ADVERSUS HOMO ECONOMICUS
1
CRITIQUE OF LESTER’S ACCOUNT OF INSTRUMENTAL RATIONALITY
Danny Frederick*
12 October 2015
Abstract In Chapter 2 of Escape from Leviathan, Jan Lester defends two hypotheses: that
instrumental rationality requires agents to maximise the satisfaction of their wants and that
all agents actually meet this requirement. In addition, he argues that all agents are selfinterested (though not necessarily egoistic) and he offers an account of categorical moral
desires which entails that no agent ever does what he genuinely feels to be morally wrong. I
show that Lester’s two hypotheses are false because they cannot accommodate weakness
of will; because they are inconsistent with agency, which requires free will; because ends,
obligations and values cannot be reduced to desires; and because maximisation is often not
possible. Further, Lester’s claim that agents are self-interested is vacuous, his attempted
reduction of moral behaviour to want-satisfaction fails, and his contention, that agents
always do what they genuinely think to be morally required, seems untenable. A defence of
freedom that depends on homo economicus is far from promising.
Keywords agency; cognitive dissonance; desires; ends; free will; Hare; homo economicus;
instrumental rationality; Lester; Leviathan; libertarianism; maximisation; morality; Popper;
self-interest; values; Watson; weakness of will; wrongdoing.
1. Introduction
In Chapter 2 of Escape from Leviathan: Libertarianism without Justificationism
(Lester 2012), the libertarian and Popperian philosopher, Jan C. Lester, offers a
defence of the ‘homo economicus’ conception of people as self-interested, utilitymaximising, rational agents. Lester sets out his connected conceptions of agency
and of instrumental rationality and he seeks to defend them against a range of
criticisms. I will show that his defences often fall down, and I will offer additional
criticisms that he does not consider. I show that the mechanical, homo economicus
conception of rationality and of agency is thoroughly mistaken. I offer instead a
conception of people as fallible, largely ignorant, often irrational and sometimes
immoral, who inherit theories about facts, values, morals and themselves which they
strive to improve in a piecemeal fashion by conjecture and experiment.
In section 2, I explain instrumental rationality and then Lester’s thesis of
instrumental rationality, which conjoins a hypothesis about rationality with a
hypothesis about agency. In section 3, I argue that the phenomenon of weakness of
will refutes Lester’s hypothesis about agency and that Lester’s attempt to explain
away the phenomenon is unsuccessful. In section 4, I show that Lester fails to make
sense of the conflict between values and desires, that his want-satisfaction account
of agency should be supplanted by an account which recognises distinct categories
of desires, values and ends, and that this scuppers his account of instrumental
rationality. In section 5, I argue that Lester’s passive conception of agency is not
really a conception of agency at all, because it does not allow for metaphysical free
will; and this entails the rejection of Lester’s hypothesis about rationality as well as
his hypothesis about agency. In section 6, I dismiss Lester’s contention that all
*
Webpage: http://independent.academia.edu/DannyFrederick
ADVERSUS HOMO ECONOMICUS
2
actions are self-interested (though not necessarily egoistic) as vacuous. In section 7,
I argue that Lester’s assumption that agents always seek to maximise the
achievement of their ends is false and that this refutes both his hypothesis about
rationality and his hypothesis about agency. In section 8, I show the falsity of
Lester’s contention that we cannot act against our moral convictions. In section 9, I
conclude.
2. Instrumental Rationality
A minimal conception of instrumental rationality is this: an action is
instrumentally rational if and only if it is a suitable means to the achievement of the
agent’s ends, and it is instrumentally irrational if and only if it is an unsuitable means
to the achievement of the agent’s ends (see Popper 1994b, 178-79, 181, 183n19;
and my 2013a). As it stands, this statement may be interpreted objectively,
subjectively, or part one way, part another. On the objective interpretation,
instrumental rationality requires that the agent’s actions are objectively suitable to his
objective ends. An agent’s objective ends are those that are assigned to him
ontologically, by his nature, or by Nature, or perhaps by God. It may be disputed
whether an agent, or at least a human agent, has any objective ends; but if he has,
then it is possible that he is quite mistaken about what they are, and he may seek to
achieve ends which are not only different to his objective ends but which even
undermine them. An agent’s subjective ends are those which he acknowledges as
ends. An action is objectively a suitable means to a given end (whether subjective or
objective) if and only if it is in fact a suitable means to achieving that end, whether or
not the agent thinks it is. An action is subjectively a suitable means to achieving a
given end if and only if the agent thinks the action is a suitable means to achieving
that end. We can therefore interpret the minimal conception of instrumental
rationality in three ways:
(o)
(h)
(s)
an action is (objectively) instrumentally rational if and only if it is a suitable
means to the agent’s objective ends;
an action is (hybrid) instrumentally rational if and only if it is a suitable means
to the agent’s subjective ends;
an action is (subjectively) instrumentally rational if and only if the agent thinks
it is a suitable means to his subjective ends.
I ignore, as uninteresting, the fourth possible case, in which an action is thought by
an agent to be a suitable means to an end which happens to be an objective end of
his but which he does not acknowledge as an end.
Lester undertakes to defend a thesis, which he labels ‘instrumental rationality,’
according to which every agent seeks to maximise the satisfaction of his wants over
time (2012, 13-16). He does not defend the egoistic version of this thesis: he
concedes that people may often want to promote the interests of others. He does not
interpret the thesis hedonistically or eudaimonistically: he accepts that people often
want things other than pleasure or happiness. He also admits that people can make
mistakes about their own long-term interests. In making this point it seems that
Lester uses ‘interests’ to talk about an agent’s objective ends, and uses ‘wants’ to
talk about an agent’s subjective ends. However, generally, when Lester uses
‘interests’ he is talking about wants or, as he sometimes puts it, ‘perceived interests,’
ADVERSUS HOMO ECONOMICUS
3
or ‘self-perceived interests.’ He regards it as a priori true that ‘we are always
motivated to satisfy our greatest desire, or want, or appetite, or perceived interest’
(2012, 16). He seems to use ‘desire,’ ‘want,’ ‘appetite’ and ‘perceived interest’
interchangeably.
Lester’s a priori thesis of instrumental rationality is a conjunction of two
hypotheses. The first is a hypothesis about rationality, namely, that instrumental
rationality requires an agent to maximise the satisfaction of his desires over time,
which entails that it is irrational not to maximise the satisfaction of one’s own desires
over time. This sounds like a hybrid hypothesis of instrumental rationality because it
concerns subjective ends but it seems to be about actions which are objectively
means to their maximisation. However, from Lester’s statement that ’Decisions are
always made on the perceived balance of costs and benefits’ (2012, 25), it seems
clear that he intends his hypothesis to be about subjective instrumental rationality,
and thus about actions which the agent thinks are means to his ends (see also 2012,
15). We can state Lester’s rationality hypothesis more clearly as follows:
(LR)
an action is (subjectively) instrumentally rational if and only if the agent thinks
it will maximise the satisfaction of his wants over time.
The content of (LR) consists of the conjunction of (s) with the following two
postulates:
(a)
(b)
an agent’s subjective ends are always the ends of his desires;
a means to an end is suitable if and only if it maximises the achievement of
that end.
I will show that both of these postulates are false and that, consequently, (s) is
superior to (LR).
The second half of Lester’s a priori thesis of instrumental rationality is a
hypothesis about agency, namely,
(LA)
all actions are instrumentally rational.
This is a bold hypothesis because it rules out irrational action. Given (LR), it implies
that an agent always acts in accordance with his strongest desire. A qualification
needs to be added. For example, an agent will not act to satisfy his strongest desire
if he is chained to a wall and unable to so act. We might say that the action in
question must be possible for him. But that is not quite right, because he still might
not act if he is not chained to the wall but mistakenly thinks that he is. So we should
say that the agent will act to satisfy his strongest desire provided the action is
possible for him and he thinks that it is. Some further tweaks may be needed to
accommodate other recherché possibilities, but I will ignore all such qualifications in
what follows. I will show that (LA) is false.
Lester defends his a priori thesis against a range of objections. I will argue
that his defences do not succeed and I will offer some additional objections. The
following discussion employs ‘instrumental rationality’ in the subjective sense
throughout.
ADVERSUS HOMO ECONOMICUS
4
3. Weakness of Will
Lester’s hypothesis about agency, (LA), implies that to act is to do what one
most desires, most wants, or thinks it best to do. Yet, cases of weakness of will – for
instance, someone wants to stop eating chocolate but succumbs to the temptation to
eat another bar – involve someone wittingly doing something other than what, under
the circumstances, he most desires, most wants, or thinks it best to do. Lester
therefore declares that there are no real cases of weakness of will (2012, 24). This is
a problem because cases of weakness of will seem familiar to us all from everyday
life and have been described convincingly in many great works of literature (for
example, Tolstoy 1911). Lester tries to dissolve the problem by explaining away
apparent cases of weakness of will. For this purpose he invokes Harry Frankfurt’s
(1971) distinction between first-level desires for actions and second-level desires to
have, or not have, first-level desires.
Lester’s exposition (2012, 24-26) is not entirely clear, but I think his
suggestion is that what is really going on in a case of weakness of will is that:
(i)
(ii)
(iii)
of all the agent’s first-level desires for action in the circumstances, the agent’s
strongest desire is the desire to perform the action he does perform;
that action has consequences, or aspects, which the agent desires not to be,
and this first- level desire is greater than any second- level desire he has to
retain the desire to perform the action;
as a consequence of (ii), of all the agent’s second-level desires about the firstlevel desires in play in the circumstances, the agent’s strongest second-level
desire is the desire not to have the desire to perform the action.
For example, a ‘smoker might like not to desire smoking so much, but given that he
does desire it so much he regards himself as better off by smoking than by not
smoking. He might want to cultivate a stronger desire to stop because, for instance,
he wants to live a little longer’ (2012, 24-25). Similarly, ‘A woman strongly desires
chocolate and feels that life without it is too miserable to forgo it.’ However, ‘She
knows that chocolate makes her fat and feels that being fat is worse than losing her
desire for chocolate’ (2012, 26).
However, this alternative account of apparent cases of weakness of will is not
convincing. Contrary to (i), the smoker who suffers weakness of will fulfils his desire
to continue smoking despite the fact that he thinks he would be better off if he did
not. Similarly, the woman suffering weakness of will does not feel that life without
satisfying her desire for chocolate is too miserable to forego satisfying that desire.
On the contrary, she feels that life without satisfying that desire would be better than
life in which the desire is satisfied. These seem to be descriptions of ordinary facts of
life and ones which many smokers and chocolate-eaters offer as a description of
their situations. Lester says that when people make such assertions they are being
insincere, just giving a sop to their critics (2012, 25). That might be true in some
cases, but it seems to be false in others. Often people make such assertions, or
have such thoughts, as a prelude to taking steps to prevent themselves from
smoking or from eating chocolate. Such steps may involve other people who can
help to restrain them from actions they strongly desire to perform but which they
think it is better not to perform. For example, someone who wants to give up
ADVERSUS HOMO ECONOMICUS
5
smoking may plead with his friends and family to prevent him from getting hold of
any cigarettes.
Contrary to (iii), the weak-willed agent might not desire to be rid of the desire
to perform the action he performs. For example, the woman who eats chocolate
even though she most wants to stop eating it, may desire to retain her desire to eat
chocolate; and her desire to retain her desire to eat chocolate may be stronger than
any desire she has to be rid of the desire to eat chocolate. For, although she would
rather not eat chocolate, she may value very highly her desire to eat chocolate,
perhaps because she values having that desire for its own sake, or perhaps because
she values having that desire frustrated, either intrinsically or as evidence of her
willpower (or for some other reason). Therefore, while her first-level desire not to be
fat conflicts with her first-level desire to eat chocolate, it need not conflict with her
second-level desire to retain her desire to eat chocolate. In consequence, the
connection Lester proposes between (ii) and (iii) does not hold. Even if, as (ii)
requires, her first-level desire not to be fat is stronger than her second-level desire to
retain the desire to eat chocolate, it need not be the case, as (iii) requires, that her
strongest second-level desire is the desire not to have the desire to eat chocolate.
The phenomenon of weakness of will refutes (LA).
4. Desires and Values
Gary Watson, invoking the Platonic distinction between reason and desire,
says that desires and values constitute two different sources of motivation, the
former being non-rational, the latter rational. There is therefore a difference between
desiring something and thinking it to be of value (or judging it to be good): some
things we desire, we do not value, or the strength of the desire may be out of
proportion to the thing’s value. Only values are, in themselves, reasons for action
(1975, 208-213). For example, a woman who has a sudden urge to drown her
bawling child in the bath does not value her child’s being drowned; and
a man who thinks his sexual inclinations are the work of the
devil…bespeak[ing] his corrupt nature…does not acknowledge even a prima
facie reason for sexual activity; that he is sexually inclined toward certain
activities is not even a consideration (Watson 1975, 210).
In contrast, Lester (2012, 28-31) thinks there is only a single source of motivation, of
reasons for action. We desire something if and only if we value it; though to say that
an agent values a thing is not to spell out the nature of his desires about it. It is,
Lester claims, absurd to say that someone is inclined to do something yet does not
have even a prima-facie reason to do it.
It seems to me that this argument about whether the woman and the man, in
Watson’s examples, value as well as desire to perform the actions in question, is a
dispute over linguistic nuance. The same goes for whether we should say that each
of those agents has a pro-tanto reason for, or a consideration in favour of, the
actions. But there is a real distinction behind Watson’s more restrictive linguistic
proposal. The woman thinks that drowning her child would not be an objectively
good or valuable action, even though she desires to do it and, therefore, in some
sense, values it. The man may, in some sense, value sexual activities, given that he
desires them; but he thinks that it would be objectively bad or wrong to satisfy those
ADVERSUS HOMO ECONOMICUS
6
desires. If, with Lester, we want to say that people always value what they desire,
then we can say that the woman and the man value something they think is not
(objectively) valuable. Lester recognises (2012, note 46, 209), that this is the point
behind Watson’s proposal; but he misstates what is at issue when he says that
Watson needs an account of objective values. That is not quite right. To make sense
of the man and the woman, what Watson needs is to attribute to them a belief in
objective values; but he does not need to endorse that belief himself.
Lester seems reluctant to accept that there are objective values, and because
he mistakenly thinks that acceptance of such is required to give the account of the
woman and the man that I just gave, he offers a different account, one which
invokes, again, the distinction between levels of desire. On this account, the woman
has a desire to drown her child and a second-level desire not to have that first-level
desire; and the man has a desire for sexual activity and a second-level desire not to
have that first-level desire. But this account fails for the same reason that Lester’s
account of weakness of will failed. The man, for instance, may strongly desire that he
retain his desire for sexual activity, for he may see it as a test of his virtue or strength
of character, so that abstaining from sexual activity despite a desire for it has much
greater moral worth than abstaining from sexual activity merely because one has no
desire for it.
We have seen that, on Watson’s view, an agent has desires and values, and
only the latter give (legitimate) reasons for acting, while on Lester’s view an agent
simply has desires, which we may also call ‘values.’ Whereas Lester seems to see
all desires as appetites (2012, 16), Watson distinguishes appetitive or passionate
desires from those which are the products of culture or habituation. The latter rank
as desires rather than values for Watson because, being merely inherited, they are
not the products of the agent’s rational judgement (1975, 214-15). Both Lester and
Watson think that action results from motivation, and that motivation depends on
desires and values. It seems to me that there are many errors in these views.
I suggest that we distinguish:
(1)
(2)
(3)
(4)
an agent’s felt desires;
things an agent thinks are valuable;
things an agent thinks are valuable for him;
an agent’s goals.
Theorists have tended to conflate these, but they are distinct. A felt desire is one that
an agent feels, such as a hungering, a craving, a yearning, a longing, a lust or an
urge. This seems to be the primary sense of ‘desire.’ An agent values something if
he thinks it is valuable for him or valuable in general. This seems to be the primary
sense of the verb ‘values.’ In these senses, I may value something that I do not
desire, and vice versa. For example, if I am not hungry and lack a desire to eat, I
may nevertheless eat because I value eating for social or nutritional reasons; or I
may go to work despite feeling no desire to do so, because I think going to work is
valuable for me. Alternatively, I may be hungry and feel a strong desire to eat, but
not eat because I am on a diet and think that eating is not valuable for me; or I may
yearn to get back to work but remain at home because I am recovering from illness
and I value a swift and full recovery. However, I may also both desire and value
something, as when I eat because I have a healthy appetite. So far I have spoken of
ADVERSUS HOMO ECONOMICUS
7
what is valuable for the agent. But all except the most self-centred of people think
that many things have value even though those things are not valuable for the
person himself. For example, someone might think that classical music is valuable
for some people, and thus valuable, even though he thinks (correctly, perhaps) that it
is not valuable for him. This person may have no felt desire for classical music,
despite valuing it. Although ‘desire’ applies primarily to (1) it is also applied
metaphorically to (2) and (3); and although ‘values’ applies primarily to (2) and (3),
the term has an analogical sense in which it applies to items in (1) too.
An agent’s goal may be to perform an action of a particular type even though
performing an action of that type is neither desired nor valued by him. For example,
when I come downstairs in the mornings, while I am waiting for the kettle to boil, I
pour and drink a glass of orange juice. I do this as a matter of routine, without
thinking about it. Sometimes I only know I have drunk the orange juice because I can
see the used glass. On some such occasions I will not have had a felt desire for the
orange juice nor will I have thought that drinking it was valuable for me or valuable
for anyone else. I just drank it out of habit. Yet the action was intentional, that is,
goal-directed: I was in control of what I was doing and of whether I was doing it (I
could stop if I wanted); it was something that I aimed to do and that I might fail to pull
off, for example, by missing my mouth and spilling the juice all over my shirt; and if,
as I was about to do it, I was asked what I was going to do, I could truly have said
that I intend to drink the juice. Thus, we can have goals that we neither value nor
desire (see my 2010a for discussion and other examples, and section 5, below, for
some more substantial examples). So, (4) are distinct from (1), (2) and (3).
Nevertheless, in practice, the term ‘desire’ or ‘want’ is extended to such cases. For
example, a person who knows my habit and who sees me go, absent-mindedly, to
the fridge in the morning, may say that I want to drink some orange juice, or that I
desire some orange juice. But, in this sense, to say what I ‘want’ or ‘desire’ to do is
just to say what I aim to do. The agent’s desire to perform the action in this sense
simply follows from the fact that the action is intentional (Nagel 1970, 29-30). This is
an even more attenuated sense of ‘desire’ than that in which it is applied to (2) and
(3).
The fact that agents may have ends which are not ends of their (felt) desires
refutes postulate (a); it also refutes Lester’s rationality hypothesis, (LR), since
postulate (a) is part of the content of (LR). The refutation goes through even if we
weaken the content of (a) so that ‘desire’ covers values as well. Postulate (a) would
be saved from refutation if ‘desire’ were weakened further, to cover also aims; but it
would then become vacuous, saying only that an agent’s subjective ends are always
that agent’s aims. It also seems clear that Lester endorses the strongest of these
three versions of (a).
5. Free Will
Lester, like Watson, employs a passive conception of agency. For Lester,
each agent has a set of wants or desires, and these, in conjunction with the agent’s
hypotheses about the facts of the world, bring about the agent’s actions. The agent
always performs the action that he thinks will most satisfy his wants. For Watson, an
agent has desires and values. So long as the agent is rational, his values, in
conjunction with his beliefs about the non-value facts of the world, bring about his
actions. When he is irrational, desires of the agent which are at variance with his
ADVERSUS HOMO ECONOMICUS
8
values, in conjunction with his beliefs about the non-value facts of the world, bring
about his actions. On these views, human action is explained in terms of ‘motivation’
which impels an agent to act.
It seems clear to me, however, that this view of agency is not a conception of
agency at all. For, on such a view, the supposed action is brought about by the
agent’s motivational factors: the supposed agent does not act, but is rather the
passive recipient of impulsions which propel him hither and thither. The agent’s body
moves in response to desires or values, but the agent does not act. Curiously, Lester
contrasts his view of the agent, as following his consciously felt, self-perceived
interests or desires, with a view of the agent as an unconscious automaton without a
spontaneous will of his own (2012, 14). But how can someone who is doomed
always to follow his strongest desire be said to have a spontaneous will of his own?
If Lester’s agent can be distinguished from an unconscious automaton, it is only
because he is a conscious automaton. Similarly, Lester says: ‘Free will seems to
require only choice, and human beings always have to choose their movements
except for certain autonomic ones’ (2012, 21-22). The problem is that, if actions are
always dictated by the strongest desire, the agent has no choice.
Lester says he uses ‘free will’ in the sense of acting without being forced by
another (2012, 20). This interpersonal sense of the term is central in some
discussions in moral, legal and political philosophy, but there is another sense of the
term which is also relevant to those discussions and which is central in metaphysical
debates about personhood or the nature of agency. This is the sense in which
someone with free will has, to borrow Lester’s phrase, ‘a spontaneous will of his
own.’ It seems clear to me, and to many other philosophers (for example, Descartes
1649, I, XLI, 350; Popper 1973; 1982, xx, 113-30; 1987, 145-52; Popper and Eccles
1977, 72-74, 540-42; van Inwagen 1983) that ‘free will’ in this sense is incompatible
with determinism. This is so because an agent’s free will is incompatible with that
agent’s actions being determined by prior circumstances, whether or not those
circumstances include desires or valuations. Lester claims: ‘The school of thought
that demands a kind of free will that escapes both determinism and mere
randomness has never given an intelligible account of a third option’ (2012, 21). This
is the infamous ‘chance objection’ to free will: if my actions are not determined, they
are a matter of chance; but if they are a matter of chance, they are outside of my
control; but if my actions are outside of my control, then I do not act freely; indeed, I
do not act at all (see Hobart 1934, 346). This objection depends for its force on
confusing an undetermined act with a random event. An act, specifically, an act of
will, is something that is inherently under the agent’s control and that is therefore
undetermined (for a full explanation see my 2013b; but the general approach is also
promoted in Alvarez 2009; McCann 1974, 1975; Steward 2008, 2009; and my
2010b, sections 3 and 4).
Free will entails that an agent is capable of acting randomly and is also
capable of acting irrationally, not only from weakness of will, but also purely for the
hell of it. But the fact that we can to a large extent manage our interpersonal
interactions and relationships shows that people do not mostly act in such ways. In
large part people’s behaviour is more or less predictable. There are two reasons for
this. The first is that much behaviour is habitual. The second is that non-habitual
actions are often informed by factual, moral and other theories that we can surmise
that the agent holds, either because we share a culture with the agent, so we hold
ADVERSUS HOMO ECONOMICUS
9
the same theories ourselves, or because we guess the agent’s theories from his
particular circumstances. However, the reason that people’s behaviour is only more
or less predictable is that it is always open to an agent to act contrary to habit, or to
act contrary to the theories he holds, and even to do so rationally (see my 2010b, 2629 and my 2013a, section 3).
The great bulk of the theories any agent holds are inherited biologically or
culturally. These theories concern not only how the physical world works but also
how the social world works. They include theories about possible and acceptable
social rules, roles and relations, about what things are valuable, what actions are
obligatory, and what goals and means to them are possible, expected, acceptable or
worthy. Many of the goals an agent pursues will be inherited in the sense that the
agent pursues them because he accepts a theory about his goals which has been
handed down to him by his family, sub-group or the broader culture. Since these
inherited theories are often vague and only partly articulated, it will sometimes be an
open question for the agent, or a question he has never asked, as to whether the
goal he is pursuing is either valuable for him or valuable; and this may be so even if
he thinks that he ought to pursue that goal. In this way, many substantial goals that
agents pursue, including in some cases getting married and raising a family, or
spending the weekends getting drunk, or working in the family business, or going to
church on Sundays, are such that the agent himself pursues them without desiring
them or thinking them valuable: he is merely acting in accord with an inherited
theory, doing the done thing. Similarly, many of the things an agent holds to be
valuable, including things he holds to be valuable for himself, will be such that he has
never questioned whether they are in fact valuable: he is merely taking on trust the
truth of a theory handed down to him from his elders and teachers by the varied
processes of cultural transmission.
We noticed earlier that Watson deems as non-rational desires all those wants
that are the products of culture or habituation rather than of the agent’s rational
evaluation, his independent judgement of what is good. Despite the fact that such
inherited attitudes are often expressed in evaluative language, such as ‘divorce is
wicked,’ and that they may generate feelings of guilt when we do not abide by them,
Watson does not count them as values because they could come into conflict with
the agent’s considered judgements of value. Thus, he says, ‘an agent’s values
consist in those principles and ends which he – in a cool and non-self-deceptive
moment – articulates as definitive of the good, fulfilling, and defensible life’ (1975,
215). However, Watson’s policy would entail that all wants are non-rational and that
no agent can have values. The reason is that a rational evaluation – an independent
judgement of what is good, a considered judgement of value – cannot simply be
arbitrary: it must involve argumentation; but an argument needs premises. These
premises in turn cannot simply be arbitrary. So, if they are not taken from some of
the agent’s inherited theories, they must be products of a prior rational evaluation.
But that prior evaluation will in turn need premises. Consequently, to insist that the
premises of the rational evaluation must not be taken from the agent’s inherited
theories generates a vicious infinite regress. But if the premises of the agent’s
rational evaluation are taken from the agent’s inherited theories, then the agent’s
evaluation will be deemed by Watson to be non-rational. Watson’s rational
evaluation which is independent of inherited theories is therefore an impossibility.
Since, for Watson, an agent has values only if he does something which is in fact
ADVERSUS HOMO ECONOMICUS
10
impossible, then, on his view, no agent has values, only non-rational desires. What
this shows is that, while critical evaluation of inherited theories is necessary if we are
to improve our views (of values, facts or anything else), such re-evaluation can be
only a piecemeal reconstruction of part of an inherited framework, which uses some
parts of that framework in criticising other parts. While we can, in principle, criticise
any part of the inherited framework, we must in doing so presuppose some other
parts, and thus we cannot criticise all of it at the same time (Hayek 1963, esp. 60-63;
1970, esp. 10-11, 18-22; Popper 1948, 129-32; 1994d, 134-39). Thus, an agent’s
inherited beliefs about what is valuable must count as his values if he is to have any
values at all.
Whereas Watson deems adherence to inherited theories non-rational if those
theories have not withstood the test of rational assessment, Lester deems
adherence to inherited theories rational because he claims they have withstood the
test of rational assessment.
people must clearly perceive certain advantages in traditions, evaluate them
as superior, if only in terms of the costs and benefits associated with those
who keep them and the costs and benefits associated with those who break
them. It is true that most people do not go in for radical criticism of all customs
or habits they practise. They often give very little consideration to some of
these, its being sufficient that they are content with them and see, on
occasional reflection, no advantage to mending, at a cost and some risk, what
does not seem broken (2012, 31-32).
This mistake seems to magnify Watson’s in that it not only assumes the possibility of
a rational assessment independent of unexamined inherited assumptions, but it also
assumes that such an assessment, in however minimal a form, has been completed
even for the traditional theories that Watson deems to be held non-rationally.
Since agents necessarily have free will, and are thus free to act irrationally,
(LA) is false. I am not claiming here that we have free will. I claim that, if determinism
is true, then there are no agents; and that if Lester’s theory of motivation were true,
we would not be agents. Since agents may choose ends either contrary to their
desires or for which they have no desire (in the primary sense of felt desire, which is
the sense of the term that Lester favours), then Lester’s postulate (a) is false and
thus (LR) is also false, since it entails (a).
6. Self-Interest
Lester (2012, 36-39) contends that everyone is self-interested in that he
pursues his own interests as he sees them (he pursues his ‘self-perceived interest’).
But Lester distinguishes this from psychological egoism. In order to avoid using the
obscure (though popular) Kantian jargon of people being ‘ends in themselves,’ I will
state Lester’s point as follows. In addition to his first-level interests, each person also
has second-level interests, that is, interests in what first-level interests are fulfilled.
When people act consciously, they are pursuing their second-level interests.
Psychological egoism states that each person’s second-level interests are
concerned only with the fulfilment of his own first-level interests. Psychological
altruism states that some people’s second-level interests are sometimes concerned
with other people’s first-level interests. Lester’s thesis of self-interest says that each
person pursues his own second-level interests. This is consistent with psychological
ADVERSUS HOMO ECONOMICUS
11
altruism because some people may have second-level interests in other people’s
first-level interests. Thus, second-level interests belong to the self but they may
concern others. Obviously, there is scope for agents to have higher-level interests
and to pursue those consciously, but the preceding can be taken as the basic case
on which further complications are built.
It seems clear to me that this talk of ‘interests’ is far too amorphous and
woolly to be useful if our aim is to obtain anything approaching a clear or precise
understanding of agency or instrumental rationality. The classification I introduced in
section 4, with another category added in the light of our discussion in section 5,
seems to be far more illuminating because it permits us to distinguish cases that
Lester’s scheme lumps together. We distinguish:
(1)
(2)
(3)
(4)
(5)
what an agent desires;
what an agent thinks valuable;
what an agent thinks valuable for himself;
what an agent aims at, his goals;
what an agent thinks he has an obligation to do.
We earlier acknowledged that items in (2), (3) or (4) may also be called ‘wants’ or
‘desires’ in an attenuated sense, and that items in (1) may also be called ‘values’ in
an attenuated sense. Since people generally think that it is often valuable to fulfil
one’s obligations, items in (5) may also be called ‘values’ and thus also ‘desires.’
We can acknowledge that any of those five things may also be called ‘interests,’ and
perhaps in the same sense, since the term ‘interest’ seems naturally exiguous, unlike
‘want’ and ‘desire’ which seem naturally applied to appetites and are only
metaphorically extended to the other cases. But to use the same term for different
cases slurs over differences that can be important. It achieves a simpler theory but,
rather than advancing our understanding, it seems a barrier to it.
For example, Fred goes out of his way to cheer up a friend, despite the fact
that it means missing a concert for which he has a ticket. We may distinguish the
following five possibilities, amongst others:
(A)
(B)
(C)
(D)
(E)
Fred has a longing for the music he would hear at the concert and an urge to
help his friend, but the urge is stronger than the longing;
as in (A) except that the longing is stronger, but Fred thinks it will be more
valuable for himself to satisfy the (weaker) urge;
Fred has a longing for the music, and no appetite for helping his friend, and
he thinks it will be more valuable for himself to go to the concert than to help
his friend, but he thinks that his helping his friend will have a value for her that
is greater than the value for him of attending the concert;
as in (C) except that Fred has no idea whether his helping his friend will have
a value for her that is greater than the value for him of attending the concert;
as in (D), but Fred thinks he has an obligation to help his friend.
In (A), Fred does what he most desires. In (B), Fred does what he thinks is most
valuable for himself, but not what he most desires. In (C), Fred does what he thinks
is most valuable, even though it is not what he most desires and not what he thinks
most valuable for himself. In (D), Fred helps his friend even though it means
ADVERSUS HOMO ECONOMICUS
12
sacrificing an alternative which he desires more strongly and thinks is more valuable
for himself, and even though he has no idea which of the two alternatives is,
objectively, most valuable. In (E), Fred does what he thinks he is obliged to do, even
though he does not desire to do it, and even though there is an alternative that he
desires and he thinks is more valuable for him, and even though he does not know
what value his chosen action has. What does Lester say about these different
cases? In each case he says that Fred acted in his self-interest. Fred’s helping his
friend was in his self-interest because it was what he most desired (in (A)), or most
valued (in (B)), or aimed to do (in (C) and (D)), or felt obliged to do (in (E)). What
does it contribute to our understanding to be told that there is a flimsy (indeed,
vacuous) level of description at which all these different cases can be described in
the same way? Nothing that I can see.
It might be objected that, in cases (C), (D) and (E) it is valuable for Fred to be
the kind of person who makes a sacrifice to help a friend in need or to fulfil an
obligation (compare Schmidtz 1995, 111-116). Well, that is one theory; and it might
be a true one. But it need not be a theory Fred holds when he makes his decision.
Fred might never have thought about whether it is valuable for him to be that kind of
person; and if asked whether it is, he might sincerely answer that he does not know;
or, even if he says that it is valuable for him to be that kind of person, he might have
no idea whether the value for him of being that kind of person exceeds the loss in
value for him of missing the concert. Whether at some deeper level, or over the long
term, it pays Fred to make this current sacrifice is also something Fred need not
have considered when making his decision, and it might be something about which
he has no opinion. In fact, he might have made that decision whatever the answer to
that question might turn out to be.
In response to a somewhat different complaint from Amartya Sen (1977) that
self-interest theories of action are vacuous, Lester asks three rhetorical questions:
How can we choose to do what we do not in some sense prefer to do? Must
not the chosen alternative be better for us in some sense? Otherwise, where
is the personal motivation? (2012, 42)
But these questions, intended as rhetorical, have more or less obvious answers
contrary to those assumed by Lester. First, I can do what I do not prefer to do, in the
sense that I do not desire to do it, or in the sense that I do not value doing it. Of
course, if one stipulates, as Lester seems to do (2012, 42), that to act is to reveal a
preference, then there will always be that sense in which whatever I do is what I
prefer to do; but that is trivial. Second, an agent may choose to do something that is
less valuable for him than an available alternative because it is more valuable for
someone else (or for some other reason or none). Third, we often act intentionally
without motivation, in that we neither desire nor value what we do. Further, when we
do desire or value what we do, the desire or valuation does not make us do it: our
intentional actions are undetermined at every time before they are begun. Lester’s
treatments of the objections of C. D. Broad and Tibor Machan (2012, 43-46) seem
similarly question-begging.
Returning to Fred, we might wonder whether, in each of the five possibilities
distinguished, Fred acts rationally, in the minimal sense of subjective instrumental
rationality, (s). However, we cannot answer this question unless we know what
ADVERSUS HOMO ECONOMICUS
13
Fred’s aim was. It is easy to construe four of the five possibilities as instrumentally
rational behaviour by attributing to Fred an appropriate aim, thus: in (A), to do what
he most desires; in (B), to do what is most valuable for himself; in (C), to do what is
most valuable; in (E), to do what he is obliged to do. In (D), we might postulate that
Fred thinks he is obliged to help his friend; but that would reduce the case to (E). If
we maintain (D) as a separate case, it seems that the only aim we could attribute to
Fred to render his action instrumentally rational is the aim of helping his friend. But
that raises the question of whether that aim is rational, given the absence of desire,
higher valuation and assumed obligation. We might say that, while Fred’s action was
locally rational (relative to his immediate aim), it is doubtful that it was globally
rational (relative to the aims that Fred regards as most important). If the question
occurs to Fred, it may prompt him to undertake the sort of piecemeal reappraisal of
his ends discussed in section 5, above.
7. Maximisation
According to Lester, as agents, we seek to maximise our want-satisfaction in
the sense that, ‘as we compare possible choices we cannot help but take the option
that in some way feels to be the most want-satisfying, or least want-dissatisfying, at
the time.’ This is what it means to feel the greatest desire for the option finally
chosen (2012, 50-51).
There are a number of problems with this view, including the following two,
which were pointed out by Herbert Simon. First, most of our decisions are habitual or
conventional rather than reasoned (Simon 1997, 102-103, 107-109), so what is
chosen is the usual rather than the best or the most want-satisfying. Second, even in
cases where we reason about options, we usually cannot identify all the options, all
their consequences or all the relevant evaluative principles (1997, 73-75, 77, 93-97),
so we often cannot identify the best, or the most want-satisfying. Simon called this
‘bounded rationality’ and he insisted that, as a consequence, decision-makers
generally ‘satisfice,’ or look for a course of action that is good enough, rather than
seeking to maximise, or look for the option which is best (1997, 118-129). In option
appraisals, managers in business and in the public-sector typically attempt to
construct very simplified pictures of reality which seem likely to enable them to
identify, and thus rule out, the more disastrous options. The choice amongst the
options that remain is then often rather arbitrary (Simon 1997, 264-65; Peters and
Waterman 1982, 29-54).
Lester retorts to Simon, in a footnote, that apparent satisficing is really
maximising because, in such cases, at some point we guess that the disutility of
search costs is likely to outweigh any other utility that we will achieve (2012, note 63,
210). This is a familiar response to Simon; but it seems to be a wholly inadequate
one. We saw this in connection with habitual and conventional actions in section 5;
but it is also so with regard to option appraisals, for the following reasons. First,
suppose that Lester were right that the decision-maker is able to say that further
search will be more costly than it is worth. This does not alter the fact that the
decision-maker would still not have enough information about the options before him
to be able to say which is better than the others. Second, when Lester says that the
decision-maker guesses that further search would not be worthwhile, he might mean
that the decision-maker just makes an arbitrary, unreasoned decision to search no
farther. But that would be to concede Simon’s point, because to decide arbitrarily to
ADVERSUS HOMO ECONOMICUS
14
search no farther just is to decide to satisfice rather than to maximise. Third, if Lester
means, not that the decision-maker decides arbitrarily to search no more, but that he
calculates whether further search is worthwhile, then we get a vicious circle. Call the
options for solving the decision-maker’s problem ‘first-level options.’ Call the options
for gathering information about first-level options ‘second-level options.’ The
decision-maker’s situation is that he does not have enough information about the
first-level options before him to be able to say that one of them is better than the rest,
and he does not know whether there are additional first-level options that would be
better than those he has before him. To decide whether to search for more
information, he must, on Lester’s view, evaluate second-level options. But he cannot
know what all the second-level options are, since there may be (and doubtless are)
sources of information he does not know about. Even for the second-level options he
can identify, he cannot know in advance what all the search costs will be. And,
crucially, he has no way of identifying the potential benefits of additional information
without knowing what that information will tell him about the costs and benefits of the
first-level options he is investigating (that is, those options he is investigating
precisely in order to discover their costs and benefits). To know the benefits of the
second-level options, he must know the costs and benefits of the first-level options.
So, on Lester’s view, the decision-maker has to solve the second-level problem
before he can solve the first-level problem; but he has to solve the first-level problem
before he can solve the second-level problem.
Since agents are often unable to identify an option which would maximise the
achievement of their ends, then, according to Lester’s postulate (b), they are often
unable to identify a suitable means to their end, which means that, according to (LR),
they often fail to exhibit instrumental rationality in their actions. That is inconsistent
with (LA).
8. Morals
We have seen that Lester thinks that every agent is self-interested in that he
is bound to want only things that he values; but Lester contends that some of the
things an agent values may be altruistically or morally valued (2012, 41). For Lester,
recall, to value something is to desire it, so he attributes to agents moral desires or
‘sentiments’ (2012, 51). Lester views morality as impartial in that moral principles are
not concerned with the interests of a specific agent, but partial in that moral
principles must motivate agents who sincerely hold them (2012, 39-40):
fully to hold a moral obligation sentimentally, not to feel it uncertainly or as a
slight pricking of the conscience, is always to act on it in appropriate
circumstances…It is possible to defend moral theories intellectually without
really feeling them. Without seeing this, one can fail to realize that one’s
‘official’ or ‘theoretical’ moral position is a sort of public recommendation that
one might not personally feel, value, or desire…we cannot knowingly do what
we feel, at that moment, is immoral (2012, 51-52).
Lester’s position here seems to fly in the face of human experience. I
seriously doubt that that there has ever been, or will be, any person who has not on
many occasions acted in a way that he is at the time of acting convinced is wrong.
One type of case involves weakness of will, but in many cases there is no such
weakness because the agent is resolute in pursuing the course he is convinced is
ADVERSUS HOMO ECONOMICUS
15
wrong and perhaps steels himself to do it. The varied situations and ways in which
this happens has been explored extensively in literature, and also in film, and its
effects have been analysed in some psychological studies of cognitive dissonance
(see Tavris and Aronson 2007). Typically, the person who knowingly does wrong
tries to dissipate the discomfort he feels by seeking a way to justify himself: ‘It is not
so bad,’ ‘the circumstances are exceptional,’ ‘I have a good excuse,’ ‘other people
would do the same in my position,’ and so on. He tries to delude himself into thinking
that he has not really done wrong at all. This, indeed, may be the truth behind
Lester’s claim: not that no one knowingly does wrong, but that most people who
knowingly do wrong succeed in deluding themselves that they have not done so.
Lester refers to Hare for further argument for his position. The view of Hare
(1952) is that the primary function of words such as ‘good’ and ‘right’ is to commend
or prescribe, but that they also have a secondary function to indicate the presence of
those non-value properties the presence of which inclines us to say that something is
good or right. On this view, there is no property of goodness itself or rightness itself;
there are just non-value properties and acts of prescription. Since moral judgements
are universal, one’s moral prescriptions for others are also moral prescriptions for
oneself if one is in the same circumstances. Thus, someone contemplating an action
which he sincerely says (or thinks) is immoral, commands himself not to perform that
action, and thus he does not perform it. One might question why someone may not
do something that he simultaneously commands himself not to do, but the objection I
want to raise here is that we can discuss dispassionately whether a thing is good or
right, without making any prescriptions; indeed, we can even prescribe what we think
is not good or not right. Hare’s response to this objection is to say that, in such
cases, we are not making genuine value-judgements, but are rather making an
‘inverted-commas use’ of the terms ‘good’ or ‘right,’ that is, we are using the terms to
refer to the non-value properties which incline some people to call things ‘good’ or
‘right’ (1952, 163-70). However, Hare is here merely re-describing the examples to fit
his theory. Perhaps this can be seen most clearly if we consider a character like the
devil, who may prescribe what he thinks to be bad or wrong. Hare has to treat this as
an inverted-commas use of ‘bad’ or ‘wrong’ (1952, 175). But that is mistaken, since
what the devil intends to prescribe is not whatever some people consider to be bad
or wrong; rather, being the devil, he intends to prescribe what is really, objectively,
bad or wrong. If he says ‘what is good is bad,’ he is using one term descriptively, to
talk of what is good (or bad) and the other term to deprecate (or commend); and
similarly if he says ‘the right thing to do is the wrong thing to do.’ We can understand
this easily enough, because there is a bit of the devil in each of us, though some
people seem to have more of it than others. Thus, while it is true that terms such as
‘good’ and ‘right’ may be used by speakers to prescribe, we should acknowledge that
this is a secondary use, their primary use being to ascribe a value property. Hare’s
view is mistaken because it omits mention of the properties of goodness and
rightness.
As so often, Lester attempts to bolster his case by asking a rhetorical
question: ‘can you recall doing anything that clearly felt immoral at the time that you
did it?’ His assumption, of course, is that the answer is ‘no.’ However, my answer is
‘yes.’ I can recall many examples of my doing things that I clearly thought were
immoral. Anyone else could do the same, provided he has not deluded himself with
the self-justifying chicaneries of cognitive dissonance.
ADVERSUS HOMO ECONOMICUS
16
On Lester’s mechanistic theory of motivation, the supposed fact of the
impossibility of doing what one knows to be wrong is explained by positing that moral
desires are always stronger than non-moral desires (2012, 52). He says:
what is felt to be immoral is what we feel no one should ever do in the
circumstances; it is a categorical sentiment…One source of confusion here is
where our general moral feelings (such as feeling that lying is usually
immoral) differ from our specific moral feelings (such as feeling that some
particular lie is moral). This seems to occur because such general morals are
usually held ceteris paribus (2012, 52-53).
Is it possible to have a feeling with so reticulated a propositional content as no one
should ever do an action of type A in these circumstances (where an appropriate
action-description is substituted for ‘A’) or lying is usually immoral or, even more
implausibly, ceteris paribus, no one should ever do an action of type A in these
circumstances (where an appropriate action-description is substituted for ‘A’)? The
supposition seems ridiculous. Of course, one might say, ‘I feel that no one should
ever do that sort of action in these circumstances.’ But there one is not using ‘feel’
literally, to talk about a desire or sentiment, but to indicate one’s uncertainty about
the thought one is expressing. It is, indeed, possible to have an attitude toward a
complex thought, for example, one might admire it or feel happy at the contemplation
of it, but the feeling itself does not have the structured complexity of the thought.
Lester seems driven to supposing that it does, not simply by his mechanistic theory
of motivation, but by that in conjunction with his attachment to construing values as
felt desires.
9. Conclusion
Lester articulates and defends two hypotheses: that all agents are
instrumentally rational whenever they act (LA); and that an agent is instrumentally
rational if and only if he seeks to maximise the satisfaction of his own wants (LR).
We have seen that (LA) is false because it cannot accommodate the phenomenon of
weakness of will and it involves a passive conception of agency which leaves no
place for the agent. We noted that (LR) can be analysed as a conjunction of a
minimal conception of instrumental rationality
(s)
an action is (subjectively) instrumentally rational if and only if the agent thinks
it is a suitable means to his subjective ends
and two postulates
(a)
(b)
an agent’s subjective ends are always the ends of his desires;
a means to an end is suitable if and only if it maximises the achievement of
that end.
We have seen that postulate (a) is false because agents often have ends that they
value but do not desire (in the primary sense of felt desire) and because they often
have ends that they neither value nor desire; and that postulate (b) is unacceptable
since agents are often unable to identify an option which would maximise the
ADVERSUS HOMO ECONOMICUS
17
achievement of their ends. This entails that (s) is a better account of instrumental
rationality than (LR) and that, if (LR) were true, then (LA) could not be true.
Lester also defends two other theses: that every agent always acts out of selfinterest and that no agent can do what he genuinely feels to be immoral. We have
seen that the first thesis can be maintained, but only in a sense in which it is quite
vacuous; and that the second thesis is false.
Lester says that we cannot understand people’s behaviour unless they are
instrumentally rational (2012, 15, 16). That is true; but we need only the minimal
conception of instrumental rationality, (s), to make sense of intentional action. The
fact that people are often instrumentally irrational means that people are often not
predictable; but this does not threaten theoretical social science because its
generalisations are either explicitly statistical or their truth requires only that the vast
majority of agents act instrumentally rationally in response to a given change (see
my 2013a, section 4).
At a number of places Lester exhibits a concern to avoid giving
encouragement to paternalism. When he admits that people can be mistaken about
their own interests, he explains why he thinks this does not provide an argument for
paternalism (2012, 13). It also seems that he regards acknowledging the possibility
of weakness of will as opening the door to coercion of people for their own good
(2012, 19, 27, 33-34). However, my argument should not be taken to be making a
case for paternalism. It is true that a person may be mistaken about his own
objective ends; but it is true a fortiori that other people can be mistaken about that
person’s objective ends. A central task of any agent’s life is to discover his own
objective ends through experiments in living; and this requires freedom. Similarly, an
agent is better placed than an observer to recognise whether an action of his is due
to weakness of will or to personal idiosyncrasy; and if he wants assistance in
overcoming weakness of will, that is his call to make. Finally, that people often do
what they think to be wrong indicates that there is a propensity to evil in each of us,
so that the most beneficial social arrangement will be one that disperses power and
gives quick feedback, positive or negative, on actions performed, which is what open
markets do and what overbearing government frustrates.
I have argued that Lester’s homo economicus accounts of rationality and
agency are false. This is not to say that they are worthless. In economics, homo
economicus can be defended as a simplifying assumption which has illuminating
consequences which enable us to make predictions. Such simplified models form
part of every science (Popper 1994b, 162-70). In contrast, an argument for freedom
requires a realistic conception of the agent, because it must show that freedom is
valuable for real agents. It does not seem to be much use showing that freedom
would be great for homo economicus, if real people are not much like homo
economicus. To defend freedom we need to show that it would be good for real
people with limited rationality who behave largely habitually, who have values and
obligations as well as desires, and aims which may be none of those, who learn by
modifying inherited theories, who often act irrationally and also in ways they think to
be immoral, who are remarkably ignorant and fallible, and who often behave
unpredictably.
ADVERSUS HOMO ECONOMICUS
18
References
Alvarez, Maria. 2009. ‘Actions, Thought-Experiments and the “Principle of Alternate
Possibilities.”’ Australasian Journal of Philosophy 87/1: 61–81.
Descartes, Rene. 1649. ‘The Passions of the Soul.’ In his Philosophical Works, Vol. 1, trans.
E. S. Haldane & G. R. T. Ross, corrected edition (329-427). Cambridge: Cambridge
University Press (1931).
Frankfurt, Harry. 1971. 'Freedom of the Will and the Concept of a Person.' The Journal of
Philosophy 68 (1): 5-20.
Frederick, Danny. 2010a. ‘Unmotivated Intentional Action.’ Philosophical Frontiers 5 (1): 2130. Available here:
http://independent.academia.edu/DannyFrederick/Papers/311267/Unmotivated_Inten
tional_Action
Frederick, Danny. 2010b. ‘Popper and Free Will.’ Studia Philosophica Estonica 3.1: 21-38.
Available here: http://www.spe.ut.ee/ojs-2.2.2/index.php/spe/article/view/93
Frederick, Danny. 2013a. ‘Popper, Rationality and the Possibility of Social Science.’
THEORIA 28 (1): 61-75. Available here:
http://www.ehu.es/ojs/index.php/THEORIA/article/view/1879
Frederick, Danny. 2013b. ‘Free Will and Probability.’ Canadian Journal of Philosophy 43 (1):
60-77. Available here: https://www.academia.edu/348757/Free_Will_and_Probability
Hare, R. M. 1952. The Language of Morals. Oxford: Clarendon Press.
Hayek, Friedrich. 1963. ‘Rules, Perception and Intelligibility.’ In his Studies in Philosophy,
Politics and Economics (43-65). London: Routledge and Kegan Paul (1967).
Hayek, Friedrich. 1970. ‘The Errors of Constructivism.’ In his New Studies in Philosophy,
Politics, Economics and the History of Ideas (3-22). London: Routledge and Kegan
Paul (1978).
Hobart, R. E. 1934. ‘Free Will as Involving Determination and Inconceivable Without it.’ In P.
van Inwagen & D. W. Zimmerman (eds.), Metaphysics: the Big Questions (34355).Oxford: Blackwell (1998).
Lester, Jan. 2012. Escape from Leviathan: Libertarianism without Justificationism,
paperback edition. Buckingham: University of Buckingham Press.
McCann, Hugh. 1974. ‘Volition and Basic Action.’ In his 1998 (75-93).
McCann, Hugh. 1975. ‘Trying, Paralysis and Volition.’ In his 1998 (94-109).
McCann, Hugh. 1998. The Works of Agency. Ithaca: Cornell University Press.
Nagel, Thomas. 1970. The Possibility of Altruism. Princeton: Princeton University Press.
Peters, Tom and Waterman, Robert. 1982. In Search of Excellence. London: Harper and
Row.
Popper, Karl. 1948. ‘Towards a Rational Theory of Tradition.’ In his Conjectures and
Refutations: the Growth of Scientific Knowledge, fourth (revised) edition (120-135).
London: Routledge and Kegan Paul (1972).
Popper, Karl. 1973. ‘Of Clouds and Clocks.’ In his Objective Knowledge, corrected edition
(206-55). Oxford: Oxford University Press.
Popper, Karl. 1982. The Open Universe. London: Routledge.
Popper, Karl. 1987. ‘Natural Selection and the Emergence of Mind.’ In G. Radnitzky & W. W.
Bartley (eds.), Evolutionary Epistemology, Rationality, and the Sociology of
Knowledge (139-55). La Salle, IL: Open Court.
Popper, Karl and Eccles, John. 1977. The Self and its Brain. London: Springer-Verlag.
Popper, Karl. 1994a.‘The Myth of the Framework.’ In his 1994c (33-64).
Popper, Karl. 1994b. ‘Models, Instruments, and Truth.’ In his 1994c (154-84).
Popper, Karl. 1994c. The Myth of the Framework. London: Routledge.
Popper, Karl. 1994d. Knowledge and the Body-Mind Problem. London: Routledge.
Schmidtz, David. 1995. Rational Choice and Moral Agency. Princeton: Princeton University
Press.
ADVERSUS HOMO ECONOMICUS
19
Sen, Amartya. 1977. ‘Rational Fools.’ Philosophy and Public Affairs 6 (4):317–44.
Simon, Herbert. 1997. Administrative Behavior, fourth edition. New York: The Free Press.
Steward, Helen. 2008. ‘Moral Responsibility and the Irrelevance of Physics.’ Journal of
Ethics 12/2: 129–145.
Steward, Helen. 2009. ‘Fairness, Agency and the Flicker of Freedom.’ Nous 43/1: 64–93.
Tavris, Carol and Aronson, Elliot. (2007). Mistakes Were Made (But Not By Me). New York,
Harcourt.
Tolstoy, Leo. 1911. The Devil. In his The Devil and Other Short Stories, translated by Louise
and Aylmer Maude, revised by Richard Gustafson, 191-236. Oxford: Oxford
University Press (2003).
van Inwagen, Peter. 1983. An Essay on Free Will. Oxford: Oxford University Press.
Watson, Gary. 1975. ‘Free Agency.’ Journal of Philosophy 72 (8): 205-220.