. Helm wagers a good deal of his case for the Calvinistic view of God on his argument that only Calvinism can consistently affirm that people are saved by grace alone. If we must exercise libertarian freedom in order to be saved, he argues, then we contribute something to our salvation. Hence we cannot say that we are saved totally by grace. Two points may be made in response to this argument.
First, if Helm is correct that God is the necessary and sufficient ground of salvation, then the ultimate reason why some are saved and some are not is because God himself chose to save some and not others. Unfortunately, this conclusion stands in tension with the uniform scriptural teaching that God's love is universal and impartial (Acts 10:34; cf. Deut 10:17-19; 2 Chron 19:7; Job 34:19; Is 55:4-5; Mk 12:14; Jn 3:16; Rom 2:10-11; Eph 6:9; 1 Pet 1:17). The Lord does not desire any one to perish, but rather he wants all people to repent and be saved (Ezek 18:23, 32; 33:11; 1 Tim 2:4; 4:10; 2 Pet 3:9). Hence Jesus died as "the atoning sacrifice for our sins, and not for ours only but also for the sins of the whole world" (1 Jn 2:2). If it was only up to God to choose who would be saved, we have every reason to believe he would choose everyone.
Second, there is simply no need to deny that humans can and must choose to accept God's offer of salvation in order to affirm that salvation is completely a matter of God's grace. Helm confuses a condition of salvation with a cause of salvation. If we in any sense caused God to save us by exercising faith, that would imply that we merit something from God and thus are not saved solely by grace. But this is not what freewill theists generally affirm or what Scripture teaches. Rather, we hold that salvation is graciously given on the condition that one places their trust in the one giving it. This no more makes salvation a merited reward than does my freely accepting a birthday gift makes this gift a reward.
Not only this, but most freewill theists believe that we fallen human beings would not even meet the condition of faith were it not for the gracious work of the Holy Spirit in our lives (see, e.g., Jn 6:44, 65; 1 Cor 12:3; Eph 2:8). We agree with Calvinists that we cannot have faith on our own. We simply reject the Calvinist view that the work of the Holy Spirit is irresistible. People can and do resist the Holy Spirit and thereby thwart the will of God for their lives (see, e.g., Is 63:10; Lk 7:30; Acts 7:51; Eph 4:30; Heb 3:8, 15; 4:7). Hence we affirm that if a person is saved, it is all to the glory of God, whereas if a person is damned, they have only themselves to blame.
A SIMPLE-FOREKNOWLEDGE
RESPONSE
David Hunt
u Helm's principal concern is to argue that Christians should embrace a compatibilist account of human freedom. (Roughly, the compatibilist affirms, while the incompatibilist denies, that human freedom is compatible with determinism.) Helm offers three arguments for compatibilism, which I consider in reverse order.
Argument 3
The third argument is simply that divine foreknowledge, which Helm rightly regards as a "given," excludes the possibility of human agents' being free in anything other than a compatibilist sense. Since I have already argued against this conclusion at length in my defense of simple foreknowledge, I will not repeat the argument here.
I should, however, note the irony that Helm and I, while taking very differ ent positions on this issue, nevertheless both claim the mantle of St. Augustine! Helm admits that he departs from the master at this point, inasmuch as neither Augustine nor Calvin thinks that divine foreknowledge (on its own, apart from considerations of divine grace and sovereignty) excludes incompatibilistic freedom. Since Jonathan Edwards does argue in his Freedom of the Will that divine foreknowledge by itself rules out incompatibilistic freedom, Helm's position is better labeled "Edwardsian." I naturally commend to Helm the genuinely Augustinian position I defend in my chapter!
Helm and I do share some common ground. We have very similar intuitions about the "Ockhamist" solution,' which we both reject. We also eschew the "Boethian" solution, though here our reasons differ. I do so because I want a solution that also solves the philosophical puzzle raised by infallible foreknowledge, and the Boethian denial of divine foreknowledge (because God knows things from an atemporal vantage point) simply exempts God from involvement in the puzzle without doing anything to solve it. Helm, however, thinks that Boethianism can't even do this much. I'm not so sure. A committed Boethian might resist Helm's "inscription argument" on the grounds that inscribed rocks and parchments aren't infallible-only God is infallible. (Ditto for cases in which God reveals the future to prophets.) But I won't press the point, since nothing I've argued gives me a stake in this controversy.
Argument 2
I found it very difficult to untangle the main thread of this argument; but I think now that its essential structure is this:
(1) It is theologically desirable that God foreknow his creatures' freely willed actions.
(2) If compatibilism is true, God can foreknow his creatures' freely willed actions.
Compatibilism is true.
The first premise follows from Helm's principle A, according to which it is theologically desirable to maximize the scope of divine omniscience. Since I accept this principle, and (1) does follow from principle A, let's turn to (2). This premise also seems true, though the situation is complicated a bit by considerations of divine responsibility. When God foreknows a good exercise of creaturely freedom, he does so by knowing his present intention to cause a good will in that creature; for God (as Helm sees it) is the source of all good, endowing us not only with wills but also with good wills (insofar as we have them). But God cannot foreknow evil actions by knowing his present intentions, since he is not the cause of evil. He must know future evils in some other way.
Helm suggests two such ways. First, he endorses universal causal determinism. For some things God is the sufficient cause; for everything else there are sufficient causes in the natural workings of this fallen world. It's the latter that must be causally sufficient for evil actions, since God cannot be their cause. So God can know future evil actions by knowing their natural (nondivine) causes. Second, Helm proposes that "God foreknows future evil by knowingly and willingly permitting particular evil actions" (p. 176). God doesn't cause evil actions and so is not responsible for them, but he permits them for reasons of his own.
Neither of these, by itself, can do the job. (The second, which presupposes foreknowledge in its explanation of foreknowledge, is circular.) But they appear to be jointly sufficient. In a deterministic universe in which only God is causa sui, God's knowledge of present causes tells him what will happen in the future if he doesn't intervene, while knowledge of what he has decided to "knowingly and willingly permit" tells him whether he will intervene (and if so, how). On this twin basis, God can predict absolutely everything. So premise (2) is true.
The problem with the argument is not the truth of its premises but the fact that the conclusion does not follow from the premises. (It commits the "fallacy of affirming the consequent.") To derive the conclusion validly, Helm needs a premise like this:
(2') If compatibilism isn't true, God can't foreknow his creatures' freely willed actions.
But Helm nowhere argues for anything like (2'). Indeed, Helm is the last of my coauthors one would expect to limit God's power without adequate explanation!
Not only does Helm fail to explain why other routes to divine foreknowledge are unavailable, but his assumption of universal causal determinism (without which Helm's God could not know future evils) is deeply problematic. The problem of how we can be genuinely responsible for our sins if determinism is true is best postponed until the next section. But two other problems bear brief mention here.
Second, if we were subject to universal causal determinism, why wouldn't God then be the ultimate cause of all our actions (and not just of our good actions)? After all, he created the initial state of the universe and the causal rules by which one state is succeeded by another state, and he foreknew just what would result from his setting things up this way. Unless he created something with the power to make an undetermined contribution to reality, God is the sufficient cause of absolutely everything-including our sins.
Argument 1
Helm's first argument against the libertarian conception of human freedom has nothing essentially to do with foreknowledge. It derives instead from an "appreciation of the plight of humankind and the power of God" (p. 169), an appreciation that I share to a considerable degree.
This Augustinian view of God and humanity rests (in good Augustinian fashion) on three supports. One is biblical authority: all the passages (far too many even to begin to cite) that speak of divine sovereignty, the corrupting and even incapacitating effects of sin, our need for salvation rather than an ethical self-help program, and so on. There are also metaphysical considerations. Whereas God is a self-existent being, I am a wholly contingent being, formed of the dust of the ground, summoned out of nonexistence into a borrowed existence; when asked, "What do you have that you did not receive?" (1 Cor 4:7), I can only reply, "Nothing." Finally, this view has the support of honest self-examination. When tracing my actions back to their source, it's hard to arrive at my own "self' as the ultimate cause-especially when endeavoring to credit myself with some good deed! "There but for the grace of God go I" expresses a rueful and salutary recognition that I am not the self-dependent captain of my own soul.
If this Augustinian view is also my view, then why am I not in Helm's camp? For the simple reason that it constitutes only part of the picture. Given just an Augustinian "appreciation of the plight of humankind and the power of God," it might be possible to think that all my actions are the result of external (divine or natural) causes. But this overlooks another important part of the biblical picture, namely, personal responsibility for the choices we make in life.
If universal causal determinism is true, then before I was even born there were already in place conditions causally sufficient (barring divine intervention) for all my sinful deeds. But I can't do anything about conditions prior to my birth; hence I can't do anything about the causal consequences of those conditions, among which are my sins.3 Augustine himself concludes the syllogism this way: "There is no blame involved when nature and necessity determine an action."4 Causal determinism, in short, is incompatible with moral responsibility.
Helm, of course, thinks they are compatible. (That's his thesis, after all.) But nothing he says helps me see how they can be compatible. Helm's distinction between GR (responsibility before God) and HR (responsibility toward humans) is particularly unhelpful. These diverge principally because the latter "depends upon fallible human knowledge of the motivations and circum stances of ourselves and others" and is therefore "necessarily rough and ready" (p. 165). The problem is that additional knowledge typically leads to diminished responsibility." If causal determinism is true, God's perfect knowledge of the "circumstances of ourselves and others" is knowledge of the very factors that undermine human responsibility. It is therefore puzzling how GR is supposed to vindicate human culpability in the face of universal determinism.
Helm's focus on one part of the larger picture leads him to endorse positions that strike me as extreme. Consider, for example, his claim that "the Augustinian account of the plight of human kind is such that people cannot want to want God" (p. 171). Is this a biblical as well as Augustinian position? That's hard to say, since Helm does not cite any scriptures. Romans 7 speaks of someone who cannot do the good but nevertheless wants to do it, while Philippians 2:13 concerns God's gracious work in the person who can neither will nor do the good on his own (but might, like Augustine himself in book 7 of the Confessions, desperately want to will the good). But I see no scriptural evidence that an inability even to want to want God is the universal human condition.
Indeed, there is a sense in which people cannot help but want to want God. After all, God made us by nature to want and find our fulfillment in him. Of course Helm would not deny this; his claim is presumably restricted to the postlapsarian human condition. But why does Helm think that the Fall completely effaced this orientation toward the divine? Classic theism (Augustine, Aquinas, et al.) consistently treats this orientation as distorted rather than lostthere's still that "God-shaped vacuum in every human heart," but instead of filling it with God we chase restlessly after various God-substitutes. (Even when pridefully rebelling against God, Augustine asserts, we are engaging "in a perverse imitation of God.")' Contrary to Helm, there is reason to believe that at some level even postlapsarian human beings may "want to want to ... want God."
While it is obvious that I am unpersuaded by Helm's arguments, I should add that free will is one of the most enigmatic issues in all philosophy, and I am prepared to acknowledge the possibility (however small!) that compatibilism might be true after all. I also have a good deal of respect for the theological tradition Helm is representing, and I regard the Calvinist position as very much of a "live option" as I wrestle with the issues covered in this book. Indeed, none of my three interlocutors seems to me to be proposing anything absolutely outrageous, from which one ought to recoil as from a serious heresy. All are responding to very real themes in the biblical witness, and doing so in a responsible way. The problem is that it's just very difficult to weave all these themes together into a single coherent story. That's what keeps theologians in business!
This is a problem for my coauthors, and it's a problem for me as well. My account of simple foreknowledge cannot in the end support an understanding of divine and human freedom that comes quite as close to our ordinary intuitions about free agency as Boyd's "open view," or a notion of divine providence that is quite as powerful as the one provided by Craig's Molinism, or a concept of divine grace quite so robust as the one offered by Helm's Augustinianism. I don't know how to get everything we want to get in just the way we want to get it-and neither do my coauthors. The most I can responsibly claim is that I can get more of it (at least with respect to the issues that lie at the heart of this book). The reader must judge whether this claim is borne out by the arguments.
A MIDDLE-KNOWLEDGE RESPONSE
William Lane Craig
Paul Helm begins his essay by specifying several scriptural "fixed points" which any adequate account of divine foreknowledge and human freedom must recognize.' He rightly emphasizes, against open views, God's universal foreordination (Eph 1:11). But Helm's formulation of the second point is an understatement: the reason people are held accountable is that they have libertarian freedom. Imagine a situation in which one succumbs to temptation. Paul's statement in 1 Corinthians 10:13 implies that in such a situation, God had provided a way of escape that one could have taken but that one failed to do so. In other words, in precisely that situation, one had the power either to succumb or to take the way out-that is to say, one had libertarian freedom. It is precisely because one failed to take the divinely provided way of escape that one is held accountable. With respect to Helm's third point, I see no reason to 2 think that God's relations to creation are unparalleled and incomprehensible. On a Molinist view, God's providential direction is not without parallel. When FBI agents conduct a sting operation to catch a child pornographer or drug trafficker, for example, they put the suspect in circumstances in which they have reason to believe he will freely commit a crime, and on that basis he can be arrested and prosecuted. Thus, there is neither reason nor necessity to cover one's flank by appeals to incomprehensibility of divine relations to the world.
Helm then presents three arguments why Christians ought to accept a corn- patibilist account of freedom, the goal being to show that divine foreknowledge is consistent with freedom so understood. Unfortunately, this stated project is misconceived. The debate between compatibilism and incompatibilism is quite independent of the problem of divine foreknowledge; and if foreknowledge is, as I think, consistent with incompatibilist freedom, then, of course, it is consistent with compatibilist freedom as well. The question is whether the incompatibilist can consistently affirm divine foreknowledge. Fortunately, some of Helm's subsidiary arguments are relevant to this issue.
His first argument, based upon the efficacy of God's saving grace, is, however, irrelevant. There is just no connection (as Calvin himself, quoted by Helm, recognized) between the question of the intrinsic efficacy of saving grace and divine foreknowledge.3 Nevertheless, Helm's discussion does raise the question, even if we are bound in things above, what about things below? Are we free to choose among sins? This question will return to haunt Helm's doctrine of divine providence.
Helm's second argument appeals to divine perfection in order to justify the doctrine of foreknowledge. This is a powerful argument, for surely the concept of God is the concept of a greatest conceivable being. Since omniscience is a great-making property, God must be omniscient.4 If the Principle of Bivalence holds for future contingent propositions (and some are true), then an omni scient deity must possess foreknowledge of future contingents. Helm's further claim that omniscience "is most at home" with compatibilism involves a huge leap in his argument, however. Why is this the case? Does Helm just assume that foreknowledge is based on foreordination? What is wrong with the view of simple foreknowledge?
Rather than address these questions, Helm turns to a potential problem for his view: God's foreknowledge of evil actions. How can God's foreknowledge of such acts be based on his foreordination without God's being the author of evil? Since the Molinist has an answer to this question, Helm turns to a critique of the doctrine of middle knowledge.
Helm mentions two areas of concern. The first involves a passing reference to an objection voiced by Robert Adams. But as this objection has been adequately answered in the literature and Helm does nothing to advance the discussion, we may leave it to the reader to pursue.' The second objection consists of merely a question: how does God actualize states of affairs involving libertarian actions, especially sinful actions? Now a question is not an objection, and one is rather surprised by Helm's abruptly terminating the discussion by concluding that Molinism is not a viable solution. In fact, the question is easy to answer, given divine middle knowledge. Imagine, using Helm's analogy, that God has before him via his natural knowledge two videos representing two possible worlds-one in which Adam sins in the Garden of Eden and one in which he does not. Suppose that God knows via his middle knowledge that Adam, if placed in the Garden, would sin. In that case God cannot actualize the video in which Adam does not sin. For if God were to play the video up to the time of Adam's decision, then Adam would freely sin. The video in which Adam does not sin thus represents an infeasible world for God. But as for the other video, that world is easy to actualize. If God knows that Adam would freely sin if placed in the Garden, then in order to actualize that world, all God has to do is place Adam in the Garden! Far from being difficult, we ourselves accomplish this sort of feat all the time. Think of the FBI sting operation. Although the criminals arrested through such operations sometimes complain of "entrapment," the courts have ruled that if the suspect did not have to commit the crime under the circumstances, then he is culpable for his action. So long as the circumstances are freedom-permitting, the person in them is determined neither by the operative secondary causes nor by God and so acts with libertarian freedom. Of course, since the circumstances themselves may be the result of contingent actions, God will have to weakly actualize them as well. But that is no problem for an omniscient deity.
Contrast Helm's own account of God's foreknowledge of evil acts. Helm wants to help himself to God's permissive will to explain how such acts fit into God's divine plan. Now on a Molinist scheme, divine permission makes sense, for God knows what libertarian agents would do were he to permit them to act freely. But without middle knowledge, how can God know what creatures would do were he to withdraw his steadying hand? Helm's answer must be that, given compatibilism, the creatures' decisions, while not determined by God, are determined by secondary causes, from which God can infer them. I have two worries about this proposed solution. First, it severely compromises the Calvinistic doctrine of divine providence, since there are events in the world not directly determined by God. Indeed, I wonder, does Helm have any doctrine of divine concurrence, according to which everything that happens is caused by God? Second, since free creatures' decisions are indirectly caused by God, he is still implicated in evil. On Helm's view the circumstances in which an evil act is perpetrated are causally sufficient for that act and are themselves directly caused by God (unless they include another evil act). Thus, just as I am responsible for hitting the eight ball into the corner pocket when I strike the cue ball with the intention of hitting the eight ball into the corner pocket, so God is responsible for evil actions of creatures when he sets up causally determining circumstances that make such actions inevitable for the creatures. Helm cannot say that God did not intend for those evil acts to occur, for this would contradict God's universal providence. They are planned by God and, while not directly caused by God, they are indirectly causally determined by God, thereby implicating God in evil. Permission is exonerating only if the immediate cause of the action is a libertarian agent who freely chooses to perform that action.
Finally, Helm's third argument endorses theological fatalism. What strange bedfellows these openness and Reformed theologians be! Together they oppose the compatibility of divine foreknowledge and human freedom. Helm provides no account of temporal necessity; but he offers two reasons for thinking God's past beliefs to be temporally necessary. First, if we could falsify God's beliefs, we should have the power to bring it about that God did not exist. Second, no one can now bring it about that God did not possess a certain belief in the past. Both these objections are based on confusions. Our ability to act otherwise than how God knows we shall implies neither the ability to falsify God's belief nor the ability to bring it about that God did not have that belief' It implies merely that we have the ability to act in ways other than those in which we shall act; and if we were to act in those ways, God's past beliefs would have been different than they were. Helm gives a passing nod to this Ockhamist-Molinist solution and, while finding no fault with it, insists that we have already found reason to reject Molinism. Not only is that not true (Helm merely posed a question), but the solution is independent of middle knowledge, as is evident from William Ockham's espousal of it long before Luis de Molina crafted the doctrine of middle knowledge. God's beliefs about the future, though past, are counterfactually dependent on the occurrence of those foreknown events and so are not temporally necessary, as any adequate account of this modality reveals. Hence, Helm's argument for theological (and logical) fatalism is unsound.
GLOSSARY
A-data. Helm's abbreviation for "all things" data, or Scripture references which suggest that God knows and ordains all things. See also D-data.
accidental necessity. What must be the case, given some contingent (or "accidental") fact. For example, it is now necessary that Smith will not graduate this year, since he failed all his classes. Because he could have passed his classes, and because the impossibility of his graduating on time depends solely on the contingent fact that he did not pass his classes, the fact that he can no longer graduate is only accidentally necessary. All hard facts about the past are supposed to be accidentally necessary on the grounds that, though it may once have been possible for them not to have been the case (if things had gone differently), since they are now the case, it is no longer possible for them not to be the case. See also logical necessity; natural necessity; temporal necessity.
actualization, weak. To bring about something that was previously only "possible" or "potential," but do so indirectly. According to Alvin Plantinga, who coined the term, God can weakly actualize a free choice C of a person P by creating the state of affairs in which P will freely do C. Weak actualization is contrasted with strong actualization where God directly actualizes C. See also actualize.
actualize. To bring about something that was previously only "possible" or "potential." To say that "God actualizes a particular possible world" is just to say that he makes that possible world actual. See also actualization, weak.
antecedent. In a conditional statement, the antecedent is the first portion of the statement (the "if' clause). E.g., the antecedent of "If you offer me one million dollars, then I will accept" is "you offer me one million dollars." See also consequent.
backtracking counterfactual. A counterfactual statement in which the truth of the antecedent requires that the past be different than it was: e.g., "If I were to pray for my son, he would not have been killed in the wreck."
Boethianism, Boethian solution. The view that God does not exist in time. Most classic theologians, such as Augustine, Anselm and Aquinas, held this position, but the Late Roman philosopher and statesman Boethius (c. 480-528) was particularly influential in its formulation. Boethius applied this conception of God to the problem of divine foreknowledge and human freedom: if God is not in time, he does not foreknow anything, and the problem dissolves.
causa sui. Latin for "cause of itself' or "self-caused." The term is commonly used to answer the question "Where did God come from?" The answer is that God is causa sui; he is self-caused. Some have taken this to mean merely that the existence of God does not need to be causally explained. Others have taken this to mean that literally God causes himself.
compatibilism. The idea that freedom is compatible with necessity, e.g., person P is still "free" with respect to choice C even though C is necessary. This term is most often used to express the idea that freedom is compatible with the kind of necessity entailed by causal determinism (see freedom, compatibilist). But this term can also be used to refer to the position of someone who holds that freedom is compatible with divine foreknowledge. Helm is a compatibilist with regard to determinism. Craig and Hunt are compatibilists with regard to foreknowledge. See also incompatibilism.
conceptualist model of God's knowledge. A theory of divine knowledge that construes God's knowledge, even his knowledge of future events, as analogous to innate ideas: i.e., God knows future events because he knows all truths and because there are truths about future events.
consequent. In a conditional statement, the consequent is the second portion of the statement (the "then" clause). E.g., the consequent of "If you offer me one million dollars, then I will accept" is "I will accept." See also antecedent.
counterfactual knowledge. Knowledge of true counterfactuals. See also counterfactuals.
counterfactuals. Hypothetical statements in the subjunctive mood: e.g., "If you were honest, you would confess." See also might-counterfactual; wouldcounterfactual.
counterfactuals of creaturely freedom. Counterfactual propositions in the following form: "If person S were in state of affairs C, S would freely do (or choose) A" where C specifies the whole history of the world prior to the time of the choice. Counterfactual propositions of creaturely freedom therefore express the content of a free choice: e.g., "If you were to offer me one million dollars, I would (freely) accept." See also counterfactuals.
D-data. Helm's abbreviation for "dialogue" data, or Scripture references which suggest that God interacts with his creation in very human-like ways. See also A-data.
disquotation principle. The principle that the truth conditions of a statement are just what the statement describes: e.g., the statement "Snow is white" is true if and only if snow is, in fact, white.
efficacious grace. Grace that is sufficiently powerful to achieve its intended, salvific effect.
enthymeme. An argument with one or more unstated but implicit steps, whether in the conclusion or in the premises: e.g., "Socrates lived in Athens; therefore Socrates lived in Greece." This argument is an enthymeme because it involves the unstated premise that Athens is located in Greece.
Evodius's argument. Hunt's label for the argument, as formulated in Augustine's dialogue On Free Choice of the Will, that divine foreknowledge is incompatible with human freedom. Augustine puts the argument in the mouth of his dialogue partner, Evodius, hence the term.
fallacy of affirming the consequent. A logical fallacy in which the consequent of a conditional is affirmed and, on the basis of that affirmation, the antecedent is affirmed. For example, "If it rains, the streets will be wet. The streets are wet. Therefore, it has rained."
fatalism. The doctrine that all things happen according to necessity.
Fixed Past Principle. A principle stating that we do not have the power to act in such a way that the past would have been different than if we had acted in that way.
free knowledge. God's knowledge of truths known by him only logically posterior to his creative decree. See also middle knowledge; natural knowledge.
freedom, compatibilist. Freedom that is compatible with some version or other of causal determinism.
freedom, Frankfurtian. A conception of freedom that rejects the Principle of Alternate Possibilities. See also Principle of Alternate Possibilities.
freedom, incompatibilist. Freedom that is the outcome of an uncaused or self-caused decision or choice. Also often called "libertarian freedom." See also libertarian freedom.
future contingents. Future events that do not have to happen.
grounding objection. The claim that there are no true counterfactuals about free decisions because there is nothing to make such statements true.
hard facts. Facts that have been settled or determined by the course of events. Any fact that is genuinely and strictly about the past (e.g., "Yesterday Smith listened to Berg's Violin Concerto") is a hard fact about the past. See also soft facts.
incompatibilism. The idea that freedom is incompatible with necessity (e.g., person P cannot be "free" with respect to choice C if Cis necessary). This term is most often used to express the idea that freedom is incompatible with the kind of necessity entailed by causal determinism (see freedom, incompatibilist). But this term can also be used to refer to the position of someone who holds that freedom is incompatible with divine foreknowledge. Craig, Hunt and Boyd are incompatibilists with regard to determinism. Helm and Boyd are incompatibilists with regard to foreknowledge. See also compatibilism.
intramundane models of causation. Accounts of causation that apply only to changes in things which exist in space and time.
Lewis-Stalnaker semantics. The most important theory of subjunctive conditionals/counterfactuals, so named after the contemporary American philosophers David Lewis and Robert Stalnaker, who are responsible for seminal versions of this theory. On their account, a counterfactual of the form "If P were the case, Q would be the case" is to be understood as asserting that Q is the case in some possible world(s), namely, the world(s) in which Pis the case which is (are) most like (or "closest to") the actual world. See also counterfactuals; standard semantics for counterfactuals; subjunctive conditional.
libertarian freedom. A position on the nature and possibility of free agency, under which freedom is incompatible with causal determinism and there exist genuine instances of free agency. Most libertarians (Hunt is an exception) believe this involves commitment to the Principle of Alternate Possibilities. See also Principle of Alternate Possibilities; freedom, Frankfurtian.
logical necessity. The strongest form of necessity. If something is logically necessary, there is no possibility at all of it being different than it is. That "nothing is larger than itself' and that "Smith is either a vegetable or a nonveg- etable" are both logically necessary. Logical necessities are true no matter how the world happens to be arranged. See also accidental necessity; natural necessity; temporal necessity.
mechanistic event-causation. When the cause of an event El is some other event .1 that, together with the relevant causal laws, necessitates the occurrence of El. Such causation is "mechanistic" inasmuch as El follows FO automatically, like the workings of a mechanism. See also teleological agentcausation.
middle knowledge. God's knowledge of all true counterfactuals about what creatures would freely do under any circumstances logically prior to his creative decree. See also free knowledge; natural knowledge.
might-counterfactual. A counterfactual proposition detailing what an agent might do in a particular situation: e.g., "If you were to offer me a job at Yale, I might take it." See also would-counterfactual.
Molinism. The school of thought, taking its name from Luis de Molina (15351600), a Spanish Jesuit, that seeks to reconcile divine sovereignty and human freedom by means of the doctrine of middle knowledge.
natural knowledge. God's knowledge of all necessary truths, including all possibilities logically prior to his creative decree. See also free knowledge; middle knowledge.
natural necessity. What must be the case, given the nature of things. That an unsupported stone will fall is naturally necessary, given the law of gravity; that it will not recite the Gettysburg Address on its way down is also naturally necessary, given that it's a stone and not a person. See also accidental necessity; logical necessity; temporal necessity.
necessity. See accidental necessity; logical necessity; natural necessity; temporal necessity.
Neo-Molinism. Boyd's newly adopted term for his brand of open theism. He uses this term because the open theist, he claims, can embrace God's knowledge of would-counterfactuals (see) and might-counterfactuals (see). Of course, this usage of the term Molinist is not likely to be embraced by "classical" Molinists (like Craig).
Neo-Platonism. A philosophical movement founded by Plotinus (A.D. 204270). Based on Platonic principles, Neo-Platonism understands all reality to be dependent on and oriented toward a supreme reality called the "One." Many early Christian theologians, like Augustine, found this the most congenial pagan philosophical system and drew upon it in their own theorizing about God.
nonstandard resolution of vagueness. See special resolution for vagueness.
Ockhamism, Ockhamist solution. An approach to the problem of divine foreknowledge versus human freedom, first conceived by William Ockham (c. 1285-1349), a Franciscan theologian and logician. According to Ockham, the fact that God knew one hundred years ago what Smith will do tomorrow is only a "soft" fact about the past; consequently, God's foreknowledge does not render Smith's future action necessary. See also hard facts; soft facts; Thomistic Ockhamism.
perceptualist model of God's knowledge. A theory about divine knowledge that construes God's knowledge, even knowledge of future events, as analogous to sense perception; i.e., God knows future events because he "sees" what will happen.
possible worlds. Different ways reality might have been. Each "world" represents a total description of reality, and each of these descriptions differs from one another. When God creates, he makes one possible world "actual."
practical belief. A belief about what to do. A practical belief is directed toward an optative proposition of the form "Would that I might do such-andsuch." See also propositional belief.
Principle of Alternate Possibilities (PAP). The idea that free agency, in the sense presupposed by moral responsibility, entails the agent's ability to do otherwise. One's choosing A is "free" in this sense if and only if one could have chosen other than A. Divine foreknowledge supposedly conflicts with human freedom on the grounds that a divinely foreknown action would not satisfy this principle. This principle is sometimes called the Principle of Alternative Possibilities.
Principle of Bivalence. A principle which states that all declarative utterances (or propositions) are either true or false.
principle of simplicity. Other things being equal, the explanation that invokes the fewest factors is to be preferred.
propositional belief. A belief about what is true. A propositional belief is directed toward a declarative proposition asserting that such-and-such is the case (e.g., "London is the capital of England"). See also practical belief.
Q.E.D. Abbreviation for quod erat demonstrandum, Latin for "which was to be proved."
soft facts. Facts that have not yet been settled or determined by the course of events. Some facts that are grammatically in the past tense are not genuinely about the past (e.g., "Yesterday it was true that Smith will go to church next Sunday"), and some that are genuinely about the past are not strictly about the past (e.g., "Yesterday Smith listened to Berg's Violin Concerto for the last time"). Such facts are merely soft facts about the past. See also hard facts.
special resolution for vagueness. A resolution which holds that in the case of backtracking counterfactuals, a nonstandard resolution of vagueness must be used to determine which worlds belong in the sphere of most similar worlds in which the antecedent is true. This resolution of vagueness must be used because such counterfactuals envision some difference in the past if the consequent clause of the counterfactual is true. For example, the truth of the proposition "If the time traveler were not to activate his time machine, he would not have appeared, as he did, in ancient Egypt" requires that the worlds most similar to the actual world in which the antecedent is true are worlds in which the past history of that world differs from that of the actual world.
standard resolution for vagueness. A resolution which holds that in the standard semantics for counterfactual propositions, a counterfactual is true if and only if its consequent clause is true in every possible world in the sphere of worlds most similar to the actual world in which the antecedent clause is true. In determining how worlds are to be ranked in terms of similarity to the actual world, one customarily resolves the vagueness involved in the antecedent by stipulating that the most similar antecedent-permitting worlds will have the same history as the actual world up until the time of the antecedent. So, e.g., the proposition "If Goldwater had been elected, he would have won the Vietnam War" is true if and only if Goldwater wins the Vietnam War in every world in the sphere of possible worlds most similar to the actual world up until the time of the 1964 election where Goldwater wins the election.
standard semantics for counterfactuals. Specifies that counterfactual propositions have the following form: they remain constant with regard to the past, and they inquire as to the changes that occur in possible worlds most similar to our own in which the antecedent is true. See also backtracking counterfactual; Lewis-Stalnaker semantics.
subjunctive conditional. A conditional statement in the subjunctive mood. The category of subjunctive conditionals includes the category of counterfactuals. In other words, counterfactuals are a subset of subjunctive conditionals. See also counterfactuals.
tachyon. A theoretical particle that is so fast it "moves backward in time," reversing the order of cause and effect. There is no consensus among physicists that tachyons actually exist.
teleological agent-causation. When the cause of an event El is an agent A who brings about El by intentionally doing something. Such causation is "teleological" inasmuch as intentional action aims at an end (or telos) that the agent has in mind and that the agent believes would be served by performing the action. See also mechanistic event-causation.
temporal necessity. The kind of necessity alleged to characterize events in the past. See also accidental necessity; logical necessity; natural necessity.
Thomistic Ockhamism. A version of the Ockhamist solution developed by philosopher Linda Zagzebski, who draws on some of Thomas Aquinas's theories about God and divine omniscience to show how God's knowing one hundred years ago what Smith will do tomorrow might constitute a mere soft fact about the past. See also Ockhamism.
truth-maker maximalism. The doctrine that every true statement has a truthmaker.
truth-makers. That by virtue of which a statement is true.
would-counterfactual. A counterfactual proposition detailing what an agent would do in a particular situation: e.g., "If you were to offer me one million dollars, I would take it." See also might-counterfactual.
Zeno's paradoxes. Paradoxes formulated by Zeno, a student of the Greek philosopher Parmenides (fifth century B.C.), to show that motion is only appearance and not reality. Zeno's most famous paradox is that of Achilles and the tortoise, cited in this book as an example of an argument that can be reasonably regarded as fallacious even if it's hard to say what the fallacy is.