AQUINAS's ACCOUNT OF FREEDOM:

INTELLECT AND WILL

1. Introduction
It is difficult to develop a comprehensive and satisfactory account of
Aquinas's views of the nature of human freedom.
For one thing, contemporary discussions of free will tend to belong
to an older, non-Thomistic tradition of thought about the topic. In this
tradition, human freedom is a property of just one component of human
mental faculties, namely, the will; and freedom consists in an agent's
ability to will autonomously in general and independently of the intellect
in particular. The influence of this tradition persists in contemporary dis-
cussion, both for libertarians and for their opponents, with the result that
Aquinas's account tends to be interpreted by its lights. Consequently, the
lineaments of his theory are obscured. For Aquinas, freedom with regard
to willing is a property primarily of a human being, not of some particu-
lar component of a human being. Furthermore, the will is not independent
of the intellect. On the contrary, the dynamic interactions of intellect and
will yield freedom as an emergent property or a systems-level feature.
Another reason why interpreting Aquinas's account is difficult is that
he gives a complicated analysis of the several acts of will he takes to be
associated with a free bodily action. Scholars sometimes pick out a subset
of these acts or even just one of them as if for Aquinas freedom of the will
were lodged in that sort of act of will alone. So, for example, it is
sometimes said that Aquinas has a particularly full treatment of free will
in De malo q.6 because in that text he discusses at length liberum arbitrium.1
But there is something anachronistic about trying to identify liberum
arbitrium with free will in our sense;2 volitions characterized by liberum
arbitrium are associated for Aquinas with only one sort of voluntary act,
namely, the sort he calls electio. De malo q.6 is therefore not about
freedom of the will as a whole but only about one of the acts of will,

"Aquinas's Account of Freedom: Intellect and Will" by Eleonore Stump,

The Monist, vol. 80, no. 4, pp. 576-597. Copyright© 1997, THE MONIST, La Salle, Illinois 61301.
 AQUINAS'S ACCOUNT OF FREEDOM 577

namely, electio, in which such freedom is exemplified. (In order to avoid

confusion, I will leave both Hiberum arbitrium' and 'electio' untranslated
and let their meanings emerge from a consideration of Aquinas's views.)
For these reasons, to understand Aquinas's account of human free-
dom, I will focus on his account of the nature of intellect and will, the in-
teractions between them, and the emergence of freedom from their
interaction. With that background, I will look in detail at Aquinas's theory
of freedom, in particular at his views of human ability to do otherwise and
the relation of that ability to freedom. Finally, I will consider what answer
can be given to a taxonomic question: Is Aquinas's theory a compatibilist
account, as is sometimes claimed, or is it a libertarian theory; and if it is
a libertarian theory, what species of libertarianism is it?
Because of his views of divine grace and its effect on the will, we
cannot be entirely clear about Aquinas's theory of freedom without also
understanding his account of grace.3 But this is a large subject, which
needs a careful treatment of its own. Furthermore, it is Aquinas's firm con-
viction, reiterated in many passages, that God moves the will only in
accord with its nature, and not against it. According to Aquinas himself,
then, his account of grace should not alter the conclusions we reach about
his theory of the nature of the will, independently considered. This paper
will therefore consider Aquinas's account of the will apart from his views
of grace.

2. Intellect and will

Contemporary philosophers tend to operate with a conception of the
will as the mind's steering-wheel, neutral in its own right but able to direct
other parts of the person. Aquinas's conception is different. He takes the
will to be not a neutral faculty, but a bent or inclination. The will, he says
in the Sumnta theologiae (hereafter ST), is a hunger, an appetite, for
goodness.4 By 'goodness' in this connection Aquinas means goodness in
general, not this or that specific good thing; that is, the will is an inclina-
tion for what is good, where the phrase 'what is good' is used attributively
and not referentially.
By itself the will makes no determinations of goodness; apprehend-
ing or judging things as good is the business of the intellect. The intellect
presents to the will as good certain things or actions under certain de-
scriptions in particular circumstances, and the will wills them because it
578 ELEONORE STUMP

is an appetite for the good and they are presented to it as good. For this
reason the intellect is said to move the will not as an efficient cause but as
a final cause, because its presenting something as good moves the will as
an end moves an appetite.5 This is one reason for calling the will a 'moved
mover' (as Aquinas notes that Aristotle does), because, in moving what is
under its control, the will is moved by an object intellectively apprehend-
ed as good, or an "intellectively cognized appetible,"6 as Aquinas puts it.
Understood in this way, the will can be seen as part of a larger
scheme. Because all things are created by a good God who wills what is
good for his creatures, all things are created with an inclination of their
own to the good, but of very different sorts. Some, like plants or even
inanimate things, have a built-in inclination to the good apart from any
cognition of the good. Aquinas sometimes calls this inclination a natural
appetite. (The sort of thing he has in mind is exemplified by plants
naturally turning toward sunlight.) Higher up the ladder of being are
animals of certain sorts which are naturally inclined to the good but with
some (sensory) cognition.7 They can cognize particular goods, although
they lack the ability to reflect on them or to think of them as good. Incli-
nation dependent on limited cognition of this sort Aquinas calls 'sensory
appetite'. Higher still are human beings whose inclination to the good is
dependent on intellect, which allows them not only to cognize particular
goods but to think about them reflectively as good. This inclination is
rational appetite, and it is what Aquinas takes the will to be.8 So close is
the association between intellect and will for Aquinas that he often speaks
of the will as being in the intellect,9 and he thinks that anything which has
intellect must also have will.10
Understood as rational appetite, the will is the primary mover of all
the powers of the soul (including itself) except the nutritive powers," and
it is also the efficient cause of motion of the body. Most important for our
purposes, the will exercises some degree of efficient causality over the
intellect. In some circumstances, it can command the intellect directly to
adopt or to reject a particular belief.12 It can also move the intellect by
directing it to attend to some things and to neglect others,13 or even to stop
thinking about something altogether. So, for example, while you are reading
a magazine, you come across an advertisement asking for money for
children, with an emotionally powerful picture of a starving child. Your
intellect recognizes that if you look at the ad for very long, you are likely
to succumb to its emotional force. Intellect sees the goodness of con-
 AQUINAS'S ACCOUNT OF FREEDOM 579

tributing to the charity, but it also recognizes that if you give money to this
charity, you won't have it for the new computer you have been coveting.
Your desire for the new computer is strong and influences intellect to rank
saving money for the computer as best for you now. In consequence of the
finding on intellect's part, and with this influence from the passions, will
directs intellect to stop thinking about the charity, and (after a further in-
teraction of intellect and will) you turn the page of your magazine.
As this example shows, in addition to control over intellect by will,
the passions—sorrow, fury, fear, greed, etc.—can also influence the intellect,
because in the grip of such a passion, something will seem good to a
person which might not seem good to her otherwise.14 The intellect,
however, typically isn't compelled by the passions in any way;15 it can
resist them, for example, by being aware of the passion and correcting for
its effects on judgment, as one does when one leaves a letter written in
anger until the next morning rather than mailing it right away. Further-
more, the passions are themselves theoretically subject to the will. In other
animals, Aquinas says, motion follows directly from the sensitive appetite's
positive or negative reaction. In human beings, however, the sensitive
appetite awaits the act of the will, which is the superior appetite. The
lower appetite, Aquinas thinks, isn't by itself sufficient to cause movement
in other powers unless the higher appetite, the will, permits that movement.16
That is why, for example, human beings can go on hunger strikes and stay
on them to the point of starvation.
I raise the subject of the relation of the passions to intellect and will,
however, only to put it to one side. Although Aquinas has many interest-
ing things to say about the moral psychology of the passions, his account
of the intellect and will and the freedom that emerges from their interac-
tion is more than enough for our focus here. I will therefore introduce the
passions into the discussion only when it is necessary to do so in order to
understand what Aquinas has to say about intellect and will.
Just as the will can affect the intellect in various ways, so the intellect
can move the will (as a final, not an efficient, cause) in more than one way.
The will can be moved to will as distinct from not willing—the "exercise"
of its act; or it can be moved to will this rather than that particular thing—
the "specification" of its act.17
There is nothing in this life that invariably and ineluctably moves
every human will to the exercise of its act, because it is always in a
person's power to refuse to think of the thing at issue.18 Since will wills
580 ELEONORE STUMP

something only in case intellect presents it as some sort of good, the fact
that will can command intellect to stop thinking about something means
that will can, indirectly, turn itself off, at least with regard to a particular
action or issue. This is only a limited ability on the part of the will,
however, since the apprehensions of the intellect can occur without any
preceding act of will and so in some cases may force the issue back on the
agent's attention. That is why, for example, the prisoner who wants not to
think about what is happening next door where other prisoners are being
tortured will find that their screams make him recur to what he wants to
stop thinking about.19
As far as the specification of its act is concerned, there is no object,
other than happiness in this life and God in the next, which by its nature
necessarily moves every human will to want that.20 Because God has
created the will as a hunger for the good, every human will by nature
desires die good. And whatever is good to such a degree and in such a way
mat a person cannot help but see it as good, the will of that person wills
by natural necessity. One's own happiness is of this sort,21 and so a person
necessarily wills happiness.22 But even things which have a necessary
connection to happiness aren't willed necessarily unless the wilier is
cognizant of their necessary connection to happiness.23 Except for
happiness and things so obviously connected with it that their connection
is overwhelming and indubitable, it is not the case that every human will
is in general determined to one thing because of its relation to the intellect.
On Aquinas's account, the will wills only what the intellect presents at
that time as good under some description. Acts of will, then, are for
something apprehended or cognized as good at a particular time in partic-
ular circumstances, as distinct from something which is good considered
unconditionally or abstractly. Besides happiness and the vision of God, all
other things are such mat they can in principle be considered good under
some descriptions and not good under others, so that mere is nothing
about them which must constrain the will of any agent always to want
them. So, for example, the further acquisition of money can be considered
good under some descriptions in some circumstances—e.g., the means of
sending the children to school—and not good under others—e.g., wages
from an immoral and disgusting job.
Finally, the will can move itself in more than one way. It can move
itself indirectly by commanding intellect to stop thinking about something,
as we've just seen. It can also move itself indirectly because in virtue of
 AQUINAS'S ACCOUNT OF FREEDOM 581

willing a certain end it moves itself to will the means to that end. That is,
the will wills a certain means because it wills a particular end and because
intellect presents that means as best for attaining that end.
But a more direct control over itself is possible for the will, too. All
the higher powers of the soul, Aquinas holds, are able to act on them-
selves.24 So, for example, intellect is able to cognize itself.25 By the same
token, the will can will to will.26 In fact, Aquinas confronts a problem that
has troubled some contemporary hierarchical accounts of the will, namely,
that there may be an infinite regress of higher-order willings. I can will
that I will something, and I can also will that I will that I will something,
and so on, apparently ad infinitum. But in such an apparently infinite
series, the will is not actually taking ever-higher orders of volition as its
object. At some point, Aquinas thinks,27 the apparently higher-order
volitions collapse, and the object of the will is just whatever action was at
issue at the beginning of the series of volitions.
If intellect does present something to the will as good, then, because
the will is an appetite for the good, the will wills it—unless will directs
intellect to reconsider, to direct its attention to something else, or to stop
considering the matter at hand. Will's doing such things, of course, is a
result of intellect's presenting such actions on the part of the will as good,
and that act of intellect may itself be a result of previous acts on the part
of the will directing the attention of the intellect. On Aquinas's view,
every act of willing is preceded by some apprehension on the part of the
intellect, but not every apprehension on the part of the intellect need be
preceded by an act of will.28
It is apparent, then, that on Aquinas's account, the will is part of a
dynamic feedback system composed primarily of the will and the
intellect, but also including the passions. The interaction between will and
intellect is so close and the acts of the two powers so intertwined that
Aquinas often finds it difficult to draw the line between them. So, for
example, he says that
. . . it happens sometimes that there is an act of the will in which something
of the [preceding] act of reason remains . . . and, vice versa, there is
[sometimes] an act of reason in which something of the [preceding] act of
will remains.29
That is why it sometimes looks as if, for Aquinas, will engages in acts of
apprehension and intellect engages in acts of willing.
582 ELEONORE STUMP

If we remember this part of Aquinas's account and also take seriously

his identification of the will as a hunger or appetite, we will be less likely
to identify the will on his account as nothing more than a toggle switch
with three positions: accept, reject, and off. Aquinas's account of the will
is more complicated than such an identification implies. Because it is an
appetite, the will can have dispositions, so that it can be more or less
readily inclined to want something; it can will something with more or
less strength.30 It can give specific commands to body parts. And under the
pull of the passions, for example, it can influence what intellect presents
to it as good by selectively directing the attention of the intellect.
For this reason, too, although Aquinas's account of the will assigns a
large role to intellect, he isn't committed to seeing immoral actions simply
as instances of mistakes in deliberation, since intellect's deliberations are
in many cases dependent on the will's influence. In cases of incontinence,
where the intellect seems to be representing something as good which the
will isn't willing, Aquinas would say that the intellect, influenced by the
will, is in fact being moved by opposed desires to represent the thing in
question as both good (under one description) and not good (under a
different description), so that the intellect is double-minded.31 In the last
analysis, what the intellect of the incontinent person represents as the best
alternative in these circumstances at this time isn't what the agent takes to
be good considered unconditionally or in the abstract.
Cases of incontinence illustrate the further complicating fact that
intellect need not present one simple, unified result to will. Sometimes an
agent is entirely of one mind about something, and what intellect presents
to will is one unified message that a particular thing is good. But what is
no doubt also often the case is that an agent's intellect is not entirely
unified. The doctor has recommended x-raying the agent's head to check
for a sinus infection. On the one hand, the agent's intellect may recognize,
the doctor is an expert in her field, and her advice for that reason should
be followed. On the other hand, the intellect may be aware that even low-
level x-rays are carcinogenic, and the intellect may wonder whether the
doctor's ordering the x-ray reflects her concern to avoid malpractice law-
suites rather than her own view about what is necessary for the health of
her patient.
Furthermore, the influence of the passions may also complicate the
case. It might be that a patient's intellect supposes some medical tests are
in fact medically required, but his passions might recoil strongly from the
 AQUINAS' S ACCOUNT OF FREEDOM 583

tests. In that case, his aversion may influence the intellect to give a divided
verdict: on the one hand, it would be good to undergo the tests, because
they are important for health; on the other hand, it would be bad to
undergo the tests because they are painful or disgusting. In such cases,
there may be considerable interaction among intellect, will, and passions,
until, in consequence of such iterated interaction, one side or another of
the divided intellect becomes strong enough to override the other. This is
a process familiar enough to anyone who has had to talk himself into
doing something he originally feared or disliked.

3. The relation of freedom to intellect and will

One of the perplexing things about the preceding analysis of the
relation of intellect to will is that it isn't immediately apparent in what
sense the will is free.
It is helpful in this connection to notice that Aquinas recognizes a
distinction between freedom of action and freedom of willing.32 He ac-
knowledges, for example, that we can lose our freedom of action while
retaining our freedom with regard to willing. Even when the will itself is
not compelled or coerced in any way, he says, the members of the body
can be impeded by some external cause so that they don't follow the
command of the will.33 While an agent might still be free with regard to
his willing in such a case, he wouldn't be free with regard to his actions,
which in the case envisaged are at least in part under some control other
than his own. In order for an agent to have freedom of action, then, it isn't
sufficient that his will be free in its willing of that action. It must also be
the case that there is no external impediment to the action of the relevant
body parts and that those parts are themselves functioning normally.
Consequently, freedom of action is not a property of just one component
of a human being. Rather, it is a property of a whole system, comprised at
least of the will and the members of the body. It emerges when the will is
freely commanding a certain sort of movement and when the relevant
bodily parts are functioning normally and are not kept by any cause
external to the agent from being under the will's control.
It is helpful to see this feature of freedom of action, because it points
us in the right direction for Aquinas's account of free will: for Aquinas,
freedom with regard to willing is also a feature of a whole system.
In explaining what constitutes a distinctly human action, Aquinas
puts the point this way. What differentiates human beings from non-rational
584 ELEONORE STUMP

animals is that a human being is master of his acts, in virtue of having

intellect and will. Consequently, no freedom with regard to willing remains
for a person who, through madness, for example, has lost the use of his in-
tellectual faculties.34 Aquinas makes the same point another way by
saying that the root of freedom is in the will as subject but in the reason
as the cause.35 That is, the property of freedom inheres in the will, which
is the subject for the property, but it does so because of the intellect; will's
relations to and interactions with the intellect are the source of the
freedom in the will. Freedom with regard to willing, then, is not a charac-
teristic either of the will or of the intellect alone.36 Like freedom of action,
freedom with regard to willing emerges from the functioning of a system,
in this case the system comprised of intellect and will.
Furthermore, Aquinas says that an agent is master of his acts or has
his acts in his own power insofar as they are voluntary, and that it is a
person's voluntary acts which make him subject to praise or blame.37 But,
in his view, whatever is voluntary requires an act of the intellect as well
as an act of the will.38 Seconding a view of Damascene's, Aquinas calls a
voluntary act "an act that is a rational operation."39 In fact, Aquinas holds
that because the will has the relation it has to the intellect, all the acts of
the will are voluntary, whether they are simple acts of will (such as willing
an end) or are commands to some other power which the will controls
(such as willing to move one's arm).40 Finally, in Aquinas's view, anything
that takes away an agent's use of her intellectual faculties also takes away
the voluntariness of her action.41
A voluntary act is thus a special case of being moved by an intrinsic
principle.42 Whatever is moved by an intrinsic principle in such a way that
it acts for an end which it cognizes as an end has within itself the principle
of its action. Some creatures act with a limited cognition of the end for
which they are acting, so that their acts are voluntary but in a limited sort
of way. The acts of young children and some animals are voluntary in this
way. Normal adult human beings can have a full cognition of their ends,
and so they can have complete voluntariness with regard to their acts.43
By the same token, and perhaps as a consequence of the same
thought about the voluntary, Aquinas thinks that anything external to the
agent which acted coercively on the agent's will would thereby destroy
voluntariness. That the voluntary movement of the will be from an ex-
 AQUINAS'S ACCOUNT OF FREEDOM 585

trinsic principle, Aquinas says, is impossible.44 This is not an empirical

claim but a conceptual one. For something to be an act of will, it has to stem
from an intrinsic source, in particular the will as informed by the intellect.
So, Aquinas says,

an act of the will is nothing other than an inclination which proceeds from an
interior cognizing principle .. . .but what is compelled or violent is from an
extrinsic principle.45

If something extrinsic to the agent were to act on the will with efficient
causation, then the tie of the will to the intellect, from which acts of will
get their voluntary character, would be broken, and so the act of the will
wouldn't be voluntary—or to put it more nearly as Aquinas seems to think
of it, in such a case it wouldn't be a real act of the will at all.
We might wonder here why Aquinas wouldn't grant that an act of
will could be voluntary even if it were caused by an extrinsic principle,
provided that the extrinsic principle produced its effects by operating
directly on the agent's intellect and only thereby, indirectly, on the agent's
will. Aquinas considers something like this question himself when he asks
whether Satan could bring it about that a human being sin.46 Aquinas sub-
scribes to the demon-possession theory of mental illness, so he supposes
that Satan can causally determine a human intellect (to one degree or
another) by possessing it. But this is to destroy it as a human intellect; an
insane person has lost his reason. At any rate, if some external agent S has
taken over entirely the intellect of some human H, then the intellect that
is operative in that human person is S's and not H's. In that case, what the
will operative in H wills might be voluntary, but it would count as S's will,
not H's, since the intellect that informs the willing is S's. In this case, there
can be an extrinsic principle S which operates on the intellect of some
other agent H, but the operation of the extrinsic principle won't give us an
act of will that can count as H's.
On the other hand, if we were to imagine Satan (or the evil neuro-
surgeon) invading H's intellect only partially, for example, by producing
a thought or a train of thoughts, H's intellect will then examine that
thought or set of thoughts and evaluate it, retaining or rejecting it according
as it seems right to H to do so. In that case, however, the voluntary acts of
will which may result will stem from the reflections of H's intellect, not
586 ELEONORE STUMP

S's. Here again, then, we will not have a case in which a voluntary act of
will on H's part is causally determined by an extrinsic principle S,
operating through H's intellect.
So, worries about grace aside, it should by now be clear that Aquinas
is not a compatibilist. The causal chain resulting in any voluntary act on
an agent's part has to originate in the system of the agent's own intellect
and will. If it originates in some cause external to the agent which acts
with efficient causation on the agent's will, what results will not be an act
of will at all. And if it takes over the agent's intellect which in turn deter-
mines the content of the agent's will, what results will not be an act of the
agent's will. So while extrinsic principles may influence human volition,
as, for example, we sometimes do when we persuade one another by
arguments, causes external to an agent cannot effect a voluntary act of will
on that agent's part, either directly or indirectly.
If Aquinas isn't a compatibilist, what sort of incompatibilist is he? It
seems clear that he must be a libertarian. And yet, although the outlines of
Aquinas's theory of human freedom are now somewhat clearer, it still
isn't obvious in what sense the will—or the system of will and intellect—
is supposed to be free. No doubt, part of what gives rise to this perplexity
is the presupposition, common enough in discussions of free will, that lib-
ertarian free will includes or even just consists in the ability to do
otherwise.47 But in what sense is it possible for the will, or the will-and-
intellect, to do otherwise on Aquinas's view?

4. Liberum arbitrium and the ability to do otherwise

Aquinas's account of a person's ability to will otherwise typically
occurs in connection with his discussion of liberum arbitrium, and it is
because of what Aquinas says about this that he is supposed by many
scholars to see human freedom just as a function of the will's ability to do
otherwise.
It is true that Aquinas makes a strong connection between liberum
arbitrium and the ability to do otherwise. In fact, although 'liberum arbitrium'
means ' free judgment', Aquinas sometimes sounds as if liberum arbitrium
is just the power of the will to do otherwise than it does. For example, he
says, "whoever has liberum arbitrium has it in his power to will or not to
will, to do or not to do."48
 AQUINAS*S ACCOUNT OF FREEDOM 587

Nonetheless, it is a mistake to suppose that 'liberum arbitrium' is

Aquinas's term for the freedom of the will in general. In fact, he explicit-
ly associates liberum arbitrium with only one of the acts of will needed to
produce a bodily movement in a free human action, namely, the act of will
which is electio. So, for example, he says, "liberum arbitrium is that in ac-
cordance with which we have electio."49 And in another place he says,
with regard to intellective appetite, will and liberum arbitrium—which is
nothing but the power of electio—are related just as intellect and reason are
related with regard to intellective cognition. . . . Now the act of electio is to
desire something for the sake of obtaining something else, so that, strictly
speaking, it has to do with means to an end.... Therefore, it is clear that will
is related to the power of electio—that is, to liberum arbitrium—just as
intellect is related to reason.50
Elsewhere he associates the act of liberum arbitrium just with the selection
of a means to an end,51 which is electio.
So although, on Aquinas's account, liberum arbitrium involves being
able to do otherwise, it is not identical to freedom of the will in general,
but instead picks out just the power of the will manifested in the act
Aquinas calls electio.
Now electio is the will's assenting to the means which the intellect
has apprehended as best for the end wanted by the wilier.52 It is therefore
a very specific sort of act of will. Furthermore, not every free action has
an act of electio in the series of acts of will and intellect producing such
an action. When the intellect finds only one acceptable means to an end,
then the act of electio collapses into the act of consent53—an act of will
which is not associated with liberum arbitrium—precisely because there
aren't alternatives available for intellect and will to rank-order.54
In addition, even understood narrowly as confined to the power of
the will producing electio, liberum arbitrium isn't a property of the will
alone. It can be understood as a property of the will only insofar as the will
itself is understood to be the rational appetite and to have a close tie to the
intellect. In some places Aquinas speaks of liberum arbitrium as if it were
in fact a power of both the will and the intellect. When he is asking
whether God has liberum arbitrium, one of the objections he raises begins
wim the uncontested remark that "liberum arbitrium is a faculty of reason
and of will."55 Elsewhere he says
588 ELEONORE STUMP

liberum arbitrium should be considered on the basis of electio. But both the
cognitive power and the appetitive power contribute something to electio.
From the cognitive power we need counsel, by which we determine what is
to be preferred to what, andfromthe appetitive power we need the desire to
accept what counsel has determined.56
That is why, he goes on to say, Aristotle supposed we ought to assign
electio either to the "appetitive intellect" or to the "intellective appetite,"
phrases meant to indicate the intertwining of intellect and will in liberum
arbitrium. (Of this pair, Aquinas opts for "intellective appetite"—the will
understood as preceded by certain acts of intellect—as the more appro-
priate candidate for the faculty to which liberum arbitrium is to be
assigned.) Although he thinks that if we take liberum arbitrium to be a
faculty rather than one of the powers of a faculty, then it is just the will
itself, he emphasizes that liberum arbitrium is the will understood as in-
terwoven with and dependent on intellect.57
Finally, that will has the ability to do otherwise even in acts of electio
does not stem from the fact that the will may simply choose not to follow
intellect,58 or may act in some other way as a homunculus independent of
intellect. It concerns instead the relations between intellect and will.
Insofar as the will has control over itself, this is mediated by the intellect.
It is also limited, since there are intellective apprehensions which are not
preceded by or dependent upon acts of will. Will may not always succeed,
for example, in getting intellect to stop thinking about something, because
something in the environment causes the thought to recur repeatedly in the
intellect, as in the case described above of the prisoner who wants not to
think about what his hearing keeps calling to his mind, namely, the torture
of his fellow prisoners. But, within a limited range, will can be effective
at controlling intellect, for example, by being able in some circumstances
to redirect the attention of the intellect, and in that way the will can also
have indirect control over itself.
Of course, the will's directing intellect in any of these ways will
depend on intellect's presenting will's doing so as good in these circum-
stances. That is why a human agent's control over her own actions is a
function of intellect and will and is an emergent power or property,
resulting from the dynamic interaction between intellect and will, rather
than a static power localized in the structure of one particular faculty.
Given the nature of Aquinas's account of freedom, it makes more sense to
attribute freedom to a human being with regard to willing or acting than
 AQUINAS'S ACCOUNT OF FREEDOM 589

to take freedom as a property of one particular component, whether will

or intellect.
5. Freedom and the will's ability to do otherwise
Aquinas supposes that human beings have control over their own
actions and that this control is manifest, perhaps even specially evinced,
in certain acts of electio involving the ability to do otherwise. It is also
important to recognize that for Aquinas the powers that give a human
being control over her actions aren't themselves grounded in an ability to
do otherwise. What produces such control is the nature of human intellect
and will and their interaction. As long as these are functioning normally
(and that, as we have seen, precludes will's being causally determined by
anything outside the wilier), an agent has control over her actions and
freedom with respect to her willing and acting, even if she cannot do
otherwise.
Some sign of this attitude on Aquinas's part can be found in what he
says about the limits of liberum arbitrium. Something can be outside the
power of liberum arbitrium in two ways. First, it can exceed the efficacy
of the motive powers. For example, flying by flapping one's arms is not
within the power of human liberum arbitrium, because flying exceeds
human powers of movement. Secondly, and this is the important point for
our purposes, acts which we do under the sudden impetus of some passion,
such as wrath or concupiscence, are outside the power of liberum
arbitrium because they occur quickly, before reason can deliberate about
them. An agent may be able to avoid letting passion have such effects by
paying careful attention; but an agent cannot always be paying careful
attention. In unguarded moments, such passions can arise without the
process of reason, even tacit reason, necessary to choice; and the agent in
acting on such a passion is consequently unable to do otherwise on that
occasion. If Aquinas supposed that liberum arbitrium were identical to
free will or if he thought that the ability to do otherwise were essential to
free will, he should go on to say here that such acts aren't sinful or blame-
worthy in any respect since they occur unfreely. What he in fact says is
that they are sinful, but constitute only venial sins since their suddenness
and their taking us by surprise provide us with some excuse.59
In De malo (QDM) Aquinas argues at length that it is heretical to
suppose that the will is moved of necessity to will whatever it wills,
because such a supposition undermines all attributions of praise and blame,
590 ELEONORE STUMP

removes the impetus to deliberation, exhortation and precept, and so on.

But in that very question he also grants that the will sometimes wills what
it wills of necessity. This happens when what is willed is so altogether
good that intellect can't find any description under which to present it as
not good—as in the case of happiness. But it also happens in other sorts
of cases, as when intellect establishes very clearly that one course of
action is in every respect superior to any other available.60 So Aquinas
ends his discussion of this point with the conclusion that although the will
is sometimes moved of necessity, it isn't always so moved.
What Aquinas means us to understand here, we might suppose, is
that what the will wills is free only when the will is not moved of
necessity. But such an interpretation would be a mistake. In arguing that
the will does will some things of necessity, Aquinas explains that there are
two relevant sorts of necessity. One is the necessity of coercion, which
would occur if some cause outside the agent causally produced in the will
a volition for some particular thing.61 This, Aquinas says, is incompatible
with freedom. (In fact, as we saw earlier, there can be no such coercion of
will for Aquinas, because he thinks that it is conceptually impossible for
any necessity of this sort to operate on the will.) But there is also the
necessity of natural inclination. This is the sort by which the will wills, for
example, things whose goodness is overwhelmingly apparent to the agent.
Necessity of this sort, according to Aquinas, is not repugnant to the will
and doesn't take away its freedom.62 Siding with Augustine, he says
"Freedom . . . is opposed to the necessity of coercion, but not to the
necessity of natural inclination."63
That is why Aquinas thinks that there can be freedom of will on the
part of the redeemed in heaven who no longer have the ability to will
evil.64 Their inability to will anything but the good stems not from any
extrinsic coercion being exercised on their wills but rather from the clear
view their intellects have of the nature of the good.
Where there is no defect in apprehending and comparing, there can be no
volition for evil with regard to those things which are ordered to the end, as
is clear in the case of the blessed.65
Their intellects can no longer find descriptions under which to present as
good things that are really evil. And so, although the blessed cannot will
evil, they nonetheless will freely whatever they will.
 AQUINAS'S ACCOUNT OF FREEDOM 591

Elsewhere Aquinas contrasts the necessity of coercion with the

necessity of the end. When someone is compelled by an extrinsic cause in
such a way that he cannot do otherwise than he does, this is necessity of
coercion, and it is altogether repugnant to the will. But necessity of the
end is different. It arises, for example, when the end desired can be
attained in only one way, as when crossing the sea requires using a ship.
This is in no way repugnant to the will, on Aquinas's view. But, Aquinas
concludes, the necessity of natural inclination is similar in the relevant
respects to necessity of the end, and so necessity of natural inclination is
also not repugnant to the will. For this reason as well as others, he
maintains that "natural necessity doesn't take away the freedom of the
will."«
Clearly, then, Aquinas doesn't suppose that human freedom even as
regards willing consists in or depends on the ability to do otherwise.
Aquinas would consequently reject what is called the Principle of Alter-
native Possibilities (PAP). PAP has many different formulations, but they
all share this claim:

(PAP) A person has free will with regard to (or is morally responsible for)
doing an action A only if he could have done otherwise than A.67

Aquinas would reject this principle not only for bodily actions but even
for acts of will.
Many contemporary philosophers also suppose that PAP is false. A
standard strategy for showing that PAP isn't necessary for free will is what
has come to be known as a Frankfurt-style counterexample.68 In such an
example, a person P does an action A in circumstances that incline most
people to conclude that P is doing A freely, but (in the example) there is
some mechanism that would have operated to bring it about that P would
have done A if P had not done A by himself. In the actual sequence of
events presented in the counterexample, however, the mechanism does
not operate, and P does do A by himself. So the counterexample is
designed to make us think that P does A freely in the actual sequence of
events although it is not the case that P could have done otherwise than
A.69 Frankfurt-style counterexamples can be constructed either for bodily
actions such as leaving a room or for mental actions such as deciding to
leave a room.
592 ELEONORE STUMP

Some libertarians defend PAP by arguing strenuously against

Frankfurt-style counterexamples.70 But Aquinas would presumably find
such counterexamples to PAP acceptable. In the actual sequence of events,
P's doing A is not brought about by any cause extrinsic to P, and nothing
in the counterexample keeps us from supposing that it is only P's own
intellect and will which are responsible for P's doing A.
These reasons for the rejection of PAP don't have the implication that
libertarian free will is never accompanied by alternative possibilities. On
the contrary, as we have seen, Aquinas emphasizes human ability to do
otherwise, for example in his account of liberum arbitrium. It may in fact
be true on his view that in most cases in which an agent acts with free will,
the agent can do otherwise. The ability to do otherwise would then be
what medieval logic calls "an associated accident," a non-essential property
that accompanies its subject much or even all of the time. Nonetheless, as
I have argued, on Aquinas's account human freedom with regard to
willing isn't to be identified with the ability to do otherwise; it is possible
for an agent to act freely even when she can't act otherwise than she does.
Even where there are alternative possibilities available, on Aquinas's
account of the will's relations to intellect they will be open to the agent
only because she can be in a different intellective state from the one she
is in fact in. For Aquinas, alternative possibilities for the will are de-
pendent on alternative possibilities for the intellect; it is not possible for
the determination of the intellect to be that doing A is what is good now
and for the will (with that determination of the intellect still in place) to
will not to do A. But it may nonetheless be possible for the will to will not
to do A because it may be possible for the agent's intellect to recalculate
and to rescind the determination that doing A is good now.
6. Aquinas among the libertarians
Although standard Frankfurt-style counterexamples are commonly
thought to show that alternative possibilities aren't necessary for freedom,
there are also analogues to Frankfurt-style counterexamples which show
that alternative possibilities aren't sufficient for freedom either.
Consider, for example, Robert Heinlein's The Puppetmasters. In that
story, an alien race of intelligent creatures wants to conquer the earth. Part
of the alien plan for invasion includes a covert operation in which indi-
vidual aliens take over particular human beings without being detected.
When an alien "master" takes over a human being, the human being (say,
 AQUINAS'S ACCOUNT OF FREEDOM 593

Sam) has within himself not only his own consciousness but the master's
as well. Since it is crucial to the alien plan that their taking over human
beings be undetected in the early stages of the invasion, they are careful
to make the behavior of people like Sam correspond to the behavior Sam
would normally have engaged in had he not been infected with the alien.
So when, under the control of the alien, Sam does some action A, it is also
true that if there had been some reason sufficient for Sam in his uninfect-
ed state to do not-A, the alien would have brought it about that Sam in his
infected state did not-A. In that case, then, there is a possible world in
which Sam does otherwise than A. Sam has the ability to do otherwise,
then; nonetheless, Sam isn't free with respect to his doing A.
In the standard Frankfurt-style counterexamples, the absence of al-
ternative possibilities doesn't preclude an agent's acting on his own
unimpeded intellect and will in the actual sequence of events. In the
analogue counterexamples, the presence of an alternative possibility
doesn't stem from any ability that the agent's own intellect and will have
in the actual sequence. What the standard and analogue counterexamples
together show, I think, is the correctness of Aquinas's position. In order to
determine whether or not an agent is free, it is important to determine
whether the intellect and will on which he acts are his own, not whether
alternative possibilities are present or absent for him. For Aquinas, human
freedom depends on human cognitive capacities and on the connection of
the will to those capacities. Consequently, as long as human acts of will
originate in those faculties, those acts count as voluntary and free, even if
the agent couldn't have done otherwise in the circumstances or the act of
will is necessitated by natural inclinations of intellect and will.
So Aquinas holds a view which is libertarian in some sense. It
maintains that human beings have free will and that free will is incom-
patible with causal determinism. Nonetheless, this is a position that will
strike some libertarians as highly unsatisfactory. For some medieval lib-
ertarians (and for some contemporaries), an act of will is free only in case
the agent could have performed a different act of will in exactly the same
set of circumstances with exactly the same set of beliefs and desires.71 For
such libertarians, the alternative possibilities available to the will need to
be available to the will simultaneously, with the agent in the same state of
mind. On this way of thinking about free will, to be free, the will needs to
be unconstrained not only by causal influences outside the agent; it needs
to be unconstrained even by the agent's intellect. On Aquinas's view,
594 ELEONORE STUMP

however, it isn't possible for the will to be unconstrained by the intellect,

and what is necessary for freedom is that an agent's will not be causally
determined by anything outside the agent. For Aquinas, there is an alter-
native possibility open to the will only in virtue of the fact that it is possible
for the intellect to be in a different state.
Such different intuitions about what is needed for free will are remi-
niscent of medieval debates about the nature of the autonomy of a free
will. Some Franciscans tended to suppose that a free will had to be inde-
pendent of the intellect as well as of all external causal influence.72
Aquinas thought that the will is free just because of its connection with the
intellect. For my purposes here, what is most important to see in this
dispute is the nature of Aquinas's position within the genus of libertarian
theories of free will. One can hold an incompatibilist theory of free will,
as Aquinas does, without accepting the principle of alternative possibili-
ties; one can maintain that the will is free in a way incompatible with
causal determinism without espousing the Franciscan version of libertari-
anism. What exactly to call Aquinas's position is not clear. It seems to me
that it is a species of libertarianism, but one which grounds libertarian
freedom in an agent's acting on his own intellect and will, and not in the
alternative possibilities open to the agent.73

Eleonore Stump
Saint Louis University
St. Louis, MO

NOTES
1. See, for example, Klaus Riesenhuber, "The Bases and Meaning of Freedom in
Thomas Aquinas," Proceedings of the American Catholic Philosophical Association 48
(1974), 99-111, esp. p. 101.
2. Cf. J. Korolec, "Free Will and Free Choice" in the Cambridge History of Later
Medieval Philosophy (ed. Norman Kretzmann, Anthony Kenny, Jan Pinborg, associate
editor Eleonore Stump, [Cambridge University Press, 1981]), p. 630; David Gallagher,
"Thomas Aquinas on the Will as Rational Appetite," Journal of the History of Philosophy
29 (1991), p. 570, n.26; and Daniel Westberg, "Did Aquinas Change His Mind About the
Will?", The Thomist 58 (1994), 41-60.
3. Cf. David Burrell, Freedom and Creation in Three Traditions, (Notre Dame, IN:
University of Notre Dame Press, 1993), and Brian Davies, The Thought of Thomas Aquinas,
(Oxford: Clarendon Press, 1992), pp. 174-78.
 AQUINAS'S ACCOUNT OF FREEDOM 595

4. 571-11 q.10 a.l and I q.82 a.l.

5. 571 q.82 a.4.
6. 57Iq.59a.lobj.3.
7. For a discussion of sensory cognition, see my paper "Aquinas on the Mechanisms
of Cognition: Sense and Phantasia," forthcoming in The Copenhagen School of Medieval
Philosophy, ed. by Sten Ebbesen.
8. See, for example, 571 q.59 a.l corpus. See also ST I q.83 a.l, I—II q.6 a.l, and De
veritate {QDV) q.xxiv a.l.
9. See, e.g., the reference to Aristotle in ST I q.59 a.l obj.l.
10. See, e.g., 571 q. 19 a.l corpus.
11. See 571-11 q.9 a.l, STl q.82 a.4, and 571-11 q.17 a.l.
12. Although faith is divinely infused, according to Aquinas, he also seems to suppose
that faith results from such an action of the will on the intellect. See, for example, QDV
q. 14 a.3 reply, ad 2, and ad 10.1 have defended Aquinas's view that will can act on intellect
in this way in "Intellectual Virtues: Wisdom and the Will to Believe" in Aquinas's Moral
Theory, ed. Scott MacDonald and Eleonore Stump, Cornell University Press, forthcoming.
13. See 571-11 q.17 a.l and a.6.
14. 571-11 q.9 a.2.
15. 57Iq.81 a.3andl-llq.l0a.3.
16. 57Iq.81a.3.
17. Cf.57I-IIq.9a.l.
18. Cf. (De malo (QDM) q.6 a.l, where Aquinas says that even as regards happiness,
the exercise of the will is not necessary since a person can always will on a particular
occasion not to think about happiness.
19. For the way in which intellect makes use of sensory cognition to apprehend partic-
ulars, see my paper "Aquinas on the Mechanisms of Intellective Cognition," Revue
internationale de philosophie, forthcoming.
20. See 571 q.82 a.2.
21. The ultimate good simpliciter is God, on Aquinas's account. Hence the sight of God
in the beatific vision also moves the will necessarily.
22. 571 q.82 a. 1.
23. 571 q.82 a.2.
24. eOVq.22a.12.
25. There is something misleading about talk of intellect's cognizing or will's willing.
Aquinas says, "We can say that the soul understands, in the same way that we can say that
the eye sees; but it would be more appropriate to say that a human being understands by
means of the soul." (571 q.75 a.2 ad 2.)
26. For Aquinas's general view that all acts of will must be preceded by an act of
intellect, see, for example, 571 q.82 a.4 ad 3.
27. 571—II q.l a.4, esp. obj.3 and ad 3.
28. 571 q.82 a.4; QDVq.22 a.12. Cf. also QDM q.6 a.l.
29. 5 7 I - I I q . l 7 a . l .
30. 571-11 q.50 a.5,1—II q.52 a.l, and I—II q.66 a.l.
31. Cf., e.g., 571-11 q.17 a.2 and a.5 ad 1.
32. This distinction is related to the distinction between external and internal actions.
For a helpful discussion of this distinction, see David Gallagher, "Aquinas on Moral
Action: Interior and Exterior Acts," Proceedings of the American Catholic Philosophical
Association 64 (1990), 118-29.
596 ELEONORE STUMP

33. See, for example, ST I—II q.6 a.4, and QDV q.24 a. 1 ad 1. Cf. also QDM q.6 a. 1 ad
22, where Aquinas says, "he who does what he does not want [to do] does not have free
action, but he can have free will."
34. STI-riq.lOa.3.
35. 5TI-IIq.17a.lad2.
36. Aquinas's position here is like his position as regards perception and intellection.
The proper sense organ for sight is the eye, and yet, Aquinas thinks, it is a human being
who sees, not the eye, and a similar point could be made about intellective cognition. See,
for example, ST I q.75 a.2 ad 2.
37. 5TI-Hq.6a.2andq.l7a.5.
38. 5TI-IIq.6a.3ad3.
39. 5TI-IIq.6a.lsc.
40. ST I—II q.6 proemium.
41. 5TI-IIq.6a.7ad3.
42. 5TI-IIq.6a.l corpus.
43. 5TI-IIq.6a.2.
44. ST I—II q.9 a.6. The exception to this claim about extrinsic principles is God, who
can be an extrinsic cause without removing voluntariness since he is the extrinsic cause
creating the will with its inclinations and its connections to the intellect. (See, for example,
5TIq.105a.4ad2.)
45. 5TI-IIq.6a.4.
46. ST I—II q.80 a.3. See also, for example, QDV q.24 a.2 where he says, "If the
judgment of the cognitive [faculty] is not in a person's power but is determined extrinsi-
cally (aliunde), then the appetite will not be in his power either, and consequently neither
will [his] motion or activity."
47. I discuss the relationship of libertarianism and the ability to do otherwise at some
length in "Libertarianism and the Principle of Alternate Possibilities," ed. Jeff Jordan and
Daniel Howard-Snyder. Faith, Freedom, and Rationality: Philosophy of Religion Today
(Rowman & Littlefield, 1996), pp. 73-88.
48. ST I q.83 a. 1 Obj.2. See also 5TI q.83 a.3, where Aquinas says, "we are said to have
liberum arbitrium when we can receive one and reject another."
49. 5TI q.83 a.3 sc.
50. 5TI q.83 a.4.
51. QDV q.24 a.6.
52. See, for example, ST I—II q. 13.
53. See, for example, 5TI-II q. 15 a.3 ad 3.
54. See also In NE III.v [434] where Aquinas says that the genus of electio is the
voluntary; on his view, although "every [act of] electio is something voluntary, electio and
the voluntary are not altogether the same, but the voluntary is in more [acts than electio
is]." One reason for insisting that electio is not identical to the voluntary is this: [436]
"Those things which we do quickly we say are voluntary, because their source is in us, but
they are not said [to be done] with electio, because they don't arise from deliberation."
55. 5TIq.l9a.l0obj.2.
56. 5TI q.83 a.3.
57. 5TI q.83 a.4.
58. Cf. QDM q. 16 a.2 where Aquinas says, "evil cannot arise in an appetite in virtue of
appetite's being discordant with the apprehension it follows."
59. QDV q.24 a. 12.
 AQUINAS'S ACCOUNT OF FREEDOM 597

60. QDM q.6a.l.

61. God's grace does operate on the will with causal efficacy, but Aquinas's account of
grace is complicated and it isn't at all clear that the operations of grace constitute an
exception to his claim here.
62. QDV q.22 a.5; see also QDV q.24 a. 10 obj.5 and ad 5.
63. QDV q.22 a.5 ad 3 in contrarium.
64. See, for example, QDM q. 16 a.5, where Aquinas says that there is no state in which
human beings lack liberum arbitriutn.
65. QDV q.22 a.6.
66. 571 q.82 a.l corpus and ad 1.
67. For different versions of (PAP) and an assessment of their strengths and weakness-
es, see, for example, Thomas Hint, "Compatibilism and the Argument from Unavoidability,"
Journal of Philosophy 84 (1987), 423-40.
68. See, for example, Harry Frankfurt, "Alternate Possibilities and Moral Responsibil-
ity," Journal of Philosophy 66 (1969), 829-39.
69. See my discussion in "Intellect, Will, and Alternate Possibilities," reprinted in John
Martin Fischer and Mark Ravizza, Perspectives on Moral Responsibility, (Ithaca, NY:
Cornell University Press, 1993), pp. 237-62.
70. See, for example, David Widerker, "Libertarian Freedom and the Avoidability of
Decisions," Faith and Philosophy, 12 (1995), 113-18, and "Libertarianism and Frankfurt's
Attack on the Principle of Alternative Possibilities," The Philosophical Review 104 (1995),
247-61.
71. For the medieval positions in question, see Bonnie Kent, Virtues of the Will: The
Transformation of Ethics in the Late Thirteenth Century (Washington, DC: The Catholic
University of America Press, 1995), esp. chs. 3 and 4.
72. For an example of a Franciscan position of the sort opposed to Aquinas's, see, for
example, John Duns Scotus. Contingency and Freedom. Ledum 139, tr. A. Vos Jaczn, H.
Veldhuis, A. H. Looman-Graaskamp, E. Dekker, and N. W. den Bok. (Dordrecht: Kluwer
Publishing Co., 1994), The New Synthese Historical Library, vol. 42, pp. 116-17.
73. I am grateful to the faculty and students at the Thomas Institute in Cologne and to
William Alston, David Burrell, Brian Leftow, and Timothy O'Connor for helpful com-
ments on earlier drafts of this paper. I am especially indebted to Norman Kretzmann for many
very useful comments and suggestions.