Abstract:It is not fashionable among Christian philosophers today to be a compatibilist about morally significant freedom and determinism. This essay sketches a case for the reasonableness of embracing compatibilism that involves both theological and nontheological considerations. This is followed by a critique of the most widely recognized challenge to compatibilism, the consequence argument against compatibilism, that attempts to show why such an argument cannot succeed. The essay concludes by noting several implications of the sort of compatibilism defended here for developing a satisfactory moral psychology.
I, an evangelical Christian philosopher, am a compatibilist about morally significant free agency and determinism. There, I’ve said it. I’ve fessed up. Compatibilism is currently not in vogue among Christian philosophers. In fact, it would not be an exaggeration, I think, to say that within the community of Christian philosophers, incompatibilism in its libertarian form is the generally recognized norm. Within our community, especially among those who self-identify as evangelicals, the libertarian view is all but taken for granted, with the consequence that much of the effort these days is spent teasing out the implications of that view for any number of other philosophical and/or theological issues. Even many philosophers whose roots are in the Reformed tradition and who otherwise generally align themselves with that tradition are happy to jettison the compatibilist part of the tradition and join the ranks of the libertarians.1 So, a philosopher like me who is both a Christian and a compatibilist seems a bit out of step with the Christian philosophical community at large. It doesn’t help that a majority of non-Christian philosophers are allegedly compatibilists of some sort or other. So, why would any conscientious Christian philosopher today embrace compatibilism?
In this essay, I wish to begin to address that question by sketching my own compatibilist view of the relation between morally significant freedom, moral responsibility, and causal determinism, and something of why I embrace it. It is not my intention in this essay to give an exhaustive treatment that interacts with all the current players in the debate or to delve into all the technical issues involved. I’ll touch on some of those issues, of course; but my aim here is simply to give a general overview of compatibilism and why I hold it. It has seemed to me that given the dominance of the libertarian position in the Christian philosophical community for some time now, it might be helpful to go back to the basics in defending the compatibilist view, and that is largely what I intend to do here. After reviewing the basic issue, I’ll briefly lay out my positive (and largely traditional) case for compatibilism. Then we’ll turn to a critical consideration of what is widely regarded to be the most important challenge to compatibilism today. Finally, I’ll conclude the essay by drawing a few implications of this discussion for developing a satisfactory theory of moral responsibility.
1. The Basic Distinction between Compatibilism and Incompatibilism
First, let’s be clear about what the problem is for which compatibilism is supposed to be the solution. The problem of freedom and determinism, as it is often called, is at bottom the issue of whether morally significant freedom (or free agency), and the moral responsibility of which such freedom is supposed to be a necessary condition, are compatible with causal determinism with respect to the acts of human agents. By “morally significant freedom” I intend simply that freedom that an agent must possess to be morally responsible for any particular act that he or she performs.2 So the question is this: can we be free in the morally significant sense if all our acts, including our choices, are causally determined by antecedent events and/or states? Compatibilists say “yes”; incompatibilists say “no.”
We may compare and contrast the basic positions on the problem of freedom and determinism in terms of the different attitudes people might take with respect to the following pair of claims:
(D) All of our acts, including our choices, are causally determined by antecedent events and/or states of affairs.
(F) We human beings are free in the morally significant sense with respect to at least some of our acts, including our choices.
Incompatibilists maintain that (D) and (F) are incompatible—that is, they affirm:
(I) It is impossible for both (D) and (F) to be true.
Note that the incompatibilist is claiming that (D) and (F) are contraries, not contradictories. That is, it can’t be that both (D) and (F) are true, but it might be that both are false. In other words, from (I) it does not follow necessarily that (D) and (F) have to have opposite truth values so that it has to be the case that one of the two propositions is true and the other false. All that’s being claimed by the incompatibilist is that the conjunction of (D) and (F) cannot be true.
There are various kinds of incompatibilists, but the most prominent kinds (and the ones who are most relevant to this essay) are libertarians and determinists. Libertarians are those who, in addition to accepting (I), take (F) to be true. Since indeterminism, the denial of (D), is entailed by the truth of the conjunction of (I) and (F), libertarians are indeterminists. On the other hand, incompatibilists who take (D) to be true are determinists (sometimes called “hard determinists”3). Determinists of this sort are logically forced to deny (F), the thesis that we have morally significant freedom, because of their commitment to both the determinist thesis, (D), and the incompatibilist thesis, (I).
Contrary to incompatibilists of either the libertarian or deterministic stripes, or any other stripe for that matter,4 compatibilists hold that (D) and (F) are compatible. They affirm the following proposition:
(C) It is possible for both (D) and (F) to be true.
Obviously, (C) is the contradictory of (I). It is impossible for both (C) and (I) to have the same truth value. Necessarily, one is true and the other is false.
A further distinction is sometimes drawn here. The simple compatibilist is only committed to the truth of (C). But some compatibilists are committed not only to the truth of (C), but to the truth of (D) and (F) as well. Call compatibilists who affirm (C), (D), and (F) substantive compatibilists to distinguish them from simple compatibilists, who might deny either (D) or (F) or both. My own view is a version of substantive compatibilism.
There are other views than the four so far enumerated, of course, but for the purpose of this essay, this will suffice. Before proceeding, I should make one more comment about my formulations of (I) and (C) above. I have used the words “impossible” and “possible” without qualification in those formulations. One might wonder precisely what sort of modality (possibility or impossibility) I have in mind.
Let me begin by saying what is not intended by “possible” and “impossible.” I take it that the issue here is not whether the conjunction of (D) and (F) is epistemically possible—i.e. possible so far as we know or so far as we can tell. Nor is the issue that of whether the conjunction of (D) and (F) is causally (or physically or nomologically) possible. That is, the issue is not whether, given the natural laws of our particular space-time universe, it’s possible for both (D) and (F) to be true.
It seems clear that the issue between compatibilists and incompatibilists is either one of metaphysical possibility or logical (conceptual or semantic) possibility.5 If the issue concerns metaphysical possibility, we can usefully think of it in terms of possible worlds.6 The question in that case is whether there are any possible worlds in which both (D) and (F) are true. Their conjunction is metaphysically possible if and only if there are some such possible worlds, whether or not the actual world is one of them. The question of logical or conceptual possibility, on the other hand, is whether the conjunction of (D) and (F) constitutes or entails a contradiction. That conjunction is logically possible if and only if it neither constitutes nor entails a contradiction.
Now, logical possibility is more extensive than metaphysical possibility. That is, anything that is metaphysically possible is ipso facto logically possible. Conversely, logical impossibility entails metaphysical impossibility. The entailment does not go the other way, however. From the fact that something is logically or conceptually possible (i.e. is not self-contradictory) it does not follow necessarily that it is metaphysically possible. Nor does metaphysical impossibility strictly entail logical impossibility.
We need do no more than distinguish these kinds of modality here. The sort of compatibilism that I’m interested in takes it that the conjunction of (D) and (F) is metaphysically possible, hence logically possible. In other words, the version of the central compatibilist thesis, (C), that I embrace is one that involves more than merely saying that the conjunction of (D) and (F) does not violate the law of noncontradiction.
2. Why I Am a Compatibilist
Why think that compatibilism is true? I’m a compatibilist for a variety of reasons. In this section, I wish to rehearse briefly some of those reasons. Again, my aim is not to offer as exhaustive or rigorous treatment as possible. My aim here is simply to give the reader some idea of why I am convinced that compatibilism is true and reasonable to accept or believe.
Before proceeding, it might be helpful to say a little something about my own more general metaphysical and epistemological commitments. My methodological orientation as a philosopher is generally that of the Scottish common sense school of philosophy. While my view of the relation between morally significant freedom and causal determinism differs from that of many common sense philosophers (for example, that of Thomas Reid, the father of the Scottish common sense school), nevertheless, I think that this general approach to doing philosophy is superior to the alternatives.
Even more significant to my own reasons for accepting compatibilism is the fact that I am a thoroughgoing theist and an evangelical Christian. I believe that God, as traditionally understood in Christian theism, exists, and that he has spoken, both in “the book of nature” (or “general revelation”) and in Scripture (“special revelation”). My belief about Scripture is especially significant for my inquiry. I embrace the Scriptures of the Old and New Testaments as not only humanly authored, but also divinely authored or God-breathed (2 Tim 3:16).7 One particularly significant implication of this view of Scripture, of course, is that it has special authority not possessed by any writings that are the products of merely human authorship.8 Indeed, because it is God’s revelation, it is absolutely authoritative for our thinking and conduct. That is, its normativity with respect to belief and conduct is such as to be nonoverridable. Nothing can trump the authority of Scripture. Thus, in forming my own views about freedom and determinism, I take Scripture to be normative in whatever it says that is relevant to our theorizing. In what follows, though I shall not go into much detail, it will be evident that the data of Scripture as I understand it and traditional Christian doctrines that are derived from or based on Scripture crucially shape my thinking.
So, why am I a compatibilist? I am a compatibilist for both theological and nontheological (merely philosophical) reasons. In what follows, I shall offer some reasons of each kind.
2.1. Theological Reasons for Compatibilism
I’ll begin with some of the theological reasons that motivate my acceptance of compatibilism. In general, it seems to me that compatibilism comports better with traditional Christian doctrines than does incompatibilism. For example, I think that compatibilism is logically compatible with a robust doctrine of God’s absolute sovereignty (including strong doctrines of divine foreordination and providence), while incompatibilism is not.9 The witness of Scripture throughout seems to be (a) that God has from eternity foreordained, and throughout history has and continues to providentially govern (in an active, not passive, way), everything that occurs10—not just some things, but everything, including the acts of human beings—and at the same time (b) that human beings are (quite often) morally responsible for their acts.
By the way, it is because of this latter point—i.e. that the Scriptures clearly indicate that human beings bear real moral responsibility—that I have no truck with theological versions of incompatibilistic determinism, such as hyper-Calvinism. Any view that does not recognize human beings to be morally free and responsible agents is simply inconsistent with the claims of Scripture. With respect to theological categories, I am a Calvinist, not a hyper-Calvinist.
But even if one balks at the claim that God has foreordained and providentially controls everything that comes to pass in this world, surely one should admit that Scripture seems to record some specific instances of human choices and action that God foreordained and was actively engaged in bringing about and for which the relevant human agents were nevertheless morally responsible. Consider just one particularly notable example: the crucifixion of Jesus Christ. In the record of Peter’s sermon at Pentecost in Acts 2, we find Peter saying the following about Jesus’s crucifixion:
This man was handed over to you by God’s set purpose and foreknowledge, and you, with the help of wicked men, put him to death by nailing him to a cross. (Acts 2:23, NIV)
We find something similar in the prayer recorded in Acts 4 of the Christians who had just heard Peter and John report on their meeting with the Sanhedrin.
Indeed Herod and Pontius Pilate met together with the Gentiles and the people of Israel in this city to conspire against your holy servant Jesus, whom you anointed. They did what your power and will had decided beforehand should happen. (Acts 4:27–28, NIV)
These early Christians seem clearly to be thinking of Jesus’s crucifixion as both something that was planned (hence determined11) by God himself and something for which the humans involved are morally blameworthy. This idea that Jesus’s crucifixion was the result both of the foreordination and providence of God, on the one hand, and the sinful actions of men, on the other hand, certainly squares with the witness of the four New Testament gospel accounts and indeed the rest of Scripture. Note that even this one instance alone is enough to show that significant moral agency is compatible with determinism. There are other instances of this coupling in Scripture as well (e.g., in the Joseph story of Genesis 37–50, the account of Pharaoh’s hardened heart in Exodus 7–14, etc.), any one of which is sufficient to establish the truth of the compatibilist’s central claim, and with that, simple compatibilism.12 That in itself is an exceedingly significant point, for if there is even one instance of morally significant human action that is causally determined, then the rug is pulled out from under incompatibilism.
Furthermore, compatibilism is clearly consistent with the traditional doctrine of divine omniscience (or more particularly, the doctrine of divine foreknowledge), whereas incompatibilism in its libertarian form is not—or so it seems to me, at any rate.13 According to the traditional doctrine, God knows all things, including the morally significant acts of human agents, before they occur.14 Such knowledge entails that there is a truth of the matter about what any agent does before he or she does it, and that in turn entails that the act is pre-determined.15
It seems to me that there is an inherent instability in the combination of libertarianism about morally significant freedom and traditional Christian theism. While I shall refrain from developing and defending this claim here, suffice it to say that it seems to me that theistic libertarians ultimately face a dilemma of either (a) giving up their incompatibilism or (b) displacing the traditional doctrine of God’s omniscience with a thinner doctrine of God’s foreknowledge—one that does not affirm that God knows absolutely everything before it exists or occurs.16 Even if that is not the case, however, it certainly seems on the face of it that the compatibilist view, at the very least, fits more readily with the traditional Christian doctrine of God’s omniscience, according to which God foreknows even the future contingent acts of human agents, than does the incompatibilist view.
Compatibilism also seems to me to square better with a traditional Christian anthropology. The biblical portrait of human nature over the span of redemptive history seems to me clearly to favor a compatibilist view. As theologians have noted through the centuries, Scripture seems to indicate that the fall in sin brought about a significant change in our agency, specifically with respect to our ability to obey or disobey God. Whereas before the fall, human beings were able either to sin or to refrain from sinning,17 after the fall we were unable to avoid sinning. In our fallenness we are dead to God and to true righteousness. By God’s grace, redemption brings about another major change in those of us who are redeemed that significantly affects our agency. Regeneration renders the agent alive to God and true righteousness, hence able not to sin. Moreover, I think that Scripture supports the claim that those who are regenerate are ultimately incapable of falling away from God again. Finally, in a future event that evangelical and Reformed theologians call “glorification,” the regenerate will be confirmed in righteousness. That is, we who are through faith united to Christ will be made personally fully holy and impeccable—incapable of sinning—by God in his grace.18 This is part of the Christian’s eschatological hope. We look forward to being forever completely free from sin—not just its penalty, but also its pollution and power. Indeed, we eagerly look forward to being perfectly virtuous and unable to sin. In other words, we look forward to being morally perfect free agents, confirmed in a personal righteousness that can never be lost.
My point here is simply that compatibilism seems to accord quite well with the biblical data concerning human nature, while incompatibilism does not. The incompatibilist, to be consistent, must take the effects of the fall and redemption on human nature to be less radical than what I’ve suggested above (which I take to be the teaching of Scripture). In our fallenness, we must still be able to avoid sinning, according to the incompatibilist. Christians (the regenerate) must be capable of rejecting God and returning to a state of fallenness in sin as they were prior to regeneration. Even in the New Heaven and New Earth, Christians must still be really able to sin, if incompatibilism is true and if we are still to be morally responsible beings. That is, incompatibilism seems to entail that impeccability is utterly impossible for us, even in the life to come after the resurrection, if we are to continue to exist as moral agents. The alternative to denying the eschatological impeccability of Christians for the incompatibilist who believes in an afterlife for Christians would be to concede that we who are Christians will be transformed so that we can never again sin, but along with that deny that we are moral agents from the moment we lose the real possibility of sinning. In other words, the price of accepting impeccability for the incompatibilist is that we lose our status as moral agents. On that alternative, not only can we no longer be morally vicious; we can no longer be morally virtuous, either. Neither the denial of impeccability for Christians in the afterlife nor the denial of morally significant freedom for Christians in the afterlife seems to me to square with the witness of Scripture.
There are other traditional Christian doctrines—that of the impeccability of Jesus Christ in his earthly life, for example—that are consistent with compatibilism but not with incompatibilism, or at the very least seem to me to fit much better with compatibilism than with incompatibilism. However, I trust that I have offered enough already to indicate something of the way in which I would contend that compatibilism is more reasonable to accept than incompatibilism on specifically theological or biblical grounds. It’s time to turn to some of the more generally philosophical (nontheological) reasons for my acceptance of compatibilism.
2.2. Nontheological Reasons for Compatibilism
The first nontheological reason I would give for accepting compatibilism and rejecting incompatibilism is that, while I am quite convinced that we are moral agents, I am inclined to think that specifically libertarian freedom19—the sort of freedom insisted on by the incompatibilist—is not really possible. It is at least far from clear to me that such freedom is really possible. According to incompatibilists, morally significant freedom requires ultimate indeterminacy of the act (or, according to some incompatibilists, indeterminacy of some relevant prior act20) by antecedent events and/or states of affairs, even if those antecedent events and/or states of affairs render the act probable in some way. The problem is that a causally undetermined event, such as an act of choice that is free in the sense required by the incompatibilist, would be ultimately inexplicable.21 There is, we might say, a certain “chanciness” about such an act. It is in some sense the “product” of chance or happenstance, an event that “just happened,” an act that was “just done.” In particular, a genuinely free act, on any incompatibilist construal, could not be explained sufficiently by the agent’s dispositions, affections, desires, intentions, beliefs, motives, reasons, etc.22 Such psychological factors cannot have necessitated or brought about the act if it was truly free in the morally significant sense, according to the incompatibilist. The idea that something—especially something as significant as an act for which the relevant agent is morally responsible—could “just happen” with no sufficient explanation defies common sense, to say the least, and I am inclined to think that such is impossible.23
Even if libertarian freedom is possible, however—and this is my second nontheological reason for embracing compatibilism—libertarian freedom would not be morally significant. There are really two problems here. First is the problem of luck (or randomness or chance). If an act is genuinely free in the libertarian sense, then as I noted in the previous paragraph, there is no ultimate explanation for the agent’s performing the act in question rather than refraining from it and/or performing some alternative act. The agent acts in a way that is ultimately not because of any motive or reason at all, even if that act accords with, and is made to some degree likely by, some particular motive(s) or reason(s). The agent’s beliefs, desires, and the like might limit his or her real or live options, to be sure; but within those boundaries, the agent’s actual act, if free in the incompatibilist’s sense, is in the end an event for which there is no sufficient explanation. All that can be said is that the agent did the act in question. The agent acted, and he or she did such-and-such. That’s all that can be said. There can be no further explanation as to why the agent performed that particular act rather than something else. But if that is the case, then how can the agent be morally responsible for the act? It would seem that indeterminacy of an act by potentially act-determining psychological facts about the agent cannot support moral responsibility, even if the agent is somehow the indeterminate cause of the act. If I am the agent in question, it is true that I might be said to cause the act in some sense; but the “I” who causes the act is not a moral “I.” Potentially act-determining psychological facts about the agent must ultimately be divorced from the act if it is to be truly free in the libertarian sense. Yet it is at least some of those very psychological facts that ordinarily enter into, and form the bases for, our moral judgments about acts and the agents who perform them.
Perhaps it would help to think about a concrete case, which I’ll call the Case of Chuck’s Choice to Cheat. Chuck, a college student, is tempted and thus faced with a choice: to cheat on Dr. Morris’s biology exam or not to cheat. Chuck struggles with the temptation as he goes through a process of deliberation during the final twenty-four hours prior to the exam. The following are some of his considerations. On one hand:
- He wants to make a good grade on the exam to keep his biology grade and his GPA in good shape.
- He believes that he is in grade trouble in the biology course, and that there is little other opportunity to improve his grade.
- He believes that it is highly improbable that he would be caught if he were to cheat on the exam.
- He believes that he would have a significantly better chance of getting a good grade on the exam if he were to cheat.
On the other hand:
- He believes that cheating is morally wrong.
- He believes that he would fail the course if he were caught cheating.
- He believes that certain people whom he loves—e.g., his parents and sister, his pastor, his best friend, etc.—would be horribly disappointed were they to know that he cheated on an exam, and that they would be disgraced were he to be caught.
- He would prefer to make a good grade legitimately rather than by cheating.
And of course there are other considerations that come into play in his thinking as he stews over this decision on the day before the exam. But we have enough here to get a good sense of the situation. Unfortunately, after much torment and vacillation, Chuck finally succumbs to the temptation, choosing to cheat on the exam.
Now, assuming that Chuck is morally responsible for his choice in this case, why did he choose to cheat? What if the explanation is that in the end, after considering the risks and so forth, Chuck preferred the potential benefits of cheating more than the potential benefits of not cheating? While he wanted the benefits of not cheating, he wanted the potential benefits of cheating even more; and it’s precisely because he wanted the potential benefits of cheating even more that he chose to cheat. What he wanted most in this case moved him to choose as he did.
This would not be at all surprising. In fact, it does not seem at all out of the ordinary. Ultimately, Chuck’s choice to cheat in this case reflects morally misplaced affections and priorities on his part. This seems on the face of it to be deeply significant from a moral point of view. Given the way I have set up the case, Chuck’s choice is not merely sad; it’s morally deplorable. We are inclined to assign blame to the agent in this sort of case. However, note that in this case as I have constructed it, Chuck was not free in the libertarian sense in his choosing to cheat. His choice was a function of who he really is.
Now modify the case slightly. What if we sever the causal tie between the psychological facts about Chuck and his choice to cheat so that he just chooses to cheat, but is not causally determined to so choose by his desires, beliefs, commitments, and the like. His choice to cheat is, in that case, a matter of chance. He happened to choose to cheat, but his choice is not because of any motive(s) or reason(s) he had. He might just as well have chosen not to cheat under precisely the same conditions. The only explanation for his choice to cheat is that, well, he just did.
Now, in this case, is the choice morally significant? It seems clearly to me that it is not. If his choice to cheat was not because of misplaced affections or something of that sort here, then that choice is not truly reflective of Chuck’s character. If his choice is not because of some reason or motive ultimately, he cannot rightly be blamed for acting for the wrong reason or for a morally bad motive. Chuck’s chance choice to cheat is no more than a sad, pitiable choice, as far as I can tell—a case of bad luck. It is not a deplorable choice. It is not a morally significant one at all. In this case, there is no basis for moral evaluation. Chuck just did it, and that’s just too bad.
The point here is that moral responsibility for an act is crucially linked to the explanation of the agent’s performance of the act. Some such explanations are morally praiseworthy, others blameworthy. Some are damning, others exculpatory. The “why” we do what we do matters. Dispositions, motives, affections, beliefs, and the like—the stuff of psychological determinism24—are the sorts of things that matter when it comes to moral evaluation. “I just did” doesn’t qualify as a morally significant explanation. Such an “explanation,” if true, would relieve the agent of moral responsibility and would constitute reason to pity the agent, no matter what he or she had done. Severing the causal tie of action to psychological determinants renders the agent out of control in some morally significant sense; the agent in such a case is a kind of loose cannon. Such could never qualify for moral praise or blame. Some psychological facts that are determinative of our acts, then, far from preventing moral responsibility, seem necessary to it!25 Unfortunately, the incompatibilist’s notion of freedom won’t allow such psychological determinacy for morally significant choices.
We should pause to note the irony for the libertarian here. Traditionally the typical motivation for incompatibilism on the part of the libertarian has been to secure moral freedom and responsibility. Unfortunately for the libertarian, it seems that the very thing that he or she was hoping to save is lost because of the demand for indeterminacy (including psychological indeterminacy). In the attempt to save morally significant freedom, the libertarian severs the very artery that supplies the life blood to moral responsibility. Or to employ a different metaphor, the libertarian is, regrettably, hoist with his own petard.
The second problem under the point that libertarian freedom would not be morally significant is what I call the problem of moral indifference. Even if freedom in the libertarian sense is possible and we actually have it on occasion, are the options in such cases ever really significant from a moral point of view? Think about those “six-of-one and half-a-dozen-of-the-other” cases like a case of which of two streets to take to walk home, where the two options are equidistant and equally qualified in terms of comparative advantages and disadvantages;26 or a case of choosing between several quarters in your pocket which one to put in the vending machine; or a case in which your friend wishes to pay for your dessert and coffee at a restaurant one evening, and you find yourself having to choose between a piece of pecan pie and a piece of key lime pie, both of which are favorites of yours. The sorts of cases I have in mind here are cases in which the particular choice that one makes from the available options is not a matter of great importance to the agent. It doesn’t matter to the agent which street he or she ends up taking, or which quarter is selected for the vending machine, or which kind of pie is ultimately chosen. In such cases, either option will do just fine.
Assume with the libertarian that in such cases we really do have libertarian freedom. We don’t have a determinative reason or motive for the particular choice we make; nothing in us or in the situation causally necessitates our making the choice that we make. But note that while there is some plausibility in thinking that such choices as these are free in the libertarian sense, they are morally indifferent. Neither option is morally better than the other in any of these cases. The question that I want to raise is this: are there any cases of choices that are plausibly taken to be cases of libertarian freedom and that are at the same time morally significant—such that the agent would be either in the right or in the wrong in what he or she chooses; such that the agent’s choice would either accord with or violate some moral duty; such that the agent’s act would manifest some moral virtue or some moral vice? I am doubtful that there are. Genuinely moral choices seem to me unlikely to be of the sort of choice that would or could ever be a “toss-up,” so to say, or a “six-of-one and half-a-dozen-of-the-other” sort of case. In fact, I suspect that we would consider one who took a significant moral choice in such a manner to be morally and/or cognitively defective in some way. Furthermore, if there are any such cases, they will inevitably run aground on the problem of chance as discussed above.
I have suggested in the last several paragraphs that at best, the sort of freedom envisioned by the incompatibilist—libertarian freedom—would allow morally insignificant choices. In cases that are potentially morally significant, such freedom would actually preclude moral responsibility. This brings me to the last reason that I shall mention here for accepting compatibilism, and that is that compatibilism accords with our ordinary experience and the way we actually live our lives. That is, we ordinarily act as if compatibilism is true, which suggests that, at the pretheoretical (i.e. common sense) level at least, we already recognize compatibilism to be true. For example, we take people to be fairly predictable. If we know Sally well, we feel confident that we can predict what she will do under certain circumstances. When Sally does something that surprises us and goes contrary to what we would have predicted, our tendency is to think either that there is some mitigating factor (i.e. that there is more to the story—e.g., she was drugged, or suffering brain damage from an injury or lesion, or …) or simply that we didn’t know Sally as well as we thought we did. The truth is that we just do not assume that her surprising act was a random or chance occurrence, free from causal determinants.
In morally significant cases, we want to know why agents do what they do, and we are particularly concerned with their motives. We take character, motives, desires, and the like to make a difference in moral evaluation of acts and agents. We praise people for their strength of character which is manifested in good deeds. We do not praise people for “just happening” to be good or “just happening” to act rightly. When we act badly, we don’t take it that we “just happened” to act badly. We don’t take sin to be a truly random or chance occurrence. Rather, we recognize a sinful act on our part as having exposed a flaw or weakness in our character. That is, we understand action to reflect the character of the agent. Parents and teachers concern themselves with the moral formation of their children, striving to cultivate character traits in children that will ultimately determine their actions for good rather than for ill. In these and many other areas of our lives, it seems to me that our attitudes, expectations, commitments, and activities presuppose a sort of psychological determinism along with a robust sense of moral significance and responsibility. As such, compatibilism seems very natural—a matter of common sense.
These, then, are some of my own reasons, simply stated and relatively undeveloped here, for taking compatibilism to be true. I think that we have good reasons to think that compatibilism is true and good reasons to think that its contradictory, incompatibilism, is false. We might not be able to give anything like a sufficient explanation of how human agents can be free in the morally significant sense and morally responsible for their choices and acts when at the same time those choices and acts are causally determined by antecedent events and/or states of affairs. Yet, it is reasonable to believe both that we are morally free agents and that our choices and acts are causally determined.
Our situation here is similar to our situation with respect to the rationality of embracing certain essential Christian doctrines concerning the Trinity and the person of Christ. Take, for example, the traditional Christological doctrine of the one person and two natures of Christ. We can’t explain it, but we have sufficient reasons to justify our belief that Jesus Christ is one person who is both fully human and fully divine. He is truly God and he is truly a human being. Yet he is one person, not two. It is rational for us to believe this even though we don’t know how it works, so to say. We have enough information to justify both our acceptance of the traditional Christian doctrine and our rejection of the skeptical claim that that doctrine is logically incoherent and utterly impossible. Analogously, if we have good reasons to believe that both the deterministic thesis, (D), and the thesis of morally significant freedom, (F), are true, as I think we do, then we have good reason to think that they are compatible.
3. The Consequence Argument against Compatibilism
In this section, I want to critically consider what many take to be the most important kind of argument against compatibilism.27 The argument, developed in various forms by several contemporary philosophers, is widely known in the philosophical community as the Consequence Argument against Compatibilism (or Consequence Argument for short).28 In this section, I shall critically consider a generic form of the argument which I take to be representative of all such arguments in the relevant respects. Nonphilosophers might find parts of the following analysis tough-going, but my hope is that the attendant benefits of conceptual clarity and argumentative rigor will make the difficulty of working through this worth it.
Taking “S” to stand for just any human agent, “A” for any particular act of S, and “t” for any particular time, we can begin formulating the argument as follows:
(1) If determinism is true, then S’s doing A at t is a necessary consequence of some set, C, of states and/or events that are antecedent to S’s doing A at t.
Since I am treating determinism generically here, I am intentionally formulating the Consequence Argument in a way that leaves open the question of what specific kinds of things constitute the causal antecedents of S’s act A. To apply premise (1) to any specific kind of determinism, the precise content of C—i.e. the specific types of causal antecedents to A that are involved—will depend on the sort of determinism that is in question. For example, if the kind of determinism in question is physical determinism (the kind, incidentally, that most philosophers who employ the Consequence Argument seem to have in view), then the relevant antecedents will be physical states and/or events plus the relevant laws of nature. For psychological determinism, the causal antecedents will be psychological events and states along with the relevant laws of nature. In the case of theological determinism, God’s eternal decrees and acts of providence will be the relevant antecedents. Whatever particular kind of determinism a compatibilist might recognize or have in view, that determinism will entail that our acts, including our choices, are causally determined by, hence the necessary consequences of, some antecedent states and/or events. That is what is important in the context of the debate between compatibilists and incompatibilists. Thus, it is acceptable here to leave the particular kind(s) of determinism, and with that the contents of set C, unspecified.
The next premise of the argument is an expression of a principle that we may call the Principle of Powerlessness with Respect to Antecedents or PPA for short.
(2) S cannot do anything at t to bring it about that those antecedent states and/or events that constitute C are other than they in fact are.
This principle is intuitively appealing, regardless of which particular kind(s) of determinism the compatibilist might have in mind. The relevant causal antecedents of S’s act A will be states and/or events in the past and the laws of nature if the brand of determinism in question is either physical or psychological determinism. The past is cemented for S, utterly unalterable by S at t. So are the laws of nature. S has no control over those laws—no power whatsoever to change them. On the other hand, if the relevant sort of determinism is theological, then the eternal decree(s) and providence of God will constitute the antecedents of S’s act at t. Whether the eternality of God’s existence and acts is atemporal or temporal, the point here will be the same; for either way, the relevant antecedents to S’s act A at t are immune to alteration or prevention by S at t.
Here is another principle that is crucial to the argument:
(3) If (a) there is nothing that S can do at t to bring it about that something, x, is not the case, and (b) something else, y, is a necessary consequence of x, then S cannot do anything at t to bring it about that y is not the case.
This is the Principle of the Transfer of Powerlessness or PTP. Like premise (2) above (PPA), PTP might well seem prima facie plausible.
Now, if we substitute “C,” the set of antecedents to A, for “x” and “S’s act A at t” for “y” in (3) above, we get the following instance of PTP:
(4) If (a) S cannot do anything at t to bring it about that C is not the case and (b) S’s act A at t is a necessary consequence of C, then S cannot do anything at t to bring it about that S does not do A at t.
Principle (4) is a straightforward instantiation of PTP. Thus, if PTP is true, so is (4). That is, given the truth of PTP, if C is fixed and unalterable to S at t and C necessitates S’s doing A at t, then S cannot refrain from doing A at t.
From (1), (2), and (4),29 we can derive the following:
(5) If determinism is true, then S cannot do anything at t to bring it about that S does not do A at t.
In other words, if determinism is true, then S cannot refrain from doing A at t.
So far, so good. But note that it is not yet clear how this is supposed to be a problem for compatibilism. The next step is crucial for doing just that.
(6) If S cannot do anything at t to bring it about that S does not do A at t, then S is not significantly free in doing A at t.
That is, morally significant freedom on the part of our agent, S, with respect to act A at time t requires that S be able at t to refrain from doing A at t. Premise (6), like premise (1), is definitional. (6) simply expresses an implication of a particular definition of “significantly free.”
With this definitional claim in place, it should now be clear how (5) is supposed to constitute a problem for compatibilism. In fact, we are now ready to draw the conclusion of the extended argument that is supposed to deliver the death-blow to compatibilism. From (5) and (6), we have the premises for a chain argument with the following conclusion:
(7) If determinism is true, then S is not significantly free in doing A at t.
Since “S,” “A,” “C,” and “t” are being used generically here to signify just any particular human agent, act, set of causal determinants, and time respectively, the argument is generalizable over all cases of human agency. If this argument is sound, then it turns out that we are never significantly free if determinism is true. In other words, morally significant freedom and determinism are not compatible after all.
How good is this argument? Does the Consequence Argument constitute a fatal objection to compatibilism? Can compatibilism be rationally defended in the face of this argument? I believe that compatibilism can be satisfactorily defended against the Consequence Argument. In fact, it seems to me that the Consequence Argument suffers from a fatal flaw that prevents it from rationally getting off the ground.
Let’s start by considering carefully premise (4). As we shall see, there is a problem with (4), and a fortiori with (3), the Principle of the Transfer of Powerlessness. That problem, in turn, affects (5) and (6) as well. Here again is premise (4):
(4) If (a) S cannot do anything at t to bring it about that C is not the case, and (b) S’s act A at t is a necessary consequence of C, then S cannot do anything at t to bring it about that S does not do A at t.
According to (4), if the causal antecedents to S’s act A at t are unalterable by S at t and S’s doing A at t is causally determined by those antecedents, then S cannot refrain from doing A at t. S cannot do other than he or she does. But it is important to note a significant ambiguity here.30
The ambiguity that I have in mind lies in the final clause of (4), the consequent of the conditional constituted by (4): “S cannot do anything at t to bring it about that S does not do A at t.” To see the ambiguity clearly, consider a distinction (suggested by Jonathan Edwards31) between two kinds of ability and inability: natural and moral.
S has the natural ability to do A if and only if S can do A if S wants to. If S cannot do A even if he or she wants to, then S lacks the natural ability to do A. For example, I have the natural ability to eat a serving of cockroach cobbler.32 I could eat some cockroach cobbler (provided some is available to me, of course) if I wanted to. It is the sort of thing that I can do in the sense of having the natural ability. I also have the natural ability right now to drive my car to the store if I want to do so. I don’t want to at the moment, but I could if I wanted to. I have the natural ability.
On the other hand, I do not have the natural ability to turn a cockroach cobbler into a blueberry cobbler merely by snapping my fingers. Although I might well want to (especially if I were hungry and cockroach cobbler was the only thing available to me at the time!), I could not do that no matter how hard I tried. Transforming a cockroach cobbler into a blueberry cobbler is a matter of natural inability for me. So is preventing an airplane from crashing in Australia while I’m sitting at my computer in north Georgia (USA) writing this essay; or leaping to the moon from my backyard, even with a running start. Only things that I could do if I wanted to do them qualify as things which I have the natural ability to do—things that are naturally possible for me, we might say.
What about moral ability? S has the moral ability to do A provided S can want to do A—i.e. only if it is possible for S to want to do A.33 If S cannot want to do A, then S is morally unable to do A, even if S has the natural ability to do A. I can order a bouquet of flowers for my wife. That is something that I have both the natural ability and the moral ability to do. Not only do I have the ability to order flowers if I want to, but it’s also quite possible for me to want to do so, and at times I do. It is also quite possible for me to want to practice playing the piano—to practice, say, Chopin’s first Ballade, or one of Rachmaninov’s Preludes—something that I enjoy immensely. It is even possible for me to want to practice playing the piano to such a high degree that I end up doing that instead of grading philosophy exams that are sitting in a stack on my desk.
On the other hand, I do not have the moral ability to intentionally cut off one of my hands, although I have the natural ability to do so. (For one thing, losing a hand would severely curtail my ability to play the piano, and with that, a great deal of the joy that I derive from playing the piano.) Nor do I currently have the moral ability to eat a serving of cockroach cobbler if one is offered to me. While I have the natural ability—I could eat some if I wanted to—given my current psychological makeup, cockroach cobbler seems so disgusting to me that I simply could not want to eat such a thing. Similarly, I could stage a hoax, misleading my dear father into thinking that I had died, if I wanted to. I have the natural ability to do such a thing. But given my affections, beliefs, commitments, and the like, I could not want to do such a thing. I simply lack the moral ability in that case.
The distinction between moral and natural ability is important, because the compatibilist need not be committed either to the claim that determinism restricts both kinds of ability for the agent or to the claim that morally significant freedom requires that the agent have both the natural and moral abilities to act differently than he or she actually does at the time of the act.34 For example, my own view as a compatibilist is that lacking the moral ability to do otherwise than one actually does—i.e. lacking the ability to want differently than one does, we might say—is compatible with morally significant freedom and responsibility. In fact, in some cases such moral inability is a virtue—deeply significant and valuable from a moral point of view! I especially have in mind cases of impeccability, such as that of Jesus, for example, or that of resurrected and glorified Christians after the consummation of Christ’s kingdom. Being morally unable to sin (i.e. unable to want to sin), while having the natural ability to do so, is a morally excellent state in which to be. It was crucial to our redemption that Jesus was so virtuous, and it is a state of character which I, as have most Christians historically, eagerly look forward to exemplifying by God’s grace in the eschatological future. The point here is that the compatibilist need not take ability to refrain from performing a particular act to be a necessary condition for morally significant freedom, especially if the sort of ability in view is moral ability. While compatibilists disagree among themselves about whether natural ability to act otherwise than we do is a necessary condition for moral responsibility, they generally agree that moral ability to do otherwise is not necessary. In other words, compatibilists generally hold that moral inability to do otherwise than we do is compatible with morally free agency.
Let’s return to our consideration of premise (4) in the Consequence Argument. Applying the distinction between moral and natural ability/inability to (4), it will become evident how that premise is ambiguous. According to the consequent (i.e. the final clause) of (4), “S cannot do anything at t to bring it about that S does not do A at t.” But in what sense is this supposed to be the case? Is the “cannot” here supposed to express moral inability or natural inability?
Note first that to be consistent with the inability expressed in premise (2) and part (a) of the antecedent of (4)—“S cannot do anything at t to bring it about that C is not the case”—the “cannot” in the consequent of (4) must be intended to express natural inability. Certainly one might want to change some of the causal antecedents of one’s action—some of one’s relevant dispositions, for example, or a relevant law of nature, or a relevant divine decree—even if one is naturally unable to do so. The problem is that we are unable to change those facts, even if we wanted to do so. But we still might want to change some of them, if only we could. Our inability to alter C is a natural inability. So, the “cannot” both in (2) and in part (a) of the antecedent of (4) is best taken in the sense of natural inability.
But notice that if we take the “cannot” at the end of (4) in the same way, (4) turns out to be false. For certainly, even if S’s doing A at t is causally necessitated by C in the sense that S is morally unable to avoid choosing to do A at t, S might still have the natural ability to avoid doing A at t. That is, S could do other than A in the sense that S would do other than A if some of his or her affections, tastes, values, desires, intentions, etc., were relevantly different than they actually are. If natural inability is what is intended in (4), then (4) is false, and so is premise (3), the principle (PTP) of which (4) is an instantiation. Moreover, if (4) is false, then so is (5).
Well, what if “cannot” in the final clause of (4) is taken to express moral inability rather than natural inability? Construed that way, (4) seems to me to be quite true (although the equivocation on the word “cannot” within the sentence taken as a whole might leave one a bit queasy). Moral inability to do otherwise than one actually does, or to refrain from doing what one actually does, does appear to be an implication of determinism—at least of the psychological and theological forms of determinism that I take to be true. And it is quite possible that S be morally unable to avoid doing A, even though naturally able to do so. Frankly, I think this describes the sort of situation in which actual human agents often find themselves.
But how is this supposed to be problematic for the compatibilist? All that I have conceded here is perfectly consistent with compatibilism. So far, the compatibilist has no cause for worry. The implication of determinism for agency so far derived is precisely what the typical compatibilist already recognizes and is prepared to embrace. I, as a compatibilist, am indeed committed to the claim that if determinism is true, the agent lacks the moral ability to avoid doing what he or she actually does, whether or not the agent has the natural ability to do so.
Left with nothing but propositions (1)–(5), the proponent of the Consequence Argument would face something of a dilemma at this point. As we have seen, for the argument to be sound up to this point and possibly convincing to the compatibilist, the “cannot” in the consequents of (3), (4), and (5) must be understood in the sense of moral inability. But again, that is a kind of inability that the compatibilist is quite happy to accept. So, to expose moral inability as a consequence of determinism will get no traction with the compatibilist, unless it can be shown somehow that this is problematic to moral responsibility. On the other hand, taking the “cannot” in the consequents of (3), (4), and (5) to express natural inability might be able to get some traction with some compatibilists—viz., those who take natural ability to do otherwise than one does to be a necessary condition for morally significant freedom. However, on this interpretation, (3), (4), and (5) turn out to be false. So, again, there is no real cause for worry for compatibilists.
Now, all of this affects the way we are to understand (6), which is crucial for connecting determinism’s implications for agency with morally significant freedom and responsibility. It is at this step that it should become clear how the reasoning represented in (1)–(5) is supposed to spell trouble for compatibilism. Here again is premise (6):
(6) If S cannot do anything at t to bring it about that S does not do A at t, then S is not significantly free in doing A at t.
Note that the antecedent of (6) is the same proposition as the consequents of (4) and (5). Thus, to succeed in connecting the implication of determinism for agency that has been derived at step (5) with morally significant freedom, the antecedent of (6) must mean the same thing as is intended by the consequents of (4) and (5).
As we have seen already, taken in one way, so that “cannot” indicates natural inability, (4) and (5) turn out to be false, leaving the argument unsound. So, again, the sort of inability to refrain from performing a particular act that is in view as an implication of causal determinism in this argument must be moral inability. Assuming that is the sense intended in (4) and (5), then to serve its bridging purpose in the argument, (6) must be understood in that sense as well. However, in that case the argument clearly begs the question against compatibilism. On the moral-inability interpretation of “cannot,” (6) clearly reflects a libertarian—i.e. incompatibilistic—understanding of what is involved in morally significant freedom. Compatibilists can (and presumably will) simply reject this construal of the notion of freedom. That is, they can deny that morally significant freedom requires the sort of ability on the part of the agent that (6) stipulates—i.e. the moral ability to refrain from doing what one actually does at a particular time. Whether such moral ability is required for morally significant freedom and responsibility is precisely what is at issue in the argument. The only way to avoid begging the question against compatibilism would be to provide good independent reason(s) for taking (6) to be true. However, such justifying reason(s) for (6) would render the Consequence Argument rhetorically superfluous. It would be the independent argument(s) for (6) that would be doing the work of rationally establishing incompatibilism, not the Consequence Argument. Thus, whether or not one has good independent reason(s) for taking (6) to be true, the Consequence Argument cannot do what it is designed to do. It cannot justify rejection of compatibilism.
My conclusion, then, is that the Consequence Argument does not, and indeed cannot, succeed as a refutation of compatibilism. Either premises (3), (4), and (5) are false, or the argument begs the question against compatibilism by smuggling in an incompatibilist notion of morally significant freedom—a notion of freedom that a compatibilist can and will simply deny. Unless there is a formulation of the Consequence Argument that avoids both questionable claims about the implications of determinism and begging the question against compatibilism (and I must confess that I have yet to see one that does), the argument presents no serious threat to the rationality of embracing compatibilism.
4. Conclusion: Toward a Satisfactory Compatibilist Moral Psychology
I hope that I have provided enough to see not only that compatibilism is eminently defensible, but that it is compatibilism rather than incompatibilism that rationally ought to be accepted, especially by evangelical Christians. The Reformed compatibilist need not be ashamed. There is, of course, much more to be considered, but it is time to take stock and draw this discussion to a close. In this final section, I want to make some brief observations based on what we have seen so far in this essay with an eye to the development of a satisfactory compatibilist theory of morally significant freedom and responsibility.35 Early in this essay, I stipulated that by “morally significant freedom” I intend simply that freedom that an agent must possess to be morally responsible for any particular act that he or she performs. This is obviously very general—schematic and relatively empty of content—but sufficient for our purpose of considering the issue of compatibilism. However, a satisfactory theory of moral psychology must ultimately yield a more specific definition. It will need to provide an analysis or account or explication of the concept of morally significant freedom that specifies the conditions that are individually necessary and jointly sufficient for an agent’s having such freedom. While I have not attempted to supply those details in this essay, our reflection on the case for compatibilism has suggested several implications for the project of filling out precisely what morally significant freedom requires.
In this essay I have given theological and nontheological reasons for thinking that compatibilism is true. In considering those reasons, two different kinds of causal determinism have come into view. First, it seems to me that we have strong theological (biblical) support for theological determinism, the view that God determines all contingent events and states of affairs. Traditional Reformed theologians think of this in terms of the doctrines of divine foreordination and providence. Theistic philosophers might unpack this in terms of God’s acts of selecting and actualizing one of the infinitely many possible worlds.36
Second, we have both theological and nontheological reasons to think that psychological determinism—i.e. the thesis that all the acts of human agents are causally determined by psychological facts about those agents—is also true. Given our circumstances, and especially certain psychological facts about us, in any case of action we are morally unable to do otherwise than we actually do. Thinking about this in terms of possible worlds, we might say that an act is psychologically determined for an agent, S, if in any possible world in which everything that is psychologically relevant to S’s act (e.g., S’s character, reasons and/or motives, perceived options, etc.) is precisely the same as it is in the actual world, S does precisely the same thing that he or she does in the actual world.37
It seems to me that any satisfactory theory of moral psychology must at least be compatible with both theological and psychological determinism, given that we have good reasons to accept both. Any theory of morally significant freedom and moral responsibility that is incompatible with either theological or psychological determinism will be less than acceptable.
Furthermore, I have suggested that not only are some psychological determinants of choices and acts compatible with morally significant freedom, but that some such determinants are actually necessary to moral responsibility. If this is correct, as I think, then to develop satisfactory theories of moral freedom and responsibility we shall need to try to pinpoint precisely which psychological factors are relevant either to facilitating morally significant action or to precluding or undercutting it. On my view, a robust sort of self-determination of choices and acts is necessary for moral responsibility. However, that cannot be self-determination of an indeterministic variety. In developing a satisfactory theory of moral psychology, then, it will be important to identify the psychological conditions that deterministically enable the sort of self-determination of acts that renders them morally significant—the factors that make the “I” who chooses and acts a moral “I.” Some such factors have been suggested in this discussion, but I have largely left this matter undeveloped.
Recognizing that both theological and psychological determinism are true and compatible with moral responsibility has a further interesting implication for reflection on human agency—and with this, I close. In any case of human agency, there will be, not one, but two sufficient explanations for the act of the agent, from a logical point of view—one in terms of God’s foreordination and providential governance of the act in question; the other in terms of other relevant facts in the actual world, including the psychological facts about the agent prior to and at the time of the act in question. Interestingly, each explanation is a sufficient explanation of the agent’s act in the sense of yielding an entailment of the act, and both are true. Moreover, the two kinds of explanation are very different, though compatible. That is, the two kinds of explanation are not redundant, but complementary. Because of this, in any case of creaturely free agency, while either an explanation in terms of God’s causal role or one in terms of the causal role of the agent’s own psychology and circumstances will be sufficient from a logical point of view to account for the fact of the agent’s performing the act, neither by itself gives a complete or whole account of the act. For that, we need both explanations and some account of how the two are related. It seems to me that an implication of this is that to be comprehensive, a theory of moral psychology must take into account both theological and psychological determination of our acts, showing (to the extent that we can) how each works and how they work together in such a way as to render us morally responsible in many cases. Obviously this is no small task, but perhaps not completely insurmountable.38