MARK MERCER
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
ABSTRACT. Weak psychological egoism is the doctrine that anything an agent does
intentionally, that agent does at least expecting thereby to realize one of her self-regarding
ends. (Strong psychological egoism, by contrast, is the doctrine that agents act always
intending thereby to realize a self-regarding end.) Though weak psychological egoism is a
doctrine ultimately answerable to empirical evidence, we presently have excellent a priori
reasons for accepting it and attempting to construct psychological theories that include it as
an organizing principle. These reasons have mainly to do with the idea that to understand
the motivation behind an action, we need to understand the force of the consideration that
motivates the agent, and the way to do this is to find a self-regarding end associated in the
agent’s mind with acting on that consideration.
I intend in this paper to defend a particular position regarding the nature
of our springs of action, one I’ve not seen explicitly formulated elsewhere,
a position I call weak psychological egoism. That I am going to defend
a position I call weak psychological egoism suggests both that I think
there is such a thing as strong psychological egoism, and that I think that,
unlike weak psychological egoism, there is something wrong somewhere
with strong psychological egoism, wrong enough that it itself cannot be
defended. And, indeed, I do think that a position robust enough to be called
strong psychological egoism can be described, and I also think both that
this position is false and that even the best arguments in its favour are
seriously flawed. I will begin my defence of weak psychological egoism
by examining strong psychological egoism and some arguments for it; this
will enable me to present weak psychological egoism by contrasting it
favourably with its sibling, before going on to support it with positive
arguments.
1.
Strong psychological egoism is just psychological egoism as traditionally
understood. It is the doctrine that all actions are self-regarding in intent.
More fully, it is the doctrine that behind any action whatever, there lies one
or more self-regarding motivations, one or more of which is the ultimate
psychological cause of that action, the real practical reason the agent had
Erkenntnis 55: 217–237, 2001.
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
218
MARK MERCER
for performing it. Another way of putting this contention is that no one
ever intentionally performs an action except, ultimately, in order to realize
some self-regarding end. Of course, part of the doctrine of strong psychological egoism is the thesis that no one ever intentionally chooses a course
of action he believes will leave him worse off, from the point of view of
his self-regarding ends, than another open to him.1
What is a self-regarding motivation? The following list, adapted from
one drawn up by C. D. Broad (Broad 1952, p. 219), seems to me to be
exhaustive, or at least close to exhaustive:
1. The desire for self-preservation.
2. The desire for one’s own happiness or contentment; the desire to avoid
one’s own unhappiness or discontent.
3. The desire to be a certain kind of person; the fear of becoming a person
of some one or more other certain kinds.
4. The desire to respect oneself; the fear of losing self-respect, the fear of
coming to loathe oneself.
5. The desire to get and keep property; the fear of losing or failing to
attain property.
6. The desire for power over others, to have them do what one wants
them to do, whether they want to do it or not; the fear of losing power
over others. (This power over others need not be directed against them
in any malicious way; team captains, teachers, bus drivers, and ushers,
as well as bosses and leaders, have a degree of it, and usually use it
benignly.)
7. The desire that particular others have certain opinions of one; the
worry that they will instead have certain different opinions of one.
(This is the desire that people have certain cognitive attitudes toward
one.)
8. The desire that particular others have certain feelings toward one, that
they love, say, or like or respect or fear one; the worry that they will
instead have certain different feelings toward one. (This is the desire
that people have certain affective attitudes toward one.)
9. The desire for one’s own pleasure; the fear of being in pain.
Corresponding one-to-one with self-regarding motivations are selfregarding ends. One’s own happiness, or that one is not unhappy, for instance, is a self-regarding end corresponding to a particular self-regarding
motivation, the desire to be happy or the fear of being unhappy. Power
over others is a self-regarding end corresponding to the self-regarding
motivation of desiring to have power over others.
Some critics of egoism, and even some egoists, suppose that, for the
psychological egoist, the desire a person has for her own pleasure is ul-
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
219
timately the only self-regarding motivation, her own pleasure the only
self-regarding end. This is a mistake, for there are many different sorts
of self-regarding motivation. Sometimes, it is true, a person’s desire that,
say, certain people hold her in high regard is a desire in the service of that
person’s desire for pleasure, in that that person wants those people to think
highly of her because believing they do brings her pleasure. But there is
no reason to think that always for everyone all desires are in the service of
the desire for pleasure. The claim that anything anyone does is motivated
by a desire for pleasure is no part of even strong psychological egoism; a
partisan of strong psychological egoism can maintain that self-regarding
motivations are irreducibly many. What self-regarding motivations have in
common is that they make an essential reference to the well-being of the
agent who has them, his well-being understood as the agent himself understands it. Thus, a person’s desire that his neighbour be free to enjoy driving
her car, a desire that might lead him to help her dig it out of a snowdrift,
is not a self-regarding motivation, but an other-regarding one. In itself it
makes no reference to any of the agent’s self-regarding motivations. Of
course, were we to ask why the agent has that desire, we might uncover a
self-regarding motivation in back of it. Another other-regarding motivation
is the desire that no child grows up in poverty.
According to strong psychological egoism, then, no one helps a neighbour dig her car out of a snowdrift simply because of some other-directed
motivation he has, such as his desire that she be free to enjoy driving
her car; an agent’s desire to help a neighbour is, rather, itself ultimately
intentionally in service to another desire, one that is self-regarding – for
instance, the desire to be thought helpful by that neighbour, or the desire that the neighbour help him at some later date. According to strong
psychological egoism, the ultimate or real practical reason for which the
agent helped his neighbour has to be along the lines of that he wanted her
to think that he is a helpful sort of person, or that he wanted to invest in his
future. Again, no one works to alleviate child poverty simply out of a desire
that no child grows up in poverty; an agent’s working to alleviate child
poverty is always intentionally in service to a self-regarding motivation. A
person’s desire that no child grows up in poverty would, then, stem from
a practical reason involving some such self-regarding motivation as his
desire to maintain his self-respect. The real reason he went to work against
child poverty, then, was his fear that not doing so would make it difficult
for him to continue to respect himself.
Strong psychological egoism is, at least on its face, extremely implausible. One does not have to look very hard to find actions that appear
to be motivated solely by other-regarding desires. Sometimes, it seems,
220
MARK MERCER
people do help other people simply out of a desire their condition improve,
without, that is, intentionally seeking thereby to satisfy any self-regarding
desire. And sometimes people attempt to alleviate child poverty simply out
of the desire that no child grows up poor, not because they intend, directly
or ultimately, thereby to maintain their self-respect or to look good in others’ eyes. Short of an argument that demonstrates that always when we act,
we act in order to satisfy a self-regarding desire, we have, in our ability to
find or construct examples of people acting solely out of other-regarding
motivations, an excellent reason for rejecting strong psychological egoism.
Those who would defend strong psychological egoism will call on one
or the other of two arguments meant to show that appearances here are
deceiving, that in fact any other-regarding action we might care to mention
was ultimately motivated by a self-regarding desire, that ultimately the
agent intended in performing that action to secure some self-regarding end.
The first argument begins with the truism that any reason a person has for
doing something will be that person’s own reason for doing it. Since to act
for one’s own reasons is to act out of one’s understanding of one’s own
interests, and to act out of one’s understanding of one’s own interests is
to act on the basis of a motivation that makes essential reference to one’s
own self-interest, to act for a reason is always to act on the basis of a selfregarding motivation. The problem with this argument lies in an ambiguity
in the phrase “one’s own interests”. Either this phrase simply means one’s
own reasons for acting, or it means one’s own self-interest, understood
in terms of self-regard. But one’s own reasons for acting need not be to
secure that which is in one’s own narrow self-interest; or, at any rate, to
say that they must be is to appeal to the very claim the argument is meant
to establish.
The second argument begins with the premise that whenever a person
satisfies a desire, that person experiences a feeling of pleasure or otherwise
realizes a self-regarding end. It concludes with the claim that whenever a
person attempts to satisfy a desire, that person does so in order to experience a feeling of pleasure or to realize some other self-regarding end.
The premise of this argument certainly itself requires support; we might
very well question whether in satisfying a desire we inevitably also realize
a self-regarding end. In any case, though, that a person realizes a selfregarding end whenever he succeeds in satisfying a desire does not imply
that whenever he attempts to satisfy a desire he does so in order to realize a
self-regarding end. His realizing that end could well be an unintended, even
if anticipated and welcomed, consequence of his action. In signalling for a
cab, I displace a series of air molecules, and I might even be aware before
signalling for a cab that in signalling I will displace air molecules, but
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
221
displacing a series of air molecules is merely an unintended consequence
of my signalling for a cab. Likewise, in helping dig a neighbour’s car out of
a snowdrift, an agent might experience a warm feeling of satisfaction, but
that experience could simply occur as an unintended effect of succeeding
in realizing his actual intention, the intention to free a car from a snowdrift. Not every consequence of our actions, not even every anticipated and
welcomed consequence of our actions, is an intended consequence of our
actions, a part of the practical reasons for which we perform our actions.
Unless there is another argument to show that appearances here are
deceiving, we have no reason to think that the counter-examples to strong
psychological egoism that we found or produced are merely apparent
counter-examples. Therefore, we can in good faith reject strong psychological egoism as inconsistent with what we can reasonably assume are
ordinary matters of fact.2 Let us now turn to weak psychological egoism.
2.
Put most simply, weak psychological egoism is the doctrine that all actions are performed in expectation of realizing self-regarding ends. A little
more elaborately, weak psychological egoism is the doctrine that behind
any action whatever that an agent performs intentionally, ultimately there
lies the agent’s expectation of realizing one or more of her self-regarding
ends, an expectation without which the agent would not have performed
the action. If weak psychological egoism is true, no one ever performs an
action intentionally except that they expect to become happy or to forestall
unhappiness, or to maintain their self-respect, or to get others to see them
in a certain light, or to realize some other self-regarding end, in or through
performing it; if an agent does not expect to forestall his own unhappiness or to promote his self-image, or whatever, in or through performing
an action of some type, then that agent will not intentionally perform an
action of that type. (I say that an agent realizes a self-regarding end “in
or through” performing an action, in order to accommodate the point that
often we take pleasure, or realize some other self-regarding end, not as a
consequence of engaging in an activity, that is, not through performing an
action, but directly as part of engaging in that activity, that is, in performing
an action. To enjoy tennis is to take pleasure in playing tennis, and not, or
not only, to attain experiences of pleasure through playing tennis.)
Before presenting an argument in defence of weak psychological egoism, I need first to explain just how weak psychological egoism differs
from strong psychological egoism, and just how it avoids the problems
besetting strong psychological egoism. I will do that in this section. But I
222
MARK MERCER
need also to explain just how weak psychological egoism, weak though it
is, nonetheless manages to remain a form of psychological egoism. I will
do that in the next section.
Strong psychological egoism, recall, is the doctrine that everything we
do we do in order to secure a self-regarding end. Strong psychological
egoism implies that always when we act, we act intending to secure a selfregarding end. Since our intentions are represented in our practical reasons,
strong psychological egoism amounts to the thesis that for any action we
perform, within our practical reason for performing it there resides a selfregarding desire. Weak psychological egoism, on the other hand, is the
doctrine that everything we do we do at least expecting to secure a selfregarding end. This is the central difference between the two. It is no part
of the doctrine of weak psychological egoism that whenever we act, we
intend thereby to secure a self-regarding end. It is perfectly consistent with
weak psychological egoism that sometimes the desire component of our
practical reason for acting is entirely other-regarding.
On weak psychological egoism, an agent can very well come to another’s assistance simply out of the other-regarding desire that the other’s
troubles be gone. That an agent who assists another out of the desire that
the other’s troubles be gone would not have done what she did had she not
expected to realize some self-regarding end in or through doing it, does
not imply that she intended to realize that self-regarding end through her
actions. It does not imply that within her practical reason for doing what
she did was a self-regarding desire.
A person’s desire to realize a self-regarding end need not be part of her
practical reason for acting; still, according to weak psychological egoism,
no one will pursue a course of action unless she does expect, should the
action succeed in fulfilling its intended objective, to realize such an end.
This means that a person’s expectation of realizing a self-regarding end is
in one way significantly unlike a person’s expectation in signalling for a
cab that air molecules will be displaced. In the latter case, the person would
have signalled for a cab even had she not expected that air molecules will
be displaced – even had she expected that they would not be displaced. The
difference between the two expectations is that our expectation of realizing a self-regarding end plays a causal role in bringing about our action.
But, again, what distinguishes weak from strong psychological egoism is
that, on weak psychological egoism, our expectation of realizing a selfregarding end might not be represented by any desire in the practical reason
we had for performing the particular action; we need not have intended to
realize a self-regarding end in or through performing that action in order
to have performed it. But if the practical reason for acting does not itself
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
223
contain a self-regarding desire, and yet the action would not have been
performed had it not been expected to serve some self-regarding desire,
then where is the self-regarding desire and how does it play its causal role?
To understand the causal role played by self-regarding desires when
they are not within practical reasons we have for acting, we must distinguish the question why an agent performed the specific action he did
perform from the question why he has the motivations he has for performing that action. The question why an agent performed the specific action
he did perform is answered when we discover his intention in performing
it or the practical reason he had for performing it. The question why he
has the motivations he has requires us to direct our attention away from
the action itself, though, and toward the practical reason the agent had for
performing it. Here we ask of the desire component of that practical reason
why the agent found in it a motivation to act. Weak psychological egoism
proposes that whenever that desire component of a practical reason is not
itself self-regarding, what accounts for the fact that it motivates the agent
to act is its relation to one or more of the desires the agent has that are
self-regarding.
The expectation that we will realize a self-regarding end through acting
on some desire, whether that desire is self- or other-regarding, is what
makes that desire motivating to us – it is, that is, what makes it a desire
in the first place. This is the causal role that expectations of realizing selfregarding ends play, when they do not function directly as components
of practical reasons for acting. Consider, for instance, an agent who believes that another person is in pain and that he could alleviate her pain
by fetching her an aspirin; moreover, he believes that it would be quick
and easy for him to fetch her an aspirin. This agent will not act on these
beliefs, he will not, say, attempt to alleviate her pain by fetching her an
aspirin, unless her being in pain matters to him. Let us suppose that it does
matter to this agent that the other is in pain; let us suppose that he wants
that she not be in pain. That she is in pain, that he can easily alleviate
her pain by fetching some aspirin, and that he wants that she not be in
pain, together gives him a practical reason for fetching her some aspirin.
We can explain why he fetches her some aspirin by citing this particular
practical reason for acting. But why, we might ask, does her being in pain
matter to him, and matter to him in the way that it does? Why does he
want that she is not in pain? Why is he not either indifferent to the fact
that she is in pain, taking no more notice of it than he does the colour of
the walls, or instead positively welcoming of the fact that she is in pain,
considering even whether to seek to prolong her pain, perhaps by hiding
the aspirin? According to weak psychological egoism, we can explain why
224
MARK MERCER
he wants that she not be in pain, why this attitude enters into his practical
deliberations, by noting its relation to one or more self-regarding desires
he has. That the thought of another being in pain distresses him might,
for instance, be a cause of his wanting her not to be in pain, as might the
thought that he would be failing to live up to his self-image if he were
indifferent to her pain. Some self-regarding motivation or other is, on this
view, causally responsible for his having the other-regarding motivation
he has. No consideration will appeal to an agent as a practical reason for
acting unless it bears a connection to a self-regarding motivation present
in that agent.
Now it is again important to note, if weak psychological egoism is not
to expand into strong psychological egoism, that the self-regarding motivation that causes an agent to have an other-regarding motivation, that causes
him to find in other-regarding considerations reasons to act, need not cause
him to acquire that other-regarding motivation as a practical reason causes
an action. An agent’s self-regarding motivations are not, or are not necessarily, practical reasons for having the other-regarding motivations he
has. That is to say, an agent can come to have an other-regarding motivation without having intended to acquire an other-regarding motivation.
Acquiring an other-regarding motivation is not, or is hardly ever, an action we perform. Of course, a self-regarding motivation can be a practical
reason for forming an other-regarding motivation; an agent who intentionally forms the desire that another not be in pain in order to promote
some self-regarding end has come to that other-regarding desire on the
basis of a practical reason involving a self-regarding desire to have it. But
the claim that other-regarding motivations are always formed intentionally
on the basis of self-regarding reasons is a claim not to be distinguished
from strong psychological egoism. (An action intended to serve an otherregarding end that, in turn, is intended to serve a self-regarding end, is
an action intended to serve a self-regarding end.) Yet weak psychological egoism allows that self-regarding motivations can engender otherregarding motivations without being practical reasons for having them.
Self-regarding motivations that do not themselves function as practical
reasons for performing actions or acquiring other-regarding motivations,
function instead as non-intentional psychological causes, both efficient and
sustaining causes, of other-regarding motivations.
Of course, in everyday speech we often use the word “reason” to refer
both to psychological causes of attitudes that involve intentions and psychological causes of attitudes that do not. We speak of a man’s childhood
experiences of hearing Beethoven’s middle quartets as a reason for his
present intention to listen to a CD of the “Harp” quartet (or as a reason of
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
225
his present intention to leave that CD on the rack), but this “reason” does
not itself involve an intention. Likewise, we might speak of an agent’s
desire not to be distressed by the thought that another is in pain as a reason
why he fetched her some aspirin, though that desire was no part of his intention to fetch her some aspirin. Strictly, the man’s childhood experiences
of the middle quartets are a psychological cause of his present fondness
for them, and it is that fondness that figures in his practical reason, the
psychological cause involving an intention, for listening to the CD. And
so, strictly, the agent’s desire not to be distressed by the thought that
another is in pain is a psychological cause of his having the standing (otherregarding) desire that his companion not be in pain, and this desire, made
occurrent by his belief that his companion is in pain, figures in his practical
reason, the psychological cause involving an intention, for fetching that
companion an aspirin.
Weak psychological egoism, then, incorporates a thesis about why the
considerations that motivate us to take action do motivate us to take action. It is not primarily a doctrine about our motivations for action as
found in our intentions, for they can be either self- or other-regarding.
Instead, it is a doctrine concerning why our motivations in our intentions
are what they are, why they have the motivating force that they do. In
back of our actions are our practical reasons, and our practical reasons
can involve other-regarding desires only. But in back of those practical
reasons involving other-regarding desires only, says the defender of weak
psychological egoism, there are always self-regarding desires.
Let us return to the question how weak psychological egoism avoids the
pitfalls bedevilling strong psychological egoism. Because weak psychological egoism does not imply that everything we do we do for self-regarding
practical reasons, it is consistent with everyday examples of people acting
for other-regarding practical reasons. Further, in allowing that not every
consequence we expect our action to produce is a consequence we intend
our action to produce, weak psychological egoism embodies no confusion between the total consequences of an action and those consequences
that constitute the intended consequences of that action. Thus, it avoids
the objections that, as we saw, rendered untenable strong psychological
egoism.
Nonetheless, no consideration lacking a tie to a self- regarding motivation could be felt by an agent to be a reason for acting, or so at least is
implied by the doctrine of weak psychological egoism. This means that,
for weak psychological egoism, self-regarding motivations have a sort of
priority over other-regarding motivations. Self-regarding motivations, and
the ends associated with them, are to be invoked in explaining why an agent
226
MARK MERCER
has the particular systems of other-regarding motivations that he does have,
when he has any such systems. Self-regarding motivations lie in back of
all other-regarding motivations, as the source of their motivational power.
These claims still very much stand in need of defence.
3.
One of the reasons strong psychological egoism attracts attention is that
it is a bold doctrine, hard-headed if not even bloody-minded, one many
philosophers, political scientists, and others have found worth getting excited about, whether they accept or reject it. Is weak psychological egoism
also worth getting excited about? Or is it an innocuous doctrine, one with
few implications for morals, politics or our philosophical or practical understanding of ourselves? These are heady questions. Let us limit ourselves
here to a somewhat more manageable question. Let us ask in what sense,
if any, is weak psychological egoism a form of egoism?
First of all, let us consider a view of the nature of other-regarding
motivation and action that has only a very small egoistic component, the
view according to which any other-regarding motivation an agent has must
have had its genesis in one or more self-regarding motivations that that
agent has or used to have. This is the view that while self-regarding motivations are necessary to the formation of other-regarding motivations, an
other-regarding motivation, once formed, can become independent of the
self-regarding motivation responsible for bringing it into existence, and
from all other self-regarding motivations as well. Other-regarding motivations, that is, can come to live lives of their own. Though Samantha came
to value playing chess only because her uncle bribed her with candy to
play, now that she does value playing chess, she values it independently of
thoughts of candy; likewise, though a person can come to take another’s
well-being as her own concern only because early on she was coaxed to
do so and then rewarded when she did, now that she has that concern, it
might very well be independent of any thought of reward.3 Clearly, the
egoistic component of weak psychological egoism is much greater than is
the egoistic component of this view. On weak psychological egoism, otherregarding motivations can never become independent of self-regarding
motivations. Since other-regarding considerations cease to motivate when
detached from expectations of realizing self-regarding ends, they can never
become capable of living lives of their own.
Now consider the more strongly egoistic view that a person will time
after time perform other-regarding actions of some particular type (say, actions intended to promote the well-being of others), only if doing so at least
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
227
sometimes has the effect of satisfying some of her self-regarding desires.4
The idea here is that having her self-regarding desires satisfied by succeeding in realizing her other-regarding ends reinforces a person in her desire
to realize her other-regarding ends, and failing to have her self-regarding
desires satisfied tends to extinguish her desire to realize other-regarding
ends. On this view, agents can, on occasion or even habitually, act on
desires to promote others’ well-being without having any expectation of
realizing a self-regarding end by doing so; realizing a self-regarding end,
however, has the effect of increasing the likelihood of their promoting
on other occasions others’ well-being. If it happens often enough that an
agent’s self-regarding desires go unsatisfied, though, the agent will cease to
perform actions intended to promote others’ well-being. Other-regarding
motivations can safely wander out of sight of self-regarding ones now
and then, according to this view, but, if they happen to remain out of
sight for too long, they will wither and eventually die. This view, then,
contains a significant egoistic component. Weak psychological egoism,
however, is a much stronger form of egoism even than this. By holding
that the agent is always aware of the self-regarding ends he expects will
be realized in acting as he does, weak psychological egoism places selfregarding ends in the foreground of our understanding of motivation, while
this view allows them to remain unnoticed in the background. In fact, by
allowing that self-regarding motivations can, on occasion, slip so far into
the background as to be entirely out of mind, this view perhaps founders
on an insuperable difficulty. It is hard to see how failing to realize a selfregarding end on some occasion could weaken an agent’s general desire
to perform actions of some type, when that agent had no expectation of
realizing a self-regarding end on that occasion. This position, then, which
might at first have seemed a reasonable compromise between the view that
self-regarding motivations engender though not necessarily sustain otherregarding motivations and the view of weak psychological egoism that they
both engender and always sustain other-regarding motivations, turns out in
the end to be less stable than either of these views.
Weak psychological egoism has, then, a greater egoistic component
than either of two other views of the relation between other-regarding
and self-regarding motivations, two other views on which self-regarding
motivations have a certain priority over other-regarding ones. But, even
so, how can weak psychological egoism really be an egoism when it is no
part of this doctrine that all actions are intended to serve self-regarding
motivations? We need to say more about what features a doctrine must
have in order to count as egoistic. I think that if a doctrine implies any one
of the following three theses, then it would be churlish not to account it a
228
MARK MERCER
form of psychological egoism: (1) that we always act at least in expectation
of reward to ourselves; (2) that altruistic selflessness is impossible; and (3)
that we never intentionally follow a course of action that, from the point
of view of our self-regarding ends, appears worse than another open to us.
Weak psychological egoism implies all three.
(1) To act in expectation of realizing a self-regarding end is to act in
expectation of reward, in expectation of a reward for oneself. If it is true,
as according to weak psychological egoism it is, that we do not act except
in expectation of realizing a self-regarding end, then it is also true that
we do not act except in expectation of reward. While weak psychological
egoism allows that it is possible for us to act without intending to reap a
reward, it nonetheless does not allow that it is possible for us to act without
expecting, should we succeed in fulfilling our intentions, whatever they are,
to reap a reward. That weak psychological egoism thus implies that no one
acts except in expectation of reward seems to me clearly to qualify it as a
form of egoism.
(2) Weak psychological egoism implies that a person never acts entirely selflessly, that is, acts without thought of reward for herself, even
though she might well be acting unselfishly, that is, acting to benefit others without intending to gain a reward for herself. If to act altruistically
is to act selflessly, without thought of reward to oneself, then, according
to weak psychological egoism, altruism is impossible. Again, then, weak
psychological egoism, weak though it is, seems clearly to qualify as a form
of egoism.
(3) Crucial to any egoism is that it contain the thesis that no one ever
intentionally chooses a course of action he believes will leave him worse
off, from the point of view of his self-regarding ends, than another open
to him. And, indeed, weak psychological egoism implies this thesis. What
makes one course of action more attractive to an agent than another is the
expectation of greater reward in or through taking that course. It seems a
mere tautology that it is never intentional of an agent that she takes a course
of action she finds less attractive than another course of action she believes
open to her; in any case, so long as this claim is true, then, according to
weak psychological egoism, it is never intentional of an agent that she
takes a course of action in or through which she expects to realize a lesser
reward than she expects she would realize in or through another course of
action.5
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
229
4.
So far I have described weak psychological egoism, and both distinguished
it from strong psychological egoism and shown how it survives objections
that render strong psychological egoism implausible. I then defended the
claim that weak though it is, weak psychological egoism is still clearly an
egoistic account of other-regarding motivations and the actions that stem
from them. What I have not yet done is given any argument in favour of
weak psychological egoism, any argument meant to incline us to think
it true. In this section I will provide two arguments in support of weak
psychological egoism.
First, though, something needs to be said about the objection that weak
psychological egoism is an unfalsifiable doctrine, and, thus, is to be rejected as unempirical and meaningless. Weak psychological egoism is an
empirical doctrine, or at least I mean to put it forth as one; but it is not
a doctrine directly falsifiable in light of particular pieces of observational
evidence. It would lie too deeply within the web of an empirical theory to
be answerable to any finite set of observations. It is, rather, ultimately to
be evaluated first according to the contribution it makes as an organizing
or constitutive principle within a larger theory of human behaviour or,
perhaps, theory of agency; and second according to the fortunes of that larger theory in furnishing investigators with predictions of and control over
agents’ behaviours. In the end, the extent to which the doctrine of weak
psychological egoism is an important part of a successful larger theory, is
alone the extent to which it commands our acceptance.
That weak psychological egoism is an empirical doctrine within an empirical theory, and, thus, is to be judged according to empirical evidence
regarding that theory, does not, however, mean that there is nothing right
now, prior to empirical investigation, to say in its favour. Arguments for
it, though, must be given in the spirit of offering reasons to think it worth
pursuing as a theory, rather than as demonstrations that we know it to be
true. The two arguments for weak psychological egoism presented below
will, then, amount to rationales for entertaining it, or for continuing to
entertain it, as a viable general hypothesis about the nature of intentional
action.6
The first argument draws on introspection. Introspection reveals to me
that whatever I decide to do, indeed I do expect that, should I meet with
success in doing it, I will realize one of my self-regarding ends. Further,
when I ask myself before acting on my decision to perform some particular
action whether I would still do what I have decided to do were I to lack any
expectation of realizing thereby a self-regarding end, I find that I answer
230
MARK MERCER
no, I would not still do what I have decided to do. Were I not to expect
to realize some self-regarding end in or through my action, I would find
myself losing the desire to perform that action. I would, I think, cease to
find important or attractive the goal I intend to achieve through that action.
Thus, I conclude, any action I intend to perform will be an action in or
through which I expect to realize some self-regarding end. This argument
doesn’t end quite yet, though. I now note that I am a typical agent in the
world, not, I think, at least at this level of abstraction from my concrete
projects and plans and likes and dislikes, appreciably different from other
agents. This fact enables me to generalize from my own case to the case
of all agents and actions. I conclude, then, that all actions are performed in
expectation of reward.
This particular argument for weak psychological egoism is beset by
considerable problems. That introspection tells me that I expect to realize
a self-regarding end whenever I meet with success in acting does not imply
that in fact I do expect to realize a self-regarding end whenever I meet
with success in acting; maybe I just think that I do. Introspection might
not be a trustworthy guide to one’s own thoughts and other states of mind.
Moreover, I cannot claim to be sure that introspection even tells me what
I say it tells me. Maybe I misinterpret what it tells me, or maybe I’m not
actually introspecting at all. Finally, generalizing from my own case might
well be improper. Maybe I happen to be more egoistic than people need to
be; this can be true even if I also happen to be less egoistic than people typically are. Perhaps the most serious shortcoming in this argument, though,
at least from the perspective of a defence of weak psychological egoism,
is that even if successful it just indicates that we all do act in expectation
of reward; it does not provide us with any insight into why this should be
so. Still, despite these flaws and shortcomings, people who rehearse this
argument using themselves as its subject often find it rather powerful. I
offer it here, then, as an invitation. Place yourself in the subject’s position,
and see whether you deem the premises true in your own case.
The second argument I will give in support of the thesis of weak
psychological egoism draws on contentions regarding the nature and conditions of understanding others and their actions. One of these contentions
is that an interpreter understands an agent’s action only if she appreciates
the motivational force of the practical reason for which that action was
performed. Earlier, in Section 2, I said that weak psychological egoism is
a doctrine that incorporates the thesis that to appreciate the motivational
force of a reason for acting is to relate the pro-attitude component of that
reason to one or more of the agent’s self-regarding motivations. According
to that thesis, those considerations that in fact can motivate us to act are
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
231
able to do so only because we see in acting on them the possibility of realizing one or more self-regarding ends. Our defence of weak psychological
egoism will hinge on a defence of that thesis.
To begin: To understand what another has done is both to have a particular sort of true description of the action he has performed, one that
reveals it to be intentional, and to know the agent’s practical reason for
performing that action. In turn, to know an agent’s reason for performing
some particular action involves understanding his motivation in doing it.
An interpreter cannot, though, really understand an agent’s motivation in
performing an action unless she sees that motivation as a motivation, unless
she is cognizant of its force as a motivation. It is not enough, that is to say,
to understand what a person who intentionally sips from a saucer of mud
has done to note merely that he had the desire to sip from a saucer of mud,
and believed himself both possessed of a saucer of mud and able to sip
from it. An interpreter has also to comprehend what in desiring to sip from
a saucer of mud was attractive to him.
Now usually, of course, there is no problem in our comprehending what
it is in the desires had by people around us that attracts them as desirable.
The people around us are more or less like us in many if not most of their
desires, wants and wishes, and few of them desire to sip from a saucer
of mud, so in our day to day life we do not often have cause to turn our
attention explicitly to the question from whence arrives the motivational
force of their desires. Still, it is not exceedingly uncommon for us, even
for those of us who are not psychologists, sociologists, or anthropologists,
to be stumped by some piece of what we take to be behaviour. How are
we to make sense of some such piece of strange behaviour? One way is
to connect that piece of behaviour to one or more of the strange agent’s
self-regarding ends. If we can see in sipping from a saucer of mud a way
of maintaining self-respect, or even a way to delight in the taste of mud,
we can understand the desire the agent had to sip from a saucer of mud. We
need not connect his self-regarding end to an intention to realize that end
in or through his action; we need only, I think, connect it to an expectation
of realizing it.
But is this the only way we can make sense of desires we ourselves
do not share and cannot, at first at least, imagine sharing? I think that it
is. Without our perceiving a connection to an intention or an expectation
of realizing some self-regarding end, we cannot see in any consideration
we attribute to an agent a motivation to act. The motivating force of the
consideration that spurred action will remain beyond our ken, the action
stemming from it unfathomable and inexplicable.
232
MARK MERCER
The argument to this point is that no action stemming from a consideration that an interpreter cannot in principle relate to a self-regarding
motivation can really be understood as a motivation by that interpreter.
The conclusion we are after, however, is that there are no motivations at
all that fail to bear close relations to self-regarding motivations. So far we
seem to have only the result that some agents’ psychologies might be so
very strange that we, with our sorts of psychologies, could never really
understand what they are doing, as we must fail to understand why they
want to do it. It might, that is, yet be possible that there exists somewhere
a truly altruistic agent, one who acts solely out of a desire that others do
well, neither intending nor even expecting by his actions to reap any selfregarding reward; all we can say is that we could never be able to recognize
such a character, or understand his behaviour.
I think, however, that the argument can be taken a couple of steps further. It seems to me that any evidence we could obtain to show that some
movement was truly an action, while yet one that, in principle, must escape
our comprehension, since we cannot fathom how the consideration behind
it was felt as a motivation, must equally be evidence that that movement
was not an action at all. We might, now, say that we cannot, in principle
and not just for lack of evidence or interest, always distinguish between
movements that are not actions and movements that are; sometimes all
possible evidence will be available to us, and still that evidence does not
support one over the other of two contrary hypotheses. Not only, then, on
this supposition, could we not know of an entirely selfless altruistic action
why the agent performed it, we could not even perceive it as unquestionably an action. We might, however, say instead that so long as we do see
it as an action, it must be possible for us to see the motivating force in
the consideration behind it. This is to reject the idea that all the evidence
could be in and still the matter go undecided. The second alternative is
to be preferred to the first, I think, because the first severs the connection
between evidence and what evidence is evidence for. Once all the relevant
evidence is in, if neither of two contrary hypotheses is better favoured than
the other, then, it would seem, either the hypotheses are not contrary to
each other after all, or the evidence is not really evidence relevant to the
matter of these two hypotheses.
What these reflections lead us to is the contention that if it is in principle impossible to comprehend the motive with which a supposed action
was performed, then actually the supposed action was no action at all.
For any action that is actually an action, the motive will, in principle,
be comprehensible to an interpreter. For a motive to be comprehensible,
though, it must be closely tied to a self-regarding motivation. Therefore,
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
233
all motives for acting bear close ties to self-regarding motivations. And
this is the central thesis within the doctrine of weak psychological egoism
we needed to defend.7
5.
Stated succinctly, weak psychological egoism is the doctrine that anything
an agent does intentionally, that agent does at least expecting thereby
to obtain a reward, if not actually intending thereby to obtain one. It is
an empirical doctrine about human behaviour, or, even, about agency in
general, though it is of such high generality or abstraction that it is not
directly falsifiable by any observational evidence. It is, rather, to be tested
by comparing those theories in which it plays a part with those theories
from which it is absent to see which are more successful in enabling
investigators to predict and control psychological phenomena generally
and action in particular. Though it is a doctrine ultimately answerable to
empirical evidence, we presently have good reasons for entertaining it as
a live hypothesis and attempting to construct psychological theories that
include it as an organizing principle. These reasons mainly have to do with
the idea that to understand the motivation behind an action, we need to
understand the force of that motivation, and the way to do this is to find
a self-regarding end associated in the agent’s mind with acting on that
motivation.
In Section 3, I raised, without answering, questions about the implications of weak psychological egoism for such things as morals and our
understanding of ourselves, and then turned instead to the simpler question in what sense weak psychological egoism is a form of egoism. I will
conclude this paper with a few reflections on the heady issues I declined to
discuss earlier.
Strong psychological egoism is a dead issue, and that fact has, I think,
encouraged some philosophers to imagine that it is possible for actions
to be entirely selfless both in origin and direction. These philosophers
champion the bold view that humans can be, if only rarely ever are, entirely selfless and giving, selfless and giving even past the point at which
they notice that their self-regarding interests are suffering unrecompensable harm. Thomas Nagel, for instance, holds that an agent’s perception
of the objective badness of a state of affairs can, by causing that agent
to comprehend that states of affairs ought be objectively good, motivate
him to try to improve things, and do so whatever his particular wants and
desires are and whatever he thinks the costs to his wants and desires will
be.8 John McDowell, for his part, holds that the virtuous person is one for
234
MARK MERCER
whom, in certain situations, all considerations involving her self-regarding
ends count for nought in face of what she perceives the situation itself to
require of her.9 James Rachels adds that people can and sometimes are
motivated not by want or desire at all, but purely and ultimately from a
sense of obligation.10
That strong psychological egoism is false does not, of course, imply
that selfless altruism is possible. Still, that it is false does, in the absence
of another viable egoistic account of motivation, render unsupported the
contention that no action is entirely selfless. Weak psychological egoism,
then, because it is a viable egoistic account of motivation, provides new
principled support for that contention. Those who, like Nagel, McDowell
and Rachels, maintain that selfless action is possible, who maintain that,
on occasion, agents can transcend or be indifferent to their self-regarding
ends, must meet the challenge of describing how and where weak psychological egoism goes wrong. This is one way in which the doctrine of weak
psychological egoism is of philosophical interest.
But what about the charge that egoistic doctrines of other-regarding
desires and actions are cynical doctrines? Is the world of the person who
accepts the thesis that everything intentionally done is done in expectation
of reward really different from the world of the person who rejects this
thesis? What does the world look like through the eyes of the weak psychological egoist? Is the egoist’s world the world of the pessimist, while
that of his opposite is the world of the optimist?
Daniel Dennett instructs us, when we set out to investigate issues regarding the possibility and nature of free action, not to feed the bugbears.11
What he means is that we should avoid inflating the significance of those
issues for everyday moral issues and our prospects of living happy, meaningful lives. This instruction applies just as well to philosophical issues
concerning the nature of motivation.12 We ought not fool ourselves into
thinking that terribly much in morals or our self-image as agents in the
world rests on whether weak psychological egoism is true or not, or
whether we believe it is or is not. The world looks pretty much the same to
both egoists and others – or, at least, differences between the two in how
it looks are better traced to other beliefs and attitudes. Still, there are some
differences in general outlook between the two sides of this issue, and it is
right that we try to get clear on them.
Strong psychological egoism is said by many of its critics to be a
cynical doctrine, for its partisans must inevitably be suspicious of their
neighbours’ designs, and themselves be somewhat if not entirely opportunistic. However, not even those who accept strong psychological egoism
need deny that sometimes people aim in their actions to do well by oth-
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
235
ers, and that sometimes they succeed in their aim. Nor need any sort of
psychological egoist take a dim view of other people’s happiness. What
psychological egoists deny is that instances of aiming to do well by others
are ever, from the agent’s own point of view and overall, instances of self
denial or self overcoming. The agent who does well by others does so
at least in expectation of, if not directly for, personal reward. Thus, the
person who does well by others is not to be applauded or commended or
cherished as embodying attitudes of self denial or self overcoming. She
might, however, be applauded or commended or cherished as one who
likes to do well by others. Between psychological egoism and cynicism,
then, stands one’s attitude toward selflessness. One who prizes it and yet,
as a psychological egoist, thinks it impossible, might well cast a jaundiced
eye on his neighbours and himself. One who does not think highly of it
in the first place, however, will be no more cynically suspicious of others
or disappointed in himself than the evidence he has of others’ or his own
dissembling and opportunism warrants.
Is this merely to say that those who accept psychological egoism can,
without contradicting their beliefs, assume attitudes toward others and the
world just as sunny as those who reject it often do? No, it is not to say
merely this, for, from the psychological egoist’s perspective, those who
deny psychological egoism have not earned the comfort their attitudes
bring them. There is something disturbingly pollyannaish about thinking
that people can on occasion set their preferences and plans, their wants and
desires, their likes and dislikes, aside, and something viscously distasteful
in the idea that it is ever appropriate that they should. To cling to the view
that entirely selfless actions are both possible and, sometimes, just what
is called for, is not so much to think that people really are capable of
right action for the right reasons as it is to refuse to grow up, to refuse
in principle to take pleasure in the world as it is, and to enjoy one’s own
contingent personality. What is admirable in the person who sacrifices his
life in assisting others is not that he acted rightly despite his inclinations,
but rather that he was so strongly inclined to be concerned for others. Perhaps it is true that sometimes psychological egoists display a knowingness
of the inner recesses of the human heart that gets annoying, just as those
who think selfless altruism possible can be insufferably smug and selfrighteous, though there is nothing in either position that makes it inevitable
that its partisans will be annoying or insufferable. Still, it seems to me, if
cynical knowingness is a risk taken by those who would put away childish
things, it is very much a risk worth running.13
236
MARK MERCER
NOTES
1 A version of psychological egoism stronger even than strong psychological egoism is
the doctrine that all actions are selfish. (Because it is not true that an action is selfish just
in virtue of its origin in a self-regarding motivation, the doctrine of strong psychological
egoism does not itself imply that all actions are selfish.) The doctrine that all actions are
selfish can be rejected on the ground that it incorporates an untenable or useless conception
of selfishness. In any case, we need not concern ourselves with it in this paper, for successful objections to strong psychological egoism must tell against even stronger versions of
egoism as well.
2 We cannot claim, however, to have shown that strong psychological egoism is false, or
at least so I argue in Mercer (1998).
3 A view with no egoistic component would be that other-regarding motivations can themselves be primitive motivations, neither sustained nor even engendered by self-regarding
ones at all. These two views of the relation of other-regarding motivations to self-regarding
ones are described by Michael Anthony Slote (Slote 1964, pp. 531–532). The egoistic
thesis that Slote defends as having empirical content, that “no human act is ever unselfish”
(p. 531, n. 2), seems to me, given what Slote means by “unselfish”, to be very close to the
thesis of weak psychological egoism. Slote does not carefully distinguish his thesis from
that of strong psychological egoism, however.
4 This is the view Hugh LaFollette defends as the truth in psychological egoism (LaFollette 1988, p. 503). LaFollette argues explicitly against strong psychological egoism before
presenting this view as the insight into human motivation hidden within it.
5 Does this mean that egoistic views of motivation exclude the possibility of weak-willed
action? Not necessarily. What makes an action weak willed is that it is performed intentionally despite the agent’s judgement that it is not, all things considered, the best action
to perform. That it is performed intentionally means that it is performed for a practical
reason, although, since it is weak willed, not for the practical reason the agent judges her
best. That it is performed intentionally does not, however, mean that it was the agent’s
intention to perform an action for other than what she judged to be her best reason for
acting. A person cannot have a practical reason that rationalizes her performing an action
on other than what she judges is her best reason. Since crucial to egoism is the contention
that an agent cannot intentionally perform an action under the description “an action that
will leave me worse off from the point of view of my self-regarding ends than another
action”, the thesis that sometimes people act in full awareness that their action will leave
them worse off than another action is compatible with egoism. What is incompatible with
egoism is the thesis that an agent can perform an action, whether strong-willed or weak,
mindless of her self-regarding ends. (The conception of weakness of will I have employed
here is from Davidson (1970).)
6 Slote (Slote 1988), however, argues of a position similar to weak psychological egoism
that it does in fact come close to being directly falsifiable by empirical evidence, and,
moreover, that certain experiments in learning theory speak clearly in its favour.
7 The second argument of this section draws substantially on arguments Donald Davidson has given in favour of the contention that there are no meaningful utterances that, in
principle, must escape interpretation by competent interpreters. See, primarily, Davidson
(1973) and Davidson (1974).
8 See Nagel (1986), especially Chapter VIII, “Value”, and Chapter IX, “Ethics”.
IN DEFENCE OF WEAK PSYCHOLOGICAL EGOISM
237
9 See, especially, McDowell (1979). McDowell thinks that an agent’s perception of the
objective facts of a situation can supply that agent with reasons for acting in a particular
way independently of that agent’s prior wants and desires. He develops an answer as to
how it is possible for perceptions to engender new motivations in McDowell (1995).
10 See Rachels (1998), p. 74.
11 See Dennett (1984), pp. 1–10. These are the pages that make up the first two sections of
the first chapter, “Please Don’t Feed the Bugbears”.
12 One philosopher who does his best to ensure that the bugbears get as fat as they can is
E. J. Bond. See Bond (1996), pp. 7–8 and 16–17.
13 I thank David Checkland and Stephen Haller for comments on an earlier draft of this
essay, and members of colloquia at Brandon University and Dalhousie University for
discussing these ideas with me.
REFERENCES
Bond, E. J.: 1996, Ethics and Human Well-Being, Blackwell, Oxford.
Broad, C.: 1952, ‘Egoism as a Theory of Human Motives’, Ethics and the History of
Philosophy, Routledge & Kegan Paul, London.
Davidson, D.: 1970, ‘How is Weakness of the Will Possible?’, Essays on Actions and
Events, Clarendon Press, Oxford, 1980.
Davidson, D.: 1973, ‘Radical Interpretation’, Inquiries into Truth and Interpretation,
Clarendon Press, Oxford, 1984.
Davidson, D.: 1974, ‘Belief and the Basis of Meaning’, Inquiries into Truth and Interpretation, Clarendon Press, Oxford, 1984.
Dennett, D. C.: 1984, Elbow Room: The Varieties of Free Will Worth Wanting, The MIT
Press, Cambridge, MA.
LaFollette, H.: 1988, ‘The Truth in Psychological Egoism’, in Joel Feinberg (ed.), Reason
and Responsibility (Seventh Edition), Wadsworth, Belmont, CA.
McDowell, J.: 1979, ‘Virtue and Reason’, The Monist 62, 331–350.
McDowell, J.: 1995, ‘Might There be External Reasons?’, in J. E. J. Altham and Ross
Harrison (eds.), World, Mind, and Ethics: Essays on the Ethical Philosophy of Bernard
Williams, Cambridge University Press, Cambridge.
Mercer, M.: 1998, ‘Psychological Egoism and Its Critics’, Southern Journal of Philosophy
36(4), 557–576.
Nagel, T.: 1986, The View from Nowhere, Oxford University Press, Oxford.
Rachels, J.: 1998, The Elements of Moral Philosophy (Third Edition), McGraw-Hill
College, Boston.
Slote, M. A.: 1964, ‘An Empirical Basis for Psychological Egoism’, Journal of Philosophy
61(18), 530–537.
Department of Philosophy
Saint Mary’s University
Halifax, NS
Canada, B3H 3C3