Free Will and Free Action in Anselm of Canterbury

Published in HISTORY OF PHILOSOPHY QUARTERLY, Volume 22, Number 4, October 2005

Free Will and Free Action in Anselm of Canterbury

Tomas Ekenberg (Uppsala University)

The discussion in this paper revolves around the distinction between freedom of the will

and freedom of action, a distinction made by many. The driving question behind the

paper is what people mean by these notoriously obscure expressions, and what they think

the difference between the two is. A fair assumption is that people who do make the

distinction and people who don’t typically differ in their ‘ontological commitments’ in

matters of ethics and philosophy of action. The aim of this paper is however much more

modest, namely, to clarify the role this distinction plays in the philosophy of Anselm of

Canterbury.1 This requires clarifying the concept of will and the concept of action in

Anselm, and relating them to the way his view of free will bears on matters like coercion,

compulsion and moral responsibility.

1 Willing and Doing

The distinction between freedom of will and freedom of action is connected to an

underlying, more basic, distinction between action and will, or between doing something

and willing something. Since the word “freedom” in the expressions “freedom of will”

and “freedom of action” presumably mean the same thing in both instances, the claim

that freedom of action and freedom of will differ commits us to the view that action and

1

will are at least conceptually distinct if not two separate things. The two notions must

also in some respect be “on par”—they need to come up in the same context of inquiry, or

belong to the same level of explanation. From this it follows that, if we subscribe to the

distinction between two kinds of freedom, we are also committed to the view that we can

at least in some contexts meaningfully contrast the action of a person with her will, or

“wills”, or acts of will.

There is something both interesting and puzzling about this distinction between will and

action. On the one hand, it is a natural distinction, bringing into focus the divide between

what we could call the inner and the outer aspects of action. While describing in everyday

terms a person’s actions—what a person does—is not so different from describing the

changes in any old perceptible object, describing what a person “wills” involves getting

clear about “softer” things—wishes, desires, motives, intentions and so on—and so it

typically requires interacting with the person who is doing the willing. It may also

involve making certain inferences and asking yourself whether the person is being honest

with you, and things like that. The methods we employ in finding out these two sorts of

facts about our fellow men and women thus differ radically—one consisting in mere

observation, the other being dependent on social interaction—and in this light, the

distinction appears warranted.

On the other hand, willing, wanting, intending and so on are certainly things we do, like

thinking, evaluating and calculating. All these things take place in time and in some

particular situation and so they should be actions, which means the willing-acting

distinction appears unwarranted.

2

Harry Frankfurt thinks action and will are distinct. In the very famous article “Freedom

of the Will and the Concept of a Person” he says:

It seems to me both natural and useful to construe the question of whether a person’s

will is free in close analogy to the question of whether an agent enjoys freedom of

action. Now freedom of action is (roughly, at least) the freedom to do what one

wants to do. Analogously, then, the statement that a person enjoys freedom of the

will means (also roughly) that he is free to want what he wants to want. More

precisely, it means that he is free to will what he wants to will, or to have the will he

wants.2

Wanting or willing, on the one hand, and doing or acting, on the other, are different

things. On Frankfurt’s account, freedom of will and freedom of action are two distinct

properties of a person.

When Anselm of Canterbury discusses freedom, he does not explicitly make this

distinction between two kinds of freedom. At least in his “official” works, the ones of

which he is the uncontested author,3 he talks exclusively about freedom of the will or,

more exactly, about “freedom of the choice of the rational will” (libertas arbitrii

rationalis voluntatis).4 Anselm is nevertheless committed to the distinction, and his

account of will and of freedom shares certain features with Frankfurt’s.

Both Anselm and Frankfurt think a person can be “unfree” in regard to her actions. In

other words a person may be overpowered by some external force and made to move in

accordance with it. A person may also be constrained and restricted in various ways so

that she is unable to do what she wants—or unable to do whatever she (thinks she) would

3

do if she weren’t under constraint. It should immediately be noted however that there is a

significant difference between Anselm and Frankfurt: while Frankfurt thinks having an

unfree will is not only possible but even quite common among people who act and react,

Anselm thinks we have a will that is both invincibly and absolutely free:

A human being may be tied up against his will (invitus), since he may be tied up

unwillingly (nolens); he may be tortured against his will, since he may be tortured

unwillingly; he may even be killed against his will, since he may be killed

unwillingly, but he cannot will against his will, since he cannot will unwillingly.5

Frankfurt thinks there can be disharmony between what he calls a person’s first- and

second-order desires. Roughly, it can happen that a person finds herself doing what she

wants (first-order desire), but not wanting to want what she wants (second-order desire).

In this case there is a kind of conflict, or mismatch, between what the person wills and

what she wants to will. By contrast, Anselm in his work De libertate arbitrii rules out

exactly this kind of state of affairs as impossible. He claims that whatever a person wills,

she also necessarily wills to will.6 The Latin word is velle: it is a rich word, and it has a

whole range of possible translations in English (it may mean to wish and it may mean to

choose), and hence comparing the two thinkers is not an altogether straightforward affair.

Still there is a real—and interesting—disagreement between the two.

4

2 To Lie or to Die

Anselm’s favourite example when he discusses issues pertaining to freedom is that of a

person who must lie in order to survive—who knows he will be killed if he tells the

truth.7 The reason examples involving lying plays such a prominent role is that they work

on several levels for Anselm and others who share Anselm’s Christian frame of reference.

In the Bible and in the Christian tradition, a morally upright person is often said to be

“doing the truth”, “standing in the truth”, and the like.8 Anselm takes these expressions

not merely as manners of speaking, but quite seriously—even literally. For him the notion

of “doing the truth” reveals something of the very essence of the goodness of an action,

and in De veritate, Anselm presents an account or theory of truth according to which

actions are true and false in much the same way as linguistic expressions are true or

false.9

On Anselm’s account, actions and utterances (enuntiationes) are both members of a more

general class of “signs”. In the case of both actions and utterances, there is a signifier and

something signified. Utterances are true just in case they signify the way they ought to

signify (debet significare).10 A sign can be correctly and incorrectly applied. When it is

correctly applied it fulfills its function of signifying and does what it ought to do. An act,

like a linguistic sign, is also true just in case it—as it were—does what it ought to do.

According to Anselm, acts signify by their very being that they ought to be done, and so

there is correct and incorrect signification also in the case of action.11

Against this background, the example of someone who is up against the choice of lying

or dying will seem especially rich with meaning. Here a person can “do the truth” in a

5

manifold way—by telling the truth, sticking to the truth, being truthful, being true to his

commitments, and so on. Since acts are signs, any evil act is a form of lie in the sense that

it is a sign being misused or abused. In this light the act of lying is a lie in more than one

sense and therefore false and wrong through and through. Our example is a paradigm

case of an evil act.

Furthermore, the example is especially well-suited for considering issues belonging to the

topic of free will. Compared to arm-raisings and door-openings, the act of lying is more

closely connected to the inner workings of the acting individual. For instance, if it turns

out a person did not really know what she was talking about, we are less inclined to call

her utterance a lie. The difference between a lie and a mistake is thus “internal to the

agent”, which means that to consider the act of lying takes us closer to the will.

In Anselm’s last complete work De concordia, he puts the lie-or-die example as follows:

Someone has it in his heart to hold on to the truth, since he understands that it is

right to love truth. […] Another [person] approaches him, and threatens him with

death unless he lies. We see now he has the choice of abandoning his life for the sake

of rectitude, or abandoning rectitude for the sake of his life. This choice, which may

also be called a decision, is free, since reason, by which he understands rectitude,

teaches him that this rectitude ought always to be preserved out of love for that same

rectitude, and that one should reject anything offered in order to lure one away from

it, and it is the will’s work to pursue and avoid things in accordance with the

judgment (intellectus) of reason. It is to this end will and reason is given to Rational

Creature. Because of this, the will’s choice to abandon that rectitude is not the result

of any necessity, although the person is under the threat of death. For while it may be

6

necessary that he abandon either life or rectitude, no necessity determines which he

keeps and which he lets go of. Indeed his will alone determines which one he holds

on to here, and the force of necessity effects nothing, where the choice of the will

alone is at work.12

Cases like this, which involves coercion, compulsion or external force, are also very

common in contemporary dicussions of free will. In the modern debate, however, it is

often assumed that freedom—or free choice, free will, free agency—is incompatible with

compulsion and constraint. To say that a person does something freely is to say that she

does it without being compelled and that there is no alien force or external constraints

involved significantly in bringing forth the action. Against this background, the moral

Anselm draws from the example will look quite peculiar, for he concludes that, in spite of

being forced and coerced, the person is still fully free. More exactly, he thinks the person

has free choice, liberum arbitrium, and that he fully enjoys “the freedom of choice of a

rational will” (libertas arbitrii rationalis voluntatis).13

From a modern point of view this may sound like mere quibbling. If Anselm wants to call

a person with a sword to his throat free, then fine. While he has then done away with a

term that might have come in handy in dealing with ethical issues, we may say: Don’t

worry; the bluntness of Anselm’s terminology can be amended by simply introducing a

new term—say, “autonomy”. The person in the example has been robbed of some or all

of his autonomy. Anselm is then free to call the person “free” but he certainly cannot call

him “autonomous”.

On closer look at Anselm’s views on freedom, we shall find that the difference is not

merely terminological. But first we must get past a stumbling-block in understanding the

7

import of this example, which looks significantly different from Anselm’s point of view

—mainly because he has certain specific views in normative ethics. People today in

general count force and constraint as extenuating circumstances: enough physical force,

and the victim is relieved of responsibility altogether. To a person who has a sword at his

throat, no blame is normally due—certainly not for such a trivial thing as a lie. Anselm,

by contrast, thinks the person in the example really ought to choose death. The “ought”

here is moral and unconditional. Of course, the lie in the example is not just any lie, but a

particularly significant one—such as the renunciation of one’s faith or the betrayal of our

Saviour. Even so, few today would agree with Anselm, even among Christians. (Martyrs

do fail to gain true appreciation nowadays, at least in our Enlightened West.)

As Anselm describes the situation, the person is free to lie or to die and is to be held

accountable in either case. While it is in this situation necessary that the person either lie

or die, it is not necessary that he lie, nor is it necessary that he be killed. He is free to “go

for” whatever he wants more (fortius vult) Anselm says.14 The presence of coercion or

force or stupefying intimidation does not threaten the person’s freedom. In fact, on

Anselm’s account, nothing threatens a person’s free will. In an important sense, people

have their freedom of choice of the will always, and no matter what.15

The lie-or-die example is intended to show just how free people are: Anselm is arguing

that coercion and constraint never remove the freedom of a person’s will. But Anselm is

also addressing another familiar issue. It is often pointed out that the a more sininster

threat to a person’s freedom consists not in external force, but rather in more subtle, more

intimate forms of coercion—threats from within that are often called “compulsion”.

Anselm thinks he is dealing with this sort too. The “inner” counterpart of coercion and

8

force he calls “temptation”, and he argues it can never overcome the will of the person of

Right Will.

Anselm does not rely solely on this example to make his case, but gives two fairly

sophisticated arguments for his claim. The first of these we have already touched on.

Anselm argues that a person’s will is always more powerful than temptation, for while a

person can be tied up, tortured, or killed against his will, he cannot be made to will

against his will.16 He claims moreover that every person who wills something, wills his

own willing, or wills that he wills what he wills: omnis volens ipsum suum velle vult.17

There is no such thing as willing unwillingly, at least not in the sense that would threaten

a person’s freedom of the will. He does admit that people will unwillingly in the sense

that they choose something reluctantly in order, say, to avoid a greater evil. A person may,

for instance, take some bitter medicine for the sake of his health.18 He insists, though, that

this kind of “willing unwillingly” has no bearing on the issue of freedom.

According to the second argument, which is more far-reaching, not even God can make a

person succumb to temptation. This argument proceeds from the assumption that

succumbing to temptation—diverting from the path of the righteous—essentially includes

failing to will what God wills you to will.19 Now suppose God wills that some person

should yield to temptation. Since yielding to temptation is the same as not willing what

God wills you to will, it follows that God is now willing the person not to will what He

wills the person to will, which is a contradiction since God both wills and doesn’t will the

same thing at the same time. Moreover, according to Anselm, since God could not will it,

he could not do it either: if God cannot will to make people sin, He cannot make people

sin. Given the idea that God has supreme power, Anselm draws a remarkable conclusion:

9

since nothing is more powerful than God who can’t overcome the will of a righteous

human being, nothing can do so, and hence the will is absolutely free.

3 Albritton and Anscombe

Anselm is not the only one who has claimed that a person’s freedom cannot be taken

away from her. Rogers Albritton has argued for a view which is in some ways similar to

Anselm’s.20 His point of departure is a worry raised by Elizabeth Anscombe’s article

“Soft Determinism”, in which she claims a person can be robbed of some—maybe even

all—of his freedom of will. For instance, if a person is chained up, he “has no ‘freedom

of will’ to walk”, or “no freedom of the will in respect of walking”.21 Albritton disagrees

with her claims and challenges her very concept of freedom of the will: he doesn’t see

that his freedom of will would be reduced in any way if he were to be chained up. “You

would of course deprive me of considerable freedom of movement if you did that”, he

says, “you would thereby diminish my already unimpressive capacity to do what I will.

But I don’t see that my will would be any the less free.”22 He thinks the will is absolutely

free and he approvingly quotes Descartes, who says in the Passions of the Soul that the

will is “so free in its nature that it cannot be constrained”.23

According to Albritton, the reason cases like the “chains case” are so significant is that

they seem to point towards a place where the distinction between freedom of will and

freedom of action might collapse.

If you chain somebody up, you seem to be getting closer to where the will is than if

you build a high wall around him at a radius of ten miles from where he sits reading

Spinoza. You seem to be working in toward the connection in him between will and

10

world, which might strike you as a good point at which to try disabling his will

itself.24

There is no such place, Albritton argues: neither physical impossibility nor constraint nor

external force can undermine a person’s free will. As long as the will is involved at all

there is always room for something like “trying” or “deciding against” or “going for

something else”.25 For instance, if a person is overpowered and brought to where she

doesn’t want to go, then it is against her (free) will, to be sure, but it is not because her

freedom of will was taken away. Her will was not involved at all; someone else’s will

was carried out, not hers. Even in the case where someone finds himself with a gun to his

head, his freedom of the will is as valid as ever, Albritton argues.26

Albritton does not provide us with a more definite characterization of freedom of the will,

such as Frankfurt does, but rather appeals to intuitions when it comes to the nature and

scope of free will. For our purposes, however, he makes two important points. The first

point is the view that freedom of the will is perfect, absolute and unconditional. The

second point is that there is a distinction between freedom of will and freedom of action

(or movement).

Although closely related, these two points are distinct, for one need not agree with

Albritton on both of them. Harry Frankfurt, for example, agrees with Albritton on the

second point, while he disagrees on the first, since he does not believe in the

unconditional or absolute freedom of the will. Like Anscombe, he thinks a person could

in one way or another be deprived of the freedom of his will. For Frankfurt this happens

if a person is the victim of some kind of clash between his first- and second-order desires.

11

Such a person finds himself unable to have the will he wants, or, to be more exact, finds

himself having a will that is not the result of “second-order volition”—that is, not the

result of a second-order desire with the content that some specific first-order desire be

one’s will. Both Albritton and Frankfurt maintain, however, that we should distinguish

between freedom of will and freedom of action.

4 Absolute Free Will and Moral Responsibility

Anselm agrees with Albritton on both points.27 Anselm claims that we—rational natures

as we are—always have free will: “semper […] habet rationalis natura liberum

arbitrium”.28 However, the idea that the will is perfectly free, or free always, or free

absolutely, raises questions. In particular the notion of absolute free will becomes

problematic once we turn to the subject of moral responsibility. It is often claimed that

freedom of some sort is a necessary condition for moral responsibility: people are

responsible only for actions performed freely, or out of their own free will, or whilst

enjoying freedom of choice. The exact meaning of these expressions is of course a matter

of much controversy, but the basic idea seems clear enough. Yet Anselm appears to

disagree with this very common and straightforward idea. If we have freedom of the will

always, absolutely—perhaps even necessarily—then it seems pointless to be testing for

free will in order to determine whether a person should be held morally responsible for

some given act: we already know her will was free because it always is. Anselm’s notion

of freedom thus seems to do no work when it comes to questions of moral responsibility.

Let us consider again the person who is going to be killed unless he tells a lie. Even if we

grant that his will is still free, some sort of freedom has been taken from him; surely,

12

Anselm must agree to this. Here it is useful to introduce the distinction between freedom

of the will and freedom of action. We can say that a person has freedom of will always,

and freedom of action sometimes, to different degrees, and depending on the occasion. In

the lie-or-die case the person may have complete freedom of will but a very small amount

of freedom of action.

It may appear that we are now in a position to solve the puzzle about moral responsibility.

The necessary condition for moral responsibility should be freedom of action—not

freedom of will. While people always enjoy freedom of will, people who do not have

freedom of action should not be held responsible for what they do. But clearly this cannot

be right.

Suppose Anselm agrees with us that the person in the example has no freedom of action.

Anselm will still maintain the person is fully morally responsible: he ought not to lie and

should be punished if he does. Consequently, freedom of action can hardly be the

condition we are looking for.

This does not mean that Anselm thinks a person’s freedom is irrelevant to questions of

moral responsibility, but it does mean the relation between freedom and moral

responsibility is more complicated on Anselm’s account than on, say, Anscombe’s, just as

it is more complicated for anyone who is inclined to distinguish free action from free

will. It is, for example, no coincidence that we get from Frankfurt a flat-out rejection of

the universal validity of the so-called principle of alternate possibilities.29 By arguing that

there are cases where moral responsibility does not require that an agent be able to do (or

13

choose) otherwise, Frankfurt calls into question one very widely held view of the relation

between moral responsibility and free will.

Anselm certainly thinks moral responsibility requires free will, but in a special sense. In

De concordia, for instance, he says that “there would be no reason for which God could

justly [punish or reward] each person in accordance with his good or bad merits, if the

person did nothing through his own free choice”.30 So freedom of the will is a necessary

condition for moral responsibility. It is, however, not necessary as a special condition that

needs to be met in the specific case. The condition concerns the agent rather than the act.

Unless you possess freedom of the will you are not the kind of being to which moral

judgments apply, and therefore not the kind of being that is to be held responsible for

your actions (or choices).31

In his discussion of freedom in De libertate arbitrii, Anselm presupposes a previous

treatment of the topic of justice, the moral good, in his De veritate. Here he argued that

justice is a special form of truth and that truth in turn is the same as rectitudo, or

rightness. More specifically, justice is same as “rightness-of-the-will kept for the sake of

that same rightness”.32 In De libertate arbitrii he defines freedom of will as the capacity

or power to keep rightness-of-will for the sake of that same rightness.33 Consequently,

having freedom of the will is exactly the same as having the capacity for justice.34 Human

beings by nature possess this capacity for justice, and therefore they are the sort of beings

which admit of moral evaluation, praise, blame, punishment and reward.35 Furthermore,

Anselm holds it is incumbent on human beings to keep rightness of will always.36

Therefore, human beings are always responsible for what they do: whatever the

14

circumstances, human beings are always able to keep rightness of will, which is exactly

what they are morally obligated to do in each and every situation.

5 Anselm’s Distinction

Let us now consider the distinction between free will and free action in Anselm. Although

Anselm does not explicitly make the distinction, the lie-or-die cases of coercion can be

seen as involving an appeal to it. There is not much the person can do in the case where

he is an helpless victim of external force, but Anselm insists he can still will freely: he

can “will to preserve the truth”, he can choose not to lie. Although he has been deprived

of freedom of action and the stakes are as high as they get, he has still the freedom of his

will. This contrast looks strikingly similar to the one Albritton draws, and so there is

reason to think Anselm would also agree that a distinction must be made between two

kinds of freedom.

Further support for such a conclusion can also be found in Anselm’s own discussion on

the nature of free will. In the De libertate arbitrii he defines freedom of choice of the will

as “the power to keep one’s rightness of will for the sake of that same rightness”, which

means that he defines freedom of the will specifically in terms of will and not in terms of

action.

Many people both before and after Anselm have claimed freedom of the will is a power

to do things rather than a power to will things. Boethius, for instance, in his commentary

on Aristotle’s De interpretatione, maintains free will (liberum arbitrium) is the power to

act or not to act:

15

For there is in us a free will, which is devoid of every necessity and is innate, and by

its power we are the masters, as it were, of what we do and do not do.37

Some fifteen hundred years later, Elizabeth Anscombe says a person has been robbed of

her freedom of will in respect of walking if she cannot walk. For these people, freedom of

the will is a power to act, or at least to do what you want, will or decide. Frankfurt, by

contrast, thinks freedom of the will is not a power to do, but the power to have the will

you want. This is conceptually distinct from being able to do what you will or to move

the way you want to. Anselm gives us something very similar in his account. When he

defines freedom of the will as the power to keep rightness of will, he contrasts rightness

of the will with the rightness of action, rectitudo actionis.38 A person may act rightly

without willing rightly. As an example, he mentions the person who gives alms not out of

love of justice, but out of vainglory.39 Justice is strictly to be understood in terms of

rightness of will, not of action, and consequently will and action are distinct on Anselm’s

account.

In this light it is clear that Anselm’s conception of freedom of the will presupposes a

distinction between action and will, and hence he is also committed to making the

distinction between two sorts of freedom. Although he never actually does so, the

omission need not be a great mystery. Given his approach to the problem of freedom,

evil, and moral responsibility, he may have simply not found the term freedom of action

useful.

Yet, it may be that he did make the distinction after all. If we allow ourselves to go

beyond the set of established works, there is this remarkable, curious little passage in a

16

single version of a work known as the De humanis moribus—a work that was always

attributed to Anselm in Medieval times although its exact origin is now considered

uncertain. In this passage, we learn that there are three sorts of freedom:

For there is the freedom of action, which everyone wills, i.e., [each person wills] to

be able to do freely the things he wills. There is also the freedom of intelligence, that

not everyone wills. Then there is also the freedom of the right will, which is always a

good thing. The last one very few will, that is, [few will] that they will the things

they ought to will.40

This text suggests in a decisive way that the distinction between freedom of the will and

freedom of action is not alien to the Anselmian frame of mind.41 Indeed, there is even a

third sort of freedom, the freedom of intelligence, or the freedom to understand, which

opens up a new area of inquiry all of its own beyond the scope of this paper.42

6 Anselm and the Will

Reflecting on the difference between freedom of action and freedom of will is tantamount

to reflecting on the nature of the will, or on the character of the activity of willing. To say

that there is an important difference between willing something and doing something is to

make room for a theory of the will. In order to get clearer about Anselm’s conception of

will, we should consider what it is to be “willing something” in the relevant sense.

A useful start is to ask whether Anselm’s account of the will involves as an essential

element a notion of an ordered structure of desires.43 The idea that will essentially

involves a structure of desires has most famously been defended by Frankfurt.44

According to him, will in the basic sense is simply “effective” desire. A person has a will

17

if she has a desire to φ and the same desire moves the person to φ:ing. But he also

proposes a second, richer conception of will, which involves the notion of a desire of the

second order. The object of a second-order desire by definition “includes” another desire,

whereas the object of a first-order desire does not. Now second-order volition is a special

sort of second-order desire. If I have the desire to φ, which is a first-order desire, and a

second-order desire the object of which is that the-desire-to-φ be my will (where will

means effective desire) I am having a volition. If I have trouble accomplishing volition,

this is to say I am having problems with the freedom of my will: I am not having the will

I want to have.

Does Anselm entertain a similar notion of an ordered structure of desires? Freedom of the

will is the power to conserve rightness-of-will for the sake of that same rightness-of-will.

Roughly, then, free will is a person’s capacity for goodness, or her power to be just.

Moreover, justice consists in rightness-of-will, and being just involves having the will

you ought to have.

We are now in a position to compare the two accounts. On Frankfurt’s picture, enjoying

freedom of the will is being free to have the will you want. On Anselm’s picture, by

contrast, enjoying freedom of the will is being free to keep the rightness of your will:

having a free will is being free to have the will you should have.45 While on Frankfurt’s

picture the notion of wanting-to-have-a-will is indicative of ordered structure, things are

not so clear for Anselm. Does someone’s keeping the rightness of her will require that she

wants to will, wants to want, or desires to desire something? Does the fact that you have

the will you should have necessarily entail that you want or choose to have this will? At

this point, eleventh century Latin terminology becomes a problem. In a word, it’s all

18

“wills” on Anselm’s account, and no translation of “vult” into “to want”, “to desire”, “to

wish”, or “to choose” will easily escape the charge of arbitrariness.

In Anselm, “willed” action, or action performed volens, is close to what we would today

call intentional action. This claim rests on the following observation: Anselm allows that

not only rational beings, but also animals are capable of willed action. Therefore terms

like “deliberate”, and even “voluntary”, in their modern use, are too narrow. For instance,

Anselm holds that a horse acts “willingly” (volens), and that it “wills” (vult) to act.46 The

concept of will—will per se—is therefore connected to freedom and responsibility only

in a mediate manner. Freedom enters the picture only with rational will, i.e., the will of a

rational being. Hence, the formula libertas arbitrii rationalis voluntatis.

What distinguishes rational willing from mere willing is for Anselm something very

particular, namely, the rational being’s capacity for a certain sort of will, namely, the will

to justice. Anselm defines justice in terms of will—willing justice is willing to will

rightly—and so the will to justice turns out to be a sort of reflexive or second-order will.

However, there is no clear-cut counter-part in Anselm’s theory for the Frankfurtian notion

of second-order volition, i.e., the idea that “full-blown willing” requires wanting that

some desire be effective, i.e., that some desire be your will. If we translate some instance

of Anselm’s “voluntas” as “desire”, and some instance of “velle” as “to want”, then it

may well be true of the morally upright person that she “wants the desire to justice to be

her will”. However, Anselm will still claim that her freedom consists exactly in her

ability to will justice rather than in her ability to form Frankfurtian second-order volitions

in general. While on Frankfurt’s picture willing freely is distinguished from doing freely

19

by means of a notion of second-order endorsement, different considerations underlie the

corresponding distinction in Anselm. The question remains what these considerations are.

On the one hand, Anselm clearly relies on a distinction between willing and doing when

he claims justice consists in willing rightly rather than acting rightly. On the other hand,

however, he appears to think willing is doing. For instance, he says in De veritate that

right willing can be deemed a subclass of right action.47 Why, then, do free will and free

action come out as separate notions? What makes will into such a distinct, and even

privileged, explanatory category?

In the first instance, will is simply the capacity for (intentional) action, shared by angels,

humans and brutes. If this were all there was to the concept of will, Anselm would have

no ground for distinguishing free will from free action. The ground comes from his

ethics: morally right action requires not only that an agent act in accordance with a

certain rule or dictate, but also that the person act with the right motive, or for the right

reason. In particular, Anselm’s ethics requires that we (or at least God) be able to

distinguish the truly willing servant from the person who merely does what he or she is

told. Anselm aims at establishing that there is a crucial difference between the structure of

the intention involved in moral action and the one involved in both immoral and amoral

action. When a righteous person acts, she is, according to Anselm, willing what God wills

her to will solely out of love of justice—only because God wills her so to will. 48 A person

who acts correctly thus acts under a very complex, and also unique, description. On

Anselm’s account every rational being has the capacity for this sui generis sort of action.

And the name of the capacity for such action is freedom of the will.

20

1 References to Anselm’s works are to F. S. Schmitt, Sancti Anselmi Cantuariensis Archiepiscopi Opera Omnia

(Friedrich Fromann Verlag, 1968). S1 and S2 refer to vol. 1 and vol. 2 of this edition. In most cases both page and line

numbers are given. The titles are abbreviated as follows: DV=De veritate, DLA=De libertate arbitrii, DCD=De casu

diaboli, DC=De concordia. All translations are my own.

2 “Freedom of the Will and the Concept of a Person”, The Journal of Philosophy, 68:1 (1971), 5–20. The quote is from

p. 15.

3 Steps were taken already during Anselm's own time to ensure that his works all be preserved for posterity. For a recent

discussion of the most important manuscript, see Ian Logan, “Ms. Bodly 271: Establishing the Anselmian Canon?” The

Saint Anselm Journal, vol. 2, no. 1 (2004), http://www.anselm.edu/library/SAJ/SAJindex.html. For the dating of

Anselm's works, see F. S. Schmitt, “Zur Chronologie der Werke des Hl. Anselm von Canterbury,” Revue Benedictine,

vol. 44 (1932), pp. 322–350.

4 DLA 13 (S1, 225:24–27)

5 DLA 5 (S1, 214:19–22) “Ligari enim potest homo invitus, quia nolens potest ligari; torqueri potest invitus, quia nolens

potest torqueri; occidi potest invitus, quia nolens potest occidi; velle autem non potest inuitus, quia velle non potest

nolens velle.”

6 Ibid. (S1, 214:22–23) ”Omnis volens ipsum suum velle vult.”

7 See DLA 9 (S1, 221), DC 1:6 (S2, 257).

8 See e.g. John 3:21, Eph. 6:14, I John 1:6.

9 For a more thorough discussion of Anselm’s theory of truth, see Sandra Visser and Thomas Williams, “Anselm on

Truth” in The Cambridge Companion to Anselm, ed. Brian Davies and Brian Leftow (Cambridge University Press,

2004).

10 See DV 2, esp. S1, 178:5–15.

11 See DV 5 (S1, 181:19–28) and DV 9 (S1, 189:2–7).

12 DC 1:6 (S2, 257:8–23) “Habet aliquis in corde ut ueritatem teneat, quia intelligit rectum esse amare ueritatem. []

Accedit alius, et nisi mentiatur minatur illi mortem. Videmus nunc in eius esse arbitrio an deserat vitam pro voluntatis

rectitudine, an rectitudinem pro vita. Hoc arbitrium, quod et iudicium dici potest, liberum est, quoniam ratio qua

intelligitur rectitudo, docet rectitudinem illam eiusdem rectitudinis amore semper esse seruandam, et quidquid

obtenditur ut deseratur esse contemnendum, atque uoluntatis est ut ipsa quoque reprobet ac eligat, quemadmodum

rationis intellectus monstrat. Ad hoc enim maxime datae sunt rationali creaturae uoluntas et ratio. Quapropter idem

uoluntatis arbitrium ut eandem rectitudinem deserat, nulla cogitur necessitate, quamvis mortis impugnetur difficultate.

Licet enim necesse sit aut uitam aut rectitudinem relinquere, nulla tamen necessitas determinat quam seruet aut deserat.

Nempe sola uoluntas determinat ibi quid teneat, nec aliquid facit uis necessitatis, ubi operatur electio sola uoluntatis.”

13 DLA 3 (S1, 212:19–20).

14 DLA 7 (S1, 220:7–9) “You see, then, that when the human being who has rightness of will under the onslaught of

temptation abandons [rightness], no alien force takes it from him but he himself turns to that which he wills more

forcefully.”; “Vides igitur quia cum homo habitam rectitudinem uoluntatis aliqua ingruente deserit tentatione, nulla ui

aliena abstrahitur sed ipsa se conuertit ad id quod fortius uult.”

15 See e.g. DLA 10 (S1, 222:5–7) “Rational nature has free choice always, since it always has the power to keep

rightness of will for the sake of that same rightness, albeit sometimes with difficulty.”; “Semper habet rationalis natura

liberum arbitrium, quia semper habet potestatem servandi rectitudinem voluntatis propter ipsam rectitudinem, quamvis

aliquando cum difficultate.”

16 See DLA 5 (S1, 214–217).

17 DLA 5 (S1, 214:22–23).

18 Ibid. (S1, 215:2–5).

19 For the argument in question, see DLA 8 (S1, 220:12–221:19).

20 Rogers Albritton, “Freedom of the Will and Freedom of Action”, Proceedings and Addresses of the American

Philosophical Association 59 (1985), 239-251.

21 See G. E. M. Anscombe, “Soft Determinism” in Collected Philosophical Papers, Vol. 2 (Basil Blackwell, 1981), 166.

22 Albritton, “Freedom of the Will and Freedom of Action”, 240.

23 For a critical discussion of Albritton’s interpretation of Descartes, see Paul Hoffman, “Freedom and Strength of Will:

Descartes and Albritton”, Philosophical Studies 77 (1995), 241–260.

24 Albritton, “Freedom of the Will and Freedom of Action”, 241.

25 Albritton admits that there may be a limit to what a person can choose: in particular, a person can't decide to do what

she knows she can't do. This is not due to any deficiency in the person's freedom of will, however, but the limit is

grammatical, conceptual or logical. See ibid., p. 244.

26 Ibid., pp. 247–248.

27 Anselm agrees also with Descartes that people possess absolute free will. For a comparative study of Anselm's and

Descartes's views on freedom with a focus on the theological aspects of the issue, see Sophie Berman, “Human Free

Will in Anselm and Descartes,” The Saint Anselm Journal, vol. 2, no. 1 (2004),

http://www.anselm.edu/library/SAJ/SAJindex.html.

28 DLA 10 (S1, 222:5).

29 See Harry Frankfurt, “Alternate Possibilities and Moral Responsibility”, Journal of Philosophy 66:23 (1969), 829–

839.

30 DC 3:1 (S2, 264:3–5) “Sed nec ullo modo esset cur deus bonis vel malis pro meritis singulorum iuste retribueret, si

per liberum arbitrium nullus bonum vel malum faceret.”

31 This feature of Anselm's account has in general been overlooked. For instance, Sandra Visser and Thomas Williams

ask in the beginning of their essay “Anselm's Account of Freedom,” with regard to Anselm's definition of freedom,

whether “the exercise of such freedom [is] a necessary and sufficient condition for moral responsibility” (p. 179) . The

authors go on to argue that Anselmian freedom is the capacity for justice and therefore the capacity for

praiseworthiness, which may be taken as a positive answer to the initial question (p. 181). It seems however that since

freedom is the capacity for justice, the exercise of freedom, on Anselm's account, cannot be a necessary condition for

culpability and so not for moral responsibility in general, since the exercise of the capacity for justice is supposedly

exactly the same as the performance of just acts. See Visser and Williams, “Anselm's Account of Freedom” in The

Cambridge Companion to Anselm, ed. Brian Davies and Brian Leftow (Cambridge: Cambridge University Press, 2004),

pp. 179–203.

32 DV 12 (S1, 194:26) “Iustitia igitur est rectitudo voluntatis propter se servata.”

33 DLA 3 (S1, 212:19–20).

34 For a discussion of the connection in Anselm's thought between the notion of freedom and the notion of justice as

rightness or truth, see Richard Campbell, “Freedom as Keeping the Truth: The Anselmian Tradition,” in Anselm Studies

II, ed. J. C. Schnaubelt, T. A. Losoncy, F. van Fleteren, and J. A. Frederick (White Plains, N.Y.: Kraus International,

1988), pp. 287–308.

35 DV 12 (S1, 192:27–193:5)

36 See e.g. DLA 10 (S1, 222:18–19) “He who abandons rightness of will casts away what it was his duty always to

keep.”; “Qui vero voluntatis rectitudinem deserit, hoc abicit quod ex debito semper erat servaturus”.

37 Carolus Meiser, Anicii Manlii Severini Boetii commentarii in librum Aristotelis Peri Hermeneias pars posterior

(Leipzig: Teubner, 1880) Bk. 3, 218:8–12: “Illud enim in nobis est liberum arbitrium, quod sit omni necessitate vacuum

et ingenuum et suae potestatis, quorundamque nos domini quodammodo sumus vel faciendi, vel non faciendi.”

38 DV 12 (S1, 193:7–13).

39 It is not clear whether the converse holds, viz. whether a person (in fact or at least in principle) may be willing rightly

without acting rightly. Although this puzzle poses no threat the point at hand—the distictness of acting and willing—it

raises interesting questions as to the relation between action and will, and the nature of (the “activity” of) willing in

particular. This issue is approached in a more straightforward manner below.

40 R. W. Southern and F. S. Schmitt, Memorials of St. Anselm, (Oxford University Press, 1969), 299. “Libertatis

siquidem tres modos dicimus esse: libertatem videlicet actionis, quam omnes volunt; id est, ut libere facere possint quae

volunt. Dicitur etiam libertas intelligentiae, quam non omnes volunt. Est etiam libertas rectae voluntatis, quae semper

est bona. Hanc paucissimi volunt, id est, ut ea velint quae debent.”

41 According to Southern and Schmitt, the manuscript is from the 1130s, and the text probably derives from second-

hand notes of Anselm’s sayings. In any case, the text is certainly Anselmian with regard to both character and content,

and maybe this is enough for our purposes. See Memorials, 297.

42 In what follows of the short passage, it is claimed that freedom of action is a good thing only if accompanied by the

freedom of the good will. The same goes for the freedom of intelligence. The freedom of intelligence is a person's

capacity to discern what is expedient form him, and as long as a person has a good will, he becomes more and more

capable in this regard with time and experience. See Southern and Schmitt, Memorials, p. 299.

43 A couple of contemporary Anselm interpreters have made remarks to this effect. See e.g. Risto Saarinen, Weakness of

Will in Medieval Thought (Brill, 1994), 29 and 45–48. See also Stan Tyvoll, Free Will and Determinism: The Anselmian

Position, (diss., Saint Louis University, 1996), Section 3.6, 178–187.

44 Harry Frankfurt, “Freedom of the Will and the Concept of a Person”, The Journal of Philosophy, 68:1 (1971), 5–20.

45 In her article “Sanctification, Hardening of the Heart and Frankfurt's Concept of Free Will” (The Journal of

Philosophy, vol. 85, no. 8 [1988], pp. 395–420), Eleonore Stump revises Frankfurt's account by making the forming of

second-order volitions a function of a person's intellect. A second-order volition always involves a judgment of the

intellect to the effect that some specific first-order desire is the good to be pursued. Stump comes close to the view just

attributed to Anselm insofar as enjoying free will, on her account, is being free to have the will one concludes one

should have. However, by introducing a distinction between effective and ineffective second-order volitions, her

account involves a more causally robust conception of will than Frankfurt's own. (See esp. pp. 400–401.) While on the

original account only acts are caused by desires, Stump's revised account includes desires that are caused by desires. In

particular, free willing is construed as the (successful) forming of a second-order volition that causes some first-order

desire to effectively move the agent to action. On such a view we appear to have two stages of effects: first the first-

order desire, then the act proper. This two-stage analysis of the causation of action is not presupposed in Frankfurt's

original account, the one that is here compared with Anselm's.

46 See DV 12 (S1, 192:22–23).

47 DV 5 (S1, 182:18–27).

48 DLA 8 (S1, 220:17–22) “We are talking about that rightness of will by virtue of which a will is called just, that is, the

one that is kept for its own sake. Now, no will is just unless it wills what God wills it to will […] Therefore, to keep

rightness of will for the sake of that same rightness is, for each [person] who does so, to will what God wills him to

will.”; “Nos loquimur de illa voluntatis rectitudine, qua iusta dicitur voluntas, id est quae propter se servatur. Nulla

autem est iusta voluntas, nisi quae vult quod deus vult illam velle. […] Servare igitur rectitudinem voluntatis propter

ipsam rectitudinem est unicuique eam servanti velle, quod deus vult illum velle”.