Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
Danish Yearbook of Philosophy, Vol. 43 (2008), 69-96
DEONTOLOGY – BORN AND KEPT IN SERVITUDE
BY UTILITARIANISM1
Asger Sørensen
Department of Philosophy of Education, School of Education, University of Aarhus, Denmark
ABSTRACT. The distinction between teleology and deontology is
today almost universally accepted within practical philosophy, but
deontology is and has from the beginning been subordinate to utilitarianism. ‘Deontology’ was constructed by Bentham to signify the
art and science of private morality within a utilitarian worldview.
The classical distinction was constructed by Broad as a refinement
of Sidgwick’s utilitarianism, and then adopted by Frankena. To
Broad it signified two opposite tendencies in ethics, in Frankena’s
textbooks, however, it becomes an exclusive distinction, where deontology signifies disregard for consequences, and it is therefore
almost impossible to think of deontology as a framework for a comprehensive ethical theory. This conception, however, is adopted by
Rawls, and in his contractarian interpretation of deontology it is in
fact no more within the sphere of ethics.
Introduction
One of the most generally acknowledged distinctions in ethics is the distinction
between deontology and teleology. The concept of deontology plays a crucial
role in analyses and discussions not just in ethics, but also in political philosophy, philosophy of law and various sciences. Nevertheless, very few scholars
have given the concept of deontology the close attention that such prominence
should merit. This article will contribute to filling out this gap by arguing that
the widespread understanding of deontology as formally opposed to – and thus,
in a sense equal to – teleology or consequentialism is not just incomplete, but
actually misleading. Through an analysis of different conceptions of ‘deontology’, I will argue that deontology as a category is and has always been subordinate to a utilitarian frame of mind; therefore, opponents of utilitarianism or
consequentialism like John Rawls should not identify themselves as deontologists. The main figures in this analysis are Bentham, Broad and Frankena, but
neither my claim nor my arguments are primarily historical. Before venturing
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
70
asger sørensen
into the main part of my analysis, I will therefore elaborate on the systematic
perspective.
In mainstream practical philosophy it is widely accepted that the roots of the
modern concept of deontology are found in Broad’s distinction between deontology and teleology in his book Five Types of Ethical Theory (1930). From its
origins as a relatively esoteric analytical term in pre-war Anglo-Saxon ethics,
‘deontology’ became institutionalized as an important ethical category in the
textbooks of P.H. Nowell-Smith (1954) and William K. Frankena (1963 and
1974). Since these publications deontological theories have been considered to
be opposed to teleological theories in Anglophone ethics, the standard example
of the former being the ethics of Kant (cf. e.g. Nowell-Smith 1954, 134; Frankena 1963, 25; Hallgarth 1998, 610), whereas the latter usually is exemplified by
some kind of utilitarianism. ‘Deontology’ in this sense was the point of departure for the political philosophy of Rawls (1971/99, 26), and it was also in this
sense understood on the continent by Jürgen Habermas (cf. e.g. 1991, 168).
However, a brief survey of ‘deontology’ in non-Anglophone philosophical
encyclopedias and dictionaries (cf. e.g. Ferrater Mora 1994, 816; Lalande
1991, 816; Canto-Sperber 1996, 401; Ritter 1972, 114) reveals an older, more
basic conception of ‘deontology’, which is derived from the Greek words ‘to
deon’ and ‘logos’. The latter should not cause us any trouble, and the former
can be translated as “that which is proper” or “what ought to be”. As a first approach to the meaning of ‘deontology’ it therefore seems reasonable to define
it as “the teachings or science of what is proper and what ought to be”, in short,
“the science of duty”. This conception can be encountered in modern ethics,
especially on the continent, as demonstrated by the dictionaries just mentioned,
but also in the Anglophone world (cf. e.g. Muirhead 1940, Carmichael 1949,
Campbell Garnett 1956 and Nowell-Smith 1954, 185).
What is even more surprising, however, is that the man who constructed the
neo-logism ‘deontology’ apparently was the same one, who invented utilitarianism, namely Bentham himself. Bentham even left an unfinished manuscript,
which was published in 1834, with the title Deontology, which was thought to
complete utilitarianism as a general worldview. The manuscript stated, for the
first and only time, the specifics of Bentham’s utilitarianism as moral philosophy.2 For Bentham, ‘deontology’ was therefore not at all intended to be opposed to utilitarianism; quite the contrary.
Originally, deontology was constructed as a part of utilitarianism; today they
are seen as opposed. What shall we make out of this shift? Is it just a curious
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
71
fact, which shows the contingency of semantical meaning and the irony of history?3 Or can it give us a hint of something more substantial? I will argue that
this curious historical fact about deontology does indeed express something of
substantial conceptual interest, namely that in the distinction between deontology and teleology, deontology is and has always been the subordinate and relatively less important aspect. From the very beginning, the meaning of ‘deontology’ has been tied to that of ‘utilitarianism’ and this is still the case, although
both the meaning of utilitarianism, of deontology and the content of the relation between the two terms have changed.
Considering the teleological aspect, utilitarianism has been criticised by
moralists ever since its conception more than 200 years ago (Schneewind 1977,
128-30), but nevertheless it has dominated Anglophone ethics for decades
(Schneewind 1993, 155). Today it seems more fitting than ever, and this is reflected in the development of the vocabulary used in textbooks and academic
discussions. In accordance with its etymological roots (telos = ‘goal’ or ‘end’),
teleological ethics is taken to stress the importance of the end in moral action
and ethical theory. From this perspective, utilitarianism can be grouped together with the ethics of Aristotle (cf. Edel 1973, 175), and this is also how
Rawls considers it (1971/99, 22, 35). As the standard counter-position of ‘deontology’, however, the traditional category ‘teleology’ is today in most cases
replaced by the term ‘consequentialism’ (cf. e.g. Stegmüller 1989, 227), a term,
which is even more closely related to utilitarianism than is teleology. Using
‘consequentialism’ implies a direct identification of the general ethical category with the utilitarian scheme of thought, since utilitarians explicitly state that
only the consequences of an action – in the classical view, the resulting happiness – should count as reasons to decide whether an action is dutiful or not.4
This terminological development is also reflected in the other aspect of the
distinction, where ‘deontology’ nowadays often is substituted by ‘non-consequentialism’, indicating – sometimes explicitly – that it is hard to give deontology a positive meaning, and therefore that it is best understood in contrast to
consequentialism (Davis 1993, 206). Furthermore, one can point to the common understanding of deontology in terms like ‘agent-relativity’ (cf. e.g. Ellis
1992, 856), ‘deontological constraints’ (e.g. Nagel 1986; Kymlicka 1988, 180),
or ‘agent-centred restrictions’ (e.g. Scheffler 1984; Brook 1991, 190), which
all tend to reduce deontology to a set of exceptions, which must be taken into
account by ethics, but which cannot be understood in a sufficiently coherent
way to constitute in itself a theoretical approach to philosophical ethics.
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
72
asger sørensen
I will argue that from the days of Bentham until today, the structural position
of deontology in relation to utilitarianism has remained the same, even though
the referential content of the two terms has changed. This way of construing
the relation reflects a more general point, which is central to the argument of
this article. Every word is one aspect of one or more distinctions, and every
distinction has two or more aspects, which are relative to each other. A distinction is, however, not just given. It is always the answer to a specific concern
and therefore expresses a special point of view. This makes it possible to question the universal validity – or applicability – of a distinction, even when it is
almost universally accepted, as it has been shown by American pragmatists in
relation to analytical distinctions like a priori/a posteriori (cf. C. I. Lewis), and
analytic/synthetic (cf. Quine).
I will, however, go one step further and follow Derrida in claiming that since
every distinction has a special point of view, it both has a focus of attention and
a horizon, a centre and a periphery (cf. e.g. Derrida 1968, 128). As such it is not
only contingent, but it is also basically asymmetrical, and the revelation of the
hierarchical pattern is at the same time an exposure of the preconditions, which
make the distinction meaningful. Consequently, a distinction can be seen as an
expression of a relation of power. In the distinction in question, I will claim
that deontology plays the subordinate part, conceived within and still dominated by the utilitarian scheme of thought.5
This argument becomes important when the intention is to formulate alternatives to utilitarianism within practical philosophy. Expressing this resistance in terms of a positive identification with deontology – as done by Rawls
– is to stay within a conceptual framework defined by utilitarianism. No matter how good one’s intentions, or how excellent one’s terminological skills
and moral insight may be, identifying positively with deontology makes it
almost impossible to counter the utilitarian programme and its domination in
ethics.
In this article, developing on an earlier suggestion (Sørensen 2003, 42-49), I
analyze three classical concepts of deontology from the perspective sketched
above. First, I present Bentham’s original concept of deontology, which is
closely connected to utilitarianism (1). Second, as the main part, I analyse the
classical teleology-deontology distinction as conceived of by Broad (2), and
developed by Frankena (3), where deontology as the disregard of consequences is considered irrational in relation to an overall conception of ethics determined by teleology and utilitarianism. Finally, I present the distinction in the
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
73
form which is accepted by Rawls in A Theory of Justice, and how it makes egoism the basis of his deontological political philosophy (4).
Even though my approach is not historical, I will present the main conceptions of deontology in chronological order. However, I will not investigate the
historical transformations of the concept, but focus on each particular conception of deontology, in order to demonstrate the specifics of each particular
subordination to utilitarianism.
1. Bentham
The foundation on which Bentham constructed the neo-logism ‘deontology’ was
the Greek words mentioned above, and its basic meaning is the ‘science of duty’.
This meaning is in accordance with what Bentham himself writes in his pedagogical treatise Chrestomathia from 1817, where ‘deontology’ is defined as
an account or an indication of that which, on the occasion in question, whatsoever it be, is
– (i.e. by him who speaks or writes is regarded as being) – fit, fitting, becoming, proper.
(Goldworth 1983, xx)
This definition (which is one of the few examples of Bentham’s use of the term
in texts printed while he was still alive) can be understood as containing both a
theoretical and a practical aspect in the Aristotelian sense of the words. Giving
an ‘account’ can be merely theoretical; but ‘indicating’ what one considers to
be proper in a particular situation is at the same time an evaluation or a recommendation, and as such an ‘indication’ is always practical, i.e. ethical or political.
Accordingly, the Deontology is divided in two parts, an “exegetical”, “expository” part, normally called the theoretical part and a more “practical part”.6
Both parts, however, are meant to be conducive to the “ultimate and practical
result” of “this work”, which is
the pointing out to each man on each occasion what course of conduct promises to be in the
highest degree conducive to his happiness: to his own happiness, first and last; to the happiness of others, no farther than in so far as his happiness is promoted by promoting theirs,
than his interest coincides with theirs. (Deont., 123)
The theoretical part relates virtues and vices to happiness or “well-being” (Deont., 130), as Bentham prefers to call it here. Virtues and vices are “fictitious
entities, imagined and spoken of as real for the purpose of discourse”, without
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
74
asger sørensen
which, however, “discourse on subjects such as this could not be carried on”
(Deont., 126). Accordingly, Bentham wants to explain virtues and vices in
terms of the only real entities, pleasure and pain. Hence, he will show that all
virtues can be seen as “modifications of two all comprehensive ones”, “prudence and benevolence”.
In accordance with the overall goal of the Deontology, the practical part is
first of all about “dictates of purely self-regarding prudence”. However, because the conduct of oneself affects the well-being of others, it is necessary
both to consider the “dictates of benevolence” and the “dictates of extra-regarding prudence” (Deont., 122-24). The second part is therefore divided into
these three subdivisions.
Seen as an art, the purpose of deontology is to promote “human welfare”.
The distinction between self-regarding and extra-regarding prudence, therefore, becomes a distinction between “self-regarding deontology”, which aims
to promote the welfare of the actor in question, and “extra-regarding deontology”, which aims to promote the welfare of “all persons concerned other than
the individual agent” (Deont., 198).
Bentham apparently considers deontology to be primarily concerned with
one’s own happiness, especially in one’s private life; but this does not mean –
or is not intended to mean – that one should be self-centred in a selfish way.
The individual achievement of particular pleasures is only a “means” in respect
to the general end, i.e. mere “subordinate ends” (Deont., 125). Among these
particular pleasures are those stemming from sympathy, and they include the
genuine pleasure of knowing that others fare well. The “business of the deontologist” (Deont., 193) is precisely to “bring to view these comparatively latent
ties” (Deont., 195) between self-regarding and extra-regarding interests, and to
show their “points of coincidence” (Deont., 193).
The purpose of the Deontology is to persuade the readers to accept the basic
principles of utilitarianism, and then to offer guidelines for how to act correctly as a utilitarian in private life – “morality made easy” (Deont. 119), as
Bentham writes on the drafts for the title page. However, if we consider the
concept of ‘deontology’ as such – as distinguished from the work called Deontology – it has a more “general end”, which is:
the same end or object which not only every branch of art or science has, but every human
thought as well as every human action has – and not only has but ought to have: the giving
increase in some shape or other to man’s well-being – say in one word the sum of human
happiness. (Deont., 125)
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
75
In so far as man’s conduct is conducive to this end, it is to be called ‘virtuous’;
‘virtue’ is then a ‘characteristic’ of man, which is manifested by “his conduct,
his actions, his deportment”. ‘Vicious’ and ‘vice’ are not surprisingly defined
as the opposite of ‘virtuous’ and ‘virtue’ (Deont., 125). Deontology, therefore,
must have as its goal the promotion of virtuous acts and virtues as such, and
with such a goal, deontology has to take obligations into account. Basically
obligations are also considered “a species of fictious entity”, but nevertheless
the ‘business’ of deontology includes:
the distribution of obligations, [...] marking in the field of action the spots upon which it is
proper that obligation in one shape or another should consider itself as attaching; and, in the
case of a conflict between obligations issuing from different sources, in determining which
should obtain and which should yield the preference. (Deont., 171)
Understanding the ends of deontology as such, Bentham considers it a
branch of the art and science which has for its object the learning and shewing for the information of each individual, by what means the net amount of his happiness may be
made as large as possible; of each in so far as it is dependent on his own conduct: the
happiness of each individual separately being considered, and thereby that of every individual among those whose happiness on this occasion an object of regard (Deont., 12425).
Deontology is in general described as a branch “of the art and science of Eudaemonics”, and at the same time ‘deontology’ is the same as “Ethics (taken in
the largest sense of the word)” (Deont., 124-25). Deontology is, we must conclude, ethics at large, and as such it is a branch of the science and art of eudaemonics.
A deontologist, then, apparently does not have to be utilitarian. In Bentham’s
terms, he can also be ascetic, or ‘ipsedixital’, i.e. basing his indication on an
‘opinion’, without reference to “happiness or unhappiness” (Util., 304-05).
The term ‘deontology’ is not meant to be exclusively utilitarian. Other worldviews can have their deontologies as well. As such, deontology is something
relatively limited, i.e. just a part of an all-inclusive faith such as utilitarianism.
‘Deontology’ in this sense apparently refers to the teaching of private morality, that is, of duties that can be deduced from the general principles. In other
words, deontology cannot be defined substantially on its own terms; its content
can only be deduced from already accepted general principles. In this case,
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
76
asger sørensen
there is nothing that conceals that deontology is dominated by something else;
deontology is simply defined as a sub-category, the teachings of duties consistent with an already accepted worldview.
*
In sum, Bentham apparently has at least two conceptions of deontology. The
first conception takes deontology – as art and science, or art with science attached – as the most comprehensive category, divided into private deontology,
which is private morality, and public deontology, which includes legislation
and government in general. This conception is consistent both with the remark
that considers deontology as ethics in general, i.e. practical philosophy, and
with the specification of deontology as a subcategory of a more comprehensive
worldview.7
The second conception takes eudaemonics – again, as arts and sciences, or
arts only – to be the most comprehensive category, and it comprises deontology, understood as private morality, politics, legislation, and government, all on
the same footing. If deontology is ethics in this sense, then ethics is concerned
with private morality, but pre-defined as eudaemonic, as an activity aiming at
one specific telos, namely happiness. As such, deontology is a subcategory
within a teleological framework, in casu utilitarianism.
The first conception of deontology is apparently relatively neutral towards
the utilitarian scheme of thought, but since it is only a formal conception, it
must get its content from the worldview of which it is a part. Consequently,
when Bentham defines the content of deontology, it is utilitarian. The second
conception of deontology adds to this point. It shows that Bentham considers
ethics as such to be inherently teleological, and that makes deontology subordinate to this scheme of thought as well.
2. Broad
Both of Bentham’s conceptions of ‘deontology’ are consistent with the general
understanding of deontology as the science of duties. As mentioned above,
however, when it comes to the modern philosophical sense of ‘deontology’,
one normally refers to Broad’s Five Types of Ethical Theory (F.T.). Here the
term deontology is introduced without any reference to Bentham in a specific
context, namely the analysis of the Sidgwick’s Method of Ethics. Broad wants
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
77
to qualify the applications of ‘ought’ in judgements, and the distinctions between deontological, teleological and logical are offered as such qualifications.
To Broad, the deontological use of ‘ought’ in a judgement means that an action should be performed in a certain type of situation, “regardless of the goodness or badness of the probable consequences” (F.T., 162). Broad notes that
many people would deny that they ever make such ‘unconditional’ judgements,
but they can probably be seen as making statements which employ ‘ought’ teleologically, meaning “that everyone ought to aim at certain ends without any
ulterior motive, e.g. his own greatest happiness, at the greatest happiness of all
sentient being, and so on.” At last, ‘ought’ can be applied logically, meaning
that if someone considers a certain end to be ultimate, “then he ought to be
consistent about it” (F.T., 162).
Broad considers whether these three applications of ‘ought’ also involve
three different meanings. He distinguishes between the narrow sense of ‘ought’,
applied to actions “which an agent could do if he willed” (F.T., 161), and the
wider sense, where this condition does not apply. According to Broad, the wide
sense is involved in the teleological application of ‘ought’, and the narrow, in
the logical application. “For we believe it is within the powers of any sane human being to be consistent if he tries.” However, “the logical ought is just a
special case of the deontological ought” (F.T., 163), and this relation is important, because the narrow sense of ought is made acceptable to those who do not
acknowledge the deontological application of ‘ought’ in general.
Broad ends up with the classical binary distinction between teleology and
deontology. What is interesting, however, is that the logical and thus deontological use of ought is based on the idea of consistency in action and not in
relation to propositions. Apparently, Broad thinks that ‘consistent’ means that
one ought to choose the appropriate means to realize an end, and to avoid actions “inconsistent with its realisation” (F.T., 163). With this concept of consistency, however, it seems strange to subsume the logical meaning of ‘ought’
under the deontological, since it is teleology which by definition should be
focused on actions as a means to an end, and not deontology, which is more
preoccupied with the actions in themselves.
To make this point more clearly, one can employ Weber’s famous distinction between means-end rationality, often called instrumental rationality, and
value-rationality. Means-end rationality is the kind of rationality that is employed by economic theory and rational choice theory, where what matters is
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
78
asger sørensen
the right choice of means in view of the optimal realization of given ends (Weber 1921-22, 12). The rationality involved in value-rationality is rationality in
another sense, i.e. the kind of rationality employed in logical reasoning, e.g. in
the deduction from premises to a valid conclusion. To Weber, it is this latter
kind of rationality that can be attributed to the protestant ‘ethics of intention’,
which like deontology is defined by its disregard for consequences (Weber
1919, 551). An ethics of intention is rational and consistent in the strictly logical sense that its particular judgement can be deduced from one or more general principles and that it is therefore non-contradictory.
One would think that it was consistency and rationality in this latter sense,
i.e. the rationality involved in making non-contradictory judgements, that must
be involved, if the logical sense of ‘ought’ is to be subsumed under deontology.
And it would be tempting to interpret Broad’s distinction between a wider and
a narrower sense of ‘ought’ as the distinction between a weak and a strong
sense, i.e. as the distinction between what one ought to do in relation to given
ends – e.g. one ought to love one’s neighbours – and what one ought to do in
relation to logical constraints – that one ought not to contradict oneself. Especially since Broad does think of rationality in this way, when he interprets Kant
(F.T., 128).
But in his analysis of Sidgwick’s utilitarianism, Broad uses the concept of
means-end rationality in his notion of consistency, although both he and Sidgwick (in contrast to Weber) are aware that rationality in the sense of “hypothetical imperatives” (F.T., 152) can be employed both in an egoistic way and
in a universal or utilitarian way. Even though consistency is a fair demand to
the deontological and logical use of ‘ought’, neither Broad’s logical application of ‘ought’, nor his conception of deontology has anything to do with logic
in a strict sense. The difference between these two conceptions of rationality
becomes even more important with Rawls’ idea of himself as an deontologist,
as we shall see below.
*
It is this distinction concerning the application and meaning of ‘ought’ that
Broads generalizes to divide ethical theory into two classes. Deontological
theories contain propositions of the form “Such and such a kind of action would
always be right (or wrong) in such and such circumstances, no matter what the
consequences might be.” Teleological theories judge the rightness or wrongDANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
79
ness of an action by its “tendency to produce certain consequences which are
intrinsically good or bad” (F.T., 206-07). Broad thinks that both types of theories can be found in monistic and in pluralistic versions, and that teleological
theories can be divided into egoistic and non-egoistic types, with utilitarianism
being an example of the latter.
Broad considers this classification to be clearer than Sidgwick’s, which distinguishes between intuitionism, egoistic hedonism and utilitarianism, since
his own classification is independent of epistemological considerations. From
a logical point of view, he is no doubt correct.8 However, one can ask why
deontological theories cannot be either egoistic or non-egoistic. Is being egoistical regardless of the consequences not an option for ethical theory?
The hidden premise that rules out this idea as senseless seems to be that
morality, and hence ethics, must be good for something, if not for the individual then for society or humanity in general. And deontological egoism does not
seem to be good for anything, neither for oneself nor for anybody else. Keeping in mind Broad’s concept of consistency and the concept of rationality implied by it, ethics appears then as a whole to be teleological, which means that
the two aspects of the distinction in question cannot be of equal value.
This conclusion is supported by the way Broad analyses Sidgwick’s intuitionism in relation to deontology. Trying to make sense of deontology, Broad
understands it as claiming that in order to determine the rightness of an actions
it is sufficient to consider “one or a few characteristics of its immediate consequences”, treating as irrelevant “the more remote consequences and the other
characteristics of the consequences” (F.T., 214). It is claims about the necessary connection between certain kinds of actions and their immediate consequences that Broad finds characteristic of deontology, and those claims are
considered to be ‘a priori’ judgements.
According to Broad, the difference between teleology and deontology is that
the former makes empirically based judgements about the relative non-moral
goodness of all of the consequences of an action, whereas the latter makes a
priori judgements about the connection between some kinds of actions and
their immediate moral consequences. Both types of judgement are based on
consequences, but teleology makes more comprehensive, empirically based
and – from Broad’s perspective – therefore better judgements. From a deontological point of view, Broad’s concession to utilitarian ethics appears inconsistent (Campell Garnett 1941, 421). Broad’s perspective, however, is teleological, at least in his analysis of Sidgwick, and, as Schneewind remarks, the idea
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
80
asger sørensen
that an “action can only be right because it produces good” is “deeply rooted”
(Schneewind 1993, 150) in ethics.
Weber made it perfectly clear that in the perspective of calculating means-end
rationality, deductive value-rationality was not rational (Weber 1921-22, 13).
Not so with Broad. His conception of ethics is not made explicit, but the result
is the same. Like the ethics of intention, deontology cannot be understood as
rational in the full sense, but appears to be a dogmatic, deficient mode of
teleology. Since ethics as such has to be teleological and rationally consistent
with regard to means-end, deontology is not an option. In the end, deontology
can only be an exception, a category which attracts irrational moralists, fanatics and the like. It is not possible to form a rational ethical theory on this basis.
This bias in favour of teleology, however, should not come as a surprise,
since the distinction between deontology and teleology is based on Sidgwick’s
tri-partition of ethics in intuitionism, egoistic hedonism, and universal hedonism. The distinction thus reflects what utilitarians themselves often consider
the main opposition to utilitarianism, egoism and intuitionism, i.e. those who
are selfish and those who on moral grounds oppose the rational calculation of
means and ends, i.e. those whom Bentham called “ipse-dixits”. Broad states in
the beginning of Five Types that he has chosen ‘men of genius whose views
differ from each other as much as possible’ (F.T., 1). The distinction between
teleology and deontology, however, only appears in the final analysis of Sidgwick, not in relation to the four preceding theories, those of Spinoza, Butler,
Hume and Kant. Kant is considered a deontologist (F.T., 207), but only within
Sidgwick’s utilitarian classification of ethical theories.
Broad’s distinction is clearly an improvement on Sidgwick’s. Through “a
slight shift in terminology” he clarifies “what is inherent in Sidgwicks position”, adding only a slight “modification”, but resting on “the same essential
principles” (Salzman 1995, 76-78). In Derrida’s terminology, one could consider Broad’s reading of Sidgwick a “displacement” (cf. e.g. Derrida 1967, 29).
It is an improvement from a logical point of view, but Broad’s new concept of
deontology is still conceived within the utilitarian scheme of thought, although
differently than in Bentham’s original conception.
*
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
81
The relation between the terms, however, is even more complex than suggested
by the analysis so far. Broad states that “purely deontological and purely teleological theories are rather ideal limits than real existents” (F.T., 207), and even
that “neither concept might be definable in terms of the other” (F.T., 278). This
way of employing the distinction shows the bi-partition to be less a classification of theories than an analytical distinction, which is used to clarify two aspects, which Broad considers to be inherent in almost every ethical theory: the
teleological focus on the ends and consequences of an action; and the deontological focus on what is considered to be intrinsically right. And in this latter
context, Broad does not mention the disregard for the consequences of an action performed in accordance with such an ideal.
Teleology and deontology can be understood as two almost independent – or
to employ Weber’s terminology, “ideal-typical” (cf. Weber 1904, 190-92) –
aspects of ethical theory. “Most actual theories are mixed, some being predominantly deontological and others predominantly teleological”. To Broad,
Sidgwick’s utilitarianism is an example of an “almost purely teleological theory”, but even in such a theory there is something considered “intrinsically
right”, namely a “mode of distribution”, and “to this extent Sidgwick’s theory
must be counted as deontological” (F.T. 207-08).9
Broad develops this conception of the distinction in more detail. In general,
he characterizes teleology as “an empirical or inductive theory”, taking probable tendencies in the overall consequences into consideration. Deontology is
claiming a priori rightness or wrongness in such judgements, whereas a teleologist demands empirical evidence. But since the distinction is to be understood as analytical, Broad’s general point is that “every Teleological theory
does involve at least one a priori judgement”, namely one that “expresses a
necessary connection between a certain non-ethical characteristic and the ethical characteristic of goodness” (F.T., 213-14).
Deontology becomes in this sense the a priori aspects of an ethical theory,
including not only the a priori claims that some actions or modes of actions are
unconditionally right, but also that some things are intrinsically good. Sidgwick, even though a hedonist, “is not a pure teleologist, since his six ethical
intuitions are deontological propositions” (F.T., 228).
This conception of the distinction between deontology and teleology is epistemological. Hence, the distinction can no longer be considered merely a classification of theories operating within the limits of normative ethics. By changing the criteria of distinction from the significance of ends and consequences
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
82
asger sørensen
to the question of a priori versus empirical evidence, the distinction becomes
meta-ethical. Broad tried to escape the epistemological premises inherent in
Sidgwick’s classification by an analytical clarification. But epistemology
seems to have sneaked in again, behind his back, so to speak!10
Since the distinction is normative in one sense and meta-ethical in another,
all kinds of utilitarian ethics can be labelled as deontological, at least to a certain extent. This conception of the distinction, however, makes it possible to
give sense to the logically constructed idea of egoistic deontological theories.
Consider for example those ethical theories that hold freedom as the ultimate
value, based on an idea of a moral sense and on a firm belief in the invisible
hand or the equilibrium theory of neo-classical economy. To a non-believer,
they often seem to regard the freedom expressed in rational market-behaviour
as intrinsically right, as a priori valuable, disregarding empirical evidence
showing the obvious inhuman consequences of the free market in society at
large.
Such ethical theories could, with this extension of Broad’s terminology, be
called predominantly deontological in both the meta-ethical and the normative
sense to the extent that they consider egoism – sometimes disguised under the
term ‘prudence’ – as intrinsically right, no matter what the consequences might
be. In contrast, such ethicists would be predominantly teleological in the normative sense, to the extent that they justify egoism by an end like the wealth of
nations or the universal happiness of mankind. However, they would be teleological in both senses only to the extent that they actually would be capable of
being proven wrong by empirical evidence. And this happens very rarely in
matters of politics and ethics, since practical philosophy not only acknowledges reality, but also ideal – i.e. non-real – matters.
Even though deontology is inherent in most actual ethical theories, it is clear
that deontology as such cannot be regarded as a rational ethical position, as
conceived of by Broad. The empiricism and means-end rationality implicit in
teleology, as well as the concept of consistency as choosing the right means to
a given end, does not admit of any good reasons for adopting a pure deontological approach, either in the normative or in the meta-ethical sense. Because
of the teleological conception of ethics inherent in Broad’s analysis of Sidgwick, deontology simply refers to the unconditional, dogmatic and hence nonjustified and irrational aspect of morality, which, however, must be accounted
for by any comprehensive ethical theory.
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
83
3. Frankena
Broad made substantial contributions to the clarification of Sidgwicks utilitarianism and made the line of thinking behind his own distinction very clear. The
distinction is today well established in practical philosophy in Broad’s original
wording thanks to Nowell-Smith and especially Frankena, whose textbooks
from the 1960s and 1970s became a widely used references within Anglophone
ethics.11 And Anglophone ethics was all there was in these two decades, while
the continental mainstream formulated normative matters in terms of politics.
Textbooks – it is usually thought – must be simple and unambiguous. And
what Frankena does is to remove all possibilities of confusion in Broad’s distinction, and place it firmly within a general conception of ethics structured by
teleology.
Broad’s last word in Five Types was a warning against the “danger of oversimplification” (F.T., 284) in ethics. In Frankena’s textbooks oversimplification is not considered to be a danger at all; quite the contrary, it is understood
as the goal to be achieved. Even though the resulting account of ethics is rather
strange, it has one great advantage. One does not have to interpret the text as
closely as is the case for Bentham’s and Broad’s texts. When the basic definitions and the overall structure are grasped, then the rest of the content can almost be deduced logically. The problem, however, is that the readers – i.e.,
students – not only might take this simple picture as the whole truth, but also
– and this is much worse – take logical simplification and rigid classification
as all there is to thinking about matters of morality and ethics. And this is, I
would claim, to a great extent what has happened to ethics in the Anglophone
world.
*
The general framework of Frankena’s textbooks is the well-known tripartite
division of ethics into “three kinds of thinking which relate to morality” (Frankena 1963, 4) in different ways, namely descriptive inquiry, normative judgements and meta-ethical thinking. But the structure of the books also reflects a
teleological conception of ethics. Two chapters are dedicated to a normative
“theory of obligation” (Frankena 1963, 10): the first to egoism and deontology,
the second to teleology, in casu utilitarianism. According to the teleological
scheme of thought as exposed by Sidgwick et al., a theory of obligation, of
what is right to do, must be supplemented by a theory of value, of what is good.
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
84
asger sørensen
Frankena writes, “a utilitarian must accept some particular theory of value”
(Frankena 1963, 15). When Broad’s teleologically framed distinction is employed within normative ethics, it is then not so much a question of regarding
or disregarding consequences, as of which kind of consequences should be
weighed, the moral or the non-moral. Accordingly, there is one chapter about
moral value and one about non-moral value, the latter being the ground on
which teleological ethics bases its judgements. The final chapter of the book is
dedicated solely to meta-ethical questions.
Within this overall teleological structure, a teleological theory is defined as
providing the “ultimate criterion or standard of what is morally right, wrong,
obligatory etc. [...] the non-moral value that is brought into being”. Frankena
argues that the justification for this definition is logical, since it would be ‘circular’ to let the moral value of something depend on “the moral value it promotes” (Frankena 1963, 13). This formally strong argument for basing moral
judgement on something non-moral is on its own premises hard to counter, but
that is exactly what deontology is supposed to do, since deontological theories
are negatively defined to “deny what teleological theories affirm” (Frankena
1963, 14).
The definition of deontology is then deduced logically by negating the statements defining teleology, i.e. by denying that that the non-moral value brought
into being is the only criterion of moral value. This implies either that there are
one or more criteria for what is right to do besides the one proposed by teleology, or that there are one or more completely different criteria. The deontologist can judge an action “right or obligatory simply because of some other fact
about it or because of its own nature” and “may adopt any kind of view about
what is good or bad in the non-moral sense” (Frankena 1963, 14). This negative definition of ‘deontology’ leaves no doubt about which side of the distinction is the centre and which is peripheral. It is like dividing the world into fish
and non-fish, the latter category being everything in the world, which is not a
fish.
It is important to bear in mind that the premises of the general teleological
conception of ethics, i.e. means-end rationality and empiricism, have direct
implications for the conception of normative ethics. The point of departure for
normative ethics is a situation where somebody is to do something of moral
relevance, but does not know what to do. It is the teleological perspective as
conceived by Broad that defines normative ethics to be about what one ought
to do in a specific and particular sense rather than how one ought to live. The
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
85
primary matter of moral importance becomes the act, not life as a whole.12 In
principle, every single act in a life can be right, and for Frankena this can be the
case if we are guided by the right ethical theory. In the attempt to find a noncircular, empirically based and means-end rational justification for what is
right to do, the non-moral consequence of the act becomes relevant.
As paradigmatic of what ethics is all about, Frankena analyses Socrates’
situation the night before his execution, considering whether to flee or to stay.
Behind Socrates’ reasoning in the Crito, he finds the following ideal:
(1) We must not let our decision be affected by our emotions, but must examine the question and follow the best reasoning. We must try to get our facts straight and to keep our
minds clear. Questions like this can and should be settled by reason.
(2) We cannot answer such questions by appealing to what people generally think. They
may be wrong. We must try to find an answer we ourselves can regard as correct. We must
think for ourselves.
(3) We ought never to do what is morally wrong. The only question we need answer is
whether what is proposed is right or wrong, not what will happen to us, what people will
think of us, or how we feel about what has happened. (Frankena 1963, 1-2)
In Frankena’s view, however, ethics is not directly concerned with solving
particular problems in specific situations. But indirectly that is what ethics
aims at all the time. Any ethical theory is assumed to presuppose this model of
an agent in a situation confronted with a specific problem. As Frankena formulates it, ethics is primarily thought to “provide the general outlines of a normative theory to help us in answering problems about what is right, or ought to be
done” (Frankena 1963, 5). In short, ethics is primarily normative ethics.
*
Summarizing the point above, I argue that ‘deontology’ is defined as ‘non-teleology’, but within a teleological conception of ethics in general and of normative ethics in particular. This being the case, even though ‘deontology’ is
negatively and thus in principle very broadly defined, in reality, because of the
general teleological framework, the possibilities of giving ‘deontology’ a positive meaning are very limited, as is obvious from Frankena’s attempt to do so.
Like Sidgwick and Broad, Frankena’s basic distinction within teleology is
between egoism and universalism, i.e. utilitarianism. Considering deontology,
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
86
asger sørensen
however, even though Frankena appears to be much stricter in his classificatory logic than his predecessors, this distinction is again avoided. Instead, we
are presented with the distinction between act-deontological and rule-deontological theories (Frankena 1963, 14-15). Shifting back to teleology again, one
would expect to find a distinction between act-teleological and rule-teleological theories, but instead we find act-utilitarianism confronting rule-utilitarianism (Frankena 1963, 30), that is, a distinction at one level below. How can that
be?
The explanation for the confusion of levels of classification is, I will argue,
found in the implicit teleological framework, or to be more precise, the inherent utilitarianism in the classificatory logic. Rule-utilitarianism is the wellknown answer to one of the most basic critiques of classical utilitarianism,
namely that it seems very impractical to have to calculate the balance of good
over evil for all foreseeable consequences every time we are to do something.
The utilitarian answer is that, if this is really the case, then we must act according to some rules of conduct, and those rules must in turn be justified in the
way specified. Hence we can distinguish between act-utilitarians and rule-utilitarians. The two aspects of this distinction, however, are not equal. Act-utilitarianism is the problem, while rule-utilitarianism is the solution.
With this background in mind, it is no wonder that things get a little complicated when this utilitarian sub-distinction is transferred to classify various
kinds of deontological theories, i.e. ethical theories that are defined as nonutilitarian. For Frankena, ethics is teleological in the sense defined above and
therefore focuses on acts and consequences. This means that both teleology
and deontology can only be understood in relation to acts and consequences.
This is not a problem for those theories that are already teleological, i.e. utilitarians of various sorts. But for those completely different kinds of ethical
thinking, which have to be classified as deontological, since Frankena’s exclusive conception of the distinction does not admit of any third possibility, the
result is rather bizarre. Act-deontological theories are ethical theories which
state that “basic judgements of obligation are all purely particular” and that
we can and must see or somehow decide separately in each particular situation what is the
right or obligatory thing to do, without appealing to any rules. (Frankena 1963, 15)
And this approach means that the ethics of Aristotle is categorized as act-deontological because of his remark that “the decision rests with perception” (Frankena 1963, 15).13
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
87
Frankena’s generalization of the distinction opens up for the transition from
teleology to consequentialism. Broad himself, however, also contributed to
both of the revisions just mentioned: He dropped the use of the term ‘deontology’ (cf. Salzman 1995, 101) and later preferred to distinguish between teleology and non-teleology (cf. Broad 1985, 229);14 and he distinguished between
hedonism as a theory of good and evil, i.e. a value theory, and utilitarianism as
a theory of right and wrong (cf. Broad 1985, 196), which is exactly the conception of utilitarianism from which the term ‘consequentialism’ is derived. What
is only implicit by Broad, however, is made explicit by Frankena in the structure of his textbooks. For Frankena, deontology is simply non-teleology, and
teleology is defined exclusively in terms of consequences. Taken together, this
position makes deontology identical to non-consequentialism; and this is indeed the most common understanding of deontology in ethics today.15
4. Rawls
It is obvious, however, that apart from the logically very strict construction of
deontology, Frankena had serious problems positively specifying what a viable
deontological theory could be. In spite of this, a theory of major importance,
explicitly referring to Frankena’s classification, voluntarily takes this burden
upon itself. In A Theory of Justice, Rawls states that his “justice as fairness”
view is not a utilitarian theory, and that it as non-teleological therefore, “by
definition” (1971/99, 26), is a deontological theory. Rawls takes an explicit
stand for deontology, but the way he develops this stand turns the concept of
deontology upside down.
Rawls accepts Frankena’s conception of ethics as structured by the distinction and the relation between what is right and what is good. Teleology simply
defines what is right to do in terms of what is good (Rawls 1971/99, 21), and
deontological theories are defined as non-teleological, as theories that either do
not specify the good independently from the right, or do not interpret the right
as maximizing the good. Rawls prefers this logical conception of deontology
as non-teleology to the view “that characterizes the rightness of institutions and
acts independently from their consequences”, i.e. Broad’s first characterization
of deontology and Weber’s conception of the ethics of intention. Such a conception of ethics is to Rawls simply ‘irrational, crazy’ (1971/99, 26).
To Rawls, the concept of rationality is embodied in teleology; to be rational
is simply to “strive for as high an absolute score as possible” (1971/99, 125).
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
88
asger sørensen
With this identification of rationality with means-end rationality Rawls thus
stays within the overall teleological conception of ethics exposed above,16 and
that strategy does not leave him much choice when utilitarianism and deontology as ethics of intention are ruled out. Staying within the classifications of
Sidgwick et al. there is only one conceptual possibility, namely ‘enlightened’
egoism (Frankena 1963, 16), or self-regarding deontology, as Bentham would
call it, and that is actually what Rawls ends up with. His concept of rationality
thus in the end becomes equal to the means-end rationality of the inherently
selfish homo economicus as construed by Weber. This should, however, not
come as a surprise given Rawls’ interest in game theory and contractual theories (cf. Rawls 1958); the weight put on rational self-interest was clearly visible
from the very first conception of “justice as fairness” (Hall 1957, 663). Contracting parties always employ the rationality of game theory, i.e. the rationality of economic man, where each is supposed to maximize rationally his own
long-term good.17
Apparently Rawls has forgotten that not every opposition to utilitarianism
“by definition” is deontological, i.e. that egoism can be considered a teleological opposition to utilitarianism (Frankena 1963, 14). To Rawls, deontology is
simply non-teleology, which is equal to non-utilitarianism. Therefore, deontology does not exclude rational egoism, and that makes it possible to let the alleged lack of concern for persons in utilitarianism – because of its impartiality
– to be filled out by selfish partiality. A contract between two rational egoists
is thus “by definition” for Rawls an expression of deontology.
Unfortunately, Rawls’ understanding of his own practical philosophy as deontological has been widely accepted, and when deontology is understood as an
expression of “agent-centered” restrictions on the strictly moral utilitarianism,
then that position easily becomes equivalent to endorsing egoism. In sum, then,
Rawls’ political philosophy does not have much to contribute to ethics. Rawls’
contract is based on a concept of man as totally selfish, whereas utilitarians at
least accept man as inherently moral in some sense. Ethics simply presupposes
that human beings do not want to be just selfish, but have sincere and serious
doubts about what systematic egoist behaviour would imply for real human life.
Economic man, however, is constructed as totally immoral – he does not even
have a moral sense – and accepting the way Rawls opposes utilitarianism thus
means as a moral-philosopher to leave the sphere of ethics as such.
In fact, Rawls could have chosen other strategies to oppose utilitarianism,
but he fails to do so. He knows from Frankena that ethics is supposed to be
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
89
rational, but he can only understand rationality in terms of rational choice theories (1971/99, 123-24), and that makes even a thoroughly logical consistent
ethics of intention irrational, since it does not calculate the consequences. He
can only understand consistency as Broad’s concept of logical consistency, i.e.
consistency in action, choosing the right means in view of the optimal realization of a given end. In contrast to Weber, Rawls cannot see that it is only from
the limited perspective of means-end rationality that deontology and the ethics
of intention are irrational, because Rawls has only this one concept of rationality.
It is precisely the concept of rationality as deductive rationality or, simply,
reason, that Kant – and the neo-Kantian, Weber – understands to be central to
ethics, and which Rawls ignores. It is precisely this idea of rationality that
makes the Kantian distinction between acting in accordance with duty and acting out of duty meaningful; ethics is concerned with how to act out of duty,
how to make such an act reasonable (Kant 1785/86, 20-22). Kant is not at all
worried about the actual actions of people; he wants to justify their actions with
reasons related to the law of reason. Kant wants to understand people as free,
autonomous, self-legislative, subjugated only to the law of reason, the reason,
which is our human nature.
In fact, the case of Rawls can be seen as tragic. Rawls explicitly wants to
oppose utilitarianism in Sidgwick’s version (Rawls 1971/99, 26), but in choosing to identify with deontology in Frankena’s sense, he actually stays within
Sidgwick’s utilitarian scheme of thought, although transmitted through the
more refined vocabulary of Broad and Frankena. Furthermore, he explicitly
wants to identify with Kant (Rawls 1971/99, xviii), but as far as I can see, he
simply does not understand the fundamentals of Kantian ethics.18 Rawls describes Kant as giving priority to the right over the good (Rawls 1971/99, 28),
and in the classification of Frankena this means that Kant is a deontologist; but,
as mentioned above, Kant is only a deontologist when considered from within
the perspective of Broad’s refinement of Sidgwick’s classification. Frankena
simply misunderstands this point, claiming that Kant “purports to be” (Frankena 1963, 26) a deontologist, and this misunderstanding is apparently transmitted to Rawls. However, even without the misrepresentation of Frankena,
Rawls would, with his very limited notion of rationality, have missed what
Kant is talking about.19
*
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
90
asger sørensen
Deontology was constructed by Bentham, but got its philosophical significance
from Broad’s classical distinction. Today, however, Broad’s distinction is
known primarily through Frankena, who codified it to be relevant only within
normative ethics. With this in mind, one final thing should be noticed, i.e. the
small shift in focus from Broad’s first conception of the distinction to the second. In the discussion of the application of ought, Broad defines deontology in
terms of the disregard for the consequences of an action. Teleology is not defined in terms of consequences, but in terms of the ends of an action. In the
second version of the distinction, both deontology and teleology are defined in
terms of consequences, the former through disregard and the latter through
both the ends and the “tendency to produce certain consequences”.
To Broad, this shift means very little, since he thinks of consequences only
as “intended consequences” (F.T., 210) as far as they can be “foreseen” (F.T.,
213); it is, however, of major significance. As long as teleology is defined in
terms of ends, one can, as Rawls does, label an ethical theory like Aristotle’s
as teleological.20 The shift of focus in teleology from ends to actions and their
consequences, however, makes both the category of teleology more exclusive
and the identification of teleology and utilitarianism much stronger. This shift
pushes Aristotle over to the side of deontology, as in Frankena’s classification.
If such a strong identification is combined with an overall teleological – or,
later, consequentialist – conception of ethics, there is hardly any sense left for
deontology, either as an ethics of intention, or as any other kind of ethics.
Today this displacement is almost complete. On the one hand, the original
teleology-deontology distinction is now normally understood as identical with
consequentialism-deontology or even consequentialism-non-consequentialism. On the other hand, discussions about utilitarianism have been reformulated into discussions about consequentialism.21 With the help of Frankena’s
classificatory skills, Broad’s ideal-typical distinction has developed into a
complete and non-arbitrary classification, which by definition – by law of the
excluded middle – covers the whole field of normative ethics.
This is clearly a logical improvement, making the distinction simple and
complete, just like Broad’s own clarification of Sidgwick’s classification was
a step forward in this sense. But by making it logically stricter, Frankena actually emptied deontology of any positive content, while at the same time making it the only possibility for opposing utilitarianism. But having an exclusive
and strict distinction between two types of ethical theories, where one of the
aspects is impossible to take seriously, amounts to having no real distinction at
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
91
all. And what is worse, the negative definition of deontology and Rawls’ influential misunderstandings have made it possible for consequentialism to monopolize the idea of impartiality in ethics and to interpret the so-called agent
relativity of deontology as partiality. This means that deontology is understood
as egoism and thereby, “by definition”, is inadmissible for most people as an
ethical position.
From the very beginning deontology has been defined within a utilitarian
scheme of thought. Deontology was constructed by Bentham, reconstructed by
Broad in a completely different, but still utilitarian sense, first as an ideal limit
and later as the marginal aspect of an exclusive distinction. Frankena generalized the logical definition, and this displacement was accepted by Rawls, who,
however, was caught in the middle on the way from ideal limits to logical completeness. Whatever the reason, he turned things upside down, constructing the
modern conception of deontology as virtually egoistic, a conception which is
actually close to Bentham’s original conception,22 but is the opposite of Broad’s
and Frankena’s. In a way, the circle is then closed. No matter which conception of deontology is chosen, accepting ‘deontology’ as a meaningful expression is submitting to the utilitarian scheme of thought. And if one finally asks
why utilitarianism has had such an attraction within ethics, part of the answer
is no doubt that it promises a conceptual development without having to bring
troublesome concepts such as the self, intention and freedom into consideration (Ashby 1950, 772-73).
Literature
Apel, K.-O. (1993), “Das Anliegen des anglo-amerikanischen ‘Kommunitarismus’ in der Sicht der
Diskursethik”, in M. Brumlik and H. Brunkhorst (eds.), Gemeinschaft und Gerechtigkeit (Frankfurt a.M.: Fischer).
Aranguren, J.L.L. (1958), Etica in Aranguren, Obras completas, vol. 2 (Madrid: Trotta, 1994).
Aristotles (Eth.Nic.), Nikomachische Ethik (Berlin: Akademie Verlag, 1956) 1991.
Ashby, W. (1950), “Teleology and Deontology in Ethics”, Journal of Philosophy, 47, 765-73.
Benn, P. (1998), Ethics (London: UCL Press).
Bentham, J. (Deont., Util.), Deontology together with A table of the springs of action and the article on Utilitarianism in Bentham, The collected works of…, ed. by A. Goldworth (Oxford:
Clarendon Press, 1983).
Bloom, A. (1975), “Justice: John Rawls Vs. The Tradition of Political Philosophy”, American
Political Science Review, 69, 648-62.
Broad, C.D. (1930), (F.T.), Five Types of Ethical Theory (London: Kegan Paul, Trench, Trubner
& Co., 1944).
Broad, C.D. (1985), Ethics (Dordrecht: Martinus Nijhoff).
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
92
asger sørensen
Brook, R. (1991), “Agency and Morality”, The Journal of Philosophy 88, 190-212.
Campbell Garnett, A. (1941), “Deontology and Self-Realization”, Ethics, 51, 419-38.
Campbell Garnett, A. (1956), “The Indicative Element in Deontological Words”, Ethics, 67, 4252.
Canto-Sperber, M. (ed.) (1996), Dictionaire d’éthique et de philosophie morale (Paris: P.U.F.,
2001).
Carmichael, P.A. (1949), “The Logical Ground of Deontology”, The Journal of Philosophy, 46,
29-41.
Davis, N.(A.) (1991), “Contemporary deontology”, in: Singer 1991.
Darwall, S. (1986), “Agent-centered restrictions from the Inside Out”, Philosophical Studies 50,
291-319.
Derrida, J. (1967), “Un hégelianisme sans réserve”, L’arc, 32, 24-45.
Derrida, J. (1968), “La pharmacie de Platon” in Derrida, La dissémination (Paris: Seuil, 1972).
Edel, A. (1937), “Two Traditions in the Refutation of Egoism”, The Journal of Philosophy, 34,
617-28.
Edel, A. (1973), “Right and Good”, in P.P. Wiener (ed.), Dictionary of the History of Ideas, Vol.
IV (New York: Charles Scribner’s Sons).
Ellis, A. (1992), “Deontology, Incommensurability and the Arbitrary”, Philosophy and Phenomenological Research, 52, 855-75.
Ferrater Mora, J. (1994), Diccionario de filosofía (Barcelona: Ariel).
Frankena, W.K. (1963), Ethics (Englewood Cliffs: Prentice-Hall).
Frankena, W.K. and J.T. Granrose (eds.) (1974), Introductory Readings in Ethics (Englewood
Cliffs: Prentice-Hall).
Freeman, S. (1994), “Utilitarianism, Deontology, and the Priority of Right”, Philosophy and Public Affairs, 23, 313-49.
Freeman, S. (2003), “Introduction” in Freeman (ed.), The Cambridge Companion to Rawls (Cambridge: Cambridge University Press).
Gauss, G.F. (2001a), “What is Deontology? Part One: Orthodox Views”, The Journal of Value
Inquiry, 35, 27-42.
Gauss, G.F. (2001b), “What is Deontology? Part Two: Reasons to Act”, The Journal of Value Inquiry, 35, 179-93.
Goldworth, Amnon (1983), “Editoral Introduction” in Bentham, Deont..
Habermas, J. (1991), Erläuteringen zur Diskursethik (Frankfurt a.M.: Suhrkamp).
Halévy, E., (1901-04), The Growth of Philosophic Radicalism (Boston: Beacon Press, 1955).
Hall, E.W (1957), “II. Justice as Fairness: A Modernized Version of the Social Contract”, The
Journal of Philosophy, 54, 662-70.
Kant, I. (1785/86), Grundlagen zur Metaphysik der Sitten in Kant, Schriften zur Ethik und Religionsphilosophie (Werke in Sechs Bänden, Band IV), ed. by W. Wieschedel (1956) (Darmstadt:
Wissenschaftlische Buchgesellschaft, 1998).
Kutschera, F.v. (1982), Grundlagen der Ethik (Berlin & New York: Walter de Gruyter).
Kymlicka, W. (1988), “Rawls on Teleology and Deontology”, Philosophy and Public Affairs, 17,
173-90.
Lalande, A. (1991), Vocabulaire technique et critique de la philosophie, 18e ed. (Paris: Presses
Universitaires de France).
Muirhead, J.H. (1940), “The New Deontology”, Ethics, 50, 441-49.
Nowell-Smith, P.H. (1954), Ethics “Harmondsworth: Penguin”.
Nagel, T. (1986), The View from Nowhere (New York: Oxford University Press).
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
93
Postema, G.J. (2006), “Bentham’s Utilitarianism” in H.R. West (ed.), The Blackwell Guide to
Mill’s Utilitarianism (Oxford: Blackwell).
Rawls, J. (1958), “Justice as Fairness” in Rawls, Collected Papers (Cambridge, Mass.: Harvard
University Press, 1999).
Rawls, J. (1971/99), A Theory of Justice (Cambridge, Mass.: Harvard University Press).
Ritter, J. (ed.) (1972), Historiches Wörterbuch der Philosophie, Band 2 D-F (Darmstadt: Wissenschaftlische Buchgesellschaft).
Ryan, A. (1987), “Introduction” in Ryan (ed.), John Stuart Mill and Jeremy Bentham. Utilitarianism and Other Essays (Harmondsworth: Penguin).
Salzman, T.A. (1995), Deontology and Teleology (Leuven University Press).
Scheffler, S. (1984), “Agent-Centered Restrictions, Rationality, and the Virtues”, Mind, 94, 40919.
Schneewind, J.B. (1977), Sidgwick’s Ethics and Victorian Moral Philosophy (Oxford: Oxford
University Press).
Schneewind, J.B. (1991), ‘Modern moral philosophy’, in: Singer 1991.
Singer, P. (ed.) (1991), A Companion to Ethics (Oxford: Blackwell).
Stegmüller, W. (1989), Hauptströmungen der Gegenwartsphilosophie, Bd. IV (Stuttgart: Alfred
Kröner).
Sørensen, A. (2003), Forskning, etik, konsekvens (København: Politisk Revy).
Weber, M. (1919), “Politik als Beruf” in Weber, Gesammelte Politische Schriften (Tübingen:
J.C.B. Mohr, 1988).
Weber, M. (1922), “Die ‘Objektivität’ sozialwissenschaftlischer und sozialpolitischer Erkenntnis”
in Weber, Gesammelte Aufsätze zur Wissenschaftslehre (Tübingen: J.C.B. Mohr, 1988).
Weber, M. (1921-22), Wirtschaft und Gesellschaft (Tübingen: J.C.B. Mohr, 1990).
Notes
1. This article has in various versions passed through many hands and been presented in many
places. I would therefore like to thank the following for valuable criticism, suggestions, and
comments on earlier versions: Anders Bordum, Asmund Born, Bent Meier Sørensen, Brian
Barry, Christine Korsgaard, Helen Korsgaard, Jacob Vestergaard, Roberto Mordacci, Robin
May Schott, Sergio Cremaschi, Steen Valentin and Thomas Basbøll.
2. Normally, Bentham’s utilitarian ethics is extracted from his An Introduction to the Principles
of Morals and Legislation (cf. e.g. Ryan 1987, 9), and in spite of the fact that Deontology is
Bentham’s only attempt to develop systematically the moral philosophy implied by utilitarianism, one rarely sees any references to it in ethical debates about utilitarianism. An exception is Edel, whose systematic conception of utilitarianism is informed by Deontology (1937,
620).
3. In a recent anthology on Mill’s Utilitarianism, Bentham’s construction of the term ‘deontology’ is noticed (Postema 2006, 28), but without any reflections about the possible terminological and conceptual consequences.
4. It is from this perspective that Kymlicka can criticize Rawls for confusing things, precisely
because Rawls thinks of teleology as including perfectionism (Kymlicka 1988, 185-88).
5. In fact, I would like to claim that the history of Anglophone ethics in the twentieth century
can be reconstructed as an on-going refinement and development of distinctions forced upon
ethics by the insistent pressure of the utilitarians with their amoral preconception of ethics.
Since this point is too general to be argued convincingly within the limits of an article, I will
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
94
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
asger sørensen
constrain myself to analyse this – I think – crucial and very clear case as an illustration of the
more general story.
This bi-partition is apparently the result of an editorial decision. Whether this also was
Bentham’s own division is not altogether clear in Goldworth’s commentary (Goldworth 1983,
xixff., xxixff.), but it seems to be in accordance with Bentham’s general line of thought.
Although, admittedly, we must then assume that Bentham forgot to qualify the definition of
deontology as concerned with private morality as a definition of private deontology.
Broad’s refinement of Sidgwick’s classification makes it possible to think of ethics as divided
into two branches, one concerned only with epistemology and questions of meaning, i.e. metaethics, and one pretending not to be concerned with these matters at all, normative ethics,
thought to be totally independent from meta-ethics. This is another instance of the more general phenomenon in Anglophone ethics mentioned above, namely that utilitarianism plays the
decisive role in the continuous development of the increasingly refined ethical distinctions.
It would, however, require much more research and analysis to substantiate my claim, and in
this context the matter is further obscured by the fact that Broad describes his own distinction
as ontological (Broad 1930, 213).
The conception of deontology as defined positively in relation to the mode of the action in
question, and the intrinsic rightness of this mode, can also be found in modern continental
European interpretation treatments of ethics (Kutschera 1982, 2; Stegmüller 1989, 231). In
Gaus’ analysis of the Anglo-American discussions, however, this conception is not given
much attention (cf. Gaus 2001a, 36 and 2001b, 183).
Salzman insists that the distinction is altogether meta-ethical (Salzman 1995, 4), but this mistake must rest on his identification of the analytical approach to ethics with meta-ethics as
such (1995, 32).
Even today, Frankena can be considered the suitable starting point for a comprehensive analysis of deontology (Gaus 2001a, 27), bypassing thereby both Bentham and Broad.
It should be noted that this focus, already implicit in Bentham’s utilitarianism, is the exact
opposite of the traditional, pre-modern ethical focus (cf. e.g. Aranguren 1958, 182).
This quote is the only one by Aristotle in Ross’ analysis of the right and the good (cf. e.g. Ross
1930, 42), and one might suspect that Frankena simply has copied it from Ross.
Instead, in the posthumously published manuscripts Broad uses the term ‘deontic’ to characterize all sentences employing words like ‘duty’, ‘obligation’ and the like (Broad 1985, 225).
In this sense, ‘deontic’ thus comes close to the general conception of deontology as a science
of duties, and such a conception makes it possible for Broad to compare anew Kant with Ross
and Sidgwick. However, since he considers the categorical imperative ‘vague’ (Broad 1985,
219), the only two possibilities are “the irreducible pluralism of Ross” and the utilitarian “theory of a single ultimate self-evident obligation”, which “have strong claims to be considered
self-evident” (Broad 1985, 242), and which do not leave deontology many chances within
ethics.
Gaus fails to find one positive definition common to the various current uses of ‘deontology’.
The only common denominator is negative, namely the opposition to consequentialism (Gaus
2001b, 190).
In fact, the teleological conception of rationality is such a commonplace in Anglophone discussions of these matters that it often passes unnoticed, as for instance in Freeman’s defence
of Rawls against the critique of Kymlicka (Freemann 1994, 313).
Freeman denies that Rawls’ contracting parties in the original position are egoists (Freeman
2003, 13), since they allegedly have a capacity for justice. Still, he describes them as “con-
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Tusculanum Press :: Universityy of Copenhagen
p
g :: www.mtp.dk
p
:: info@mtp.dk
@
deontology – born and kept in servitude by utilitarianism
18.
19.
20.
21.
22.
95
cerned only with promoting their own interests” (2003, 14), which is precisely what egoism is
about.
Allan Bloom points to the same differences in the understanding of rationality between Rawls
and Kant, and argues convincingly that the thought of Rawls “has nothing to do with that of
Kant” (Bloom 1975, 657).
The contrast between the Hobbesian concepts of rationality common in Anglo-American
practical philosophy and Kant’s concept of reason is underlined strongly by Apel (Apel 1993,
152-54).
Aristotle takes as his point of departure what man strives toward, i.e. an end, and later – after
various analyses of the dynamic character of moral and intellectual virtues – defines this end
as the good life, where happiness consists in actions in conformity with virtues (cf. Eth.Nic.,
1176). Broad, however, does not categorize the ethics of Aristotle as teleological; in fact,
Aristotle is not mentioned in Five Types at all.
With almost all the moral weight put on the consequences as such, the intended end is now
treated as only of minor moral importance, and this creates new ethical problems, like those
of acts and omissions and the so-called double-effect (cf. Benn 1998, 74, 78).
The biographer of Bentham, Elie Halévy, simply concludes that to Bentham “egoism [was]
installed at the very basis of morality” (Halévy 1901-04, 477).
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301
Museum Tusculanum Press :: University of Copenhagen :: www.mtp.dk :: info@mtp.dk
DANISH YEARBOOK OF PHILOSOPHY No. 43, E-JOURNAL © Museum Tusculanum Press 2011
ISBN 978 87 635 3834 3 :: ISSN 1901 9114
http://www.mtp.hum.ku.dk/details.asp?lang=e&eln=300301