FREE BELIEVERS
Pascal Engel
University Paris IV-Sorbonne
0.
Introduction
We know, or at least we suppose that we know, what freedom to act means,
even when we deny that there is such a thing. But do we know what freedom of
belief is, and is there any such thing? It is much less clear. In this paper, I
want first to try to characterise what it means. And I want also to sketch an
answer to the question whether we do enjoy such a freedom.
There is, indeed, a relatively
straightforward sense of freedom of belief, which is freedom of speech and of
opinion. But it is not the sense which concerns us, for freedom of opinion is
the freedom to express one’s beliefs and opinions, and not the freedom
to have beliefs or to form them. It is a species of political freedom,
and it involves the public sphere, whereas freedom of belief concerns, at least
prima facie, the individual. Of course the two senses are not unrelated,
but it is not my purpose here to say how. A free thinker, or someone who enjoys
free thought, is, at least on one familiar conception of political freedom,
someone who is not constrained by others to conceal his opinions, in particular
in religious and political matters. Everyone is familiar with the classical
enlightenment doctrine that each of us has a special responsibility for her own
beliefs, and a right to maintain them against various authorities. But it’s
precisely because the individual is supposed to form them freely, in her own
private sphere, that the question arises of the possible intrusion of others in
this sphere, and hence the question of the possibility of how expressing her
beliefs to others, which is therefore derivative with respect to the first
sense. Now in what sense do we take ourselves to have freedom of thought in
this sense, and do we have a right to say so?
To be free, as I just said, seems at
least to be responsible. In the sphere of action, to be responsible is to be
able to control one’s actions by one’s will, to have free will. But beliefs are
not actions that one takes, and we do not seem to have the capacity to choose
to have any belief that we please, in the sense in which we have the capacity
of going out for a walk if we please. Beliefs, to take up Hume’s famous phrase,
seem to belong to the passive, not to the active part of the mind. This is why there is a traditional
philosophical problem of freedom of the will for the philosophy of action, but
there is no comparable problem of freedom of belief for epistemology. It is often said that the threat that the
will might not be free is analogous to the threat of scepticism in epistemology,
but the claim that we might not be free believers does no seem to pose any
special threat to our concept of knowledge. So there is an important asymmetry
here, one which prevents us, prima facie, from applying to belief the
ordinary model that we have of free action. Or at least, if we do not stop
here, and simply renounce to bring together the two kinds of spheres, the
cognitive and the practical, we should develop a sense in which beliefs can be
actions and active states of the mind, and thus subject in some sense to the
will.
A second kind of thought which can lend
belief to belong to the sphere of responsibility is this: beliefs are subject
to norms. For instance someone who believes that P, and who believes that if P
then Q, ought to believe that Q. If she recognises the force of the modus
ponens rule (and even if she does not recognise it), and if she believes
that not-Q, then there is something wrong. In general, people suppose
that there are certain things that they ought to believe or that ought not to
believe, and they assume that they are responsive to what they think they ought
to believe, when they form such and such beliefs (Pettit & Smith 1996:
429). In this sense they can be subject to blame or reprobation in their
beliefs, when they fail to comply by these norms. Can this give us a plausible
sense of our responsibility towards our beliefs? It depends on how we
understand the ought here and the relevant notion of norm. For is we
understand it in the practical sense, as pertaining to things that we ought to do,
then we are back to the previous model which conceives freedom of belief as a
kind of action. But are the norms which are in play when we say that we ought
to believe, say, the conclusion of a certain inference when it is forced on us
by the premises and a certain rule of inference, the same kind of norms as the
norms (social, moral) which pertain to the practical sphere? We might instead
think of them as epistemic norms. But are epistemic norms the same kind of
norms as practical norms? We are back again to the previous question: is
believing the same kind of thing as acting, and if it is not can we say that we
are liable to responsibility in the cognitive or epistemic sphere? If the norms
of epistemic reason are forced upon us, we can hardly say that we can be free
believers in this sense, or we have to explore a sense in which our
responsibility towards epistemic norms can be put on a par with our
responsibility towards practical or ethical norms.
It is at this point that my problem
crosses the problem of mental causation. The problem of mental causation is the
problem of how states or events which have a semantic content can cause states
of events in the physical world. Our problem here is how reasons,
whether or not we understand them as based on psychological states or not, can
have a causal power. In the practical domain, the question pertains to reasons
to act. In the epistemic domain, which concerns us here, the problem pertains
to reasons to believe. Can our reasons to believe be causal in the sense of
causing us to believe in certain ways, and thus making us free ?
In the first part of this paper, I
discuss the idea that we could be free believers in the sense in which we could
have a voluntary control over our beliefs. The upshot is negative. So I turn,
in the second part, to the idea that to be a free believer is someone who is
responsive to certain norms, but responsive in the sense that she is aware or
conscious of these norms, hence that she has a certain capacity of reflection
upon her reasons to believe. In many ways, this is a Kantian proposal, one
which locates freedom of belief in a certain form of autonomy, which the
Kantian associates to agency and the capacity of first personal thought with
respect to reasons to believe, which in this sense is parallel to the kind of
autonomy that is claimed to be a condition of Kantian freedom. But I find this
model inadequate, if it implies that freedom of belief is tied to the capacity
of reflecting, through second-order beliefs, upon one’s first order beliefs.
Mere reflection cannot move us to believe, and cannot move us to move ourselves
to believe, in the appropriate sense of power and control which seems to be
required by genuine agency.
1.
Freedom
of belief as believing at will
The first sense, then, in which we could be
said to be free believers is this: we enjoy freedom of belief when our beliefs
are controlled by our will, just as we are free agents when our actions are
controlled by our will. I shall not question, for the moment, this conception
of freedom of the will, but shall simply ask whether, if true, it can be
transposed to the purportedly similar phenomenon of freedom of belief.
But as I said in my introduction, the
suggestion is implausible, for familiar reasons (Williams 1970)[1].
Normally our beliefs are not under the direct control of our will. Of course it
is in general admitted that they can indirectly be under such control.
Through hypnosis, auto-suggestion, the formation of habits or other stratagems,
we can produce in us the state of believing. But these are all forms of mediated
belief productions. Moreover these are not cases of beliefs as actions, but of
actions which have, as their effects or products, states of believing. Certainly the fact that I can do an action,
described by its product, through performing another kind of action, which
leads to this product as the result of a causal process, does not prevent us
from characterising the whole process as an action. For instance hitting the
target is an action, which I initiate by bending the bow, and sending the
arrow. In this sense, believing that I can fly, through taking a drug, can be
called an action. But if the formation of the belief can be in this
sense voluntary, through the voluntariness of the act which initiates it
(taking the drug), process and the state which is the result is not under the
control of the will. It no more shows that beliefs are actions than the fact
that by bending the bow and sending the arrow shows that speed of the arrow,
its trajectory, or the resistance of the air are actions. The contribution of
the world to the success of the action is not something which is up to me,
although I take it into account. What does seem impossible is the formation of
a conscious intention to believe that p which would result in the act of
believing that p in the same sense as that in which a conscious
intention to act directly produces the act in question. We can call this the uncontrollability
feature of belief (Noordhof 2001). Familiar explanations of this feature
include psychological ones and constitutive or conceptual ones. A psychological
explanation can appeal either to the ordinary phenomenology of belief or to the
functionalist thesis that beliefs are normally caused by environment inputs and
result in outputs in behaviour. If beliefs could be caused by intention and be
actions, this would run counter to this normal role. A conceptual explanation
consists in saying that the very nature of belief, unlike desire and other
motivational states, is to “aim at
truth”, and to have the “mind to world” direction of fit. One can also combine
the two sorts of explanations, by saying that we could not, in full
consciousness, acquire a belief irrespective of its truth and think of
it as something purporting to represent reality, for I could not think of this
belief as mine, since I would mean that I could both believe that not p
(since I have no evidence for its truth, or disregard this evidence) and
believe that p(if my action of getting this belief is successful)
(Williams: 148). So here again we hit upon the fundamental categorical
difference between beliefs on the one hand, and desires, intentions and
motivational states on the other.
It is not clear, however, that these
psychological or conceptual arguments for the uncontrollability of belief are
satisfactory. For the psychological argument do not rule out the psychological
possibility of creatures who would, unlike us, not aim at truth nor the
possibility that the normal function of belief might dysfunction. And the
conceptual argument does not undermine the possibility that after our forming
the conscious intention to believe p irrespective of its truth, there
might be later evidence for its truth. There is also, of course, the
possibility of self deception or of wishful thinking. And there is the
case of self-fulfilled beliefs such as the belief that I am courageous which
makes me courageous. Can’t these count against the thesis that there cannot be
direct voluntary belief formation? In the case of self deception, it’s not
clear that we are in presence of direct belief formation. But in these as in
the other cases, such as wishful thinking, it is not clear that the beliefs are
rational. And here, as with the conceptual argument, we hit upon an
important feature of belief: that it is normally shaped by evidence for what is
believed, and that a believer is rational to the extent that she is responsive
to the evidential reasons for her belief. This is not to say that the concept
of irrational belief – such as believing contrary to evidence - is a psychological or conceptual
impossibility – on the contrary it is most common – but that the irrational
beliefs, or the less than rational ones, do not seem to be free. A
self-deceived person, or a wishful thinker, are rarely described as free in
their beliefs. On the contrary, the very fact - if it is a fact – that they have been able to manipulate their
beliefs in some way, seems to show that they are unfree. This why the
classics described motivated belief formation as a kind of flaw in the human
agent. For instance Pascal said, famously :
(Nobody ignores that there are two ways
through which opinions are received in the mind, which are its two main powers,
understanding and will. The most natural is that of the understanding, for one
should never give one’s assent but to proven truth; but the most ordinary,
although it is against nature, is that of the will. For men in general are
almost always led to believe not through evidence, but through pleasure. This
way is low, indignant, and foreign; this is why everybody disapproves it.
Everyone professes that he believes, and even loves only what he knows to be
worth it. ) [2]
Believing at will is against “nature”. And
nature is such that we should believe only what is worth believing. A belief
which would be willed and acquired irrespective of evidence, then, would not be
free: it would be under the lead of the lowest part of the mind. This is
conform to a most common intuition that we have about freedom, which is well
expressed by Christine Swanton when he says: “ The freedom phenomena are
characterised by the absence of various flaws, breakdown and restrictions on
human practical activity, namely those which limit the potential of
human beings as agents”(Swanton, 1992, quoted by Peacocke 1999: 308).
Believing freely, then is believing for
reasons which are independent of us, which are appropriate to an evidence that
we do not create, nor manipulate. It is believing according to nature, nature
been such that the understanding, or the cognitive part of the mind is not
under the control of the will, which rules the practical part. But what kind of
freedom is this? If the claim is that a good belief system is one which
conforms to nature, which functions well, by registering evidence, then
certainly someone can have a belief system in good functioning order, or the
normal potential, but not be free for that. Animals, in so far as they have
informational states which register evidence, and if we are prepared to call
them beliefs, could be free in this sense. But are we prepared to take them as
paradigms of free believers?
What apparently needs to be added, in
order to fit our intuitions, is that the subject must be aware of his
functioning well, that his being responsive to evidence must imply some
consciousness of it. And this does not seem to involve the will or any sort of
mental act of this kind. But before developing this suggestion, I want to
discuss briefly another one, which could, on the face of it be compatible with
finding a place for the will in rational beliefs.
It is often said that even if belief
are not under the direct control of the will, there are certain epistemic
activities which help the formation of our beliefs. I can be more or less
attentive to evidence, desire more or less to investigate the matter at hand,
form various hypotheses, accept them or not, and take them more less seriously[3].
I can make judgements, or suspend them. All these expressions betray certain
kinds of actions or decisions. Judgement, or acceptance, in this sense if often
said to be under the control of the will, and this gives plausibility to
Descartes’ doctrine that judgement is voluntary, and that it is the source of
error. In this sense, cognition is under the control or practical activities,
and hence under the control of free will. But this proposal pertains obviously
to indirect voluntarism about belief, not to direct voluntarism. And although
these activities of belief formation can be more or less rational, they do not
for that make the beliefs which are the products of these activities rational.
Whether or not the activities which help to fix belief are amenable to
practical norms, and in this sense rational, it does not follow that the beliefs
that they produce are rational. For the norms of rationality to which belief
obeys are epistemic norms (evidence, coherence, truth): the fact that I can
arrive, but more or less rational means to a belief does not make the belief
more or less rational in the epistemic sense (Owens 2000: 85-87).
The failure of our attempts to base
freedom of belief in the will and in other motivational states thus shows to
things. First it shows that we cannot hope to base epistemic freedom upon
freedom of the will. For whether the will is free or not is irrelevant to
whether we are free in our beliefs. For beliefs and other cognitive attitudes
are not actions. So we cannot found freedom of belief in freedom to act.
Second, and by implication, it shows that our intuition, we I have noted at the
beginning, that we incur certain epistemic obligations, and that belief is
subject to oughts, should not be cashed out in aligning this kind if ought
with the practical ought. If there is an “ought to believe”, it is not of the same kind as the “ought to
do”. If there is a normative requirement here, it is not a practical, but an
epistemic one. This means that the kind of agency and responsibility that we
hoped to find in the cognitive realm is not the kind of agency and responsibility
that we can find in the realm of action.
We could stop here, and simply conclude that the hope
to define a clear concept of freedom of belief is a non-starter. But we can try
also to redefine it. For there is an important feature of the activities, such
as accepting, hypothesising, or judging which are said to help the fixation of
belief: in so far as these are actions, they are conscious, and they imply that
the subject is aware of his beliefs, able to stand up behind them. This
suggests that our responsibility towards our beliefs emerges from our awareness
or them. So we might try to locate freedom of belief in this very feature.
2. Freedom of
belief and the power of reflection
I have said that the reason why we are reluctant to locate freedom of belief in the presence of motivated mental states and in acts of the will or intentions is that we feel that such a mode of belief formation is irrational. If the subject were really rational, in the epistemic sense, it seems that she should be aware of her beliefs. The awareness, in cases of irrationality, such as self-deception, that she is forming intentionally beliefs which are contrary to those that she already has, is also the awareness of this irrationality. But to be aware of one’s beliefs, that is having second-order beliefs about one’s first-order beliefs, seems precisely to be what we mean by “being responsive to the reasons” of our beliefs. So we can suggest that this is the proper source of our feeling of responsibility towards our beliefs. But this responsibility does not lie in our having first-order beliefs – for instance, and paradigmatically, perceptual ones – but in our being capable of being conscious of them, and to reflect upon them, hence to form second-order beliefs.
So the suggestion is that the proper source of our responsibility towards our beliefs, hence our freedom of belief, lies in our power of reflection. This view has been espoused by many recent writers. For instance McDowell writes:
“Judging, making up
one’s mind what to think, is something for which we are, in principle,
responsible – something we freely do, as opposed to something that merely
happens in our lives. Of course a belief is not always, or even typically, the
result of our exercising this freedom to do what we think. But even if a belief
is not freely adopted, it is an actualisation of capacities of a kind, the
conceptual, whose paradigmatic mode of actualisation is in the exercise of
freedom that judging is. This freedom, exemplified in responsible acts of
judging, is essentially a matter of being answerable to criticism in the light
of rationally relevant considerations. So the realm of freedom, at least the
realm of freedom of judging, can be identified with the space of reasons.” (McDowell
1998a: 434, quoted in Owens 2000: 1).
In similar fashion, Christine Korsgaard identifies the capacity to think about our beliefs, which only humans can have, to the capacity of having reasons, which is the proper domain of the normative (Korsgaard 1996 a: 92-93, quoted in Owens 2000: 9-10). Tyler Burge (1996, 1997) tells us that to be intellectually responsible requires the capacity to understand the way norms govern our thinking, and that this capacity requires our acknowledging this responsibility. This, he argues, requires in turn having the concept of the first person, of a person who he able to endorse his own beliefs and to stand behind them as commitments. In turn again this intellectual responsibility requires “conceptualised self-consciousness.[4] Robert Brandom also has defended such a view. This account can be called Kantian, for the reason that Brandom indicates:
“Kant’s reconciliation
of us as free in virtue of being rational, with us as bound by norms in virtue
of being rational – and so of freedom as constraint by a special kind of norm,
the norms of rationality – accordingly involves treating the normative status
of moral obligation as instituted by normative attitudes.” (Brandom 1994: 51)
This is the familiar Kantian concept of freedom as autonomy. But the point made by these writers is that freedom as autonomy does not pertain only to the practical sphere, but also to the theoretical one: being a free believer, and being intellectual responsible, involves being responsive to the norms of theoretical reason. If, however, we appreciate that the locus of responsibility is not in the will as such, but in the rational will, as it exercised in judgement, we can reconcile the demands of practical and theoretical rationality. What they have in common is that in both freedom is tied to our reflective awareness of norms.
In their paper “Freedom in belief and desire (1996) Philip Pettit and Michael Smith apply this model both to belief and to desire. They call it the ideal of “orthonomy”(1996: 442) :
“Responsible believers
and desirers are orthonomous subjects, in the sense that their recognise
certain yardsticks of right belief and right desire and can respond to the
demands of the right in their own case.”
The principle which underlies these analyses can be formulated as a principle :
(PR) An agent is a free believer, when
i)
she is rational or responsive
to reason and to the norms of belief
ii)
she is responsive to reason when she is able to reflect on her
own reasons for believing
Two things at least speak in favour of such an account. First the fact, already noted that it give pride of place to our consciousness of our own beliefs, which seems essentially for rationality. For instance in his book The moral problem, Michael Smith proposes the following principle for rational belief (which has a parallel for rational desire):
C2 If an agent believes that she has (most) reason to believe that p then she rationally should believe that p
Smith cashes out the reflectivity requirement thus:
“Suppose an
agent believes that she would believe that p if she were fully rational and yet
fails to believe that p. Is she irrational? She certainly is. And by her own
lights! For she fails to believe something she believes she has most reason to
believe. Indeed this must surely be a paradigmatic case of irrationality. Moreover,
note that we can also explain why someone who believes that p is most credible,
but who also finds herself believing that not p, rationally should get rid of
her belief that not p and acquire the belief that p instead.” (Smith 1994: 178)
Second the location of intellectual responsibility in second-order judgements seems to agree well with another familiar concept of freedom in the practical sphere, namely Frankfurt’s (1971) concept of freedom as based on second-order motivational attitudes. As well known, Frankfurt ties freedom of the will to the capacity to have second-order desires about our first order desires. A free agent is one who has the will that he wants to have. We could try to model our definition of freedom of belief after this by saying: a free believer is one who has the beliefs that he thinks he should have. But this wouldn’t do, for this definition might fit an individual who is self-deceived or a wishful thinker. What has to be added is that the beliefs that a believer should are those which are required by the norms of belief. It does not matter what they are for the account to be plausible, but we can think of such norms as truth ( believe that p iff p ), of evidence ( believe that p iff the evidence points to p) and coherence ( believe that p iff p is coherent with your other beliefs, or entailed by them). But, as Pettit and Smith remark, such norms are not like conventional norms of behaviour, or like rules of etiquette: they are not up for choice, but “inescapable” ( Pettit & Smith 1996: 433-434). In this sense, freedom of belief cannot be like what free will is said to imply: the capacity of the agent to do otherwise. These norms are constraining. But awareness of them is supposed to make us free believers.
Still this account leaves something unclear. If to be a free believer is to be a rational agent, the agent must in some sense have a control over her beliefs. This does not mean, as we saw, that she must be free to adopt any belief that she pleases, but at least that she must be able to accept those beliefs that she finds rational, and reject those which she doesn’t find rational. So in this sense the agent must be able to be moved to believe certain things, and not moved to believe others. She must be in some sense the cause of her beliefs, and their source must be in her, and not in some other person or in some part of her mind to which she does not have access. But the neo-Kantian account is either silent upon this, or unable to explain this feature.
There are in fact two versions of the neo-Kantian account as an answer to this question, a weak one and a strong one:
a. Weak version of PR : reflection is a precondition for the rational control of beliefs
b. Strong version of PR: reflection is constitutive of rational control.
Let us examine first the strong version. It seems to be the one which is held, for instance, by Burge. In “Reason and the first person” (1997) he tells us that “having reasons and having some capacity to be moved by them are necessarily connected” and that one must be “immediately moved by reasons”:
“One must be
susceptible to the force and implement normative evaluations in guiding thought
and other acts that fall under such those evaluation and (to understand reasoning),
one must regard reasons as effective in one’s judgements. Doing so amounts to
acknowledgement of one agency.
… in recognizing
the effect of reasons on one’s judgements and inference, one cannot think of
oneself as powerless.” [5]
The same view is expressed by Bill Brewer about our beliefs reached by logical inference:
“ There is more to grasping the
laws of logic or mathematical argument than simply being disposed to have one’s
beliefs mirror the moves they prescribe. Epistemologically productive reasoning
is not a merely mechanical manipulation of belief, but a compulsion in thought by
reason, and as such involves conscious understanding of why one is right in
one’s conclusion. (Brewer 1995: 242)
Brewer talks here of a special form of causation, “causation by reason”. But how is this kind of causation supposed to operate? How can the mere fact that I am aware of my rational reasons for believing have a direct influence on my beliefs? I see snow falling, and I come thus to believe that it is snowing. I reflect on my belief and say to myself : “I believe that it is snowing”. On the reflection model, my attending to my reasons (I perceive that it snows) and my attending to my belief is supposed to show that I am responsible for this belief. But why is this power of reflection supposed to show that I am in control over this belief? Could it be because I attend to the norms of belief, and reflect that the norm “believe what you take to be true according to the evidence at hand” forces me to entertain the belief? But being aware of the norm does nothing to move me to belief. Remember here the famous case of Achilles and the Tortoise. The Tortoise agrees that the modus ponens is a valid rule of logic. She agrees that if one believes that P and that if one believes that if then Q, one ought to believe Q. But she does not believe Q.
Certainly the Tortoise is irrational, just as Smith’s person who fails to believe what she has most reason to believe. But the point is that her awareness of what reason demands does nothing to force her to believe otherwise, if she does believes otherwise. She could be epistemically akratic, and there is nothing impossible in this.[6]
A more plausible description of the
situation is this. We have first-order beliefs, which are rational or not. We
do not have any control over them. They are just forced upon us. Neither do we
have control over the norms of rationality. We do not become rational just by
being aware our of beliefs and of the norms which govern them. Certainly if a
conflict arises between her first order beliefs and what she believes that she
ought to believe, then she should try to change her beliefs. But she does not
do that because she is aware of these norms and of her beliefs. She does it
because she has to comply by these norms, and is forced to do that.
The neo-Kantian account, then,
misdescribes the situation when it says that awareness of the norms involves a
sort of power of the agent, or a “causation by reason”. But this causation, of
compulsion by reason can hardly justify the claim, made by McDowell, that we
exercise here “the freedom of judging”.
The weaker version of PR, however,
seems to be much more plausible. It says that awareness of our own beliefs and
of the norms of reasons is a precondition of rational control, and that if we
have this awareness, then we can be in position to accept certain beliefs and
reject others. But it does not say that control is effected by this very
awareness. It says that it is a necessary, but not a sufficient condition.
Certainly to be able to maintain a certain belief, to stick to it, or to reject it, I must be conscious of it. And
to asses it, it seems that I must be aware of the norms which govern it. This
is just what is required by the activities which help to fix belief, and which
I have earlier described as a kind of indirect voluntary control.
2. Responsible belief and knowledge
Still, it is not clear what rational
control of belief means. Is the principle of reflectivity even a necessary
condition for rationality?
The neo-Kantian claim implies than we have
a strong authority over what we believe, whether it is a precondition of
rationality or constitutive of it. But do we have this authority?
Belief involves a claim to knowledge. When I say “I believe that
P”, I claim to know that P. This does not mean, of course, that belief is
knowledge, but as Williamson (2000) says to believe that P is to have an
attitude towards P ,which for all one knows , is knowledge. “Belief is botched
knowledge.”
It
is not possible here to argue or of this view, which reverses that traditional
account of knowledge in terms of true belief plus something, justification or
warrant, and which takes the notion of knowledge as a primitive mental state.
It has two important features. First if knowledge is primitive, it does not
imply belief. Second, it does not imply what Williamson calls “transparency” or
“luminosity”, that is the claim that in order to know that P, one must know
that one knows that P ( the “KK thesis” in epistemic logic). So if knowledge is
rational belief, rational belief does not imply that one is aware of one’s
rational belief. Williamson’s argument for this claim is extremely
sophisticated, and I can only indicate its upshot here. It relies on the fact
that our powers of discrimination are limited. If we are in case A, and if a
case A’ is close enough to A, then for all we know we are in A’. To take a classical example, sensations,
such as pain of feeling cold or hot, are typically “luminous” in the sense that
if one feels cold, then one knows that one feels cold. Now suppose a situation
where you feel chilly in the morning, and the whether very slowly warms up,
until you feel hot by noon. Since the change is gradual and extremely slow, our
feelings of heat change but we are not aware of any change over one
millisecond. Suppose that our sensation is reliable (reliability being the
condition for knowing something through
sensation), and that one knows that this sensation is reliable. But by a
familiar kind of sorite argument we can show that although we can be in
position to know that we feel cold in situation A, and to know that we feel
cold in an imperceptibly gradually distinct situation A’, we may end up knowing
that we feel cold in a situation An
where An is
a case where it is hot. Hence when we feel cold, and know that we feel cold, we
can end up in a situation where it is not cold, but hot, hence where we do not
know that we know that it is cold.[7]
Hence, if one generalises, knowledge is not knowledge that one knows; these
imperceptible differences can happen all the way through our knowledge.
The
upshot is that we can know something without being in condition to know that we
know. If we transfer this to rational belief, and take belief as a claim to
knowledge, this will apply to rational belief. There is an important asymmetry
here between the case of belief and the case of desire, between the conditions
of free action and the conditions of purported free belief, which is well noted
by Pettit and Smith: failures to exercise free will are matters of everyday
experience, but failures to exercise free thought are elusive.( 1996: 449). In
the practical sphere, we cannot say: “I am willing to f but I ought not to”. My reasons not to f do not lose
their probative force because I have decided to f. But in theoretical sphere, there is
nothing paradoxical in “I believe that P, but I ought not to “.
As Pettit and Smith say, you may be denied
any experience of evaluating contrary to reasons, but you are by no means
denied the experience of evaluating contrary to evaluation. So I do not
understand why they still insist, in the 1996 paper to say that there is a strict
parallel between freedom of desire and freedom of belief.
I think that the reason why they defend the
parallel is that they adhere to the neo-Kantian account. But if the objections
that I have adressed to this kind of account are correct, there is not reason
to maintain the parallel.
Conclusion
Where are we left, then? If we cannot have a reflective awareness
of our own reasons, can we still maintain that there is such a thing as freedom
of belief? There is a parallel issue in the case of knowledge, and it should
not be surprising if we accept the close relationship between belief and
knowledge. There is a certain account of justification according to which a
justified belief is knowledge if the agent obeys certain epistemic obligations
that he acknowledge. This is called the deontological account of justification.
But it presupposes, in some versions, voluntarism about belief, and at least
something like the principle of reflectivity. Opposed to it is the reliabilist
account of justification, according to we have knowledge only of we are
reliable. But the reliabilist view seems to sever the link between knowledge
and our claim to know that we know. We are not free believers at all.
To
restore this link, and to allow space for our responsibility towards our
beliefs, we should adopt, I think, another alternative, which I can only
suggest here. To be responsive to reasons means adjusting one’s beliefs in the
right way. But this kind of adjustment is not to be conceived as complying to a
rational “ought”. It means taking the appropriate attitudes, the appropriate
habits of thought. to say such things involves adopting a form of what is
called “virtue epistemology”. It ties epistemic responsibility to the having of
certain epistemic virtues, which call for merit, and to some epistemic vices,
which call for epistemic blame. There are more or less reliabilist views of it.
But if one accepts the idea, which I have expressed earlier, that there are
certain activities which help fixing belief, without belief itself being
voluntary, then I think we can understand our responsibility for belief without
endorsing the claim of autonomy that the neo-Kantian view presupposes. *
Brandom, R. Making
it Explicit, Harvard University Press
Brewer, B. “Compulsion by Reason”, Proceedings of
the Aristotelian
Society, Supp vol. LXIX, 237-254
Burge, T. 1998
“Reason and the First Person”, in C. Wright, B. C. Smith and
C. McDonald, Knowing our Own Minds,
Oxford: Oxford University Press,
243-270
Engel, P. 1999 “Volitionism and voluntarism about
belief” in A. Meijers ed.
Belief,
Cognition and the Will, Tilburg University Press
(ed.) 2000 Believing and Accepting,
Dordrecht: Kluwer
Frankfurt, H. 1969 “Alternate possibilities and Moral
responsibility”, Journal of philosophy, 66
1971 “Freedom of the will and the concept of a person”, reprinted in
Watson, ed.Free will, Oxford,
Oxford University Press 1982
Koosgard, C. The sources of normativity
Cambridge Univesrity Press
Mc Dowell, J. 1998 “Having the World in view”, Journal
of Philosophy, 95, 431-50
Noordhof, P.2001 “Believe what you want”, Proceedings
of the Aristotelian
Society, CI, 247-266
Owens, D. 2000 Reason without Freedom, London:
Routledge
Pettit P & Smith M. 1996 “Freedom in Belief and
desire” , Journal of philosophy , XCIII, 9, 429-449
Smith, M. (1984) The Moral Problem, Oxford:
Blackwell
Swanton, C. 1992 Freedom: a Coherence Account,
Hackett, Indianapolis
Williams, B. 1970 “Deciding to believe”, in Problems
of the Self, Cambridge, Cambridge University Press
Zagzebski, L. 1996
Virtues of the mind, Cambridge
[1] The literature is large. Apart from Williams 1970, see Winters 1979, Pojman 1983, Elster 1978 , Bennett 1990, Engel 1999, Noordhof 2001.
[2] Pascal , De l’esprit géométrique et de l’art de persuader, Oeuvres Complètes, II, Pléiade, ed. Le Guern, p.171)
[3] On acceptance, see for instance Cohen 1992, Engel 1998, Engel 2000.
[4] See also Larmore 2001, unpublished.
[5] see also Larmore 2001 (Attending to reasons”, ms p. 29) : “How can reasons operate as causes, as they certainly must do, if their existence is to explain our coming to grasp them?” “The snag is conceiving how reasons as such are able to move us”
[6] for a defense of the possibility of epistemic akrasia, see Hookway 2001
[7] Williamson 2000, 1é-13, 93-113.
A version of this paper was read at the Departement of Philosophy , Oslo University , in january 2002. I thank Olav Gjelsvik, Jennifer Hornsby, Bjorn Ramberg, Jonathan Knowles and participants to the seminar “Freedom” for their comments and criticisms. The present version was read at the conference on Mental Causation , Universidade Federal de Paraiba, in March 2002. I am much indebted to André Leclerc insightfull comments, to Sandy Golberg’s and Lynne Baker’s encouragements and remarks, as well as to the participants of this memorable event.