“No-Risk” Libertarian Freedom A Refutation of the Free-Will Defense

PhilosoPhia Christi

Vol. 10, No. 1 © 2008

“No-Risk” Libertarian FreedomA Refutation of the Free-Will Defense

Walter J. SchultzDepartment of PhilosophyNorthwestern CollegeSt. Paul, Minnesota

Some Christians believe that when people suffer apparently pointlessly or excessively as a result of the actions of others it is because God could not prevent it. Others believe it is because God would not prevent it. One’s con-cept of God’s omnipotence requires revision if one believes God could not prevent the suffering. Similarly, God’s goodness requires clarification if one believes God would not prevent it. The process theologian John Cobb writes, “Parents who remained in absolute bliss while their children were in agony would not be perfect—unless there are such things as perfect monsters.”� The free-will defense and related theodicies take the first alternative and ex-plain God’s apparent inability by postulating that some cases of suffering are the unavoidable consequences of creature-freedom. The idea is that God is not to blame because it is not possible for God both to achieve his purposes in creation and to prevent such moral evils.�

This paper shows that it is possible for God to achieve his purposes (as libertarian free-will theists describe them) and to prevent every moral evil. In other words, apparently pointless or excessive suffering as a result of the actions of others is not an unavoidable consequence of granting libertarian freedom. The argument for this appears to be corroborated by recent work in the neuroscience of decision-making and action theory. If successful, al-

AbstrAct: Free-will defenses and theodicies reason that since God’s purpose in creation re-quires libertarian free will, God cannot prevent every event which occurs as a consequence of the misuse of freedom. However, given libertarian free will and free-will theistic accounts of God’s purpose in creation, I describe (in terms of a dispositions/powers ontology) how it is logically possible for God to achieve his purposes while preventing moral evil. This, then, is a refutation of the free-will defense and related theodicies that should shift the focus of Christian theorizing from God’s ability to God’s goodness.

�. John B Cobb and David Ray Griffin, Process Theology: An Introductory Exposition, (Louisville, KY: Westminster John Knox, �976), 47. Also, Greg Boyd, “If every evil event could have been avoided had God so willed, how are we to avoid thinking of God as a conspirator in evil?” (Is God to Blame? Beyond Pat Answers to the Problem of Suffering [Downers Grove, IL: InterVarsity, �003], 46.

�. By “moral evil” I mean suffering caused by the actions of agents. I do not mean to include such actions themselves or other sins.

166 PhilosoPhia Christi

though it refutes the free-will defense and related theodicies, it also demands an account of God’s goodness that justifies God’s unwillingness to prevent moral evil when he could have.

The Freedom Postulates

God’s being unable to prevent cases of moral evil by virtue of his com-mitment to libertarian free agency is what I am calling a freedom postulate. The freedom postulate is meant to morally justify God in view of apparently pointless or excessive suffering because it is supposed to be impossible for God to prevent such evil. There are two types. The first type treats the is-sue as a conceptual truth and offers minimal analysis, if any. For example, Michael Peterson calls it a “seldom-noticed conceptual connection between free will and the possibility for gratuitous evil.”3 He continues, “God cannot eliminate evil without somehow curtailing or eliminating human freedom.”4 David Griffin concurs, “it is logically impossible for God unilaterally to pre-vent all evil.”� Likewise, Jerry Walls claims, “since (God) has granted man freedom, he cannot eliminate all evil.”6 Each of these theorists7 expresses a direct version of the freedom postulate:

(FP�) It is not possible that some creatures have libertarian freedom and God prevents every moral evil.

The second type, on the other hand, locates the issue in the nature of libertar-ian freedom as it functions relative to a view of God’s purpose in creation. The idea is that God cannot both prevent the suffering and achieve his pur-poses in creation.� Alvin Plantinga alludes to the role of freedom in God’s purposes in his claim that “It is possible that God . . . wasn’t able to create (free) creatures in such a way that they always exercise their freedom to do good; for if he causes them always to do only what is right, then they don’t do what is right freely.”9 Since Plantinga’s claim is intended to defend clas-

3. Michael Peterson, Evil and the Christian God (Grand Rapids, MI: Baker, �9��),�0�.4. Ibid., �4. �. David Griffin, God, Power and Evil: A Process Theodicy (Philadelphia: Westminster,

�976), �0�. 6. Jerry Walls, “The Free Will Defense, Calvinism, Wesley, and the Goodness of God,”

Christian Scholar’s Review �3 (�9�3): �9–33, xxx (emphasis added). 7. See also John Hick, Evil and the God of Love (New York: MacMillan, �96�), xxx: “There

is a necessary connection between personality and moral freedom such that the idea of the creation of personal beings who are not free to choose wrongly as well as to choose rightly is self-contradictory.”

�. William Hasker, “The Necessity of Gratuitous Evil,” Faith and Philosophy 9 (�99�):�3–44, says that gratuitous suffering is conditionally necessary.

�. Alvin Plantinga, “Self-Profile,” in Alvin Plantinga, ed. James E. Tomberlin and Peter van Inwagen (Dordrecht, Holland: D. Reidel, �9��), 4�.

sical theism by showing its logical compatibility with the existence of moral evil, we may paraphrase Plantinga’s claim as:

(AP) It is possible that it is not possible that God achieves his purpose in creation and God prevents every moral evil.

Given the characteristic axiom of S� modal logic (p ⇒ p), (AP) en-tails:

(FP�) It is not possible that God achieves his purpose in creation and God prevents every moral evil.�0

Similarly, Clark Pinnock writes, “(a) God created for the sake of loving rela-tionships. (b) This required giving real freedom to the creature that it not be a robot. (c) Freedom, however, entails risk in the event that love is not recipro-cated. (d) Herein lies the possibility of moral and certain natural evils—those which appear irredeemably malicious and demonic.”�� Greg Boyd agrees, “God created the world out of love and for the purpose of love. . . . Cre-ation doesn’t have to have actual evil, but it must allow for the possibility of evil—if the possibility of genuine love is to exist.”�� We may summarize these developments of the freedom postulate by noting that, whereas for Plantinga, God’s purposes essentially involve agents’ freely chosen, morally right actions and the consequences that follow, for Pinnock and Boyd, God’s purpose is essentially agents’ freely-chosen attitude of love toward God. It should be clearly acknowledged at this point that what is at stake here for God, according to free-will theorists, is libertarian free, morally right actions and attitudes. Libertarian free, morally wrong actions and attitudes are not valuable and are not aspects of God’s purpose in creating the world. The central notion is this: in order to provide for the possibility of libertarian free, morally right actions and attitudes, God cannot prevent the morally wrong ones.

Libertarian Freedom Revisited

In order to see just exactly how this freedom postulate is supposed to work, let us begin with Alvin Plantinga on libertarian freedom:

�0. Letting p represent the conjunction, God achieves his purpose in creation and God pre-vents every moral evil, the argument formalized is this:

(�) ~p AP (above)(�) ~~~p � def exchange(3) ~p � double negation(4) p ⇒ p 3 S� axiom(�) ~p 3, 4 contraposition��. Clark H. Pinnock, Most Moved Mover: A Theology of God’s Openness (Grand Rapids,

MI: Baker, �00�), �3�–�. ��. Boyd, Is God to Blame? 63.

Walter J. sChultz 167


If a person is free with respect to a given action, then he is free to per-form that action and free to refrain from performing it; no antecedent conditions and/or causal laws determine that he will perform that ac-tion, or that he won’t. It is within his power, at the time in question, to take or perform the action and within his power to refrain from it.�3

Thus, we have the following semiformal definition of libertarian free will:(LFW�) S is libertarian-free regarding action A at t if and only if at t: [no

conditions determine: S performs A & no conditions determine S refrains from A].

It follows that for any person S and action A, if (a) at t God causes: S per-forms A, or (b) at t God causes: S refrains from A, then S is not libertarian-free regarding A at t, because God’s causing either one constitutes a set of conditions which at t determine S’s action by determining S’s decision.

In view of recent advances in action theory and in the neuroscience of decision-making, we may now see how it is possible for God to prevent every moral evil without in any case overriding the essential aspects of liber-tarian freedom. In order to understand this, we must be more specific about the nature of libertarian freedom involved in either S’s performing A or S’s refraining from A-ing. What is it about, for example, raising one’s hand from a relaxed position or pulling a loaded gun’s trigger that would make them libertarian-free in the sense of (LFW�)? Toward greater specificity of the concept of libertarian free action, we should note the following pertinent and necessary features of any libertarian free action A performed by any person S (abbreviated as “S’s A-ing”):

(�) the physical movement in S’s A-ing is an event occurring over some duration δ, and

(�) S forms the proximal intention�4 or wills�� to A prior to δ but after t, and

(3) forming the proximal intention to A is causally independent of situ-ations prior to δ.

Every libertarian free action meets (at least) these three conditions. Let us now examine some of the most pertinent aspects and consequenc-

es of these conditions. First of all, beginning with the second condition, ev-ery libertarian free action requires and follows temporally from an event of deciding or choosing which initiates the action. This is crucial: the time involved in deciding is not a duration-less instant. The event which initiates

�3. Alvin Plantinga, God, Freedom and Evil (New York: Harper and Row, �974), �9. �4. See Alfred R. Mele, Springs of Action: Understanding Intentional Behavior (New York:

Oxford University Press, �99�), “Free Will and Neuroscience,” in Free Will and Luck (New York: Oxford University Press, �006), 44–69, and Alfred R Mele and Paul Moser, “Intentional Action,” The Philosophy of Action, ed. Alfred R. Mele (Oxford, UK: Oxford University Press �997), ��3–��.

��. Jing Zhu, “Intention and Volition,” Canadian Journal of Philosophy 34 (�004): �7�–94.

the physical movement is one of “forming a proximal intention.” Thus, a libertarian free action involves both.�6 This can be graphically represented below in figure �:

| ← libertarian free action → || ← proximal intention → | ← physical movement → |

Fig. �

If the physical movement aspect of S’s A-ing is not preceded by a proximal intention (or decision, or volition) to A, S’s A-ing is not an action, but a mere event.�7

We should distinguish, therefore, between two kinds of intention. The first kind is a settled determination to take some action at a future time. For example, on Wednesday one may determine to go for a long run on the com-ing Saturday no matter what the weather. Some theorists have called this a “distal intention.” The other kind of intention is a proximal intention. The latter is what initiates and sustains an action.��

Second, acquiring or forming a proximal intention is not itself an act. If it were, then one must form a proximal intention to form a proximal inten-tion—which is itself an act requiring another proximal intention resulting in an infinite regressive succession of proximal intentions and no act is ever taken. Agents do not decide to decide. As Stewart Goetz claims, “a choice is the exercising by an agent of his mental power to choose, where . . . the exercising of a mental power is essentially an uncaused event.”�9

Third, the process of the formation or acquisition of a proximal intention is subliminal for the agent. The choice that initiates a physical movement does not come into being in a durationless instant, but rather comes into be-ing or consciousness over some amount of time and that amount of time is not perceptible by humans; it is subliminal. In other words, the time it takes to form a proximal intention is so short that humans cannot be aware of the process of its formation; they cannot introspect the coming into being of their choices. They just choose and experience such choosing as instantaneous.

�6. Later, we will modify this to account for mental acts. �7. See Alan Donagan, The Theory of Morality (Chicago: University of Chicago Press,

�977), 37–��; see ��–�� on the terms “action” and “act.” ��. It may be that what has been represented as sequential is possibly overlapping. In other

words, it might be that the physical movement begins before the choosing or willing events is properly completed. This alternative makes no difference to the argument. All that is needed is that the event of choosing, volition, or conscious initiation of movement precede the move-ment.

�9. Cp. Stewart Goetz, “Libertarian Choice,” Faith and Philosophy �4 (�997): �96.



Yet every physical action must be preceded by proximal intention. (These considerations seem to find support from Benjamin Libet’s experiments in neuroscience,�0 which have shown that measurable brain activity precedes awareness of the intention to act.)

Taken together, these consequences (that is, that proximal intentions are not acts and the process of their formation or acquisition takes time and is subliminal for the agent) may explain why it is incorrect to say that persons cause their choices. Persons choose to A. Persons do not cause their choice to A. The forming of the proximal intention is the choosing. It takes time and is subliminal.

Fourth, since forming a proximal intention to A is subliminal, humans cannot be aware of the distinction between a completed proximal intention and the beginning of its associated action. Humans, of physical or biological necessity, experience their intentional actions (that is, their acts) as undif-ferentiated wholes.

Fifth (as a further consequence), once the formation of S’s proximal in-tention to A begins, the formation cannot be interrupted by S, even though the action may be “vetoed” after it has been initiated. But such vetoes result from a choice at a subsequent decision “fork.”�� A baseball pitcher’s unplanned balk illustrates my point. Once the formation of the pitcher’s proximal inten-tion to throw a curve ball, say, begins, the formation of the intention can-not be interrupted by the pitcher. Only after the proximal intention has been formed and the physical action begins is the pitcher able to “veto” his deci-sion. The earliest possible veto opportunity is after the physical movement has begun—after, that is, the decision had been made. We may veto a distal intention before the physical movement, but we cannot veto the formation of a proximal intention after it has begun. This is illustrated below in figure 2:

| ← libertarian free action → || ← proximal intention → | ← physical movement → |

↑earliest possible veto opportunity

Fig. �

�0. Benjamin Libet, “Do We Have Free Will?” The Oxford Handbook of Free Will, ed. Robert Kane (Oxford: Oxford University Press, �00�), ��–64, and “Unconscious Cerebral Initiative and the Role of Conscious Will in Voluntary Action,” Behavioral and Brain Sciences � (�9��): ��9–66. See also Benjamin Libet and Anthony Freeman, ed., The Volitional Brain: Towards a Neuroscience of Free Will (Thorverten, UK: Imprint Academic, �999).

��. See Mele, Springs of Action, and Mele and Moser, “Intentional Action,” on the role of proximal intentions in sustaining an action.

Sixth, during the time a proximal intention is being formed, its converse is impossible. In other words, from the moment S begins to form the proxi-mal intention to A, it is not possible for S to refrain. That is, S is libertarian-free re: A at t, but not after t while the proximal intention is forming. That is, S is not free to take the alternative while the proximal intention is forming. This is a crucial point. It is also not physically possible that S be libertarian-free at every moment throughout the completion of the physical action as-sociated with A, because there is a subliminal interval of proximal intentional formation and subsequent awareness of what one is doing. This should be underscored: during the time the agent S is forming the proximal intention to A, S is not free to refrain. Therefore, whenever an agent completes a physical action, that agent has already gone through a sequence of moments of liber-tarian freedom and lack of it.

Finally, when one freely refrains from A-ing one is “doing” something else—even if, say, remaining motionless absorbed in contemplation. Let us, then, expand the analysis of libertarian freedom to include these elements (see figure 3):

t0 t� t� t3 t4 t�

| ← libertarian free action → || ← proximal intention → | ← physical movement → || ← minimum perceptible duration → |

↑earliest possible veto opportunity

Fig. 3

(LFW�) For any action A: S is libertarian-free with respect to A at t, just in case

(�) at t: [no conditions determine: S performs A & no conditions determine S refrains from A], where

(�) S performs A just in case:(a) A-ing is an event (physical or mental) occurring over some

duration δ = t�–t�, and(b) S forms the proximal intention to A prior to δ but after and

beginning at t0, and(c) forming the proximal intention to A is causally indepen-

dent of situations prior to and up to t0, and(3) while S is in the process of forming the proximal intention to

A, S cannot form an intention to refrain from A, and the con-verse.

So, for example, S is libertarian-free with respect to pulling the gun’s trigger at t if and only if



(�') pulling the gun’s trigger is an event occurring over some duration δ, and

(�') S forms the proximal intention to pull the gun’s trigger prior to δ but after and beginning at t0, and

(3') forming the proximal intention to pull the gun’s trigger is causally independent of situations prior to and up to t0, and

(4') while S is in the process of forming the proximal intention to pull the gun’s trigger, S cannot form an intention to refrain from pulling the gun’s trigger (and vice versa).

With this more complete analysis of libertarian free will in mind let us return to the issue at hand.

The Teleological Suspension of the Dispositional

Let us first initially suppose the world to be a causal network of com-plete determinism.�� (We will modify this conception of the causal nexus later to make room for libertarian free choice.) Secondly, let us assume the Christian doctrine of God’s sustaining the universe in existence moment by moment. Notice also that whatever length of time one assumes a moment to be, a nanosecond or a discrete moment of Planck time (�0–43 sec.) at the least, God—by virtue of his sustaining a neutrino in existence—is aware of the time it takes for a neutrino to travel the distance of the diameter of a neutrino. It follows that God is aware of the moment when any action, physi-cal or mental, will become feasible for any agent. Thus, at that moment, if God decides neither to cause nor to prevent one of the possible actions, he may decide to allow it. But the question is how this allowing could make sense in a deterministic world. To conceptualize God’s allowing a libertarian free action in an otherwise deterministic universe, let us assume an ontology of dispositions and powers. In this ontology, persons are constituted fully by a complex configuration of dispositions, capacities, and powers—which, in turn, are fully dependent on God’s sustaining them at every moment.�3 (Perhaps agents are, at some very basic level, simply God-sustained, con-stant powers of intention formation.) In granting freedom, God suspends the relevant, freedom-overriding aspects of the causal nexus and permits the

��. We note that quantum physicists disagree as to the nature and reach of quantum inde-terminacy.

�3. However, even if such dispositions and powers have categorical bases, the point remains the conceptual possibility of God’s exhaustive control over every event. See James H Fetzer, “A World of Dispositions,” in Dispositions, ed. R. Toumela (Dordrecht, Holland: D. Reidel, �977), �63–�7; James Franklin, “Are Dispositions Reducible to Categorical Properties?” The Philo-sophical Quarterly 36 (�9�6): 6�–4; Rom Harré, “Is there a Basic Ontology for the Physical Sciences?” Dialectica �� (�997): �7–34; and George Molnar, Powers: A Study of Metaphysics (New York: Oxford University Press, �003), and “Are Dispositions Reducible?” The Philo-sophical Quarterly 49 (�999): �–�7.

expression of an agent power/disposition cluster that previously had been suppressed.

Putting this all together, we may imagine that, for any feasible action that God has decided neither to cause nor to prevent, he grants libertarian freedom at that moment when—were the agent not thus determined—he or she will form a proximal intention either to take the action or to refrain.�4 God then allows the formation of a proximal intention to begin. At time t God does not determine which alternative proximal intention will begin to form. In other words, agents are thereby granted power of choice at that sublimi-nal moment of suspension at which proximal intentions begin to be formed or acquired. Since God’s allowing the formation of proximal intentions is for his purposes, every such event that he neither causes nor precludes is an instance of (what may be termed) God’s “teleological suspension of the dispositional.”

Since the forming of an proximal intention takes time, we can see how God could ascertain what a person will do beforehand, primarily because forming a proximal intention (that is, the act of willing) is—at least, con-ceivably—much slower to God than to a person. Hence, God (by virtue of sustaining the creation moment by moment) can be thought to see where the proximal intention is going. In other words, it is conceivable that God orches-trates the occasion of each event in creation except the formation of those proximal intentions he permits. He can observe their formation in “slow mo-tion” (as it were); he observes the formation of every intention in increments. If God does not want to happen what could momentarily happen (if he does not intervene), then (in a fraction of a humanly perceptible moment) just before the intention is completely formed and consciously available to the person whose it is (t� in figure 4), God withholds the necessary conditions for the completion of the formation of that intention, thereby precluding the act that would have otherwise followed. These ideas are graphically represented below in figure 4:

t0 t� t� t3 t4 t�

| ← libertarian free action → || ← proximal intention → | ← physical movement → || ← minimum perceptible duration → |

↑ ↑point of divine intervention earliest possible veto opportunity

Fig. 4

�4. Maybe it is redundant or somehow mistaken to think that every time an agent refrains that agent forms an intention to refrain. Maybe there is deliberate refraining and simply that absence of forming a proximal intention to A.



It should be emphasized at this point that what has been shown is only the possibility of God’s preventing suffering as a result of an agent’s action (stip-ulated as “moral evil”) without contravening free will. Free will is not con-travened because at t God did not cause the person’s decision. Furthermore, completed morally wrong actions that cause suffering are not claimed to be aspects of God’s purpose in creation. They are claimed to be the unavoidable consequences of the way God achieves it. The aim here is at most to show what might possibly be the case, because what is argued against is an impos-sibility claim. It is not even claimed that this is true; only that it is possible that God prevent every moral evil and achieve his purpose in creation as claimed by free-will theism.

Objections and Rejoinders

Objection 1. Haven’t such teleological suspensions of the dispositional with subsequent preventions violated libertarian freedom as required for the achievement of God’s purposes in creation?

Rejoinder. No. Such acts of God do not violate libertarian freedom be-cause every necessary condition of the definition is met: at time t, no condi-tions determine: S performs A and no conditions determine S refrains from A. Secondly, as I have described it, God is not causing them always to do only what is right. He permitted them to choose the right, but they were decid-ing contrary. To object that God’s intervention somehow is not “genuine” freedom is to change the definition as it is stated and as it functions in the achievement of God’s purposes as given in free-will defenses and theodicies. We must keep in mind that, for free-will theists, what is at stake in God’s pur-pose in creation is this: libertarian free, morally right actions and attitudes. Libertarian free, morally wrong actions and attitudes are not an aspect of God’s purpose. Thirdly, humans cannot themselves interrupt the formation of a proximal intention. Once the formation of a proximal intention has be-gun, a threshold of “no return” has been crossed. In this sense, the deciding has begun and the agent is no longer free to refrain. God acts after the fact (so to speak), but before the effect. It is crucial at this juncture to reconsider what is required of the condition “at the time in question.” Must “at the time” be throughout the duration of the performance of the action? Or could it be for some shorter duration—say, whatever duration is required to ensure un-determined initiation of the formation of a proximal intention? The achieve-ment of God’s purposes does not require that after t—after S’s choice has begun to take shape to do what is contrary to God’s will and over which S has no control—no conditions prevent the completion of S’s proximal intention to do what is contrary to God’s will. Freely inflicting apparently pointless or excessive suffering on others is not claimed by free-will theists to be an as-pect of God’s purpose in creation, but rather an unavoidable consequence.

Objection 2. But even so (the objection may continue), does not libertar-ian freedom involve the agent’s expectation to carry out his or her intentions such that to interrupt the process is to frustrate the agent? Shouldn’t this condition be added to the analysis of libertarian free action?

Rejoinder. Frustration follows from consciousness of having been thwarted in one’s efforts. Humans cannot perceive the interruption of the formation of a proximal intention, because, since the time it takes to form a proximal intention is subliminal, so is any segment of time in its formation. Hence, there would be no consciousness of the interruption and no feeling of having been frustrated. Besides, it is normal experience to have imag-ined taking some action, but never actually having performed such actions. In these cases, there is no sense of having been frustrated. This experience would be indistinguishable from the experience of God’s thus preventing the fruition of their intentions.�� In our examples, one experiences oneself as having never gotten around to raising one’s hand, even though one (dis-tally) intended to. One experiences oneself as never having brought oneself to actually pull the trigger. These types of experiences are common. It would be impossible to tell whether we simply failed to choose to do so or whether God intervened.

To motivate this a bit further consider the example of Abimelech and Sarah recorded in Genesis �0:6. Abimelech saw Abraham with Sarah and Abimelech took Sarah from Abraham after having received the false infor-mation that Sarah was not Abraham’s wife. Genesis narrates God’s revealing himself to Abimelech in a dream: “Yes, I know that you (Abimelech) have done this (taken Sarah from Abraham) in the integrity of your heart. It was I that kept you from sinning against me.” Presumably, Abimelech formed the distal intention to have sexual intercourse with Sarah as his wife. Given his justified false belief, he was not culpable in his beginning to form the proximal intention to fulfill his desires. How might we conceptualize the mechanism whereby God prevented the actual sin? It is possible that when-ever Abimelech formed a proximal intention to go to Sarah, God precluded its completion. Abimelech could not have been conscious of such preclu-sions—only that he never carried out his original distal intention. It is like the person who says, “I always wanted to get physically fit, but never got around to doing anything about it.” These cases have been thought to be cases of akrasia or weakness of will. In general, agents form a distal intention to carry out some plan of action, but when the time comes they find themselves unable to do it. What has been described shows that some cases classified as instances of akrasia may be cases of God’s intervention. One could not introspect the difference.

��. Something similar is described by David Deutsch, The Fabric of Reality: The Science of Parallel Universes—And Its Implications (New York: Penguin, �997), ��4, ��.



Objection 3. But isn’t there a legitimate intuition that freedom requires noninterference in an agent’s completing the physical movement aspect of A-ing?

Rejoinder. Completing the physical movement aspect of A-ing is not a necessary condition of a libertarian free action as it has been defined in ver-sions of the free-will defense and various free-will theodicies.

Secondly, there is a subliminal interval of proximal intentional forma-tion and subsequent awareness of what one is doing during which the agent is not free. That is, during the time the agent S is forming the proximal inten-tion to A, S is not free to refrain. Even if one changes one’s mind and refrains, such a change of mind is later and involves a different choice. Therefore, whenever an agent completes a physical action, that agent has gone through an alternating sequence of moments of freedom and lack of it. Each subse-quent moment of freedom, the relevant action A is not identical to the previ-ous action, even though the physical movement aspect of A is the same.

Moreover, on a biblical account of the moral assessment of actions, both motives and intentions count. The difference between manslaughter and murder in Leviticus and Deuteronomy turns on the motives and aims the agent had in mind when acting. Similarly, Jesus notes that one need not per-form a sexual act to be guilty of adultery.

Since, as is often claimed, “Moral responsibility entails moral freedom,” let us pursue this line just a bit further. An agent S is said to be morally re-sponsible only if that agent is morally accountable. The conditions of hold-ing someone morally accountability are generally held to be these:

(�) S able to categorize actions into types,(�) S is aware of all pertinent norms, (3) S can use norms in the assessment of his or her behavior, and(4) S is able to control his or her actions accordingly.

Persons who fail to meet any of these conditions of accountability are typi-cally placed under some type of supervision or are institutionalized. By the same token, when an accountable agent’s proximal intentions are to perform evil actions, that agent is morally guilty before God even if God intervenes to prevent the physical movement and its harmful consequences. That agent is morally responsible because though able to control his or her actions, S began to form a proximal intention to act contrary to God’s will. In other words, in such cases, agents meet the relevant conditions of moral responsi-bility—even when God prevents the completion of the formation of a proxi-mal intention.

Finally, the freedom that is said to be required for the achievement of God’s purposes is not undermined by God’s preventing the completion of proximal intentions to act contrary to his purposes. Freedom to complete a morally wrong action is not necessary for the achievement of God’s pur-poses. Such actions are supposed to be unavoidable.

Objection 4. Another objection may be raised on grounds of what may be called, “integrity of agency.” Integrity of agency (so the objection may go) requires not only libertarian freedom in practical decision making and conscious initiation of intentional action, but also the executive control of intention implementation throughout the course of an action.�6 If God were always to intervene in the sequence of someone’s executing a morally wrong act, the “integrity of agents” would be either abrogated or at least severely diminished.

Rejoinder. The objection misses the points on two grounds. First, form-ing a proximal intention to act takes time, is subliminal and uncontrollable after it is initiated. God’s interrupting what is already beyond one’s control to prevent moral evil is not a violation of integrity of agency. My point in this paper is that although it is possible for God to do such and it seems that he has and does prevent some moral evil, he often does not. Second, and most crucially, such integrity is not a necessary condition of God’s achieving his purpose in creation as stated by free-will theists. The choice offered, as de-scribed by free-will theists, is the choice freely to love God. Free-will theists claim that God cannot offer the choice freely to love God and prevent actions contrary to his will. But such a claim is mistaken given the possibility I have described. We now understand how it is possible for God to prevent actions contrary to his will without taking the primary choice off the table.

Objection 5. A related objection may be raised on the grounds that, when God prevents the completion of the formation of a proximal intention, it is not the agent S who decides, but God. Therefore, since libertarian free agency requires at least the completion of a proximal intention, when God prevents the completion of the formation of a proximal intention the agent is not free.

Rejoinder. Incompatibilists (libertarian free-will theorists) assert that an agent is morally responsible for A-ing only if he or she could have done oth-erwise, that is, only if an alternative was possible. This is called the principle of alternative possibilities (PAP). Harry Frankfurt offered counterexamples to PAP which aim to show that a person can be morally responsible even in the absence of alternative courses of action.�7 In Frankfurt counterexamples, some imagined mechanism is able to prevent an action after the decision, because such mechanisms require that the decision be known. In our sce-nario, the action is prevented by preventing the completion of the proximal intention. Yandell explores various moves that an incompatibilist can make against Frankfurt counterexamples. Yandell imagines that Ann can anticipate and control Mary’s thoughts by virtue of her planting a microchip in Mary’s brain. Yandell writes:

�6. See Zhu, “Intention and Volition,” �79.�7. Harry G. Frankfurt, “Alternative Possibilities and Moral Responsibility,” Journal of

Philosophy 66 (�969): ��9–39.



Move 5: If Ann interferes, it is Ann, not Mary, who is doing the de-ciding—Ann decides that Mary shall not send the letter, and perhaps also that it shall seem to Mary that Mary has decided this. But if Ann interferes then Mary has not decided this. She has made no decision at all if Ann interferes. . . .��

Yandell emphasizes that Frankfurt counterexamples emphasize the alterna-tives of “freely choosing to send the letter versus freely choosing not to send the letter while ignoring such things as freely choosing to send the letter versus not freely choosing to send the letter.”

I have a twofold rejoinder. First, this objection does not apply unam-biguously to our case because, in our more detailed scenario, S has freely crossed a threshold of no return and is in the irrevocable process of forming a proximal intention (that is, deciding). Thus, since the process is irrevo-cable for the agent, the agent has—in a legitimate sense—decided. But due to the time it takes for physical processes, God intervenes and prevents its becoming causally potent. Had not God intervened S would have begun the physical movement associated with it. In other words, since S could not have stopped after the moment of its commencement, as far as any consequence is concerned, S has decided. In short, God did not do the deciding. God did not cause the “direction” of the formation of the proximal intention. God prevented the decision from being causally potent. The objection does not consider the time involved in deciding.

Furthermore, it bears reiteration: the freedom that is said to be required for the achievement of God’s purposes is not undermined by God’s prevent-ing the completion of proximal intentions to act contrary to his purposes. Freedom to complete a morally wrong action is not necessary for the achieve-ment of God’s purposes according to those free-will theorists quoted above. But the objection logically requires that libertarian free actions per se are a value God sought to achieve in creating the world. Such a view, if held at all among free-will theists, is not widely held.

Objection 6. Another possible objection is that the exercise of libertarian free will requires a stable natural order and that God’s continually interven-ing in the natural order renders free and rational action impossible.�9 The idea is that if agents could not acquire a knowledge of the regularity of the natural order, then they could not deliberate, plan and take courses of action with the reasonable expectation of success. God’s continual intervention would render the natural order unpredictable.

��. Keith Yandell, Philosophy of Religion: A Contemporary Introduction (New York: Rout-ledge �999), 3�9–30.

�9. See Peterson, Evil and the Christian God, �0�, �09, and Bruce Reichenbach, “Natural Evils and Natural Laws: A Theodicy for Natural Evils,” International Philosophical Quarterly �6 (�976): ��7.

Rejoinder. This objection fails to give full appreciation to two factors. First, the forming of a proximal intention itself is not only causally immune (otherwise it would not be libertarian-free), it is also causally impotent and imperceptible to the agent when God prevents its completion. The natural order is not, therefore, rendered irregular whenever God intervenes. The ob-jection requires God to interrupt the physical movement aspect of the ac-tion.30 But when God intervenes by precluding the formation of the proximal intention, he does so before such an intention is causally potent; before, that is, there is any physical movement. Thus, God need not create an irregular or unstable natural order to prevent such actions.

Objection 7. A further objection is that neither morally-evil proximal intentions nor morally-evil “refrainings” from what one ought to do can be prevented.

Rejoinder. This objection trades on an equivocal use of the term, “moral evil.” We have stipulated that “moral evil” means suffering resulting from the actions of agents. Neither the intentions nor sins of omission or commis-sion are moral evils in this sense.

Objection 8. Still, one might object that cases of suffering that result from someone’s inaction seem plentiful. Cases of someone’s suffering as a result of someone else’s non-actions are instances of “moral evil.”

Rejoinder. Such cases are not moral evils. Here is why. Such cases of suffering have a causal history and someone’s inactivity is not a contributing cause of the suffering. Furthermore, other agents’ actions cannot be unavoid-able contributing causes either, because God could have prevented them on our analysis. Such cases, then, are cases of natural, not moral evil.

Objection 9. One might object by reference to the debates over how to interpret Libet’s research. Alfred Mele, for example, has argued that what Libet refers to as the beginning of the formation of choice to act, maybe be nothing more than an urge while the formation of the proximal intention begins much later.3�

Rejoinder. It must be underscored that whatever turns out to be the cor-rect interpretation of the results of Libet’s experiments regarding this point will not make a difference. The concept of a libertarian free action and our experience of choosing to act indicate that free action requires choice, choos-ing takes time even though it seems invariably to be instantaneous. Since one does not choose to choose, but chooses or forms a proximal intention to A,

30. In “The Irrelevance of the Free Will Defense,” Analysis 3� (�97�): ��0–��, Steven E. Boër also argues that God could interrupt the physical movement aspect of action. Frank Dil-ley, “Is the Free Will Defense Irrelevant?” Religious Studies �� (�9��): 3��, objects similarly to Boër’s thesis.

3�. Mele, “Free Will and Neuroscience.”



the choosing must begin prior to reflexive consciousness of having formed the intention.

Objection 10. Finally, it might be objected that the idea of persons being in some sense constant powers of proximal intention formation is conceptu-ally opaque. That is, the idea cannot be analyzed. Moreover, the idea of free will as a constant power of proximal intention formation seems difficult to reconcile with God’s sustaining the universe in existence moment by mo-ment. The objection, then, is that the possibility argument itself fails.

Rejoinder. If this objection cannot be defeated, then the idea of libertar-ian free will falls with it. For libertarian freedom itself is finally conceptually opaque. Neither introspection nor neuroscience seems to be able to adju-dicate its standing. The objection, when offered by a free-will advocate is, therefore, self-defeating.

Summary Remarks

God is aware of every action before its proximal intention is completed by virtue of his sustaining the universe in existence moment by moment. Since the formation or acquisition of a proximal intention takes time, God knows what each creature will do before they do. If, as we supposed, God’s Sovereignty (as God’s control over every object and event) can be concep-tualized in terms of a dispositions/powers ontology, then since God sustains creation moment by moment, God may (whenever he wants to) suspend the causal nexus to permit an agent:

First type: To “decide to act” (construed as the acquisition or formation of a proximal intention to take or to refrain from a physical act of A-ing or a mental act of directing one’s attention to X), or

Second type: To “be willing or not (absent correlative desires)” to have a mental disposition produced in oneself.3�

God is also able to bring about an action by virtue of agential and physical dispositions. These options represent God’s freedom both to prevent or allow some libertarian free actions and to bring about other actions. Hence, here are two possible types of “no-risk” libertarian freedom.

Furthermore, agents (under this construal) are libertarian-free to per-form morally good actions (as construed by free-will theists) and achieve what results from them, or freely to take an attitude of love to God. But these things are, according to the theorists we quoted, just what God aims at in creation. Furthermore, since God can prevent any action by the teleological

3�. Such willingness is not meritorious. Further discussion will have to wait for another day.


suspension of the dispositional, God can prevent every case of suffering at the hands of agents without failing to achieve his purposes. Therefore,

It is possible both that God achieves his purposes in creation and God prevents every moral evil.

Therefore, the free-will theodicies and defenses depending as they do on either (FP�) or (FP�) fail to exonerate God regarding cases of apparently pointless or excessive suffering as a result of the intentional, libertarian free action of agents. Therefore, even granting libertarian freedom,

(�) God cannot fail to “foreknow”33 libertarian free actions (if there are any), and

(�) God is able to allow, prevent or cause any event, and, thus(3) God is responsible for all the necessary conditions for the existence

of suffering.This result shifts the theological focus of God’s relationship to moral evil from theorizing about God’s ability to theorizing about his goodness. This God could, but would not position demands an account of God goodness just as the God would, but could not presupposes such an account.34

33. “Foreknow” here is nonstandard because it applies only during the formation of the proximal intention. However, it does not contradict the possibility of God’s knowing (from all eternity) what action will occur. What he does not know on such occasions is whether the action will be free or not.

34. I want to thank my students over the years who have endured my course in philosophy of religion and who have raised important and probing questions on these issues. I also want to thank David Natwick, Timothy Miller, Timothy Johnson, Courtney Friesen, Wayne Mayhall, Paul Helseth, John Milliken, Justin Daeley, Justin Cope, William Hasker, Daniel Johnson, and an anonymous referee for insightful and helpful comments on earlier drafts of this paper.

“No-Risk” Libertarian Freedom A Refutation of the Free-Will Defense

Documents