Self-Deception, Action, and Will

Author(s): Robert Audi

Source: Erkenntnis (1975-), Vol. 18, No. 2, Action, Agency, and the Will (Sep., 1982), pp. 133158
Published by: Springer
Stable URL: http://www.jstor.org/stable/20010803 .
Accessed: 09/02/2015 11:39
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Springer is collaborating with JSTOR to digitize, preserve and extend access to Erkenntnis (1975-).

http://www.jstor.org

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

ROBERT AUDI

SELF-DECEPTION,

ACTION,

AND WILL

it has
has been widely regarded as paradoxical.
Perhaps
Self-deception
someone else typically in?
that since deceiving
seemed to many writers
volves

getting
oneself would

knows
most

him to believe

something

one knows
oneself

is not

to believe

true, deceiving
something one

typically require getting

true. It is not clear how this is possible,
if indeed it is; and
have tried to avoid conceiving
philosophers writing on self-deception
is not

to resolve the
have been widely differing attempts
in
this
of
the
way
conceiving
Perhaps
paradox
implicit
self-deception.1
most important respect in which treatments of self-deception
tend to differ
is the extent to which they take it to be cognitive2 or, on the other hand,
accounts construe self-deception
volitional.3 Cognitive
primarily in terms
it in this way.

There

of the subjects (5"s) beliefs or knowledge;

primarily in terms of actions or patterns

accounts

volitional
of action,

construe

and give S's beliefs

it
a

subsidiary role.
to be developed here is closer to the cog?
The account of self-deception
and it gives a central role to unconscious
nitive end of the continuum,
to do
notions and attempts
belief. But the account also uses volitional
to
treatments
data
that
the
make
volitional
of
justice
self-deception
the account, I have been guided by the assumption
plausible. In developing
that it should be capable of solving, or helping us solve, at least the follow?
ing problems: How should we resolve the paradox of self-deception? How
can we account

for the analogy that presumably exists between self-decep?

tion and two-person
deception? How are we to square the existence of
access ac?
unconscious
beliefs with certain plausible views of privileged
cording to which, if S has a belief, he either knows he does,or at least does
not falsely believe he does not?4 Moreover,
if self-deception
embodies
unconscious belief, how can self-deceivers bd morally
for get?
responsible
or
or
at
out
if
into
in
least
for
it
of
it?
S
has an
it,
(For
ting
remaining
acting
belief, how can he be expected to give it up or to avoid acting
on it?) Is self-deception
in some way under the subject's control and if so
how does that bear on the self-deceiver's moral responsibility?
unconscious

In approaching
Erkenntnis
Copyright

these questions

18 (1982) 133-158.
1982 by D. Reidel

I shall be extending

0165-0106/82/0182-133
Co., Dordrecht,

Publishing

and clarifying

$02.60
Holland,

and Boston,

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

U.S.A.

the

134

ROBERT AUDI

I have sketched in earlier papers,5 exploring the

of self-deception
and "the will," and specifying some major
between self-deception
self-deceivers may be morally
for their self
ways in which
responsible
or
I
for
behavior
connected
with
it.
Section
presents my account
deception
account

relation

of self-deception.
Section II addresses the relation between self-deception
and the will. Section III applies the results of the previous sections to self
section will summarize
deception and moral responsibility. The concluding
and extend

some of the main

points.

I. THE NATURE

OF SELF-DECEPTION

term 'self-deception'
is vague, and, at least in philosophical
literature,
in quite different ways. Itmay not be possible for a univocal account
to capture all its admissible uses. Some cases, however, appear to be central
to our use of the term, and I believe that if we can explicate
these ad?
we
can
uses.
In any event, the cases I have in
understand
the other
equately

The

used

are a quite adequate field for testing an account of self-deception

aimed at solving the problems cited above. The cases I refer to often occur
in the lover who cannot bear the thought that his beloved is drawing away,
the alcoholic who cannot admit that he is unable by himself to stop drink?

mind

ing, the terminal patient unable to face his death, and the athlete unable to
reconcile himself to his waning powers.
Imagine, e.g., that Ann is dying of cancer and is aware of many facts,
to this outcome,
such as her long, steady decline, pointing
though no one
has told her that her case is terminal and she has avoided letting her doctor
Suppose further that she talks of recovery and dis?
give her a prognosis.
cusses her various plans for the long future. If the facts pointing
to her
are not unmistakably
and her talk of recovery is ap?
prominent
to
it
be
wrong
suppose she is lying when she says
parently sincere,
might
she will recover. But if we know that she has better than average medical
death

and if (among other things) we notice that her talk of recovery

knowledge,
and that
lacks full conviction
(or exhibits too much apparent conviction),
or anxiety,
to
it might also be wrong
it is often followed by depression
suppose

that she believes

she will

recover. No

doubt
and

some people in this

in a candid moment

position might be just keeping up appearances

that they believe they are dying. Others might simply
would acknowledge
fail to perceive the drift of the evidence. But (1) ifAnn is in self-deception,

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

SELF-DECEPTION,

ACTION,

135

AND WILL

she has some sense of her dying; it is partly in virtue of this that she may be
said to deceive herself about it. (2) She also appears in some way to be
sincere in saying she will recover; it is partly in virtue of this that she may
it is largely
be said to be deceived. If these two points are uncontroversial,
they are vague. My aim in this section
self-deception which clarifies them.

because

is to provide

an account

of

but of S" s being in self-decep?

Let us speak not merely of self-deception,
p, e.g. that he will recover. Initially, I
respect to a proposition,
as a state in which S uncon?
propose that we conceive such self-deception
tion with

sciously knows that/?, but sincerely avows, or is disposed to avow sincerely,

seems to hold for paradigm
that not-p.6 Though
this rough formulation
that p,
cases, I believe the self-deceiver may have, in place of knowledge
an
true
that
also
that
sometimes
unconscious
belief
grant
people
only
p.71
say 'You're deceiving yourself without
implying that the addressee has an
unconscious
belief. But there what is referred to is not a state but an action,
seems to me non
such as making a statement. This use of 'self-deception'
literal, or at least peripheral, but I believe that on the basis of the account I
as the sort characteristic
am developing
such acts can be understood
of
a
people in state of self-deception.
further by allowing that S
One might wish to liberalize the conception
need only suspect that not-p, provided not-p is indeed true.8 But suppose S
merely suspects, and does not even weakly believe, that not-p. May he be in
self-deception with respect to pi Consider whether S can deceive someone
else with respect to p, when S tells him that p and thereby gets him to
believe p, which is false, yet S only suspects, but does not believe, that not
in believing p and also deceives
p. No doubt S causes him to be deceived
him into believing that S knows (or believes) p. But it is at least not obvious
that S deceives him with respect to p. This is probably not so if S believes
that p ismore likely to be true than not-/?. But suppose that not-p is, in S"s
eyes, at least as likely to be true asp. Here it is again quite unclear whether
S deceives

him. These

S does not believe

and other considerations

then even

if he meets

lead me

to conclude

the other criteria

that if
for self

not-p,
deception with respect to p, we have a borderline case of such self-decep?
tion. Cases of this sort should be decided individually
in the light of the
to accommodate
the belief condition
facts, and I see no need to weaken
them.

Let us now consider

whether

a motivational

condition

is needed.9

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

For

136

ROBERT AUDI

results in part because S wants something, say to see

typically self-deception
himself as a certain kind of person, to be loved and respected, or to avoid
painful thoughts. As this suggests, self-deception may arise from wishful
thinking.10 But must a want (or other motivational
element) occur in every
case of self-deception?
If Ann
is deceiving herself about her impending
death from cancer, must her self-deception
be accompanied
by, e.g., a want
not to die of cancer? This is not entirely clear. Perhaps, e.g., a person can
have an "instinctual"

tendency not to think of his death when he perceives

it near. This might explain self-deception.
It is difficult, however,
to imagine cases of self-deception
in which no
want explains why S is deceiving himself and why the relevant belief is
believes
Moreover,
normally, at least, where S unconsciously
a
the
belief
is
want
unconscious
because
of
in
virtue
of
which
he
has a
p9
stake in repudiating
the belief. He may, e.g., want to believe he is not the
sort of person who believes that, or want that not-p be the case instead. I
unconscious.

am not sure that the relevant concept of unconscious

belief requires that it
in part by appeal to at least one of S"s wants; but that it
be explainable
does require this is a plausible hypothesis.
If it does, then a motivational
element

in self-deception
would follow from my unconscious
belief con?
In any case, normally part of what explains why a belief of S's is
unconscious
is one or more of his wants,
such as his wanting
to avoid
certain painful thoughts; and typically the very wants which explain in

dition.

also explain what motivates

part why S believes not-p only unconsciously,
him to do the things he does in becoming
self-deceived with respect to p.
Ann's wanting not to die of cancer, e.g., might explain both why it is only
that she believes she will die of it and why she deceives
unconsciously
herself with respect to it.
I believe, then, that at least in the full-blooded
cases of self-deception
with respect to p S has one or more wants which explain, in part, both why
the belief that p is unconscious
and why S behaves or tends to behave in a
manner

characteristic
of self-deception.
is not im?
Perhaps this condition
am
in
I
of
the
but
to
inclined
think
it
is and to
concept
plicit
self-deception;
add it to the sketch of self-deception
above.
For
given
paradigm cases,
certainly, the condition holds. These are the kind I have described, in which
we seem to have both a deceiver and a dupe, as we in fact do in ordinary
I shall proceed, then, on the hypothesis
that at least
two-person deception.
the paradigm cases of self-deception may be understood
as follows:

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

SELF-DECEPTION,

S is in self-deception
(1)
(2)
(3)

with respect

ACTION,

AND

to p if and only

WILL

137

if

knows that not-p (or has reason to believe,

S unconsciously
and truly believes, not-p);
and unconsciously
or
is disposed to avow sincerely, that/?; and
S sincerely avows,
S has at least one want which explains in part both why the
and why S is disposed
to dis?
belief that not-p is unconscious
avow a belief that not-p, and to avow /?, even when presented
with what he sees is evidence against it.

So far, I have talked as if it were clear that there can be unconscious

I have argued for this elsewhere.11 On my view, unconscious
beliefs
are very much like conscious ones, apart from two major differences:
(i) if

beliefs.

in 5"s consciousness,
he is very unlikely, without
or
someone
from
else, to attribute these
special self-scrutiny
prompting
to them; and (ii), he is, with the same exceptions,
manifestations
very
unlikely to explain actions of his which are due to them, as due to them.
But - and this is the important point here
they do tend to manifest
and behavior, and in essentially the same way
themselves in consciousness
they manifest

themselves

as conscious

beliefs, though usually less frequently. S may even tend to

them and may actually do so through slips of the tongue.
More must be said, however, about the force of calling a belief uncon?
if and
scious. Roughly, we might say that S"s belief that p is unconscious

avow

only if (1) he does not know or believe he has it, and (2) he cannot come to
know or believe he has it without
either outside help (such as that of a
or
some
friend)
special self-scrutiny.12 Note that this does not preclude the
itself in consciousness,
the prop?
e.g. in S" s entertaining
or his drawing certain inferences from p. To be sure,
in which (1) but not (2) holds, e.g.
there is a related use of 'unconscious'

beliefs manifesting
osition in question

one has been unconsciously

in a debate. Here
assuming something
one will normally
realize one believes the proposition,
upon clearly for?
course
cases
like this shade off into the kind I am
it; but of
mulating
where

I might add, is tied to Freudian psychology,

notion,
describing. Neither
to
I
be
take
the
latter
similar to some Freudian concepts of uncon?
though
scious belief.
Should we allow, as a Freudian might, that S can unconsciously know or
believe he believes, /?, where the latter belief is unconscious?
Perhaps this

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

138

ROBERT AUDI

be ruled out a priori. If not, we need a way to describe such uncon?

beliefs so that we can protect the above characteriza?
scious second-order
that
tion of unconscious
beliefs from circularity. We must not say merely

cannot

are those S does not know or believe he has, except

or unconscious
unconscious
belief. We might
knowledge
perhaps through
thus say something
like this: An unconscious
belief that /? is a belief such
that, apart from outside help13 or some special self-scrutiny, S is disposed
(1) sincerely to deny that he believes/? if asked whether he does, and (2) to
unconscious

beliefs

explain typical manifestations

havior as due to some factor

of this belief
other

in his consciousness

than the belief.

The

notion

or be?
of self

is vague, but must not be taken too strictly. We need not

scrutiny
construe unconscious
beliefs as so buried that S cannot bring them to
and fantasy; and people
consciousness
by, e.g., analyzing his behavior
here

this.
scrutiny they need to accomplish
that the want(s) explaining why S"s belief that p is
be, however,
unconscious
would prevent S's coming to know or believe, even uncon?
believes that he
that
he believes /?. But if S only unconsciously
sciously,
in how much

differ

It may

believes p, perhaps
him

all

the

"protection"

he need not "face" believing

he

needs

against

consciousness

it, and this may

of his

belief

provide
that/?.

I am not sure this question can be settled by reflection alone. Fortunately,

my purposes here do not require resolving it.We can simply work with the
belief just suggested.
of unconscious
wider characterization
It does not appear to me that 5"s being in self-deception with respect to p
- albeit con?
and unconsciously
believing not-p entails his also believing
that/?. All my view requires regarding 5"s positive attitude toward
sciously
is
S
that
be disposed sincerely to avow it. Since S"s sincerely avowing/?
is
p
generally excellent reason to think he believes it, we may want to say, in
believes" it. Should
virtue of his sincerely avowing it, that he "consciously
we say, however,
that he actually does believe /? when he unconsciously
or believes not-pl I think not.14 In this respect, my account differs
from those requiring beliefs that p and that not-p.
Let me offer several reasons in defense of the view that self-deception
does not embody beliefs that p and not-/?, beginning with some negative
knows

(i)We surely need not take sincere avowal to entail belief.15 (ii)
for one to have beliefs of incompatible
it seems possible
propo?
one
one
two
to
not
believe
for
it
is
sitions,
propositions
clearly possible
are
are
that
normal
and
such
believes
person
any
not-p
incompatible. Yet/?

points,
While

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

SELF-DECEPTION,

ACTION,

AND WILL

139

them incompatible,
(iii) Even supposing that in self-decep?
tion S could believe, e.g., both that he will die of his disease and that he will
recover from it, it is preferable to explain the data - as I shall argue we can
- on a
hypothesis not implying this irrational and, I suspect, inexplicable
condition,
(iv) The pattern of behavior and thought required to warrant

would

believe

to him a belief
that /?, counts against attributing
even if one of the beliefs is unconscious;
if, e.g., we may at?
to Ann the unconscious
belief that she will die of cancer
say

attributing
that not-p,
tribute

to S a belief

she asks about funeral arrangements

"just in case," rewrites her
in
her
do not pull through
and
that
condition
will,
grasps
normally people
- can we also find
adequate grounds to attribute to her the belief that she
because

will pull through? She says this, but "actions speak louder than words,"
and her overall behavior will not support the attribution of this belief nor,
I think, adequately explain her holding obviously
incompatible beliefs. It is
not as though she had the sorts of grounds for both that one might have
for incompatible beliefs upon discovering Russell's Paradox.
That a person self-deceived with respect to /? does not believe it, is also
The behavior that seems to
supported by a different kind of consideration.
a
to
of
is
such
belief
S
typically due to the wants, needs,
justify attribution
etc.

belief is unconscious
in the first
that explain why the unconscious
cancer
can
to
not
Ann's
of
die
Thus,
place.
wanting
explain her only
can
that
and
also
she
is
of
it
realizing
dying
explain why she
unconsciously
talks with

others about what

a belief

suggests
to the deluded

that she will

Such talk superficially

as opposed
in the self-deceived

she will do when well.

recover. But

in Ann's position there will be contrary indications,

to act on the ostensible belief, say by making vacation
e.g. little willingness
plans that are not easily undone.
it is true that if S is self-deceived with respect to /?,
Supposing, however,
he knows not-p, and does not also believe p, may we really say he sincerely
person

avows pi I think we may, though my account would be largely unchanged

ifwe said only that (e.g.) S non-lyingly
avows/?. Consider Ann's saying she
will recover. This statement differs in several respects from an insincere
one. (a) She apparently
lacks a belief that she is speaking falsely. With an
belief
there is normally a quite conscious
statement, however,
that one is speaking falsely, (b) Similar points hold regarding an intention,
the truth from, the hearer(s):
hope, or effort to deceive, or to withhold
In either case we have a
are
or
unconscious.
these
either she lacks
they
insincere

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

140

ROBERT AUDI

contrast

with

normal

insincere

(c) Ann is not appropriately

friends she trusts, that she will not
tend?
it is outweighed
by conflicting

statements,

to affirm to herself or intimate

disposed
recover.

If she has

this disposition,
this count, too, her statement differs from an insincere one. (d)
like 'You don't believe that', she is
If her hearer responds with something
to
think
she
is
being falsely accused of lying, deceitful
strongly disposed

encies. On

This does not apply to normal

ness, pretense, or perhaps self-deception.
we expect such a state?
cases of insincere statements. With self-deception,
ment
with
more

whereas
sensibilities and honest protestations,
wounded
a
we
of
decent
normal
expect (in
person) signs
guilt, and
insincerity
more
a
Far
about
could
said
be
if
there is defensive reply.
insincerity
to produce

these points at least show that the notion may be

to
the
range of avowals I have in mind.
plausibly applied
To be sure, self-deception may pass over into simple delusion, and the
these things; and
transition may be gradual. But we need to distinguish
a
we
at
which
it
is
clear that S con?
reach
at
when
least,
point
normally,
to genuine
sciously believes /?, he has, I think, passed from self-deception
delusion and no longer believes that not-p. There is a natural tension here:
sincerity,

but I believe

against p pulls one away from the deception and

that
and weakness
to
one's
down
break
threatens
defenses; the motivation
on
the evidence and threaten
sustain the deception pull against one's grip
of the truth. Self-deception
to overthrow one's perception
resides, I sug?
the sense of the evidence

gest, only where neither of these forces is paramount. Thus, if S"s behavior
and discourse are sufficiently free of conflict to justify saying he believes /?,
then at the time in question we cannot reasonably
reject his sincere dis?
in attribut?
a
not
be
warranted
We
then
would
avowal of belief that non-/?.
or
to S; for if S does not know
truly believe not-p, he
ing self-deception
cannot be deceived

by himself with
so that moments

the evidence may

respect to /?.Granted
convinced
S
after we are justifiably

rapidly,
simply believes/? we can justifiably conclude that he unconsciously
not-p. But this shows that self-deception must be distinguished
change

tain oscillations
believe

in belief,

not

that in self-deception

S may

believes
from cer?

believe p and

not-p.

is right, however, how can we do justice to

If this view of self-deception
cases
of two-person deception? There one person
the analogy with typical
knows that not-/?; the other believes that /?. So shouldn't the self-deceiver
believe both p and not-/?? After all, itmight be thought, just as we need to

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

ACTION,

SELF-DECEPTION,

AND

141

WILL

suppose S knows not-p to account for how he can be said to deceive

himself about /?, we must suppose he believes p to account for his being
deceived about it.
any more
My reply is that the analogy simply does not hold completely,
than does the analogy between self-denial and denying something to anoth?
someone else. But why should
er, or between self-control and controlling
it is as if we had two; for S
it?We have one person, not two. However,
level, at which he
operates, as it were, at two levels. On the unconscious
from which he ma?
/? is false, or at least at a kind of metalevel
or
own
to "believe" /?,
himself
he
consciousness
his
behavior,
gets
nipulates
whether he aims at precisely that or not; and since he sincerely avows/? and
knows

is not conscious that he believes not-p, it looks quite as if he did believe/?. I

to account for the term
think we thus preserve enough of the analogy
we
avoid
the
in
both
cases;
yet
paradox implied in saying the
'deception'
is not true. My
self-deceiver believes something he knows (or believes)
point is not that such a belief is impossible; but how it is possible, if it is, is
without
such
not clear, and we can account for self-deception
positing
beliefs.
If
The analogy with ordinary deception
suggests a further problem.
someone

gets
falsely but quite justifiably believes not-p and intentionally
in some sense de?
another person to believe p, the former has apparently
belief
ceived the latter. Now suppose S has a justified but false unconscious
cases of self
else goes as in paradigm
that not-p, and that everything
I do not
so that S sincerely avows /?, etc. Is this self-deception?
deception,
with respect to p, which is the sort that concerns
think it is self-deception
me; for if/? is true, S is not deceived in believing it. But S is indeed being
deceiving toward himself. Perhaps what we have is a kind of abortive self
The kinship to ordinary
deceptiveness.
I would not call the case self-deception

is obvious, and while

self-deception
in the light of
it can be understood
a
to
with
respect
self-deception
proposition.

the points made above about

One may yet object that knowledge
can be. But why not? Not because

cannot be unconscious

even if belief

that one
entails knowing
knowing
even
an
not
for
first-order
Not
for
that
is
because
so,
knows;
knowledge.16
unconscious
belief cannot be justified. For it surely can be.17 If one takes a
one may readily countenance
unconscious
reliability view of knowledge,18
is not required for a belief to have the rele?
knowledge. For consciousness
vant kind of reliable connection

with

the fact in virtue of which

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

it is true. An

142

ROBERT AUDI

unconscious
belief might, e.g., arise from excellent evidence which S simply
does not consciously
register as evidence for /?.
This is a good place to emphasize
that I have been giving an account of
I have already suggested that there
not
of
oneself.
self-deception,
deceiving
appears to be, for normal persons, at least, no act of deceiving oneself, if
at a stroke, e.g. by a resolu?
that entails putting oneself into self-deception
tion to enter it. There are, to be sure, various kinds of behavior to which
'deceiving oneself applies, and I shall consider some shortly. Even if there
are no acts of deceiving oneself at a stroke, itmay be that in fact persons
acts, such as concentrat?
get into self-deception
only through self-deceptive
on
not
one-sided
evidence.
But
this
is
ing
required by the concept of self
I
not
and
have
therefore
built
it
into my account. The account
deception,
it easy to see why this should be so; for given the
does, however, make
we would expect such be?
want which partly causes the self-deception,
on one-sided, welcome
havior as concentrating
evidence. But suppose a
to
conditions
the account simply by
S
the
of
caused
neurosurgeon
satisfy
Would we then have a state of self-deception without
brain manipulation.
I think so; but the case is problematic,
in
5" s having deceived himself?
to call a person brave who (after,
much the way it would be problematic
say, a religious

conversion)

is ready

to act bravely

yet has never

been

tested.

In closing this section, letme make explicit some of the

I have examined, (a) The main subject of analysis has been
with respect to a proposition,
in self-deception
/?. From
acts
of
(b)
(c)
deceiving
being
self-deception,
tinguished,

central concepts
a person's being
this I have dis?
toward

oneself,
not
(but
self-deceived)
as a result of self-deception,
i.e.

deceived

i.e., merely being

(d) simple deception,
with respect to /?, and (e) being deluded
being deceived in believing /?, as a result of having been in self-deception
are not acts of putting oneself into
with respect to it. Acts of self-deception
a
are
acts
to
at
it or conducive
stroke; they
manifesting
self-deception
or
certain kinds of evidence out of mind,
one
to
true.
knows
be
The
unconsciously
something
sincerely denying
itself is
existence of such acts does not, of course, imply that self-deception
more
acts
than
ever an act. They do not constitute
of
any
self-deception,
love. As to being deceiving toward oneself, this is the sort of
love constitute
behavior by which one gets into self-deception,
though it does not entail

producing

that;

a person

it, such as putting

may,

e.g.,

manipulate

evidence

in a way

characteristic

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

of

the

SELF-DECEPTION,

ACTION,

AND WILL

143

self-deceiver, but in the end sincerely avow what is true and unconsciously
believe its negation. By contrast, the person who is simply deceived with
respect to p actually (and falsely) believes it. Since one can cause this in
oneself, e.g. by too hastily judging a complex matter, we must be careful to
from simple deception
that is self-caused. Fi?
distinguish
self-deception
- inwhich
can
one
to
since
from
pass
nally,
self-deception
genuine delusion
one is simply deceived in believing/?, without
the ambivalence,
oscillation,
- we
or the like characteristic
of self-deception
anxiety, self-manipulation,
must distinguish being self-deceived from being simply deceived as a result
of being self-deceiving
offered in this section

The account of self-deception

(or of self-deception).
us
meant
to
is
these
preserve and understand
help

distinctions.

II. SELF-DECEPTION

AND THE WILL

as a kind of
talked as if they construed self-deception
behavior,19 even if not as an action; and self-deception has frequently been
Is it in a sense voluntary? And if it is
regarded as in some way motivated.20
not voluntary,
under the person's control? These two
is it nevertheless
Some writers

have

questions will be the main concern of this section.

We might first ask whether self-deception
is in any sense something
does. If one notes that we say such things as 'Don't deceive yourself,
may

suppose

that

in at

least

some

cases

self-deception

is an

act.

But

we

one
one
can

'deceive your?
explain such apparently behavioral uses of 'deceive oneself,
is action or behavior.
self, and the like without supposing that self-deception
Note that we also say such things as 'Don't believe that' and 'All right, I'll
In
believe you', where clearly no action of believing need be postulated.
both sorts of context, there may be actions the addressee can appropriately
the evidence or reminding himself of his strong
perform, e.g., reconsidering
nor even necessarily
biases. But these actions neither are self-deception
manifestations

of it.

The analogy with belief indicates why some of the points most likely to
is a kind of behavior do not establish
appear to show that self-deception
that. But we still need to explain the appropriateness
of such imperatival
as 'Don't deceive yourself
and 'Stop deceiving
yourself. We
are
in
to
that
effect
reconsider
the evidence,
suppose
might
they
injunctions
face one's biases, or the like. But since it is generally realized that such acts
locutions

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

144

ROBERT AUDI

this view does not do justice

may not prevent or eliminate self-deception,
to an apparent presupposition
of the relevant locutions, namely that in
some sense self-deception
is voluntary. What could this mean? It does not
of course, that there is a sub-personal
mean,
agent, the will, which de?
termines whether S enters or remains in self-deception. Nor need itmean
in the way actions are. We say of certain
is voluntary
that self-deception
such as absences, that they are voluntary, where the force of
non-actions,
this is roughly that they are in some appropriate way a result of the agent's
action and in that sense under the control of the will. Let us
voluntary
pursue this notion of control in relation to self-deception.
here. First, we must
It will help us to draw two sets of distinctions
a terminology
some
in
what
is
between
(or,
directly voluntary
distinguish
prefer, under the direct control of the will) and what is
It seems plausible to say that only actions are directly
indirectly voluntary.
on
one conception
of action, this is a truism. There are
Indeed,
voluntary.

philosophers

may

several ways we might unpack the notion that an action is directly volun?
tary. On one view, to say this is to say that S can, under favorable con?
ditions (e.g. non-paralysis)
just perform it, i.e., perform itwithout doing so
some
other action. I have inmind, of course, what are called
by performing
basic actions; and I am here characterizing
directly voluntary actions in a
that is neutral on the question whether their occurrence requires some?
I shall not adopt), one
(in a terminology
thing like a volition. Alternatively
are
if
and
actions
that
only if one can, under
directly voluntary
might say
at
where
this implies that one
them
favorable conditions,
will,
perform
way

need not have any belief to the effect that one performs them only through
one's finger at will, but
else. Thus, one can normally move
something
one
a
not
like an
one
did
have something
if
not
switch
could
flip
normally
instrumental belief, e.g., that one can do so through moving one's finger.21
We might

characterize

consisting
favorable

of actions,

is only indirectly voluntary,

by contrast, as
one
can, under
events, and states of affairs which
some basic
or
about
performing
by
bring
perform
what

conditions,
and
action,
only in that way.
A second set of distinctions we need cuts across the first set. Control may
be negative or positive. Let us say that a state of affairs (or type of state of
con?
affairs) is under 5"s negative control provided S can, under favorable
a
we
Let
further
distinction.
find
With
it.
control,
ditions, prevent
positive
us say that S has partial positive

control

over a state of affairs provided

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

he

SELF-DECEPTION,

ACTION,

145

AND WILL

either bring it about or, if it obtains, alter

can, under favorable conditions,
can
not
If
but
both.
S
do
both, then (and only then) he has full control
it,
over the state of affairs. Clearly, control of each of these three kinds has
so that even full control

degrees,

the less S needs

things equal,
is his control

greater
Does

need not be complete. For instance, other

to do to bring about a state of affairs, the

over it.

in any of the above respects, have control over one's being in

self-deception? My first point here is that there may be no general answer
to this implicit in the concept of self-deception. The question may be essen?
one,

account of self-deception
and if so a philosophical
should
answer
we
it. To be sure, when
say such things
by itself, to
as 'Don't deceive yourself we appear to be conceiving
as
self-deception

tially empirical,
not be expected,

under S's negative control. But surely such locutions apply to self-decep?
If S has control over unconscious
tive acts, not states of self-deception.
can
intro?
beliefs
e.g.,
always render them conscious by an appropriate
then my account of self-deception
spective inventory of what he believes
For in prin?
imply at least partial positive control of self-deception.
rid
it
S
could
himself
of
the
relevant
belief
conscious.
ciple
by rendering
belief.
Perhaps something like this is implicit in the concept of unconscious
But rather than explore that possibility
directly I want to suggest some

would

in which

respects

self-deception

itself, as I conceive

it, is under the subject's

control.

It is at best doubtful
directly

voluntary,

that there is anyone

anyone

who

can

become

for whom
self-deceived,

self-deception
or

is

unself-de

one cannot

intelligibly will, intend, or

self-deceived
want, to become one of these. But surely becoming
cannot,
for any normal person, be achieved like raising an arm. On the other hand,
each of the kinds of indirect control sketched above is possessed by some
ceived,

at will. This

people

for some kinds

becoming

is not because

self-deceived

themselves
reminding
about another person's

of self-deception.
Some people can prevent their
about their chances of winning a contest by simply
of the evidence.
Some can become
self-deceived

affection for them by dwelling on the evidence for

view and constantly
evidence for
avoiding or disregarding22
can
some
the dreaded view. And
rid themselves of self-deception
people
about their achievements when they look carefully at their careers to come
the favored

themselves simply as the individuals they are.

In none of these cases need S aim at avoiding, getting into, or getting out

to terms with

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

146

ROBERT AUDI

For instance, whether or not one suspects that one is

of, self-deception.
one might
set out to reassess the
about one's achievements,
self-deceived
evidence with a view to being honest with oneself. An autobiography might
to root out self-deception.
also provide an occasion
Clearly people will
in the kind and degree of control they have over the
differ considerably
sorts of self-deception
to which they are liable. Joan's self-deception may
be readily uprooted by her honest reappraisals, while John's, tottering ego
it hard for him even to begin self-scrutiny.
makes
are other ways self-deception
for instance
may be eliminated,
and consciously,
the prop?
through S's gradually embracing, spontaneously
or through his forget?
osition he previously believed only unconsciously,
There

say as changes in him make it irrelevant to

altogether,
There are also factors other than the depth of the
in?
unconscious
belief (i.e. how deeply repressed or otherwise

ting that proposition

his main concerns.
constituent
sulated
tion

it is) which affect how entrenched

self-decep?
to
the strength of S"s motivation
is perhaps
this
where
conflicts
with
the
conception
self-conception,

from consciousness

is. The most

maintain

notable

a certain

evidence supporting the constituent belief. One merit of viewing self-decep?

is that if it does embody unconscious
tion as I propose
belief, we can use
our theory of unconscious
belief to account, in part, both for the entrench?
ment of self-deception
and for the various ways S can uproot it, e.g. by
careful

self-observation.

are in some
it is clear that at least many instances of self-deception
to
it is natural
ask whether self-deception may be
way indirectly voluntary,
of
of will. One might think that weakness
construed as a kind of weakness
Once

therefore is not a
only in actions and that self-deception
of will (for
for it. But surely intentions may manifest weakness
reasons I have given elsewhere).23 Perhaps,
we
also
then,
may
speak of
ac?
with incontinent
it. These might be, analogously
beliefs manifesting

will

ismanifested

candidate

beliefs against S's better judgment; roughly, S will hold

tions, uncompelled
to hold
them - without
(say, by psychosurgery)
having been compelled
them - despite his judging that, on the basis of the evidence he has, the
in question
is false (or improbable). Here we must be very
proposition
if we speak of a belief which is against one's better judg?
careful, however:
we
not
must
ment,
imply that normally beliefs are positively and directly
voluntary.

They

and certainly

are,

to be

such actions

sure,

sometimes

as dwelling

under

on one-sided

one's

negative

evidence

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

control;

are directly

SELF-DECEPTION,

ACTION,

AND WILL

147

voluntary. When S forms a belief against his better judgment, however,

where it is under his negative control, or when he forms a belief by con?
evidence, the belief he forms
centrating on what he realizes is one-sided
- or
a
sense:
in
be
akratic
derivative
it
arises
may
persists
through the
or permission of a weak will.
But must self-deception
embody or arise from such beliefs, or even from
beliefs formed through similar self-manipulation
but in the absence of a
one
not
that
to
do
the
the manipu?
judgment
ought
things constituting
lation? At this point we can see a major difference between my account of
action

and inconsistent-belief
accounts.
Since I maintain
that S
self-deception
does not believe/?,
i.e., the proposition which, on some accounts, he con?
sciously believes or believes to some degree,241 need not take self-deception
to embody akratic belief. This avoids a difficulty in accounting
lor cases
in which S
(which some writers have held to be central in self-deception),
believes
in the face of what he sees as preponderant
counterevidence.
Inconsistent belief accounts, as well as those which simply maintain
that S
believes p, must explain how, despite his awareness of evidence against p
which S regards as superior to his grounds for it, S comes to believe, or
continues to believe, /?. Such belief is by no means clearly impossible; but if
we can understand

self-deception

without

positing

it, that is surely pref?

erable.

On the other hand, if self-deception

need not embody akratic belief, it
may often arise in part from, and then later cause, akratic acts of a not
unusual sort. How are we to explain this? There are many ways weakness
of will can contribute to producing
In each case, there is an
self-deception.
action, intention, want, or perhaps belief, formed or maintained
against
one's better judgment, or afailure to perform an action, or form an inten?
tion,

want,

or

belief,

when

one's

judgment

requires

it. Here

are

some

ex?

his better judgment regarding what he ought to do, S may

evidence against
against p out of his mind, (b) manipulate
so
as
to
reduce
its
evidence
p
apparent plausibility,
(c) manipulate
for/? so
as to increase its plausibility,
new
seek
evidence for/? knowing that he is
(d)
biased toward inflating the evidence, and (e) do things - such as rejecting
advice that one write a will - appropriate only on the belief that/? (say, that
one will recover) is true. Similarly, S might exhibit weakness
of will in
the
or
whether
not
he carries them out.
forming
intentions,
corresponding
amples. Against
(a) put evidence

It will be clear that these sorts of weakness

can also result partly from self

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

148

ROBERT AUDI

actions - such as putting

deception, most commonly where it motivates
which help to perpetuate
it.
evidence out of mind
can manifest weak?
These and related examples show that self-deception
ness of will

in a variety of ways. If so, we can see more clearly how self

to
is often under 5"s control; for if S exercises enough will-power

deception
avoid the sorts of things just described, he is less likely to get into self
deception, and less likely to stay self-deceived when he has gotten into it. In
a way, then, volitional elements are essential to understanding
self-decep?
tion. But it is not itself a kind of action, any more than the deception we
is an action. We do, however, become
or
largely through actions, and act as a result of being
partly
are
accounts
These
among the truths that make volitional
in others we deceive

produce
deceived
deceived.

self
self
ap?

pealing.
A major

of these reflections is this: once we see how self

consequence
of will and may be under S" s control, we
weakness
involve
deception may
can begin to see how to assess self-deceivers.
Indeed, we are now in a good
a
is
to formulate
number of questions about how self-deception
position
related
the

to moral

next

III.

Given

responsibility.

Some of these questions

are the concern

of

section.

AND

SELF-DECEPTION

the results

of Sections

I and

about moral

MORAL

RESPONSIBILITY

II, we can deal with a number of

I shall consider
and self-deception.

responsibility
questions
for
three sorts of problems under this heading:
(1) moral
responsibility
for
in
and
moral
into
it;
remaining
responsibility
(2)
self-deception;
getting
for acting out of self-deception.
(3) moral responsibility
there are many
one's getting into self-deception,
important
Regarding
that not-p, or does
variables. Does S initially know or believe consciously
he simply act so that he comes to know or believe not-p only uncon?
sciously? Does S realize, or may he be reasonably expected to realize, that
evidence, e.g. putting evidence against/? out of mind? Is
in which S does or should see that much depends on
whether or not /? is true? How reasonably does S respond to others who
regarding /?, or try to show him that he is un?
give credible testimony
he ismanipulating
the situation one

reasonable
To

about

illustrate,

it?
consider

Tom,

who

is self-deceived

with

respect

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

to the

SELF-DECEPTION,

ACTION,

AND WILL

149

of a paper of his. If he intially believes that it is mediocre,

yet
to repress this belief by such things as manipulating
the evidence,
manages
for getting
then other things equal he bears greater moral responsibility
in which he sincerely avows that the paper is
into a state of self-deception
merits

For one thing,

first-rate, than if he never consciously believed itmediocre.
a person generally has to do more to repress a belief than to acquire one as
and have it remain so; and some of the things one may need to
unconscious
do are reprehensible.
This takes us directly to a second point. Whether
Tom's belief that the
or not, the more he manipulates
is initially conscious
paper is mediocre
or
evidence to repress it, or to keep the proposition
from consciousness,
seem
more
to
the
the
make
he
other
first-rate,
is,
paper
just
reprehensible
still if he knows, or can be rea?
things equal. He is more reprehensible
the relevant evidence.
sonably expected to know, that he ismanipulating
On the other hand, if the thought that the paper ismediocre
is truly dam?
we
an
to
his
other
would
have
ego, then,
aging
things equal,
extenuating
Indeed, perhaps the thought or belief could be so damaging
to Tom that at least for a time it would be reasonable for him to try to
deceive himself. Itmight be different if someone's well being depended on
to stay with his family only if his
his assessment, e.g. if he were determined
circumstance.

talent did not warrant

his leaving to develop it. It also matters how he deals

with advice. Does he deceive himself uninfluenced
by credible attempts to
help him see his errors, or do such attempts at least delay him or lead him
to find further evidence for/? before letting himself espouse it as true? The
latter is preferable,
prepared to believe

things equal; for it indicates that Tom is not

the paper is first-rate without having reasons for the
belief. This, in turn, suggests that his self-deception
ismore nearly rational
than it might have been, where the degree to which self-deception
ap?
as
is
determined
such
how
reasonable
proaches rationality
by
things
nearly
itwould

other

be for S to believe

the sorts of efforts he makes

/?, how much he is prepared to stake on /?, and

to see that he has evidence for p.

Similar points

hold for one's remaining in self-deception.

For instance,
itwould be to S to be conscious that not-p, the less rep?
damaging
rehensible it is for him to be in self-deception with respect top, other things
equal. A related factor is the degree to which he takes evidence into ac?
the more

count. At one extreme

sidering,

is indifference. An

but readily manipulating,

intermediate

the evidence

case would

he encounters

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

be con?

against

/?.

150
Near

ROBERT AUDI
the other

extreme

is S's

evidence

so seriously

that finding
/?weakens his self-deception.
There are, however, some important differences between moral respon?
and moral responsibility
for remain?
sibility for getting into self-deception
one
someone
can
who is not in self-deception
ing in it. For
thing, whereas

evidence

taking

against

intentionally act in order to get into it, or to avoid it, someone who is in it
can rarely if ever intentionally act in order to remain in it, or to get out of
it. He can do things with these effects, e.g., keep his mind off evidence
to act intentionally
in order to remain in, or to get out of,
S would have to want to remain in it, or to get out of it, and
self-deception
to believe certain actions would (ormight) contribute to this. That seems at

against

/?. But

to be consistent with 5"s being in self-deception,

best unlikely;
it would
that he is in it and his unconsciously
require his unconsciously
believing
to remain so or to cease to be in it.We may talk this way; but the
wanting
facts admit of a simpler explanation,
and we should avoid postulating
any
unconscious

elements

not required

to explain
difference

the data.

The second morally

between getting into self
noteworthy
and remaining in it is less problematic. Getting
into it requires
deception
that S undergo change and, typically, that he do a variety of things. But 5"s
does not entail his changing. For instance, no
remaining in self-deception
to act on it, need arise. Here he
evidence regarding /?, nor any occasion
for
bear
little
might
responsibility
remaining self-deceived, perhaps none if
to review critically, from time to time,
people do not have an obligation
certain of their beliefs, or what they take to be their beliefs. (This is an
obligation which S would fail to fulfill if his apparent belief that/? is of the
sort which he ought to have reviewed.) By contrast, there is at least typi?
even if no blameworthiness,
for getting
cally some responsibility,
some
sense one's cognitive system is not adequately
in
deception:
sive to the evidence.

into self
respon?

Let us now consider moral

for acting out of self-deception.

responsibility
To begin with, self-deception may
of distinctions.
intentional and non-intentional
actions. For the latter, S

Here we need a number

bring about both

or may not bear moral

may

the seriousness

responsibility,
of the actions and

on, among other

depending
the degree to which S could
control over behavior of the

things,
reasonably have been expected to exercise
kind in question under the circumstances. With
tions arising from self-deception,

respect to intentional ac?

there are at least three sorts: those stem

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

ACTION,

SELF-DECEPTION,
from

the unconscious

AND WILL

151

belief

component
(typically
representing
from the "conscious belief
those arising ostensibly
compo?
knowledge);
nent; and those emerging from some unconscious
want(s) connected with
the self-deception.
Let me take these up in turn.
knows his paper is not first-rate,
that
Tom, who unconsciously
Suppose

ming

that it is not, by declining an offer to publish it. He might

give as a reason that he is a perfectionist, which, let us suppose, he is. This
trait might explain why he wants to publish only first-rate work; and that,
acts on the belief

belief, would
together with his unconscious
seems
Here
action
the offer.
the
rational;

largely explain why he declines

and if it is, this shows that an
in a common way by) an unconscious

arising from (and motivated

state can be rational.25 To be sure, this is not a case one wants to
call acting out of self-deception.
But the action does arise from self-decep?

action
mental

tion because

of the way

by a crucial element in that decep?

want is
and the motivating
knowledge,

it ismotivated

tion. Since that element represents

rational, it is not surprising that an action arising from this element is
rational. The point is important because there has been too little appreci?
involves reason and can as a
ation of the extent to which self-deception
result play a role in rational action.
An

action arising

tion may,

from unconscious
be neither

rational

in self-decep?
embodied
knowledge
nor such that *Sbears moral
re?

however,
sponsibility for it. Imagine that Ann foolishly believes her family plans to
cremate her and cannot bear to entertain this thought. Her unconscious
that she will soon die may then lead her to write her attorney
knowledge

because she should

insisting on ordinary burial. This could be unreasonable
see that her family plans to respect her wishes and that they will be hurt by
the act; yet she might not be morally
responsible for it because she does it
under a strong morbid compulsion.
It is actions

belief
element in
ostensibly
arising from the "conscious
or arising from an unconscious
want connected with the
I say
that we think of as actions "out of
self-deception,
self-deception.
I
from
the
'conscious
of
think
because
course,
belief,"
"ostensibly arising
a
there is no such belief in self-deception.
There normally
is, however,
self-deception,

conscious

want

regarding the proposition

sincerely avowed
though not
to
act
in
accordance
with
and
there
it;
may be one or
(/?), namely,
conscious wants operating,
such as a want to put evidence against /?

believed
more

out of mind.

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

152

ROBERT AUDI
first cases in which

S ostensibly acts on the sincere avowal of/?

to act accordingly.
cancer patient may do
The self-deceived
this by talking of plans she has for the distant future, by explaining away
evidence against her recovering, and in other ways. There is no reason why
Consider

and the want

she cannot be morally

for such things. (Some, of course, may
responsible
be quite harmless.)
It appears, however,
that these actions' arising from
an
is
the degree of extenuation
circumstance,
self-deception
extenuating
on
the situation. It is not that S cannot help himself. Rather,
depending
because

he is not conscious

of the truth of the situation

and is at least
errors
his
are, on the first
belief,
acting
on
an
or
based
at
least
count,
abnormal,
partly
potentially
debilitating,
condition,
and, on the second, based partly on a prima facie commedable
desire to act consistently with one's avowed beliefs. The extenuation would
in accordance

with

his avowed

if S" s remaining
in self
however,
or
If
S
deception
highly reprehensible.
got
intentionally
knowingly
himself into the self-deception,
this would also seem to reduce the degree of
be diminished

and perhaps

eliminated,

were

other things equal.

extenuation,
We come now to actions out of self-deception
an

unconscious

want

of

5"s.

The

writer

we

which

imagined

arise mainly

might

from

unconsciously

want

to put out of mind evidence that his paper is not first-rate. In some
people, of course, such a want might be quite conscious. But the kind of
person we are imagining could not face having such a want. He might
consciously want to refute claims he sees as evidence against his statement
that his paper is first-rate; but that is a quite different want. Suppose, then,
that S does avert his gaze from evidence against his statement, e.g. by leaving

a discussion,
to get fresh air, when his paper is under scrutiny.
ostensibly
responsible for doing this, even though he does
Clearly, S may be morally
not (at least consciously)
know why he is doing it. He is not compelled; and
to consider evidence re?
his self-deception
does not cancel his obligation
his
The
views.
is
For
it
illustrates
that S may be
garding
point
important.
even
to him, he
if, unbeknownst
morally
responsible for doing something
want. I believe that a good
is doing it in order to satisfy an unconscious
have tended to assume the opposite.26 That is certainly en?
much
of what Freud has said, though there are indications
couraged by
the metaphor
that he would reject the assumption.27
Consider
of the mind

many

people

as an iceberg nine-tenths of which is below the surface. This suggests that a

and that we lack
great deal of what determines our behavior is unconscious
to control much of it.
the self-knowledge

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

SELF-DECEPTION,
In response

to this I want

ACTION,

to make

AND WILL

153

a point crucial to assessing moral

about by unconscious
factors. One

for behavior brought

responsibility
does not have to know or even suspect one has a desire in order to feel
oneself impelled or drawn toward the actions itmotivates.
One may there?
fore try to resist these pushes and pulls. This applies to both conscious and
unconscious

desires

from unconscious

(and other motivators).

motives
need not be

inclinations arising
Moreover,
irresistible. That an inclination

them may often reduce, but it does not in general eliminate,

moral responsibility
for yielding to it.We must consider the strength of the
for example, as well as how much
unconscious motivation,
resistance S
arises from

might
matter

be reasonably
expected
of 5"s own psychology.

to muster.

This

It is also a matter

is not just a
expectation
of what a morally
sound

person may be expected to do in the circumstances.28

causes him to A
Let me carry this further. Suppose S's self-deception
he
be
still
for
non-intentionally.
Might
morally responsible
^4-ing? Imagine
Tom in a seminar, sitting next to a person setting out, from notes, argu?
ments

want to ignore evidence

against Tom's paper. Tom's unconscious
against his paper's being first-rate might agitate him. That in turn might
cause him to knock over his coffee, thereby soaking his colleague's papers.
for this, though the circumstances
are
Surely Tom ismorally
responsible

For we may still expect that he perceive his agitation

and
extenuating.
exercise additional care with his coffee. Indeed, even if he did not perceive
his agitation, he should exercise care. Not all accidents are excusable, and
elements need not excuse the agent.
by unconscious
a
not
is
state that simply comes upon one
morbid
then,
Self-deception,
like an illness.29 At least typically it is indirectly voluntary,
to some degree.
This applies particularly
to negative control, but S may also have con?
siderable positive control over his self-deception. Moreover,
self-deception
their causation

often occurs or remains only by virtue of a kind of voluntary complicity.

One may, through intentional actions, get oneself into it, whether one aims
at becoming
self-deceived or not. One may do things that keep one in self
aimed at maintaining
it.
deception, perhaps including things unconsciously
One may thus be morally
or for
responsible for getting into self-deception
remaining in it. One may also be morally
kinds of actions arising from it, though
for all of them.
tenuating circumstance
even when

the action

is performed

for any of the several

tends to be an ex?
self-deception
It may extenuate
little, however,
responsible

to satisfy an unconscious

want. We may

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

154

ROBERT AUDI
even if we cannot make

be quite responsible for doing something,

origin of the force inclining us to do it.

out the

IV. CONCLUSION
Recent

seen much concern, by both scholars and lay people,

should diminish an agent's legal responsibility.
Psycho?
have led the list of defense attorneys' favorite extenuators.

times have

about what

factors

logical factors
Since our discussion

of moral

and self-deception
has some
of legal responsibility
by psy?
to bring some of them out. Most

responsibility
for the extenuation

interesting implications
chological factors, I want, in concluding,
are not confined to self-deception.
I am not aware
of these implications
that it has been cited to reduce anyone's
legal responsibility,
except per?
a
as
or
of
of
mental
emotional
part
pattern
haps
problems. There is a
to construe unconscious
for defense attorneys
in?
tendency, however,
or even eliminating,
fluence on action as strongly extenuating,
the agent's
for it. Given
moral responsibility
this tendency, they naturally also take
such influence

to diminish

On this tendency
that

an

action's

his legal responsibility

our discussion has considerable

being

unconsciously

motivated,

for it.30
bearing.

It indicates

or

in

even

some

(1)

sense

caused by an unconscious
element, need not eliminate, and need not even
for it; and (2) how a
reduce, the agent's moral responsibility
substantially
in cases of
number of the variables determining
the degree of responsibility
are to be understood
and, in some sense, measured.
the degree of an agent's moral re?
the
variables
determining
Regarding
an
action influenced by unconscious motivation,
whether
it
sponsibility for
or not, we have identified at least five
is connected with self-deception
factors. First, clearly, other things equal, the stronger any irrational un?
self-deception

the
elements, the less the moral responsibility. Much
motivating
same applies, secondly,
to what we may call their depth, i.e., their "dis?
in terms of how difficult itwould
tance" from 5"s consciousness,
measured
that he has them. Third, other
be for S to come to know (consciously)
conscious

factor makes S ignorant of, or blind

things equal, the more an unconscious
to the importance of, facts relevant to morally
the action, the
evaluating
more
or pre?
the
deliberate
for it. Fourth,
less his moral
responsibility
for it, other
meditated
the action is, the greater 5"s moral
responsibility
things equal. Fifth,

the better S's opportunities

to disabuse

himself

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

of self

SELF-DECEPTION,

ACTION,

155

AND WILL

or actually getting help

for self-scrutiny,
e.g., opportunities
deception
from others in understanding
the relevant behavior - the greater his moral
or the unconscious moti?
for acting out of the self-deception
responsibility
other

vation,

things equal.
or unconscious mo?
similar points seem to hold for self-deception
in relation to legal responsibility
conceived from the moral point
of view.31 Apart from strict liability statues, liability to punishment
should
one would think, primarily on the basis of the degree of
be determined,
Quite
tivation

moral

for the action in question. Much more needs to be

responsibility
said, however, about what determines degrees of moral responsibility. The
one of fact-finding;
is not merely
and the same holds for de?
problem

In both cases, the

appropriate
degrees of legal responsibility.
a
can
of
what
sound person
be reasonably expected to do
question
morally
in the circumstances
is crucial for determining whether S fell short of his
moral responsibilities.
termining

If self-deception
is conceived as I have proposed,
and if it is indirectly
I
in
the
have
it
is
ways
voluntary
suggested,
altogether reasonable to expect
to
some
avoid it and to expect
to reduce or
self-deceivers
many people
eliminate their self-deception.
So far as this case is a fitting model for the
law, it appears

that recent

trends have

often

led us to go too far in ex?

tenuating legal responsibility.32

To summarize briefly, I have argued that self-deception with respect to /?
is a state in which S unconsciously
knows (or has some reason to believe,
and unconsciously
and truly believes)
that not-/?, sincerely avows, or is
so to avow, that /?, and has at least one want which
in part
the
that
is
belief
unconscious.
On
the
account
explains why
not-p
proposed
of self-deception,
it need not be paradoxical,
and it can be understood
as a
not unnatural response to the tension between certain of one's desires and
disposed

In
response to evidence that dims the prospect of their realization.
and the will, we have seen a
examining the relation between self-deception
can be indirectly voluntary.
In part
variety of ways in which self-deception
because it can be, agents can be morally
responsible for getting into self
for remaining in it, and for actions arising out of it.We have
deception,
one's

noted

some of the main

for

actions

variables

that determine

S's degree

of moral

from

re?
this

sponsibility
arising
self-deception.
Against
I argued that an action's arising from unconscious
moti?
background,
vation does not necessarily
the
eliminate, or even substantially
diminish,

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

156

ROBERT AUDI

for it, and that an action stemming from self

agent's moral responsibility
not
need
be
irrational.
Similar points apply to responsibility
deception
before the law. It is perhaps not surprising that an account of self-decep?
unconscious
knowledge central should have these impli?
entails justified belief, and if uncon?
For knowledge
presumably
scious factors can be justified, it becomes more natural to suppose that one
tion which makes

cations.

for acquiring or retaining them and for ac?

might be morally
responsible
tions stemming from them. The account of self-deception,
then, far from
the
and
moral
of the
apparent
diminishing
rationality
responsibility
human agent, encourages us to enlarge the domain of both.33
The University

of Nebraska,

Lincoln

NOTES
1

For a good discussion

of most
of what are perhaps
the major
to self
types of approach
A Study of Self-Deception
see M. R. Haight,
Press and Humanities
(Harvester
deception,
Canadian
Press, Sussex and Atlantic
Highlands,
1980), and Bela Szabados,
'Self-Deception',
4 (1974).
Journal of Philosophy
2
F. Gustafson,
and Donald
23 (1962
See, e.g., John V. Canfield
'Self-Deception',
Analysis
American
1 (1964);
'Pleasure and Falsity',
Penelhum,
63); Terence
Philosophical
Quarterly
and Jeffrey Foss,
'Rethinking
Self-Deception',
Sartre's view is not easily clarified. For good

American
17 (1980).
Philosophical
Quarterly
short discussions
of his views on bad faith see
'Bad Faith and Lying to Oneself,
Philosophy

E. Santoni,
op. cit., Ch. 5, and Ronald
Haight,
and Phenomenological
Research
38 (1978).
3
and Kegan
See Herbert
Paul, London,
Self-Deception
1969).
Fingarette,
(Routledge
4
see William
For
discussion
of various
of privileged
doctrines
P.
access,
illuminating
American
'Varieties of Privileged
8 (1971).
Access',
Alston,
Philosophical
Quarterly
5
In my papers,
'The Epistemic
of the First Person',
The Personalist
56 (1975),
Authority
and 'Epistemic Disavowals
and Self-Deception',
The Personalist
57 (1976). Some of the
are similar to views of D. W. Hamlyn
in those papers
in 'Self-Deception',
central views
Vol. 65 (1971), a paper I read after
Proceedings
Society,
of the Aristotelian
Supplementary
they were completed.
6
of the First Person',
'The Epistemic
p. 12.
Authority
7
Ibid., p. 13.
8
and Self-Deception'
I acknowleged
this use in 'Epistemic Disavowals
(p. 384) and have
related issues there.
discussed
9
in which one
As Szabados
argues, "One can be self-deceived
only about matters
plausibly
one can be self-deceived
link up with one's wants,
has a personal
stake. What
about must
needs". Op. cit., p. 67.
hopes, fears, and emotional
10
in 'Epistemic Disavowals
and Self-Deception',
This
is stressed
p. 382. That paper also
suggests

how wishful

thinking

can generate

self-deception.

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

SELF-DECEPTION,
11
My
believing
12
This

ACTION,

The Personalist
'The Concept
of Believing',
paper
on which beliefs can be unconscious.

Person',

is essentially
p. 10. The

here

is not

that of

taking it to express a conceptual

impossibility.
13
come from
The outside help would
normally
limited.

It could

be a story meant

to be about

53 (1972),

in 'The Epistemic

introduced

the formulation
'cannot'

AND WILL

logical

impossibility,

defends

157
an account

of

of the First
Authority
but I am tentatively

a person; but its sources are virtually

someone
else, but giving S the relevant

un?
self

understanding.
14
on this difficult
in n. 5 for some comments
issue.
See my papers mentioned
15
I have argued for it by implication
This has been argued by a number of writers.
in 'The
4 (1974).
Limits of Self-Knowledge',
Canadian
Journal of Philosophy
16
I have defended
of the First Person'.
this claim in 'The Epistemic
Authority
17
Ibid., pp. 11-12.
18
statements
For important
of this view, see D. M. Armstrong,
Belief, Truth and Knowledge
I. Goldman,
'Dis?
Press, Cambridge,
12-14; Alvin
1973), esp. Chs.
University
and Perceptual
Journal of Philosophy
73 (1976); and Fred Dretske,
Knowledge',
'Conclusive Reasons',
Journal of Philosophy
Australasian
51 (1973).
19
See, e.g., Fingarette,
op. cit., pp. 28-29, 62-63, 72, and 98-99; on 72 he goes so far as to
as he understands
"as disavowal"
it should
(but perhaps disavowal
speak of self-deception
(Cambridge
crimination

not be called

Lance
behavior).
though he does not

deception,

and the Functionalist

Deception
155. Cp. Michael W. Martin's

Factor

at one point speaks of calling

seem to conceive
it as action

certain

behavior

self

or behavior;
see 'Self
of Mental
The Personalist
Processes',
58, 2 (1977, p.
Theory
remark that "intentional
is both purposeful
and
self-deception
in general

in 'Factor's Functionalist
in knowingly",
Account
of Self-Deception',
The Personal?
engaged
ist 60, 3(1979),
p. 341.
20
note illustrate this, and I shall shortly indicate how self
The works
cited in the previous
can be motivated.
deception
21
reason for introducing
the second characterization
is to avoid relying on the
My main
are correct
controversial
notion of a basic action. Quite apart from whether my formulations
can be made either within a fine-grained
in detail, my distinctions
theory of act individuation
or within
the Anscombe-Davidson
discussion
of
coarse-grained
theory of it. For penetrating
both

see Hector-Neri
and Identity in Human Action
and
Casta?eda,
Tntentionality
Nous
and
Method',
13, 2 (1979). Alvin Goldman's
reply, 'Action, Causation,
in the same issue. For a view intermediate
between Goldman's
appears
fine-grained

theories

Philosophical
Unity',

and the coarse-grained

to recent literature on the topic,
references
one, with many
approach
see Lawrence Davis,
Theory of Action (Prentice-Hall,
Cliffs,
Englewood
1979), Ch. 2. Another
treatment of the topic is Irving Thalberg's
valuable
Emotion
and Action: A Com?
Perception,
ponent Approach
(Basil Blackwell,
Oxford,
1977).
22
a similar distinction
in connection
one between
Michael Martin makes
with self-deception,
and disregarding.
See op. cit., p. 341, and 'Ignoring and Self-Deception',
read at
distracting
the Western
Division
of the American
Association
in 1979.
Meetings
Philosophical
23
In my
of Will
and Practical
'Weakness
13, 2 (1979).
Judgment', Nous
24
See, e.g., Penelhum,
op. cit., p. 88; and Foss, op. cit., esp. p. 241.
25
I have argued for this in 'Psychonalytic
and the Concept
of Rational
Explanation
Action',
The Monist
56 (1972).
26
John Hospers
that
suggests
Sidney Hook
(ed.), Determinism
University
Hospers
a version

there

in 'What Means
this Freedom?',
in
tendency
in the Age of Modern
Science
(New York
Section
3. It is not clear to me
to what
extent

is this

and Freedom

Press, New York,

1957), esp.
to this tendency.
In this article
is sympathetic
of the view under discussion.

he appears

somewhat

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions

inclined

to hold

ROBERT AUDI

158
27

one must hold oneself

says at one point, e.g., "Obviously
for the Content
of one's dreams".
See 'Moral Responsibility

Freud

impulses

1957), p. 156.
Papers, Vol. V, (Basic Books, New York,
28
I have discussed
in detail the relation between
the notion
morally
'Moral

sound

person

Responsibility,

can

be expected

reasonably
and Compulsion',

Freedom,

to do

for the evil

responsible
of Dreams',
Collected

of compulsion
in the relevant

American

and that of what

circumstances,

in

Quarterly

11

Philosophical

(1974).

29

has no evil or sickly

self-deception
Nationalism:
in 'Philosophical

That

Farlow,

(1978).
30
For

e.g.

some detailed

number

of citations

discussion
of literature

essence

forcefully
argued
and Self-Direction',

by John King
Dialogue

17

and a
between
legal and moral
responsibility
S. Moore,
of
the topic, see Michael
'Legal Conceptions
Illness: Law and
and H. Tristram
(eds.), Mental
Engelhardt
of the relation
on

in B. A. Brody
Illness',
and Boston,
Public Policy
(D. Reidel, Dordrecht
31
and judicious
discussion
For a wide-ranging
before the law, see Stephen J. Morse;
sponsibility

Mental

has been

Self-Deception

1980).
of moral

responsibility
'Diminished
Capacity:
2 (1979).
and Psychiatry

and diminished
A Moral

re?

and Legal

Journal of Law
International
Conundrum',
32
inMoore,
and evaluated
Some of these trends are discussed
op. cit.
op. cit. See also Morse,
33
was
at
the University
of North
read
of
this
version
An
earlier
paper
on Action, Agency,
and at the University
of
and the Will,
Symposium
Carolina/Greensboro's
for helpful comments.
Alfred Mele,
I especially want to thank my Commentator,
W. Martin,
John Heil, Michael
from comments
benefited
by Robert M. Gordon,
and
Richard
Louis Pojman,
Norman
Malcolm,
Snyder, Bela Szabados,
Reilly, A. Aaron
Richmond.
I have

Irving Thalberg.

Manuscript

received

25 February,

1982

This content downloaded from 155.54.213.4 on Mon, 9 Feb 2015 11:39:34 AM

All use subject to JSTOR Terms and Conditions