Tuesday, November 9, 2010

Excerpt from That Damned Book: Efficacious Grace and Rational Freedom

Since participants on this blog seem to have an ongoing interest in issues of free will, I thought I'd share something from the chapter I'm working on right now from That Damned Book. The excerpt is from the first of two chapters in which we critically assess the argument that God will not save all because to do so would require Him to trump the freedom of the unregenerate in a morally unacceptable way. In the chapter from which this excerpt is drawn, we consider the first of two responses to this "liberal" justification for the doctrine of hell--namely, the response which invokes the concept of "efficacious grace" (roughly, a divine act which guarantees the salvation of the sinner by ensuring that the sinner responds favorably to God's offer of loving union with Him). In this excerpt, we begin to consider the case for the view that it is possible for God to bestow efficacious grace. Once we narrow in on the "Thomistic" view on this matter (meaning the view endorsed by followers of St. Thomas Aquinas), we offer an account of the Thomistic view of freedom. It is this part of the excerpt that I think may be of special interest to readers of this blog.

That God can bestow efficacious grace was assumed by most older dogmaticians, whether Protestant or Catholic, who discussed the matter. But the Catholics were divided over the nature of efficacious grace. Adherents to the older Thomist and Augustinian tradition took such grace to differ in kind from so-called “merely sufficient grace” (which gave sinners all they needed for salvation other than the appropriate subjective act of will). Followers of Molina, however, held that efficacious grace does not differ in kind from sufficient grace, but differs only in virtue of the fortuitous situation the creature is placed in when receiving sufficient grace.

While volumes of scholastic theology have been written on this dispute, for our purposes a brief overview is sufficient. According to the Thomists and their Protestant followers, when God grants efficacious grace, what He does is guarantee conversion and regeneration by putting creatures in a state that influences their motives such that they have every reason to respond favorably to the offer of salvation and no reason not to. But if this is what efficacious grace involves, it raises important questions about the nature of creaturely freedom. Most significantly, one may wonder if efficacious grace is consistent with libertarian freedom—by which we mean, roughly, the power to act or not act on motives that incline but do not determine the will. Freedom in this libertarian sense exists only if, when one makes a choice, one could have chosen otherwise—that is, there is some possible world in which one chooses otherwise.

In affirming that God can grant efficacious grace, the Thomists did not mean thereby entirely to deny creatures freedom in something like this sense. Rather, they meant simply to limit its scope. Specifically, the Thomistic view is that what we call libertarian freedom is a coherent understanding of freedom only when the creature confronts conflicting motives for action. It does not extend to circumstances in which the creature has every reason to pursue a given course of action and no reason not to. Under such circumstances the Thomistic view is that the will of the creature is not merely inclined towards the given action but determined to do it. The action remains wholly voluntary, but there is no possible world in which an agent who has every motive to do A and no motive not to nevertheless refrains from doing A—and so, it seems, the action is determined even though voluntary, and so conforms to what is usually labeled “compatibilist freedom” by contemporary philosophers.

But it strikes us that this contemporary language implies something Thomists did not mean to imply—namely that there are two kinds of freedom, compatibilist and libertarian. We find it more in tune with Thomistic ideas to say that freedom simply operates differently under conditions of uniformity of motives than it does under conditions in which motives conflict. In the latter case, free choice looks like what we think of when we speak of libertarian freedom, whereas in the former case it looks like what we think of when we speak of compatibilist freedom.

If this is right, God could guarantee that the unregenerate freely-but-inevitably make the subjective choices necessary for salvation. For Thomists, this is the essence of how efficacious grace works: it brings all the creature’s motives into conformity with the choice of pursuing loving union with God above all things.

The Molinists, however, objected to the Thomistic view of efficacious grace because they took it that freedom has a libertarian character even under conditions of uniform motives. Granted this strong notion of libertarian freedom, it would initially seem impossible for God to give efficacious grace without first extinguishing the creature’s freedom. But the Molinists argued, on the contrary, that (i) God has middle knowledge, and (ii) for every rational creature there is a possible world in which she would freely (in the strong libertarian sense) respond favorably to God’s offer of salvation. That God has middle knowledge means He knows, for any creature X He might create, what X would freely do in any circumstance God might put X in. Thus God can give efficacious grace to X by creating that world in which He knows, by middle knowledge, that X would favorably respond to grace.

For a number of reasons we are unconvinced by this Molinist doctrine. First, we are not convinced that divine omniscience entails middle knowledge. It is not clear to us that, in the absence of an actual (libertarian) free choice made under an actual set of circumstances, there is any truth of the matter with respect to what the agent would freely choose in the libertarian sense. Furthermore, as William Lane Craig has shown, supposing that God does have middle knowledge does not, by itself, demonstrate that He can give efficacious grace. Hence, we will not argue here that efficacious grace is possible on the Molinist view of freedom. Given the strong libertarian perspective of the Molinists, we think a different argument for universalism (which we discuss in the next chapter), is more compelling than the argument from efficacious grace developed here.

In this chapter, then, we will argue that God has available to Him a morally permissible means of bringing it about that all a creature’s motives uniformly favor conversion. If so, then on a Thomistic view of freedom there is a morally permissible means for God to guarantee that all freely make the choices necessary for salvation.

So, is there a means whereby God could, without moral fault, bring about in a creature uniformity of salvation-inducing motives? To answer this, we think it may be helpful to develop a fuller picture of the Thomistic view of freedom by way of a contemporary philosopher—Thomas Talbott—whose thinking is very close on this matter to that of the Thomists.

Like the Thomists, Talbott insists that one cannot imagine anyone freely choosing what they have no motive to choose and every motive not to choose. Such a choice, for Talbott, is incoherent. If one is in a condition such that all of one’s motives converge on a single choice, then Talbott thinks this choice becomes inevitable.

On Talbott’s view such a choice may nevertheless be truly free—but only if certain conditions are met. Talbott argues that if ignorance or deception entails that one chooses based on misrepresentations of the alternatives (such that what one thinks one is choosing is different from what one is actually doing), then one’s freedom is impeded. And if controlling affective states entail that an agent is determined to choose one option even if informed deliberation would come down in favor of another, then the agent is “in bondage to desire” and, again, not truly free. But if someone is “freed from all ignorance, deception, and bondage to desire,” the agent’s choice is free even if all motives converge on a single option, thereby making the choice of that option inevitable.

But if all of this is right, then there will be different ways to produce conformity of motives, not all of which should be assessed in the same way. Consider the following case. Suppose Jenny grows up in a dystopian future where all children are fed a highly addictive drug from infancy. They are taught (falsely) that the drug is a medicine they need to stay healthy—while in fact it is used by a tyrannical regime to control the people. Given her addiction and beliefs, Jenny’s motives all converge on the choice to continue taking the drug. But insofar as this choice is governed by deception and addiction, it is not free in Talbott’s sense.

But suppose a resistance group reveals to Jenny the truth, so that she now knows the drug is harmful but remains addicted. She now has reason-based motives to stop taking the drug, but they are impotent because she is in bondage to her addiction. Perhaps the resistance gives her a counter-drug that weakens the strength of her addiction but does not shut down the cravings. Now, whenever she is in the vicinity of the drug, she faces an inner struggle. Sometimes, with the right help and support (and a bit of luck) she can resist her craving; but usually she falls prey to it, weeping in horror at her own weakness. At this point we might say she has some measure of freedom—but it remains constrained by the hold the drug continues to exert.

But then imagine the resistance group finds a way to break her addiction. Now she neither craves the drug nor thinks taking it is a good idea. Let’s suppose, further, that she has no other motive to continue taking it but many reasons not to: concern for her health and continued freedom from addiction, gratitude to her liberators, a desire to oppose the unjust regime, etc. Suppose, in other words, that once she is finally freed of her addiction all her motives converge on a single choice: not to take the drug. Even if this means (as Talbott and the Thomists believe) that her rejecting the drug is inevitable, we wouldn’t want to say her choice isn’t free. Rather, we’d say that, with respect to this issue, she is truly free for the first time.

This example shows, we think, that the “libertarian” and “compatibilist” labels are both inadequate for the sense of freedom that Talbott champions. Prior to help from the resistance, Jenny’s choice to take the drug would be free in the compatibilist sense—but not free in Talbott’s sense. After her final liberation, her choice to refuse the drug is free in Talbott’s sense—but not in the libertarian sense (which assumes the possibility of having chosen otherwise). What makes the choice free in the one case but not in the other is that reason is no longer impeded from playing the role it ought to play in decision-making. Hence, we think the best label for this conception of freedom is “rational freedom.” And given Aquinas’s emphasis on the natural ordering of the will to follow reason, we think such “rational freedom”—which treats only some cases of uniform motives determining one’s choices as free—best captures the Thomistic view.

Underlying this view of freedom are several presuppositions, which can be summarized as follows: (i) Values are objective, such that there are objectively good or best choices and objectively bad ones; (ii) the rational faculty makes judgments in accord with its finite grasp of this objective order of values; (iii) the will can be controlled by non-rational forces (such as addictions, childhood coping mechanisms, entrenched habits); (iv) the will is naturally ordered to choose in accord with rational judgments, such that in the absence of non-rational controlling factors the will always chooses in accord with reason.

These presuppositions imply, in brief, that the will is naturally ordered to follow reason (its “default setting,” if you will, even if the will can be reprogrammed), and that reason in turn is naturally ordered to discern the objective good. On this view, choices are free to the extent that both will and reason can operate in accord with their nature—that is, there is nothing (such as “ignorance, deception, or bondage to desire”) that impedes them from acting on their natural teleology. This, we think, captures the essence of freedom as it is understood both by the Thomists and, more recently, by Talbott.

In any event, what all of this shows is that one can bring about uniformity of motives both in ways that impede “rational” freedom and in ways that do not. As such, if efficacious grace is a divine act of producing uniformity of salvation-favoring motives in the unregenerate, this divine act may or may not impede freedom in the Thomistic sense—depending on whether this uniformity is produced by inducing false beliefs and/or affective states at odds with reason, or whether it is produced by revealing truth and removing affective barriers to acting on what reason discerns.

Given this perspective, the question is not only whether God can bring about uniformity of salvation-favoring motives in the unregenerate, but whether He can do so in a way that promotes rather than impedes rational freedom. In fact, we think it falls within the power of an omnipotent being to do both. Hence, not only do we think it is within God’s power to bestow efficacious grace. We think it is within God’s power to do so in a way that does not violate the freedom of the unregenerate—at least if “freedom” is understood in this Thomistic sense.


  1. Eric

    You may be able to hep clear something up for me here, because I don't understand this idea of possible worlds at all.

    You write of the kind of freedom where there is some possible world where one could have chosen otherwise. I assume this means that the physical inputs are the same across the worlds. So, to stop in at the shop for an ice cream or not? In both worlds my hunger, body image, climate, hurry, hormone balance, income, concern for animals etc are identical. And my various motives thus entwined incline me towards but do not determine my stopping. What can this mean?

    It sounds as if this means the inputs are fed into some sort of probability function, but in the different worlds the die can roll as it will (so if the inclination to stop is strong, we might define the function such that a one means drive on, but any other number means stop, for example). The problem here is clear, the rolling of a die doesn't feel at all like freedom to me. It is entirely random, so although I have freedom, I have no actual power in the exercising of it. Is this what is meant by libertarian freedom? It seems an odd definition.

    Sometimes people speak as if they think of a self that floats free of the probability function but can nevertheless guide it in a non-random way. But this is perhaps contradictory, for any non-random guiding of that function's resolution is itself part of the function I would have thought, it can be built into the calculation of the strength of various competing inclinations and triggers.

    Freedom might mean something more limited, as used by Dennett say, which is the ability of a self aware agent to anticipate various outcomes according to the actions taken (duck, don't duck) and then choose a path depending upon the desired result, but this would appear to remain static across all possible worlds, because the key things, expectations, desires and weightings, are presumably part of the background conditions.

    So, I'm confused as to what the Thomist tradition is saying. Is there a way of framing the possible world stuff without relying upon randomness becoming the deciding factor? And if not, in what sense is being beholden to random fluctuation freedom?


  2. Eric,

    I’d like to propose the following premises:

    1. As a matter of fact our current condition is such that there exist “conflicting motives”, such as beliefs, values/desires, feelings, etc. (Further, unless we achieve some future condition of theosis, it is reasonable to assume that our condition will always be one of at least minimally conflicting motives. This further qualification is not required for the point I am here trying to make though.)

    2. These conflicting motives determine the probabilities of how we shall freely choose to guide our life towards God or else away from God.

    3. The free choice among these probabilities looks from the outside exactly like a random choice. (This premise avoids the complications inherent in the libertarian versus compatibilist debate about free will.)

    4. A choice that brings one closer to God is on average experienced as good, and thus as a personal gain. Or, more precisely, a choice that brings one closer to God will on average produce more personal gain than any other choice.

    5. A minimally rational being will tend to choose in a way which maximizes their personal gain. (In terms of motives this says that a choice that produces a personal gain will tend to affect one’s motives - beliefs, values/desires, feelings etc - in a way that will make it more probable that in the future one will choose in a way that maximizes one’s gain.)

    6. We are minimally rational beings.

    7. The opportunity we have for coming close to God extends far beyond our current condition in this life. (That’s the view of the Irenaean theodicy, and is the most direct interpretation of the Gospel parable according to which God is like the shepherd who does not tire to look after the one lost sheep.)

    I think that virtually all theists would agree with premises (1) to (6). If one also accepts (7) then universal salvation becomes a necessary outcome, even though there will always be possible worlds where at least one person will freely (or “randomly”) choose not to come closer to God. The reason is that given the limited number of humans the probability of at least one freely choosing to remain far from God becomes vanishingly small with the passage of time.

    The claim that an outcome is necessary even though there is a possible world in which it does not obtain may sound incoherent at first, but in fact is not. Consider something as simple as continuously flipping a fair coin. Even though there is a possible world in which one will never get heads, it is necessarily true that after some limited time one will get heads. Similarly, giving the universal attraction of God’s love, no matter how “lost” or “perverted” or “misguided” we are, given enough time we shall all in the end freely find our way towards God. (Indeed, the coin flipping analogy is too strong; one flip of the coin is independent from the others, whereas our choices, and the motives that guide them, are dependent on our past choices as well as on the people around us. That’s why I think salvation is not only universal, but communal.)

    My point here is that all theistic thought about various types of grace, and various types of omniscience, and various types of freedom – unnecessarily complicates matters. One can see the reality of universal salvation without first resolving such difficult theological issues. Indeed grace is here understood as referring to an inherent, and thus universal, property of creation, namely the property which guarantees universal salvation by free choice. Which is indeed the best possible outcome among all possible worlds.

  3. Dianelos,

    Your argument is VERY close to what we develop at length in the chapter after the one from which this excerpt is taken. We argue, in effect, that if libertarian freedom is assumed even under conditions of uniform motives--or if it is assumed that conditions of uniform motives cannot be attained PRIOR to salvation--then we can still make the case for universal salvation assuming an infinite timeline in which creatures are free to make choices pertaining to union with God. Given such a timeline, the probability of anyone being damned become infinitesimally small as the limeline approaches infinity.

  4. Bernard,

    You have described very nicely the conundrun of so-called libertarian freedom. In a recent article in Faith and Philosophy, I come to the same conclusion you do--that in order for it to be possible that I choose otherwise than I in fact do under the same input conditions, there must be some random element at work in my choice. In fact, the same line of argument is nicely developed by the 19th Century American theologian, Jonathan Edwards.

    One way to characterize the basic dilemma is as follows: Either I choose X for a reason, or I choose X for no reason. If I choose X for no reason, then I choose at random. So, either I choose X for a reason or I choose at random. But suppose I have conflicting reasons leading to opposing choices. In that case, it might seem that what I am choosing is which reason for action will motivate me. So why isn't this--deciding among reasons--the locus of freedom? The problem is that it only moves the difficulty up one level. Do I settle on THIS motive rather than that one for a reason, or for no reason?

    Now whether you choose on the basis of a reason or at random might itself be based on a reason (you don't want to fall into a rut, say, and for this reason allow some chance to shape your choices). Or you might at random choose to let randomness decide. But eventually, go back far enough, and either randomness or some ultimate reason will lie at the root of the choice.

    But I wonder if there is a difference between a choice BEING random and a choice being made AT random. I don't see an alternative between either choosing for a reason or choosing for no reason and hence at random. But perhaps an agent BRINGING ABOUT a result (being its causal source--what is called "agent causation"), even if for no reason, is different from a result being produced directly by chance. In the former case, we might say there is an intermediary between the action and the forces of chance, and this intermediary (the agent) is directly causally responsible for the action (as opposed to chance being directly responsible).

    I'm not at all sure how helpful this distinction is, but at the moment it's my best hope for trying to get at how and why libertarian freedom--which includes the possibility of having done otherwise under precisely the same circumstances--might be construed as different from mere randomness.

    Ultimately, however, my approach to this dilemma is surprisingly like your approach to the alternatives I present as flowing from the cosmological argument: Maybe there is some third option that we just can't get our brains around because of our finitude.

    I do tend to be a Thomist when it comes to situations of uniform motives. But when it comes to cases of conflicting motives, there does SEEM to be something more going on, when I go with THIS motive rather than THAT one, than mere chance at work. Perhaps this is a mistaken impression, but I'm not ready to leap to that conclusion. In short, I don't have an answer to your problem--but I keep looking, and in the meantime keep thinking there is something more going on than mere randomness (simply because I can't help it).

  5. Hi Eric

    Thanks for your very clear and honest response. It links nicely to the bigger theme doesn't it, of what to do in the face of the uncertain and the irresolvable? And this comes down in part I think to personality/instinct.

    Some will prefer to jettison consistency in favour of other values, like aesthetic appeal, just as some will prefer mystery over commitment to a particular hypothesis, and some will be more interested in pursuing research programs which will not be completed in their lifetime over building a personal metaphysical superstructure. In a way I suspect we must all, for pragmatic purposes, close down some avenues of enquiry, and in doing so accept our own philosophical achilles heel.

    For myself, I must operate as if I and others around me have free will, I can find no way around this, while on some other level believing there is no such thing. A slightly shabby, make-do compromise.