The Libertarian Foundations of Scotus's Moral Philosophy

The Thomist 62 (1998): 193-215

Contemporary libertarians typically claim that their conception of freedom is necessary to safeguard our commonsense understanding of moral responsibility, but beyond that claim, little is said about the implications of libertarianism for moral philosophy.  Perhaps philosophers generally do not think it has any other such implications.  Duns Scotus, however, made his libertarianism the cornerstone of his system of ethics.  Unfortunately, commentators have failed to show how his theory of freedom unites various elements of his thought.  They have failed to trace (and consequently, they have failed to defend) the inferences that Scotus drew from his account of freedom.  They have, in short, failed to treat Scotus's moral philosophy as a system at all, and have written as if Scotus had nothing more to offer than disjointed observations about the will and a few other subjects of interest to moral philosophers.1

A single paper can, of course, only begin the task of exhibiting Scotus's moral philosophy as a system based on a libertarian conception of freedom.  I shall not say anything at all in this paper about the implications of Scotus's libertarian view of divine freedom, or about the intricate interrelationships between those implications and the views I shall discuss below.  Instead, I shall concentrate entirely on his libertarian view of human freedom.  After setting out in part 1 Scotus's libertarian account of the will, I shall discuss two of the most important implications Scotus understood his account to have.  First, according to Scotus, the Thomist understanding of the will as intellective appetite is inadequate to provide a libertarian account of freedom.  Scotus therefore rejects that understanding and offers an alternative moral psychology.  In part 2 of the paper I therefore draw attention to the passages in which Scotus offers his reasons for rejecting Aquinas's account in order to show that they arise directly out of the libertarian account of the will I stated in part 1.  I then ask whether Scotus is in fact justified in supposing that Aquinas's conception of will is incompatible with freedom as Scotus understood it.  In parts 3 and 4 of the paper I shall argue that he is, since Aquinas's conception of possibility at best allows him to make room for diachronic alternatives, whereas Scotus insists on synchronic alternatives.

The second implication of Scotus's libertarian understanding of freedom is his distinctive conception of choice and of rationality in action.  In part 5 of the paper I explain this conception and show why Scotus associates it with a libertarian understanding of freedom.

1. Scotus's Libertarianism2

 According to Scotus, the fundamental distinction to be made among active powers has to do, not with their objects, but with the way in which they elicit their acts.3  There are only two possibilities.  First, a power might be determined by its very nature (ex se) in such a way that it cannot but act, so long as it is not impeded by any external object.  Second, a power might not be determined by its very nature.  Such a power can do this act or that; it can even act or not act.  The first sort of power is called a natural power, or simply 'nature', and the second is called a rational power, or 'will'.

In one respect, at least, the two sorts of causes are alike: one cannot sensibly ask why they behave as they do.  For example, if one asks "Why does heat heat?" the only sensible answer is "That's just the sort of thing heat does."  And similarly, if one asks "Why does the will will?" the only sensible answer is "That's just the sort of thing wills do."  In Scotus's terminology, "Heat heats" and "The will wills" are immediate propositions; they are not derived from any more basic propositions that explain or account for their truth.

Of course, an act of will might prompt a slightly different kind of question.  Since the will is not determined to this or that act, one could well ask not merely why the will wills, but why the will wills this as opposed to that.  But even here Scotus insists that the same sort of answer applies:

Hence, the will's acts are contingent.

In order to understand what Scotus means by 'contingent', we must look at Scotus's understanding of modality.  In Modalities in Medieval Philosophy, Simo Knuuttila argues that diachronic and statistical-frequency models of possibility dominated medieval discussions of modality well into the thirteenth century.  According to Knuuttila, it was Scotus who first systematized a theory of modality that involved synchronic alternatives, a theory that is in many respects similar to contemporary possible-worlds semantics.  Scotus's explanation of what he means by ‘contingent' encapsulates this new insistence on understanding modal expressions in terms of synchronic alternatives: "By 'contingent' I do not mean whatever is non-necessary or non-sempiternal, but a thing of which the opposite could have been brought about when that thing itself was brought about."5  Scotus clearly affirms that this sort of contingency is characteristic of our volitions:

With this understanding of possibility in mind, we can characterize more precisely the fundamental difference between rational and natural powers.  A rational power is such that, at the very moment at which it acts, it can act otherwise.  A natural power is such that, at the moment at which it acts, it cannot act otherwise.  Note that this understanding of natural powers does not imply that a natural power always acts in the same way (as it would if Scotus adopted a statistical-frequency model of modality) or that a natural power cannot at one time act otherwise than it acts at some other time (as it would if Scotus adopted a diachronic model of modality).  This point will be of considerable importance when we examine Scotus's reasons for rejecting Aquinas's account of the will.

2.  Why Scotus Rejects Aquinas's Account of Freedom

For according to Scotus, an account according to which the will is intellective appetite does not preserve contingency in this strong sense.  Indeed, such an account makes of the will a merely natural power rather than a rational power.  So if the will is to be free, it must be more than merely intellective appetite.

Scotus's name for that "something more" is affectio iustitiae.7  The affectio iustitiae is the "ultimate specific difference of a free appetite"8; in other words, it is what distinguishes a free or rational appetite from an unfree or natural appetite.  In addition to the affectio iustitiae the will possesses another inclination, the affectio commodi.  Scotus repeatedly insists that if the will possessed only the affectio commodi, apart from the affectio iustitiae, the will would be merely intellective appetite.  Moreover, he claims that intellective appetite as such cannot be free.  For example, consider his discussion in the Ordinatio of the fall of Satan:

He puts the same point more economically in the parallel passage in the Reportatio: "Hence, an intellective [appetite], if it lacked the affectio iusti, would naturally desire what is suited to the intellect, in just the same way that the sensitive appetite desires what is suited to the sense, and it would be no freer than the sensitive appetite."11  Similar discussions can be found at Ordinatio 2, d. 25, nn. 22-23 (W 13:221-223), where intellective appetite is said to act per modum naturae and is identified with the affectio commodi; 2, d. 39, q. 2, n. 5 (W 13:415-416); and 3, d. 26, n. 17 (W 15:340-341).  In all these passages Scotus illustrates his point about intellective appetite by appealing to a comparison with the sensitive appetite: if the will were merely intellective appetite, he says, it would be no freer than the sensitive appetite.  And since the sensitive appetite is a natural power, not a rational power, intellective appetite would also be a natural power.

It is here that the defender of Aquinas's account would surely demur.  Aquinas, after all, goes to no little trouble to differentiate between the sensitive appetite and the intellective appetite, and to show how the intellective appetite is free.  And so the Thomist could well argue that when Scotus puts the will on the same level as the sensitive appetite, he misrepresents what is distinctive about intellective appetite.

If this charge could be made to stick, Scotus's moral philosophy would be in real trouble.  For Scotus's rejection of intellective appetite is, as I am arguing in this paper, a central feature of his system.  Scotus uses his understanding of intellective appetite as one of his principal arguments against eudaimonistic ethics.12  So if Aquinas can establish that intellective appetite is free, one of Scotus's major weapons against eudaimonistic ethics is disarmed.  Furthermore, as I shall show later in this paper, Scotus uses his attack on intellective appetite in order to make room for his own positive moral psychology.  A successful Thomistic defense here would therefore make Scotus's positive views seem both unappealing and unmotivated.

I shall argue that Scotus need not do violence to Aquinas's careful explanation of the difference between intellective and sensitive appetite in order to establish that intellective appetite as such is not free.  A close look at the ways in which Aquinas tries to differentiate the intellective from the sensitive appetite will show that there is nothing in Aquinas's account to guarantee that the will is free in the libertarian sense.  Indeed, much of what Aquinas says on this score suggests that our volitions are determined by antecedent intellectual cognition, which would mean that they are free, if at all, only in a compatibilistic sense.

3.  How Intellective Appetite Differs from Sensitive Appetite13

Bear in mind that appetite is supposed to be a very general feature of Aquinas's view—and of Scotus's too, for that matter.  The medieval universe is teleologically rich.  Everything in it has an end, and corresponding to that end is some sort of inclination.  In some things this inclination functions in the absence of any cognition.  Stones "seek" the center of the earth, and plants draw water and nutrients from the ground, without in any way realizing what they are doing.  But in other beings the appetite for their end is consequent upon some sort of cognition.  When appetite follows upon sense cognition, it is sense appetite; when it follows upon intellectual cognition, it is intellective appetite.  So obviously the most basic way of distinguishing intellective from sense appetite will be to examine the different sorts of cognition upon which they follow.

Sense perception is limited to the concrete particular, whereas intellectual cognition involves the understanding of universals.  This at first looks rather unhelpful, as Aquinas admits in raising an objection to the distinction between sensitive and intellective appetite: "But this distinction has no place in the appetitive power.  After all, since appetite is a movement of the soul towards things, which are singulars, every appetite seems to be for a singular thing."14   What Aquinas means, though, is that intellectual cognition, although directed at particular things, involves a certain apprehension of a universal.15

3a.  Intellective Appetite and Universal Intelligible Features of Objects

We can look at three complementary ways of putting this difference and see that in each case the difference between intellective and sense appetite is not such as to guarantee the freedom of intellective appetite.  First, intellective appetite is aimed at things insofar as they are apprehended as having a certain intelligible feature, namely goodness.  In more modern terms, we can say that it is aimed at things under the description 'good'.  Sense appetite, by contrast, simply takes things (or refuses to take things) as it finds them, without recognizing them as falling under any particular description.  On this understanding of the difference between the two sorts of appetite we must attribute to human beings the possession of some purely formal concept of the good.

Now our question is this: does the fact that human beings can desire things under a description—that is, can desire things as good rather than simply desiring them—guarantee freedom in the libertarian sense?  Clearly not.  Even supposing that Aquinas is right to claim that we have a formal concept of the good, and that we can therefore desire things as falling under that concept, it could still be that our desire for those things is causally determined by our cognition of them.

This point becomes quite clear when we consider two cases, one in which intellective appetite is at work and one in which only sensitive appetite is at work.  I eat a hot fudge sundae; my dog eats his bowl of Alpo.  We will suppose, as Aquinas does, that the dog's sense perception of the Alpo in a given set of circumstances is causally sufficient for the dog's being moved to eat the Alpo.  Now suppose I do not engage my intellect in my pursuit of the sundae; only sense perception is involved.16  By parity of reasoning, this perception is causally sufficient in a given set of circumstances for my eating the sundae.  Somehow, though, when my intellect gets involved and cognizes the sundae, freedom is supposed to enter the picture.  How?

We get different answers from different passages of Aquinas.  In one place he explains it in this way.  Sense perception involves a form that is particular.  The appetite that follows upon it can therefore be directed only toward a particular object of desire.  Intellectual cognition, however, involves a universal.  Since a multiplicity of objects falls under this universal, intellective appetite can be directed to any of a number of different objects.  It therefore has alternative possibilities available to it, which sense appetite cannot have.17

This response does not establish the point at issue.  Let us grant, for the sake of argument, that I cognize the sundae as good, whereas the dog does not cognize the Alpo as good.  That does not even begin to show that I have alternative possibilities and the dog does not.  Obviously the dog's desires will be directed to objects insofar as they have certain features, even though (on our assumption) it will not be directed to objects on the grounds that they have those features.  This is enough to guarantee our dog a variety of objects to choose from: this or that bowl of Alpo, the Alpo or a yummy dog biscuit, and so on.  But no one supposes that the dog is free in the libertarian sense.  So mere multiplicity of objects is not sufficient for freedom.
 Aquinas puts this same argument somewhat differently when he claims that the will tends primarily not to the object that is desired, but to the reason for its desirability.18  When I will that hot fudge sundae under the description 'good', what I primarily will is not the sundae itself but goodness, which I find to be instantiated in the sundae.  And since goodness is instantiated elsewhere as well, my will is not necessitated to will the sundae.  It is necessitated only to the good in general.  In other words, if I will anything at all, the primary object of my willing will be the goodness of that thing, and I will will it as a good; but there is no particular good thing such that I necessarily will that thing.

As Aquinas explains it, the sensitive appetite is not necessitated to any thing before that thing is apprehended under the formality (sub ratione) of pleasant or useful.  But once the pleasant thing has been apprehended, the sensitive appetite is drawn to it necessarily.  A brute animal, on seeing something pleasant, cannot fail to desire it.  But the pleasantness of the object, as opposed to the object itself, is something that does not enter into the animal's mind.  That is, the animal's attention is completely captured by the particular object; it is not aware of any general feature in virtue of which it is willing the object.  By contrast, a being with reason perceives not only the desirable object but also the formality under which it is desirable.  The will is necessitated only to that formality; it is not necessitated to any particular thing that is apprehended as good.

Once again, it is a mistake to suppose that this distinction makes any relevant difference to the freedom of the agent.  Let us take a particular sense appetite—the appetite for food—as a comparison.  Under certain circumstances, a dog will eat his bowl of Alpo, and under other circumstances he will not.  Sometimes he will go for a doggie biscuit, and at other times he will not.  We do not (or at least the medievals did not) take this as evidence that the dog is free.  We could perhaps say, if the fancy took us, that the dog is necessitated to food in general but not to any particular food, but we would hardly be tempted to suppose that this implied anything about canine libertarian freedom.  In fact, since we are committed to the view that the sense appetite operates deterministically, we would have to say that in any given set of circumstances there are causally sufficient antecedent conditions for the dog's doing whatever he does.

Does it make any difference that we human beings, unlike our dog, know what we are doing?  I do not see why it must.  It is perfectly conceivable that we are in exactly the same position as the dog, except for the fact that unlike him we can say to ourselves, "Ah, that's a bit of food right there," as the laws of nature take their invariable course.  Mere awareness of the fact that the hot fudge sundae falls under a more general class (food, source of pleasure, good thing) does not imply libertarian freedom.  Libertarians would not be such a disheartened bunch if their pet position could be established as easily as all that.

3b.  Intellective Appetite and Conceptions of the Good Life

Aquinas of course resists the thought that our eating of the hot fudge sundae is on a par with a dog's eating of his bowl of dog food.  One reason for this resistance is that he tends to think of the dog as overwhelmed by the canine equivalent of "My, doesn't that look yummy," while he pictures us seated in our recliner thinking over the question "What sort of life shall I lead?" when the butler brings in a hot fudge sundae.  Then, in the light of a general plan for our lives, we can either dismiss Jeeves with a haughty gesture or dig into the sundae while calling imperiously for more whipped cream and an extra cherry.

It is true, I suppose, that dogs do not form conceptions of the good life, and that some of us human beings do.  But the question recurs: Does this difference in itself show that we are free in a way that dogs are not?  Aquinas certainly thinks so, since he uses our ability to form a general conception of the good life as a second way of distinguishing intellective from sensitive appetite and (as he thinks) of showing how the intellective appetite is free and the sensitive appetite is not.  Here is the argument: a human being is necessitated to will happiness, where happiness is understood as the purely formal concept of a complete and perfect human life.19  Consequently, if the intellect conceives of the life of, say, aesthetic experience as the complete and perfect human life, then the will necessarily wills such a life.  But the intellect does not necessarily conceive of this or of any other sort of life as embodying happiness, and so the will is free to the extent that the intellect is not necessitated to any particular concrete conception of happiness.20

This argument, I am afraid, leaves the libertarian unsatisfied, and for two different reasons.  First, the argument again rests on a confusion.  It shows only that there is no object (this time the "object" in question is a plan of life rather than a particular thing) such that the will is necessitated to will that object.  But the libertarian wants a stronger claim.  The libertarian wants to say that even in a completely specified set of circumstances it is sometimes the case that the will is not necessitated to an object.  Not only does Aquinas's argument not establish this stronger claim, it actually presupposes the very opposite.  For it presupposes that, once the intellect has presented an object as the concrete instantiation of the formal concept of happiness, the will cannot help but will it.

Second, the argument simply pushes the problem back a step.  As Aquinas argues, actions are determined by inclinations, and inclinations are determined by judgments.  So we must ask whether the intellect itself is free with respect to its judgment about which of the available conceptions of happiness it will adopt.  The answer is "Of course not."  The intellect, as everyone in this debate would admit, operates deterministically.  To put it in more modern terms, in a given set of circumstances, we have no control over how things look to us.  If in a given set of circumstances my intellect presents the life of aesthetic experience to me as the perfect and complete human life, it is not physically possible for it in that set of circumstances to present any other life to me as embodying happiness.  Now Scotus would here insist that you cannot build freedom out of a deterministic agent; you cannot turn the intellect into a sort of super-will.  If in fact we are free in the libertarian sense, it will not be because we control how things appear to us, but because, however things appear to us, we control how we act on that information.21

3c.  Intellective Appetite and Relating Ends to Means

We can therefore deal very quickly with the third contrast Aquinas draws between intellective and sensitive appetite.  The intellect, unlike the senses, can relate ends to means.  Thus, having willed one's plan of life, one can examine further objects and actions in the light of that plan and determine which of them are suitable means to, or constituents of, happiness as one conceives it.  Now unless there is only one available means to realizing one's end, one has a variety of alternatives open to one, and so one can freely choose among them.  For example, having determined that the aesthetic life is the best, I can then decide whether to take organ lessons, join a choir, become an art buyer, or sign on as a newspaper theatre critic.  My will is not necessitated to any of these, since none of them is the sole possible means of attaining happiness as I conceive it.22

The same response is in order here.  First, mere multiplicity of objects is not sufficient for freedom.  Second, this account of the contrast between sensitive and intellective appetite still leaves open the possibility that the will is necessitated to an object, not tout court, but given the results of deliberation.  That is, once the intellect has deliberated and seized on one of the available options as the best, the will cannot fail to will that object.  And since the intellect operates deterministically, there is no room for freedom here.

In fact, given what we have seen already, it should be clear that Aquinas does in fact think of the will as necessitated to the option that the intellect presents as best.  Recall that he claims that in any volition the primary object is not the concrete particular but its goodness.  So if from among a number of options the intellect presents one as the best, the will must choose that one, since it is in that particular that the will's universal object is most fully realized.  It could fail to choose that particular only if it could will in view of something other than goodness—which Aquinas of course denies.

4.  Some Complicating Factors

There is one other feature of Aquinas's view that is worth mentioning.  According to Aquinas, the will can always turn the intellect away from considering a given object.  If the will is free to avert the intellect, then obviously the will is free not to will that object.23

The libertarian might wonder about this averting of the intellect.  Averting the intellect is, obviously, an act.  So one can ask about this act, as about any other, why it takes place.  Now if the will is indeed intellective appetite, it would seem to follow that the will can only avert the intellect if the intellect judges that this course of action is best.  Obviously this just moves the problem back a level.  If the averting of the intellect is intellectually determined, then the will's not willing the object is intellectually determined, although at one remove, so to speak.  But perhaps in this case Aquinas could say that the will can act on its own steam.  The will can simply avert the intellect at its discretion.  If he can say that here, however, why could he not say it anywhere else?  That is, why must he restrict this libertarian freedom to one sort of volition?  One gets the picture of a will that can only avoid being determined by the intellect if it asserts itself first and prevents the intellect from doing its job.

These arguments do not show that Aquinas was really a compatibilist.  What they show, I think, is that his understanding of the will as intellective appetite does not entitle him to regard the will as free in the libertarian sense.  Its association with the intellect cannot make the will free; it can at best make the determination of the will more elaborate and interesting.  So if Aquinas wants to be a libertarian, he must do so by postulating something in the will itself that allows for such freedom, rather than trying to build freedom out of the intellect.

But does Aquinas want to be a libertarian?  Unlike Scotus, he takes no pains to safeguard a notion of freedom that would satisfy the libertarian.  And it is arguable that the attempt to categorize him as a libertarian is hopelessly anachronistic because full-fledged libertarianism involves modal concepts that differ significantly from Aquinas's.  In particular, libertarianism involves a conception of possibility as involving synchronic alternatives.  When a libertarian claims that it was possible for an agent to act otherwise, she typically means that it was possible for the agent to act otherwise at that very time and in those very circumstances.  As we have already seen, there is nothing in Aquinas's discussion of intellective appetite that would permit us to attribute such a conception of possibility to him.  In fact, his arguments that intellective appetite is free suggest that he thinks of possibility diachronically rather than synchronically.  The alternative possibilities to which he appeals are not actually available to the agent at the very moment of choice, but only prospectively.  Since there is nothing in the nature of the various human potencies, independently of the actual circumstances of deliberation and choice, that constrains the will always to choose one thing rather than another, the will's choices can be regarded as contingent or free.  Nonetheless, the will's choice could still be necessary given the various causal factors at work in a particular situation of choice.  As Knuuttila says in discussing Aquinas's view,

So in order to show that the will's activity is free, Aquinas need only argue that the will is the sort of cause that has alternative possibilities open to it prospectively, that it is not of its very nature necessitated to every object that it in fact wills.  And that is exactly how he argues in the passages I have already discussed.  What he does not show, and given his own conception of possibility does not need to show, is that those alternatives are open to it synchronically.

 So it seems that Aquinas could not have been either a libertarian or a soft determinist, since in order to adopt either of these positions one must have a conception of possibility as involving synchronic alternatives.  Scotus, by contrast, has such a conception, and so there is nothing anachronistic about calling him a libertarian.  Like most libertarians, Scotus regards any non-libertarian account of freedom as wrong-headed, quite apart from the details of the account.  He is therefore satisfied with pointing out that Aquinas's account of the will as intellective appetite is not a libertarian account; whether Aquinas's account is soft determinist or not is of no importance for his purposes.

5.  Libertarianism and Moral Theory

Nevertheless, when Scotus says that the will is not merely intellective appetite, he does not mean to imply that the possession of intellect is irrelevant to the exercise of our freedom.  If you try to imagine a dog that possessed just the cognitive faculties that dogs generally possess, but unlike other dogs possessed libertarian freedom as well, this becomes quite clear.  Any exercise of this freedom—if indeed it is conceivable that he could exercise it—would have to be totally arbitrary.  He could not choose one thing over another in view of a plan, or because of a reason, or as a means to something else, or on the grounds that it possessed a certain desirable feature.  He could do nothing but choose, pointlessly and inexplicably.

This understanding of the intellect's contribution to choice may seem rather minimalist; it is just this sort of conception that gives rise to the objection that libertarianism makes choice inexplicable.  To a certain extent libertarians themselves are responsible for the prevalence of this objection.  Since the crux of the dispute over freedom is the role of causal determination, libertarians are tempted to concentrate their energies on explaining why they think freedom is incompatible with causal determinism.  They thus devote all their attention to closing off the most obvious possible explanation for choice, and so it seems that the inexplicability of choice is not just an embarrassing consequence of their view but the whole motivation for holding the view in the first place.

If the only point of libertarianism were to secure a place in our ontology for actions that are not determined by antecedent causal conditions, this charge would surely have some merit.  In fact, however, libertarians are looking to get more out of their theory than just that.  For Scotus, free actions are valuable because in them we express our likeness to the Creator, whose "superabundant sufficiency" is mirrored, though imperfectly, in our own freedom.25  The paradigmatic instance of freedom is God's creating the universe.  As every medieval Christian philosopher agreed, there was nothing about this universe that constrained God to create it.  And as Scotus takes pains to emphasize, there can be no finally adequate explanation of why God willed to create as he did.

Freedom thus conceived is a pure perfection, and like every other pure perfection it can be predicated univocally of God and creatures.  So for Scotus, free creatures (that is, creatures who have wills) are free in exactly the same sense in which God is free.  It is their likeness to God's unconditioned creative activity that makes free actions valuable and noble on Scotus's view.  And for those free actions, as for God's, there can be no fully adequate explanation.

It is important here to point out two implications that the libertarian does not wish to draw.  First, a free action is not an unintelligible, arbitrary, or random action.  The fact that I freely chose to write this paper does not imply that there were no reasons why I chose to write it.  There were any number of reasons.  The libertarian simply wishes to insist that those reasons can provide only a partial explanation for my choice, since it was possible for me, even in exactly the same circumstances, with exactly the same reasons, to choose differently.  Second, notice also that such a view does not imply that just anything can count as a reason.  We do not find ourselves in a position in which we can regard just any old thing as valuable.  Being creatures of a determinate sort, we will (so long as we are not pathological) draw our reasons from a fairly limited pool of possibilities.

Scotus endorses this conception of freedom in a number of places.  The most striking, perhaps, occurs where he is asking whether the will necessarily enjoys (frui) the end when that end is apprehended by the intellect.26  He considers an argument for the affirmative:

Against this Scotus argues thus: Here again we see the contrast between the sensitive appetite and the will.  The sensitive appetite has no power over what will be actually suitable to it; therefore, whatever is aptitudinally suitable to it is also actually suitable to it.  The same stricture would apply to a purely intellective appetite, as Scotus makes clear at Reportatio 2, d. 6, q. 2, n. 9: "An intellective [appetite]...would naturally desire what is suitable to the intellect, just as the sensitive appetite desires what is suitable to the sense, and it would be no more free than the sensitive appetite."30  The will, by contrast, has power over whether what is aptitudinally suitable will also be actually suitable.

Commentators unsympathetic to libertarianism often charge that on this sort of view the will is wrenched apart from the rest of human nature and left dangling in an abyss of untrammeled choice.  For example, Patrick Lee compares this view unfavorably with that of Aquinas, who unlike Scotus "keeps the will integrated with the rest of man."31  Joseph Incandela says that on Scotus's view "the will is truly isolated from and independent of prior attachments or commitments,"32 and he implies that Iris Murdoch's complaint about the "giddy empty will" of modern moral philosophy applies to Scotus's view.33

In fact, however, the will's power to set its own ends does not imply that its choice is completely unfettered.  As I said before, not just anything can count as a reason; or, in Scotus's terminology, not just anything is aptitudinally suitable to the will.  The class of  aptitudinally suitable things is delimited, not by the will's choice, but by the nature of the agent and the nature of the object.  The will, however, does determine which of these aptitudinally suitable things will count as actually suitable.  Libertarians are apt to find such a claim perfectly obvious; non-libertarians are apt to find it perfectly obviously false.

This libertarian conception of freedom leads Scotus to hold a distinctive view of what it means for action to be reasonable.  "Reasonable" action might be contrasted with self-frustrating action, or with chaotic action, or with arbitrary action, or with action undertaken on the basis of insufficient deliberation or incomplete information.  In any of these senses, Scotus can agree that there is such a thing as reasonable action, and that we have some interest in acting reasonably.  But there is a stronger and more morally loaded sense of 'reasonable' that we expect from medieval philosophers working in the Aristotelian tradition.  In this sense of the word, the reasonable action is the one that has reason on its side, in something like the following sense.  While one might have reasons to commit adultery, and reasons to refrain from committing adultery, the conflict between these two sets of reasons could not be resolved within reason by anything other than a judgment in favor of refraining from adultery.  In this sense of 'reasonable', I think Scotus would have to deny that there is any such thing as reasonable action.  Where there are competing considerations in favor of incompatible courses of action, the conflict cannot be resolved by reason.  It can be resolved only by an act of will by which I decide to regard certain considerations as having a claim on me.

Bear in mind that the paradigmatic case of freedom is God's decision to create this world.  Was that a reasonable decision?  The concept of 'reasonable' in the morally loaded sense does not even seem to apply here.  It was not as if the divine intellect pointed out that creation was the reasonable thing to do, and the divine will fell in line.  Now it would certainly have been unreasonable for God to create this present world if his aim had been to produce a world full of unicorns, say.  But apart from questions of divine self-frustration, does it even make sense to ask whether God's free volition to create was a reasonable act on his part?

In just the same way, I think, 'reasonable' falls out of moral theory in its distinctive use as a term of all-things-considered commendation.  To recur to the case of adultery, Scotus must deny that there is any morally significant sense in which it is unreasonable to commit adultery.  It could lead to bad consequences, certainly: disease, illegitimate children, eternal damnation.  And of course it would be morally wrong.  On the other hand, it could be a lot of fun, and one could very well decide to do it for that reason.  Reason points out that it would be fun; reason points out that it would be dangerous.  The conflict is not resolved by reason, and so neither committing adultery nor refraining from it could properly be called unreasonable.  If you say that reason tells you that such an action would not be in accordance with the human good, Scotus would simply say that you are mistaken.  Reason tells you nothing of the sort.  For the human good is a loving union with the Triune God, and it is perfectly possible to have such a union even if one commits adultery.  And Scotus does not simply mean that adulterers can repent and be forgiven.  He means (indeed, he explicitly says) that God could easily have set up the moral law in such a way that adultery was not forbidden, and his doing so would in no way have diverted us from the attainment of our ultimate end.34

To give another example that Scotus throws out in perfect seriousness: Since God created us in the first place, he would have been well within his rights to impose upon us obligations extending to the whole of our conduct.  But he did not do so.  Instead he confined himself to imposing the Ten Commandments.  So long as we do not violate those commandments, we are free to do as we please.  Scotus immediately proceeds to derive from this the conclusion that one is free to sell oneself into slavery.35  This sounds, and is meant to sound, quite extreme.36  Such a use of one's freedom would be stupid, Scotus admits,37 but it is no more a violation of the moral law than is marriage, which similarly involves giving up certain rights to one's own body.38

6.  Conclusion

There is much to be gained by thinking of Scotus as attempting to work out the implications for moral theory of a libertarian understanding of human freedom.  We can understand more fully why Scotus rejected Aquinas's account of the will as merely intellective appetite, and consequently why he felt the need to posit an additional inclination, the affectio iustitiae, in virtue of which the will could be free in the libertarian sense.  We can also understand how Scotus leaves room for the will to choose its own ends, without thereby falling into that caricature of libertarianism according to which the will's choice has no anchor in human nature or the moral order.  And finally, we can understand why Scotus adopts his distinctive and un-Aristotelian conception of what it means for an action to be reasonable.


1.  Not only have commentators sometimes written as if they believed this, they have occasionally stated it outright.  In the recent Scotus number of the American Catholic Philosophical Quarterly (67:1993), for example, Mary Elizabeth Ingham says flatly that "It is well known that Scotus presents nowhere in his writings a full-blown ethical theory" (p. 128).  Gilson says of his book on Scotus, "On n'y trouvera pas non plus un 'systéme' de Duns seule raison est que nous ne l'avons pas trouvé nous même" (Jean Duns Scot: Introduction a ses positions fondamentales.  Paris: Libraire Philosophique J. Vrin, 1952, p. 7).

Other interpreters, most notably Allan B. Wolter, have insisted strongly on the systematic character of Scotus's moral thought; Wolter indeed identifies Scotus's account of freedom as the key to the system, just as I propose in this paper.  But as Hannes Möhle rightly notes, a good deal of Wolter's work has involved editing, translating, and commenting on discrete passages rather than substantiating his claim that those texts present a systematic moral theory.  Speaking in particular of Wolter's collection Duns Scotus on the Will and Morality, he writes, "Die ausführliche Einleitung, die Wolter den Texten voranschickt, ist wesentlich davon geprägt, die von ihm edierten Texte im einzelnen einzuleiten und zu kommentieren.  Die als wichtig bezeichnete innere Systematik der scotischen Lehre als Ganzes kommt also nur bedingt in den Blick und ist deshalb weiterhin als ein Desiderat der Scotusforschung zu begreifen" (Ethik als scientia practica nach Johannes Duns Scotus: Eine philosophische Grundlegung, in Beiträge zur Geschichte der Philosophie und Theologie des Mittelalters: Texte und Untersuchungen, Neue Folge, volume 44 [Münster: Aschendorff, 1995]).  Möhle himself argues at length for a systematic understanding of Scotus's moral thought, but he takes practical science rather than freedom to be the central notion.  Back.

2.  Not everyone agrees that Scotus was a libertarian.  Douglas C. Langston, in God's Willing Knowledge: The Influence of Scotus' Analysis of Omniscience (University Park, PA: Pennsylvania State University Press, 1986), argues that Scotus was in fact self-consciously a compatibilist.  I cannot in this paper respond in detail to Langston's arguments; the interested reader should consult Wolter's review of Langston in Theological Studies 48 (1987): 182-185, as well as Simo Knuuttila, Modalities in Medieval Philosophy (London: Routledge, 1993), pp. 144-145.  For a helpful survey of the dispute over Scotus's libertarianism, along with interesting suggestions about how to adjudicate between competing interpretations of Scotus, see Joseph M. Incandela, "Duns Scotus and the Experience of Human Freedom," The Thomist 56 (1992): 229-256.  Despite the many merits of his essay, Incandela is obviously not at all sympathetic to libertarianism, and his picture of Scotus's theory is an unkind caricature; in section 5 of this paper I explicate what I take to be the real significance of Scotus's libertarianism.

In any case, I shall in this paper assume the standard interpretation of Scotus as a libertarian.  This interpretation is confirmed by the reasons Scotus offers for rejecting Aquinas's account of the will as intellective appetite, since they are precisely the sort of reasons that only a libertarian would find persuasive.  Back.

3.  Quaestiones subtilissimae super libros Metaphysicorum Aristotelis 9, q. 15, n. 4 (W 4: 797b).  References to the Wadding edition (Lyons, 1639) are indicated with a 'W' and references to the Vatican critical edition with a 'V'.  The translations of Scotus throughout the paper are my own.  Latin texts are reproduced exactly as they appear in the Vatican critical edition.  Wherever possible, I have edited the Wadding edition on the basis of manuscripts as well as Wolter's edition in Duns Scotus on the Will and Morality (Washington, D.C.: The Catholic University of America Press, 1986).  I have indicated those sources in the notes.  Back.

4.  Ibid., n. 5 (Wolter, p. 152).  ". . . vult Aristoteles quod contingit opinari propter quid, scilicet per immediata, et quia, per mediata; ita in proposito, voluntas vult A.  Si non est causa inter extrema, habetur propositum.  Si est causa, puta voluntas vult B, procedetur ulterius.  Alicubi stabitur, ubi quare voluntas illud volet nulla est alia causa nisi quia est voluntas."  Back.

5.  Ordinatio 1, d. 2, part 1, q. 1-2, n. 86 (V 2:178).  [N]on voco hic contingens quodcumque non-necessarium vel non-sempiternum, sed cuius oppositum posset fieri quando illud fit.  Back.

6.  Lectura 1, d. 39, q. 1-5, nn. 50-51 (V 17:495).  Haec autem possibilitas logica non est secundum quod voluntas habet actus successive, sed in eodem instanti: nam in eodem instanti in quo voluntas habet unum actum volendi, in eodem et pro eodem potest habere oppositum actum volendi....Et huic possibilitati logicae correspondet potentia realis, nam omnis causa praeintelligitur suo effectui, — et ita voluntas in illo instanti in quo elicit actum volendi, praecedit natura volitionem suam et libere se habet ad eam; unde in illo instanti in quo elicit volitionem, contingenter se habet ad volendum et contingentem habet habitudinem ad nolendum; non quia prius habuit habitudinem contingentem ad volendum, quia tunc non fuit causa, sed nunc —  quando est causa eliciens actum volendi — contingentem habet habitudinem ad actum, ita quod 'volens in a, potest nolle in a.'  Back.

7.  Or, in the Reportatio parisiensia, "affectio iusti."  Back.

8.  Reportatio parisiensia 2, d. 6, q. 2, n. 9 (W 11.1:289a).  ". . . affectio iusti est ultima differentia specifica appetitus liberi."  Back.

9.  Commoda are whatever things the affectio commodi wills.  More precisely, 'commodum' is the description under which the affectio commodi wills whatever it wills.  I leave the word untranslated so as not to beg any questions about just what these commoda are, since this is a matter of dispute among interpreters of Scotus.  See John Boler, "Transcending the Natural: Duns Scotus on the Two Affections of the Will," American Catholic Philosophical Quarterly 67 (1993): 109-126, and Thomas Williams, "How Scotus Separates Morality from Happiness, American Catholic Philosophical Quarterly 69 (1995): 425-445.  Back.

10.  Ordinatio 2, d. 6, q. 2, n. 8 (W 6.1:539-540, Wolter 468, Codex P 162va, Codex Q 121va-b).  "Si enim intelligeretur—secundum illam fictionem Anselmi De casu diaboli—quod esset angelus habens affectionem commodi, et non iustitiae (hoc est, habens appetitum intellectivum mere ut appetitum talem et non ut liberum), talis angelus non posset non velle commoda, nec etiam non summe velle talia; nec imputaretur sibi ad peccatum, quia iste appetitus se haberet ad suam cognitivam sicut modo appetitus visivus ad visum, in necessario consequendo ostensionem istius cognitivae et inclinationem ad optimum ostensum a tali potentia, quia non haberet unde se refraenaret."  Back.

11.  Reportatio 2, d. 6, q. 2, n. 9 (W 11.1:289a).  "Unde intellectivus, si careret affectione iusti, ita naturaliter appeteret conveniens intellectui, sicut appetitus sensitivus conveniens sensui, nec esset magis liber quam appetitus sensitivus."  Back.

12.  On this point see the articles cited in note 9.  Back.

13.  My account of intellective appetite in Aquinas owes a great deal to David Gallagher, "Thomas Aquinas on the Will as Rational Appetite," Journal of the History of Philosophy 29 (1991): 559-584.  Back.

14.  1a.80.2, obj. 2: Praeterea, cognitio intellectiva est universalium, et secundum hoc distinguitur a sensitiva, quae est singularium.  Sed ista distinctio non habet locum ex parte appetitivae: cum enim appetitus sit motus ab anima ad res, quae sunt singulares, omnis appetitus videtur esse rei singularis.  Non ergo appetitus intellectivus debet distingui a sensitivo.  Back.

15.  1a.80.2, ad 2: Ad secundum dicendum quod appetitus intellectivus, etsi feratur in res quae sunt extra animam singulares, fertur tamen in eas secundum aliquam rationem universalem; sicut cum appetit aliquid quia est bonum.  Back.

16.  In fact Aquinas denies that this is ever the case.  Although a human being can of course act in the way that a sense desire prompts him to act, he can do so only if his will consents.  See ST 1a.81.3 and 1a2ae.77.1 and 2.  I mean to introduce this as a thought-experiment, not as representing Aquinas's own view.  Back.

17.  For this way of distinguishing the intellective from the sensitive appetite, see in particular De malo 6.1: "Forma intellecta est universalis sub qua multa possunt comprehendi.  Unde cum actus sint in singularibus, in quibus nullum est quod adequet potentiam universalis, remanet inclinatio voluntatis indeterminate se habens ad multa."  See also ST 1a2ae.13.2.  Back.

18.  There is an extended discussion of this way of distinguishing the intellective from the sensitive appetite at De veritate 25.1.  Back.

19.  ST 1a.5.1; 1a2ae.1.5, 1.7, 5.8.  Back.

20.  ST 1a.82.2; 1a2ae.1.1, 2; 14.  Back.

21.  David M. Gallagher, "Free Choice and Free Judgment in Thomas Aquinas," Archiv für Geschichte der Philosophie 76 (1994): 247-277, argues that Aquinas does leave room for just this sort of freedom.  It would require a whole article to deal adequately with Gallagher's intriguing exposition, so unfortunately I cannot do so here.  I will, however, make two points relevant to our present purposes.  First, it seems to me that the view Gallagher attributes to Aquinas in this paper does not cohere well with his arguments in "Will as Rational Appetite."  In fact, in "Free Choice and Free Judgment" the will turns out to be something rather different from intellective appetite.  If the earlier paper is correct about Aquinas's view, the criticisms of that view that I have offered on Scotus's behalf would be cogent.  But if the later paper is correct, those criticisms would appear to leave Aquinas's view unscathed, and the dispute between Aquinas and Scotus would have to be fundamentally recharacterized.  In this paper I have directed Scotus's criticisms against what I take to be a fairly standard interpretation of Aquinas's understanding of the will, one that is presented in "Rational Appetite."

Second, even in "Free Choice and Free Judgment" it is not altogether clear that Aquinas's understanding of the will would count as a libertarian one, since it is not clear whether the alternative possibilities Gallagher allows for are to be understood synchronically or diachronically.  Back.

22.  ST 1a.18.3, 1a2ae.6.2; De veritate 22.4.  Back.

23.  See, for example, ST 1a2ae.10.2 and De malo 6.1 ad 15.  Back.

24.  Knuuttila, p. 133.  I have omitted his references.  Back.

25.  See QQ in Meta. 9, q. 15, n. 5 (W 4:798a).  Back.

26.  Ordinatio 1, d. 1, part 2, q. 2.  For similar passages see J.R. Cresswell, "Duns Scotus on the Will," Franciscan Studies 13 (1953): 147-158, especially pp. 154-156.  Back.

27.  Ibid., n. 77 (V 2:59).  Avicenna VIII Metaphysicae: «Delectatio est coniunctio convenientis cum convenienti»; finis necessario convenit voluntati; ergo ex coniunctione eius cum voluntate est delectatio, ergo fruitio.  Back.

28.  The expression "natural or sensitive appetite" might need explaining, since Scotus's usage differs from Aquinas's.  Aquinas typically uses "natural appetite" to designate an inclination that does not follow upon cognition of any sort, and so it is to be distinguished from sensitive appetite, which follows upon sensitive cognition.  For Scotus, however, "natural appetite" can be used to designate any appetite that operates deterministically.  In other words, it includes any appetite that is a "natural" rather than a "rational" power in the sense explained in part 1 of this paper.  The expression "natural appetite" therefore encompasses natural appetite in Aquinas's sense, sensitive appetite, and even intellective appetite.  The only rational appetite is the will.  Back.

29.  Ibid., n. 56 (V 2:106).  [D]ico quod aliquid est aptitudinaliter conveniens, vel actualiter conveniens.  Conveniens aptitudinaliter est quod convenit alicui ex se et quantum est ex natura rei, et tale convenit actualiter omni ei in cuius potestate non est quod ei actualiter aliquid conveniat vel disconveniat; et ideo quidquid convenit alicui naturaliter vel aptitudinaliter, appetitu naturali vel appetitu sensitivo, convenit etiam actualiter.  Sed in potestate voluntatis est ut ei aliquid actualiter conveniat vel non conveniat; nihil enim convenit sibi nisi quod actu placet.  Propter hoc nego minorem, cum dicitur ‘finis necessario convenit voluntati'; hoc enim non est verum de convenientia actuali, sed aptitudinali.  Back.

30.  W 11.1:289a.  "Unde intellectivus...ita naturaliter appeteret conveniens intellectui, sicut appetitus sensitivus conveniens sensui, nec esset magis liber quam appetitus sensitivus."  Scotus uses the subjunctive because he does not in fact believe that any merely intellective appetite exists.  The affectio commodi, which is intellective appetite, is always associated with the affectio iustitiae, in virtue of which the will is free.  Back.

31.  "Aquinas and Scotus on Liberty and Natural Law," Proceedings of the American Catholic Philosophical Association 56 (1982): 76.  Back.

32.  "Experience of Human Freedom," p. 233.  Back.

33.  It is no accident that both Lee and Incandela wish to defend Aquinas's account of freedom as superior to Scotus's and that they do so on the basis of exactly the sort of criticisms that non-libertarians typically raise against a libertarian conception of freedom.  I take this fact to confirm my thesis that Scotus is doing no injustice to Aquinas by rejecting his account precisely on libertarian grounds.  Back.

34.  Ordinatio 3, d. 37, q. un., n. 5 (W 7.2:898).  Back.

35.  Ordinatio 4, d. 26, q. un., n. 10 (W 9:583).  Back.

36.  While I grant that Scotus did not have quite the attitude toward slavery that we have, he certainly had a profound distate for it.  See Wolter, pp. 114-123.  Back.

37.  Ordinatio 4, d. 36, q. 1, n. 2 (W 9:755).  "Talis subiectio esset fatua."  In this passage Scotus is speaking specifically of "that vile servitude" in which the master can sell his slave like cattle.  Back.

38.  Ordinatio 4, d. 26, q. un., n. 10 (W 9:583).  In the context of this passage the comparison between marriage and slavery is not as striking as it seems here.  Scotus is considering the argument that God would have to give explicit approval of marriage, because marriage involves giving over one's body into someone else's control.  Since by right of creation every body belongs to God, God would have to approve of any such transfer of dominion.  In response to this, Scotus argues that it is licit for someone to sell himself into slavery even though Scripture gives no special divine approval for such an action.  Now selling oneself into slavery involves a transfer of dominion over one's body just as marriage does.  So if it is licit to sell oneself into slavery even though there is no special divine approval for doing so, no special divine approval is required for entering into marriage.  Back.