The Neapolitan region in southern Italy bred one of the great figures in the history of philosophy: St. Thomas Aquinas (1225-1274). Aquinas’s theories of action, morality, and law, all reflect a conception of human beings as essentially rational. It is in virtue of their reason that human beings have control (dominium) over their own actions. Non-human animals, which lack reason, act by natural instinct. They are “not masters of the movement of their appetite, for such movement in them is by natural instinct.” (ST I-II, 15.2. Resp.l )[1] Their appetitive movement is controlled by nature, and not by the animal itself (cf. ST I, 83.1). By contrast, the human soul has a rational part, and therefore, human actions are not determined by nature, but by the human agent herself. She is morally responsible for them because she has some sort of freedom of decision, which non-human animals lack. Thus Aquinas claims that “...the very fact that a human being is rational makes it necessary that a human being be characterized by free decision (liberum arbitrium).” (ST I, 83.1).

The rational part of the soul, which Aquinas normally refers to as ‘reason’ (ratio), has two faculties, one cognitive and another appetitive. Aquinas uses the word ‘intellect’ (intellectus) to denote the cognitive faculty, although he sometimes uses ‘reason’ in this more specific sense as well. The appetitive faculty, however, is invariably referred to as ‘will’ (voluntas).[2] It is through the complex interaction of will and intellect that distinctively human actions occur, according to Aquinas.

Aquinas claims that human beings are characterized by free decision (liberum arbitrium, see ST I, 83), and it is important that his account of human action should provide a plausible explanation of what this means. However, because of the great emphasis Aquinas places on intellect's role in human action, will seems at times to be left without anything one could plausibly regard as freedom, and to be reduced instead to the mere status of an obedient servant, who can only choose actions that intellect has judged approvingly.

If will is thus determined by intellect, Aquinas seems to be faced with at least two problems: First, it becomes difficult for him to claim that the will is free, or that human beings are morally responsible for their actions in virtue of having a will. If will is ultimately a mere executor of intellect’s deliberative judgments, it seems to be in virtue of their intellect and not their will that human beings have freedom and moral responsibility. Second, even if, on reflection, we agreed that such determination of will by intellect might fit a plausible conception of freedom and responsibility, we would still need to consider whether it fits the phenomena of human action. Is it plausible to think that all distinctively human actions conform to the agent’s intellective judgment, or do we know from experience that human agents sometimes act in defiance of their own intellective judgment? Aquinas himself seems to acknowledge the phenomenon of willful defiance of better judgment in his treatment of deliberate malice, where he claims that “a person is said to sin purposely and with resolute malice when he knowingly chooses evil” (ST I-II, 78.1). Surely, Aquinas wants to say that an agent who deliberately chooses evil is responsible for that choice, so here is an instance where someone seems to choose freely despite his better judgment. For these reasons, it would seem to be a serious problem for Aquinas, both as regards the internal coherence of his views and as regards their plausibility, if his theory of human action committed him to intellective determinism.[3]

I will eventually be arguing that, appropriately understood, will's dependence upon intellect's deliberation and judgment is not a liability but a necessary condition for freedom and moral responsibility, plausibly understood. Furthermore, I argue that initial appearances notwithstanding, Aquinas’ theory of action can accommodate the phenomenon of willful defiance of the dictates of practical reason. Before presenting an interpretation of Aquinas along these lines, however, I will begin by describing the interpretive and philosophical problem at hand a bit more fully.

The threat of intellective determinism

According to Aquinas, a human being shares with the rest of natural substances inclination toward what is good. But a human being's possession of reason makes her unique in nature. Her reason enables her to have “cognition of the very nature of the good” as opposed to either having merely a cognition that some particular thing is a good (as non-human animals do), or inclining toward what is good without any cognition at all (like plants) (ST I, 59. lc). Will is inclination toward what is good with cognition of good in general. At this level, the good as object of will is determined by will's own nature, although intellect enables that object to take the form of a universal concept rather than only a particular representation.

Contrary to what one might initially suspect, the real problems for Aquinas' libertarianism do not arise from the claim that will has the good in general as a necessary object. That claim does not entail necessity with respect to any particular object of will. The will still has a “free range” among all the particular things which may be regarded as good from a particular point of view, and can thus be willed in virtue of their apprehended relationship to the good as such. [4] The problems begin to appear once it is recognized that in order for will to choose among the particular objects within that range, intellect must come to the scene first and present one of these objects as the good to be chosen. Consequently, will never seems to be free to pick for itself its own object; its fate is rather to take or leave what intellect offers. And this seems inconsistent with the familiar idea, which Aquinas himself explicitly embraces, that a free will should have the power to choose among alternate possibilities.[5] This power seems compromised if the will is only capable of choosing the particular alternative which intellect has judged to be the best.

Even though Aquinas never explicitly acknowledges this tension within his view, it is readily apparent from the text. On the one hand, he claims that a human being chooses freely (ST I-II, 13.6) and he very often gives the impression that will itself has the power to choose among alternatives:

Man is master of his own acts because he can deliberate about them, for when deliberating, reason is related to opposite alternatives, and the will can tend to either. (ST I-II, 6.3. ad 2., emphasis added)

From this description, it appears that intellect merely lays out the various alternatives and then leaves it up to will to pick as it pleases. Furthermore, in explaining that no particular good is a necessary object of the will, Aquinas claims that since any such thing can be regarded as a non-good, it “can be refused or accepted by the will which can tend to one and the same thing from different points of view.” (ST I-II, 10.2 R, emphasis added) Aquinas also describes the will as “an active principle which is not determined to one thing, but is related indifferently to many” (ST I-II, 10.4 R, emphasis added) and claims that there is a specific “way of causing, which is proper to the will as master of its own acts apart from the way which is proper to nature as determined to one thing” (ST I-II, 10.1. ad.1, emphasis added).

On the other hand, Aquinas also claims that “reason precedes the will in a certain way and orders its act to the extent that the will tends to its object according to the order of reason, for the power of knowing presents the appetite with its object” (ST I-II, 13.1. R, emphasis added). He also states that “man by his reason determines himself to will this or that thing” (ST I-II, 9.6. ad. 3, emphasis added). These passages suggest that when intellect has come up with an ordering among available alternatives, presenting one as clearly better than the others, will is not capable of choosing anything else. After all, will's only preference is for the good in general, and its only source of information about what constitutes the best alternative in a given situation is intellect's judgment. So how could it possibly act in defiance of that judgment? And if will cannot defy the dictates of practical rationality, how can it be considered free?

Aquinas’ arguments for the freedom of will

Aquinas offers at least three arguments for the freedom of will. To appreciate them, we need to consider the relationship between will and intellect a bit more closely. Aquinas distinguishes between two kinds of movement involved in an act of a power of the soul, (a) that of acting or not acting and (b) that of doing this or doing that (ST I-II 9.1). Whatever moves a power of the soul moves it in two respects: (a) in respect of the exercise of that power's act, and (b) in respect of the determination, or specification of the act. What moves in respect (a) is the subject of the act, and what moves in respect (b) is the object of the act. Will is the source of the movement of the act of all powers of the soul considered as a subject, including itself (9.3). Intellect, however, moves will only in the sense that it provides it with its object. The object gives an act its specification and operates in the way of a formal principle. Hence “the will moves the intellect as to the exercise of its act...but with respect to the determination of the act...the intellect moves the will” (9.1. ad. 3).

These distinctions allow Aquinas to hold that will is undetermined by intellect with respect to the exercise of its act. This means that even if intellect has presented will with an option as clearly the best, it is in will's power to withhold its consent and remain unmoved. The power of determining whether or not to act remains with the will, and in that respect, will is free from determination by intellect’s particular understanding of the good.

However, this first argument for will’s freedom doesn’t sound very reassuring to someone who is wondering about will’s freedom to choose among alternatives. Even though intellect doesn’t determine will’s acting or not-acting, will doesn’t seem to have any means of controlling how to act when it does. As if to meet this worry, Aquinas offers a second argument for will’s freedom, based on his assertion that will “moves intellect and all the other powers of the soul” as an efficient cause (ST I, 82.4). Since will moves not only itself but also intellect with respect to the exercise of its act, will has a certain means of influencing what it is that intellect presents as a choiceworthy object “for no matter what the object might be, it is in man’s power not to think of it, and hence not actually to will it” (ST I-II, 10.2. R). According to this second argument, will’s influence over intellect consists in its power to turn off intellective activity that seems headed for a practical conclusion that will doesn’t like.

The problem with this second argument appears once we consider what might lead will to turn off an intellective activity that might lead to an unwanted intellective judgment. Two possibilities suggest themselves: Either will pulls the plug on these thoughts because intellect itself suggests that thinking about this object would not be good (perhaps on account of a suspicion that such thoughts would lead to an erroneous practical judgment), or will pulls the plug independently of any cognition concerning the goodness or badness of thinking about this object. The latter option seems ruled out by the very definition of will as inclination to what is good with cognition of good in general. The act of turning off thoughts about an unwanted object would not be an act of will unless accompanied with cognition of that act as a good thing to do. The former option, however, seems to imply the same sort of dependence on intellect as before. Will is not free to cause intellect to think of the unwanted object as long as intellect itself determines that thinking about it would not be good. So if the will turns off these thoughts it is because intellect has judged that it should do so. The problem of will’s determination by intellect has therefore not been solved but rather pushed back one level.

Aquinas’ third argument for will’s freedom has already been mentioned, and it is of a rather different sort. This argument appears as his official answer to the question of whether the object of an act of will moves will with necessity, and the answer involves more than just pointing out that will moves itself (and intellect) to the exercise of its act:

Now because the lack of any good whatsoever has the aspect of a non-good, consequently only that good which is perfect and lacking in nothing is such a good that the will cannot not will it, and such a good is happiness. Any other particular good, insofar as it lacks some good, can be regarded as a non-good, and in this respect can be refused or accepted by the will, which can tend to one and the same thing from different points of view. (ST I-II, 10.2 R)

Any object, which intellect presents to will as good, might also have been presented to it as bad, because nothing is good in absolutely every respect except the ultimate good. It is therefore not out of necessity that intellect presents this object as good and not some other; intellect could have presented any of a vast number of actions as the thing to be done. Aquinas applies this argument specifically to the question of whether a human being chooses out of necessity or freely (ST I-II 13.6), and concludes that since choice is always of something particular for an end, and never of the ultimate end (which couldn't possibly be regarded as bad in any respect), no choice is made out of necessity.

Still, this argument doesn't seem much more satisfying than the other ones. Once it has happened that intellect comes up with an object and presents it as good, it is of no consolation to will that it is a contingent fact that intellect presents it with this object and not some other, if will has no means of changing that fact. Contingency is a necessary, but not a sufficient condition for freedom. Given the contingent fact of intellect's particular presentation of an object, will seems unable to move itself toward any other object. Consequently, there is no room for the idea that the will can choose among alternatives, or that “any other particular good ... can be refused or accepted by the will” (ST I-II, 10.2. R, emphasis added). In order to live up to that idea, it seems that will would have to be able to either push intellect to change its understanding of the good, or cause it to present various alternatives from which will would then be free to pick. But as we have just seen, Aquinas’ attempt to attribute these abilities to will does not seem very convincing.

Does Aquinas attribute freedom to will or reason?

In characterizing the problem, we have been assuming that will must be able to choose among alternatives, and the question has been whether intellect's role in decision prevents will from having that ability. On Aquinas’s part, it might now be asked in turn why we will, considered as entirely distinct from intellect, should be able to choose at all. Choice among a number of possible options has to involve not merely an arbitrary grabbing of one out of many possible objects, but also some awareness and consideration of these various alternatives. It would seem wrong to attribute this aspect of choice to a non-intellective feature of the human soul. Therefore, the demand, that will must be able to choose on its own from alternatives, seems to be misguided. This realization may point the way towards a more adequate understanding of Aquinas' conception of free will, and eventually enable us to evaluate its coherence and plausibility more fairly. With this in mind, let us consider further the theoretical and terminological context within which Aquinas asserts that a human being chooses freely.

According to Aquinas, any human act, i.e. an act for an end, consists of two steps, that of determination and that of execution. The act of choice (electio) is the last element of the determination, and it also causes the executive powers to start acting (cf. ST I-II 16.4).[6] Choice is always preceded by deliberation, which belongs to intellect, but presupposes a given end. Deliberation is undertaken only when will desires the end, and aims at finding the most appropriate thing to be done in order to attain it. Deliberation terminates in a judgment (iudicium), which is then followed by will's consent. Consent is an act of will which we can for present purposes regard as identical with the act of choice (for the precise distinction, see ST I-II 15.3 ad. 3).

Interestingly, although choice is an act of will according to Aquinas, he also claims that it is a distinguishing characteristic (proprium) of free decision (liberum arbitrium, cf. ST I 83.3; ST I-II 1.1 and 1.2), and he regards arbitrium as belonging not just to will, but rather to the combined activities of will and intellect.[7] Arbitrium, then, belongs to reason, the faculty that is jointly constituted by will and intellect. The activity of arbitrium must therefore engage that faculty as a whole, even though it terminates in a specific act, electio which, when considered by itself, belongs to will rather than to intellect. It is important to notice that Aquinas talks about freedom as a characteristic of arbitrium generally and not of choice in particular. What has freedom in a human being is her capacity to determine her act by her reason, and that capacity seems to be actualized when the last step of determination is achieved, viz. when the act is chosen. A human being has this capacity in virtue of being a rational agent, i.e. in virtue of possessing both intellect and will. The locus of human freedom, for Aquinas, is therefore not will in particular, but reason in general.

Aquinas comments on this relationship between choice and free decision when he claims that “we are said to be characterized by free decision in virtue of the fact that we can accept one thing while refusing another—which is what choosing is” (ST I 83.3. reply). This explains why choice (electio) is a proprium of free decision (liberum arbitrium). It is also clear that Aquinas takes intellect to play an important role in choice, even if he attributes the act of choosing to will in particular. Choice is “made up of something from the cognitive power as well as something from the appetitive power” (ibid.), and it is through deliberation that one “discerns (diiudicatur) what is to be preferred over other things” whereas “what is needed from the appetitive power...is that one appetitively accept (appetendo acceptetur) what is discerned through deliberation” (ibid.). The act of choice thus engages both will and intellect, because without intellect we would not consider and rank our alternatives with a view to our ends, and without will we would not accept any particular action based on those considerations, or commit ourselves to it.

This elucidation of Aquinas' terminology should give us a more accurate picture of his overall view. However, it may not by itself rescue the plausibility of his arguments for the freedom will, since they seem to presuppose that it is will, and not the combination of will and intellect, that enjoys freedom. Instead of offering a fresh and salvaging interpretation of these arguments, the textual evidence just considered seems to reveal a tension within Aquinas’ views. The arguments themselves attempt to predicate freedom of will and not reason, while also leaving us with the discomforting suspicion that will's role is confined to something like the mere timing of acceptance, while intellect is the independent source of that which makes decision distinctively free.[8]

The first argument identified will’s freedom with its ability to determine whether or not to act, but if freedom is now to be seen as an attribute of decision generally, a free determination of whether or not to act would have to be “made up of something from the cognitive power” (ST I 83.3) as well as something from the appetitive power. It would have to be a choice and therefore, like all other choices, directed by intellect.

The second argument for will’s freedom remains problematic for the same reason: Even if the supposedly free act of will consists in determining whether or not to think about an unwanted object, the freedom of that determination requires it to be a choice and thus dependent on intellect’s direction.

The third argument (from ST I-II 10.2 and 13.6) was that any act of will toward a particular good is contingent and therefore not necessary. This sort of answer seems no more satisfying now than it was before. The contingency of intellect's presentation to will of a particular object is only a necessary, and not a sufficient condition for will’s ability to bring about a different presentation. Aquinas’s answer affirms only that the current state of affairs (where intellect presents a particular course of action as best) could have been otherwise, and does not contain the stronger claim that it is in will's power to change it. Nevertheless, the stronger claim somehow hangs in the air, and it is almost as if Aquinas doesn’t recognize the important distinction between it and the weaker one. What we have to find out, then, is whether, and how, Aquinas could incorporate the stronger claim into his theory, thus allowing will to retain freedom to choose among alternatives, as opposed to suffering the fate of having intellect do all the work, in virtue of which a human being can be said to choose freely.

Explaining willful defiance of intellective judgment

The problem of intellective determinism can now be stated as follows: Suppose intellect presents will with only two options, A and B, ranking A as clearly better than B. It would require a special explanation on Aquinas's part if will could somehow command intellect to take a second look at the objects it has presented, in order to find some way of presenting as better the object, B, which has already been presented as worse as a result of the first round of deliberation. Usually, will’s preferences are explained by intellect’s presentation to it of its object, together with its natural inclination to (intellect’s understanding of) the good. But if will is now supposed to stand up against intellect’s proposal and ask for a different one, it seems that will must have some preferences of its own prior to intellect’s presentation of A as clearly better than B. Unless Aquinas’s theory can somehow account for how will could prefer B despite intellect’s presentation of A as better than B, his theory seems unable to give a satisfactory account of the experience of choosing something other than what one had previously judged to be the most rational. Furthermore, free will is often understood to include the ability to reject as well as comply with the dictates of practical rationality.[9] So the view, to which Aquinas appears to be committed, that we don't have that ability, is not only contrary to phenomenological experience, but also contrary to what many, including himself, would consider to a necessary condition for freedom of will.[10] Finally, we have seen that Aquinas regularly speaks of freedom as will’s choosing among contingent alternatives, even though his own arguments for the freedom of will do not seem to succeed in fending of the suspicion that his theory of action entails that once intellect has judged one alternative as best, will is forced to take it or remain inactive.

It would be a significant step in the direction of defending Aquinas’ view against these charges of incoherence and implausibility if it could be shown to have the resources to deal with the alleged phenomenon of willful defiance of intellective judgment. A natural place from which to mount such a defense is the idea that when we take the “worse” option, our choice is in fact motivated by considerations that are just less obvious, even to ourselves, than those we are defying in our choice. Aquinas's theory does easily allow that will should (for whatever reason) hesitate to act on intellect's presentation of A. Suppose intellect then continues to deliberate, and eventually comes up with a different judgment as to which option was best. Our will is energized as this new option is presented, and it grants its consent. It may have seemed to us, experientially, that we were willfully defying reason by taking the new option while in fact, our will only waited indifferently until our intellect offered something that excited our (rational) appetite. A particular will may have its own idiosyncratic inclinations and dispositions that explain why it responds more energetically to some intellective considerations than others, or more energetically under one type of circumstance than another. Such robust idiosyncrasy is more easily seen as a feature of will than intellect, even though intellect itself need not operate in exactly the same way in all individuals. At any rate, when a human being’s practical intellect issues a judgment, it is an open question how strongly or weakly the will of that particular human being is going to respond, because this depends on the contingent constitution of his character and the desires he has at that time.

How could will exercise its idiosyncrasy and still result in characteristically human actions, i.e. actions done for an end? Suppose A is the option of putting on my thick coat before going out for a walk, and B is the option of putting only on my thin jacket instead. I desire the end of taking a walk without freezing, and my intellect figures that putting on the big coat serves that goal better than putting on the thin jacket. Nevertheless, I surprisingly choose to put on my thin jacket instead. How could this happen, given Aquinas’ theory of action? If this was indeed a human act, as opposed to a mere reflex on a par with unwittingly scratching one’s beard,[11] I must have chosen the jacket under a different description from the one under which I would have chosen the coat. My intention cannot have been exactly the same. The coat-option was presented by my intellect as part of the best way to achieve the end of taking a walk without freezing. The jacket-option, on the other hand, was clearly never presented as the best means toward that end. This does not mean, however, that it wasn’t presented by intellect at all, as the thing to do. As we have seen, it isn’t open to Aquinas to allow that something could be chosen without having been presented that way. So instead of being jumped on in pure defiance of intellect’s presentations, the jacket-option must have been presented, as the thing to do, relative to different considerations (of ends and facts) from the ones that formed the basis for intellect’s presentation of the coat-option. The jacket-option could for example have been presented as part of the best way to take a walk without freezing OR having too much trouble getting prepared OR having to feel the annoying itch which accompanies wearing the coat. This more complicated end is clearly different from the one that the coat-option was presented as the best for.

If the jacket option is chosen for this thick sort of end, the example hardly represents an irrational or rebellious choice. On the contrary, it may be explained by positing further considerations by intellect, subsequent to the judgment that the coat-option is best. However, this example does point out a feature which might be relevant for seeing how to handle cases that phenomenologically appear to involve true willful rebellion: The change from the presentation of A as the thing to do, to the presentation of B as the thing to do, does not involve merely further reflection on how to achieve a given end, but a shift from one subsidiary end to another. And this suggests that cases of rebellious choice could be explained on Aquinas's theory by regarding them as resulting from a shift between subsidiary ends.

Aquinas's theory of action clearly requires that every human action, and thus every voluntary action, is done for an end (cf. ST I-II 1.1, 1.2, 6.1, and 6.2). Part of the problem he appears to have in accounting for cases of rebellious choice boils down to the problem of identifying a sense in which such actions are done for an end. The seriousness of this problem depends on how permissive Aquinas is with respect to “ends” that are nothing more than scattered and perhaps sudden thoughts or intellective representations, without any semblance of genuine deliberative process connecting them together. If he can allow these as limiting cases of ends for human actions, he is thereby able to view actions resulting from rebellious choices as limiting cases of human actions themselves. These would be cases in which will exercises its executive power in response to isolated mental representations rather than organized thought.

As we have seen, Aquinas attributes to will the power of determining acting or not-acting, and claims that will is an efficient cause of the operation of all the powers of the soul, including itself. By attributing this brute power to will alone, Aquinas implies that acting or not-acting is ultimately to be explained without reference to cognition. This enables him to account for our felt ability to “go crazy”, or do something that goes against what we take to be our rational thoughts. This would also enable him to hold that we are in control of our action in virtue of the fact that our practical reason, which supplies will with its objects, provides something over and above isolated representations; but a necessary precondition for our control is that we have the power to force ourselves to action, and when we consider that power in isolation we see that it is not to be explained by our possession of intellect. The brute coercive force we have over our mind and limbs is, when considered just as such, frighteningly independent of our intellect, and can therefore be applied without our considered judgment agreeing. Even though will could not move at all in the absence of an object provided by intellect, it can in principle move in response to any practical thought that pops up in our mind, however foolish and ill-considered, and it can be lethargic to our most carefully considered judgments.

A sane person doesn’t normally exercise her brute will-power in a random and arbitrary way, although Aquinas's theory does, on this interpretation, allow that she is able to do so.[12] Still, her brute will-power can force intellect not to think further, in cases where she has somehow gotten in her mind a unique representation of an object (or a subsidiary end) to be willed, and then actually to will that object (or a means toward that end), no matter how irrational it may be given the totality of her desires and beliefs at the time.

Avoiding intellective determinism

On the explanation of willful defiance just offered, will is independent of intellect in the sense that it may be moved by intellective judgments that are superficially inconsistent with judgments that intellect has just issued—superficially because the inconsistency disappears once we realize that these are not competing judgments about how to achieve a single given end, but rather judgments about how to achieve two different ends. The remaining questions are how well this explanation fits into Aquinas’ account and whether this sort of independence of will from intellect removes the threat posed by intellective determinism.

Since every human action is done for an end, acts of willful defiance must also be done for an end, and this end cannot be identical to the end that was presupposed by the defied judgment. But exactly where could such a shift from one subsidiary end to another fit into Aquinas's account? The answer may lie in the role that intention plays in Aquinas's theory. In ST I-II 12.4, Aquinas says that when that which is for an end is considered simply as being for an end, and not considered in itself, the movement of will toward the end and toward what is for the end is one and the same in reality. So the intention on the one hand, and the consent/choice on the other, is one and the same movement of will. Now, clearly, choice/consent does not occur until after some deliberation has taken place. So it is obvious that Aquinas is referring to what happens at the moment of choice/consent. The movement of will at that point is both at once, choice of that which is for the end and intention of the end itself. Strictly speaking, then, intention is always to be described as willing to achieve a specific end E by means of taking a specific action A. Let us call the time when the end E is first desired t₁, and the time at which intellect issues its judgment, that A is to be done for the sake of E, t₂. If Aquinas holds—as he would if the charge of intellective determinism were correct—that the act of will occurring at (or after) t₂ is determined to exactly one outcome in virtue of the way intellect presents A as clearly the best means toward attaining E, he would be committing himself to saying that the intention of the end E is somehow determined by the deliberation which occurred between t₁ and t₂. Is this plausible? Well, since deliberation always presupposes a desire for the end, it would follow that the intention also presupposed that desire. But as I understand Aquinas, the intention is simply this desire when it takes a certain form, namely the form of desiring the end and willing what is for the end at the same time. In his own words: “The will is moved by one and the same movement to the end, as the reason for willing the means, and to the means themselves. But it is another act by which the will is moved to the end absolutely” (ST I-II 8.3). The desire for the end, then, seems to come to the scene twice, first as a desire for the end, regardless of what it will take to bring that end about, and second, as a desire for the end as something to be brought about in a certain specified way, presupposing a particular ordering of something to be done for that end. If intellective determinism is assumed, we have to conclude that when desire comes to the scene for the second time, it is somehow determined to be exactly the same as before, when the will was moved to the end “absolutely”. But since all movements of the will are dependent upon intellect’s understanding of the good, it would indeed be strange if this second movement of the will, to the end “as the reason for willing the means”, could not be affected by the information gained through the deliberative process of finding out how the end may be achieved. A more plausible interpretation is one that takes this second entry of the desire for the end to play an actual role, viz. that of a double-check.

A closer examination of Aquinas’ account of intention (at ST I-II 12.4) reveals that ‘intention’ does indeed refer to two separate acts of will. First, “we can intend the end without having determined the means, which relates to choice” (12.4. ad 3). Presumably, this movement of the will is motivated by the understanding that it would be good to achieve this end, which on Aquinas’ view means in turn that the agent views this end as ordered toward the ultimate good, perhaps via a series of intermediate ends. But second, after the intellect has determined the means by which the end can be attained, that means can be chosen, and insofar as this choice is motivated by a desire for this particular end only, “willing the end and willing the means to the end are one and the same movement...because the end is the reason for willing the means”[13] (12.4. R). More precisely, “insofar as the movement of the will is to the means as ordered to the end, it is called choice, but the movement of the will to the end as acquired by the means is called intention.” (12.4. ad 1) All choices are intentions, but not all intentions are choices. ‘Choice’ and ‘intention’ can be used to denote different aspects of the same movement of will, but in addition, ‘intention’ also sometimes refers to “another act by which the will is moved to the end absolutely” (ST I-II 8.3).

Let us now recall the general problem we are considering. The problem was to see how Aquinas could account for a change in a subsidiary end for the sake of which a choice can be made. Having desired a refreshing walk, and having deliberated to the effect that putting on my coat is the thing to be done for that end, how can I suddenly choose my jacket, for an entirely different and perhaps obscure sort of end? It seems that on Aquinas's account, the role of intention in human action guarantees that no choice is “trapped” within a particular piece of deliberation. No particular presentation by intellect of something to be chosen for a particular end can sufficiently determine what will be chosen. The reason for this is that the deliberation presupposes a desire for an end, at a time when it still remains to be seen whether will has sufficient inclination to act on that desire, given what attempting to achieve the end would actually require.

It seems natural to think of intellect's task, during the time interval from the beginning of deliberation, t₁ to the moment of the resulting judgment, t₂, as of a merely technical nature. An end is presupposed, and intellect's task is to find out the easiest way to achieve that end. But technology requires someone to be in charge of it; to utilize its work. In this case, the utilizing subject must be identical to the subject who desires the end which is presupposed by the deliberation. It would be odd for someone who desires an end, without knowing how to achieve it, to plunge into the relevant calculations without any reservations whatsoever about whether their result, as yet unknown, could turn out to require more sacrifice than the originally desired end is worth. Any rational agent would pause when the calculations are over to ask herself: “Well now, should I take this action for the sake of achieving that end?” It seems plausible to suggest that the role Aquinas wants the second entry of intention to play in human action is analogous to a response to such a question. There is no reason to think that this response is determined to be ‘yes’ just because the original desire was strong enough at t₁ to set in motion a process of deliberation.

It has now become possible to explain how action B could in principle occur at any time when action A has just been judged to be the thing to do.  Instead of intending E and choosing A (in a single act of will), I now choose B for some different end. At t_l my intellect presumably apprehended E as tentatively a good thing to do (or achieve) for some superordinate end, F. At t₂ my intellect apprehends A as the best thing to do for the sake of achieving E. Still, my will is not bound to consent to A at t₂, because I may be unable to form the intention of E-by-A. I will only be able to form that intention if E-by-A seems to me (or “is apprehended by my intellect”) to be a good thing for the sake of achieving the superordinate end F, and that is in no way guaranteed. It is for instance perfectly possible for me to realize, at t₂, that option A is likely to involve side effects, which will be independently destructive of my prospects for attaining F, for the sake of which E was willed in the first place (or, for that matter, of my attaining the superordinate end G, for the sake of which F is desired, and so on). Suppose, for example, that F is my end of leading a rich and rewarding social life, and E is my end of becoming a member of a certain gang of people, an end which I desire for the sake of F. When my deliberation involves purely the question of how I can be accepted as a member of the preferred group, it seems clear that in order to do so, I will have to engage in activities, call them A, involving drug use and unsafe places. However, I may find myself unable to form the intention A-for-the-sake-of-E, because this requires an assessment of A in light of the reasons I had for wanting E in the first place. I may realize that trying to lead a rich and rewarding social life by means of drug use and other dangerous activities would be self-defeating. Consequently, my rejection of A will be sufficiently motivated, even though intellect had just presented A as the best means toward the end I desired.[14]

Alternatively, A may undermine some other end, H, which is otherwise unrelated except that it forms part of my conception of the good and is thus, according to Aquinas, desired by my will as a means to the ultimate good. The most sure and efficient way to get my cat to be silent may be to kill it, but this purely intellective deliberative conclusion is going to conflict with a firm aspect of my conception of the good, even though I still would like the cat to be silent. My will is therefore not going to consent to the cat-killing option, unless I have suddenly gone insane.

Finally, our conceptions of what our superordinate ends consist in need not be stable or self-conscious. Hard decisions are not typically decisions about which of two possible actions one should perform in order to achieve a proximate end. Hard decisions typically involve uncertainty about whether certain more remote ends should be striven for—an uncertainty which often takes the form of wondering whether the end is really worth the trouble that is required in order to attain it. Such uncertainties can easily be seen as resulting in episodes where deliberative judgments are followed by motivated and explicable shifts between subsidiary ends.

We now seem to have a sound explanation of how intellect's presentation of A as clearly better than B for the end E can be followed by a choice of something other than A. The act of will which is called ‘choice’ is always also an intention of an end. Different choices involve different intentions. If A is not chosen, it means that the agent doesn’t intend E-by-means-of-A. The new intention can be of E with a slight change, or it can be of something entirely different. But changing one’s mind so as to choose B rather than A, in spite of intellect's judgment at t₂ that A is clearly better, always involves becoming motivated by something other than E, and that change in motivation can be explained on the basis of the double-checking role of intention in human action.

The formation of a new intention must itself be motivated by something apart from the deliberation from E to A. I cannot intend a subsidiary end without first stepping outside the narrow deliberation which presupposes desire for that end. If my character is firm and stable, and if the circumstances are not particularly novel or challenging, the new motivation may be some modification of the motivation for desiring E in the first place. But if I am in a state of upheaval and deep practical uncertainty, the new motivation may be radically different from the old one. And the blind and brute executive power of my will is in fact relevant to determining what sort of considerations I will entertain, since every single consideration is susceptible to being cut off by that switch. This does not mean that will has a mind of its own, because the choice not to think of something is always motivated by some other thoughts. Accordingly, a person whose practical intellect was always perfectly harmonious and consistent without ever entertaining contradictory thoughts, and whose experiences never put her in a deliberative crisis, could not in practice act willfully. Her will would never be presented with an object which, if taken, would mean choosing contrary to some other judgment of her intellect. Such a person could in principle act willfully, since it is only a contingent fact about her that her intellect is so harmonious and her circumstances so fortunate. But in practice she could not. We should not regard this person as any bit less free for this fact. To the contrary, she is presumably more fully in control of her acts than most of us, since her choices are less affected by accident.

The role of intention in Aquinas’ theory of action enables him to avoid the troublesome charge of intellective determinism, because it guarantees that after intellect has judged a certain option to be best for a given end, will has yet to be motivated to form the intention of that end given what it takes to achieve it. This is will’s escape route from every particular bit of deliberation, and it seems to restore the sense that will typically continues to have alternate possibilities even after intellect has ruled them out in a particular judgment.[15]