. 20
( 38 .)


Finally, purity. The intended sense of purity is also related to the con-
cept of the radically voluntary, for by the purity of morality Williams means
“its insistence on abstracting the moral consciousness from other kinds of
emotional reaction or social in¬‚uence.”27 This sense of purity expresses
an ideal that even Williams the critic of morality calls “one of the most
moving: the ideal that human existence can be ultimately just.”28 For the
purity of morality holds out the hope that human agents can, through their
own efforts to create and sustain a moral world, transcend luck and the

23 Williams (1995f), pp. 242“243. See also Williams (1993), p. 67.
24 Williams (1995b), p. 25.
25 Williams (1981a), p. 29.
26 Williams (1985), p. 194.
27 Williams (1985), p. 195.
28 Williams (1985), p. 195.
112 Robert B. Louden

myriad natural lotteries of life. In real life, some human beings are born
into communities with abundant natural resources and hospitable climates,
while others are not. Some are born in periods of great cultural and techno-
logical progress, and others are not. After the initial space and time lotteries
are held, everyone is subject to further lotteries associated with the class
system, the race system, and the gender system (systems, we might add,
whose practical effects, taken together or even singly, are often far more
destructive than anything dreamed up by the morality system). Additional
lotteries of natural talent, good or not-so-good looks, and psychological
temperament are also held at their appropriate times. The odds of one per-
son™s drawing a winning combination for all of these lotteries are incredibly
slim. But the purity of morality shields us from these contingencies of luck
and misfortune. Behind the shield we create a realm of freedom, where
moral agents are viewed as more than mere playthings of biology, history,
and social force.
However, even though Williams readily concedes that the ideals
expressed by this purity “have without doubt . . . played a part in produc-
ing some actual justice in the world and in mobilizing power and social
opportunity to compensate for bad luck in concrete terms,” he also believes
that we should jettison purity.29 For unfortunately, “the idea of a value
that lies beyond all luck is an illusion.”30 As a liberal (albeit a pessimistic
one), Williams does endorse the social justice aims of the morality system.31
More broadly, he also embraces the Enlightenment ideals that provide the
cultural setting for the morality system, in so far as they are identi¬ed with
“the criticism of arbitrary and merely traditional power.”32 But he wants
justice (or a reasonable facsimile thereof) without the multiple illusions of
the morality system.

29 Williams (1985), p. 195“196.
30 Williams (1985), p. 196.
31 In contrasting his own views to those of Charles Taylor and Alasdair MacIntyre, Williams
offers the following compact summary: “Taylor and MacIntyre are Catholic, and I am not;
Taylor and I are liberals, and MacIntyre is not; MacIntyre and I are pessimists, and Taylor is
not (not really),” Williams (1995g), p. 222, n. 19. Taylor and MacIntyre are also noted critics
of the morality system. However, Williams™ own brand of pessimistic, secular liberalism
sets him apart from these intellectual neighbors.
32 Williams (1993), p. 159. Cf. p. 11. On this particular point, Williams™ stance doesn™t seem
terribly different from Richard Rorty™s. Both are secular liberals who endorse the moral
and social ideals of Enlightenment, but they reject the metaphysical and epistemological
assumptions that traditionally accompany these ideals. Rorty, for instance, summarizes his
recent work as follows: “Most of what I have written in the last decade consists of attempts to
tie in my social hopes “ hopes for a global, cosmopolitan, democratic, egalitarian, classless,
casteless society “ with my antagonism towards Platonism,” Rorty (2000), p. xii.
The Critique of the Morality System


Now that we have a better idea of what Williams means by “the moral-
ity system,” what are we to make of his complete dismissal of it? What
should our own attitude toward it be? One ready sociological response is
simply that the morality system is more a philosopher™s idea than a reality
in today™s world. That is to say: it is highly doubtful that very many peo-
ple today actually do believe that morality is only about obligation, that
obligations cannot ever con¬‚ict, that there is no sense of practical neces-
sity outside of contexts of moral deliberation, that there exist no options
between the extremes of reason and force, and so on. As one reviewer of
Ethics and the Limits of Philosophy remarked, in questioning the ¬t between
Williams™ depiction of morality and the contemporary context: “I think
it unlikely that ordinary conceptions of morality are so highly developed
in the one direction de¬ned by Kant.”33 However, this does still leave us
with the problem of Kant. Again, Kant is allegedly “the philosopher who
has given the purest, deepest, and most thorough representation” of the
morality system.34 To what extent does Kant himself articulate and defend
“morality, the peculiar institution?” In the present section, I shall explore
this question, with speci¬c reference to the four philosophical mistakes of
morality analyzed in the previous section.35
Obligation. For Kant, obligation or the sense of acting under rational
constraint is indeed the central phenomenological feature of human moral
experience. For creatures with greater cognitive powers than us (or who
have different (e.g., less egotistical) psychological make-ups than us), the
story will be different. As he remarks in the Groundwork: “no imperatives
hold for the divine will and in general for a holy will: the “ought” (das Sollen)
is out of place here, because volition (das Wollen) is of itself necessarily in
accord with the law.”36 But meanwhile, back on earth, so to speak, as human
beings are creatures who can both be aware of the importance of moral
principles and yet oppose them because of contrary inclinations, morality
will confront them as an imperative. Morality™s demands and goals always

33 Wong (1989), p. 722.
34 Williams (1985), p. 174.
35 Needless to say, Kant™s moral theory is very complex, and a thorough investigation of all
of its myriad mysteries is far beyond the scope of this essay. Rather, my aim is the more
manageable one of examining brie¬‚y those speci¬c aspects of it that are targeted in Williams™
depiction of the morality system.
36 Kant (1785/1996d), Ak. 4: 414; p. 67.
114 Robert B. Louden

remain an ought that we must bring ourselves to strive for; for creatures like
us they are never reducible to an automatic is.
However, for Kant the moral ought that confronts humans is much
broader and more multidimensional than is the case with typical moral obli-
gations. Typically, an obligation is always something owed to others rather
than to oneself. Williams endorses this common usage, calling the very idea
of a duty to oneself a “fraudulent” item and an “absurd apparatus.”37 But
for Kant duties to oneself are the most important and fundamental kind of
obligation “ he sees them as necessary presuppositions of every other kind
of duty. As he states in the Metaphysics of Morals:
Suppose there were no such duties: then there would be no duties what-
soever, and so no external duties either. For I can recognize that I am
under obligation to others only insofar as I at the same time put myself
under obligation, since the law by virtue of which I regard myself as being
under obligation proceeds in every case from my own practical reason; and
in being constrained by my own reason, I am also the one constraining

On Kant™s view, it is our ability as rational agents to act on ends that we have
chosen that makes us moral agents in the ¬rst place. Only creatures who
can constrain themselves to act according to self-chosen principles can have
moral obligations. Only by working on ourselves “ making ourselves into
certain kinds of people “ can we carry out moral projects from the requisite
motivational structure.39
Kant™s view that duties to oneself “are the most important [duties] of
all” means that his own position differs from the morality system in several
important respects.40 First of all, blame will not play nearly as big a role
in the former as it does in the latter. According to Williams, blame “is the
characteristic reaction of the morality system.”41 In Kant™s ethics, to the

37 Williams (1985), p. 182; Williams (1972), p. 75. See also Williams (2002), where he dismisses
Kant™s “unhelpful vocabulary of duties to oneself,” p. 107.
38 Kant (1797/1996h) Ak. 6: 417“418; p. 543.
39 For further discussion, see my “Morality and Oneself,” Louden (1992), pp. 13“26. Williams™
own arguments against duties to oneself do not seem to me to be relevant to Kant™s position.
He views them simply as licenses to do what one already wants to do, under the guise
of a moral reason. Williams (1972), p. 75; Williams (1985), p. 182. But for Kant it is the
possibility of self-constraint and self-direction (regardless of what one may happen to want)
that generates duties to oneself.
40 Kant (1784“85/1997a) Ak. 27: 341; p. 122.
41 Williams (1985), p. 177. Similarly, in Williams (1995a) he states: “there is a special form
of ethical life, important in our culture, to which blame is central: we may call this special
form of the ethical ˜morality™,” p. 15.
The Critique of the Morality System

extent that we can and do criticize ourselves for failing to live up to our
own commitments and ideals, there will certainly be a place for self-blame.
But strictly speaking, blaming others (which for Williams is the primary
kind of blame) has no proper place within Kantian ethics.42 Secondly, the
strong self-regarding orientation of Kant™s ethics opens up the possibility
that morality may not after all be guilty of alienating agents from their
own projects and emphasizing impartiality at the expense of the personal.43
Granted, it may still alienate them from their nonmoral projects. But strictly
speaking, the self-regarding core of Kant™s ethics means that it is intensely
personal. Finally, proper acknowledgment of the centrality of duties to
oneself in Kant™s moral scheme moves it much closer to the virtue ethics
tradition “ a tradition that Williams himself often points to as a promising
alternative to the morality system.44 Kant™s ethics is in fact much more
about long-term character development (and much less about generating a
decision-procedure for determining speci¬c obligations) than many of his
foes as well as friends acknowledge.45
A second basic way in which Kant™s moral ought differs from the obli-
gations of the morality system concerns the broadness of its scope. For
Williams (and this is also true of many contemporary authors), obligations
are always about the ful¬llment or non-ful¬llment of speci¬c acts (e.g.,
keeping one™s promise to repay a debt by a speci¬ed date; not injuring other
people). But for Kant, some of the most fundamental moral obligations
concern the promotion of general ideals or ends such as our own perfec-
tion and the happiness of others. The obligations of the morality system
are narrow; those often emphasized within Kant™s ethics are wide. Williams,
however is opposed to such “general and indeterminate obligations,” on the
ground that they provide (too much) work for idle (as well as not-so-idle)
hands.46 There are a lot of unhappy people out there, and if the happiness
of others really is an end that is also a duty, it would appear that we also
have a moral duty never to rest, even for a second.

42 Christine Korsgaard also questions Williams™ emphasis on blame in describing Kant™s ethics
in several of her essays. See, e.g., her observation that the Kantian duty of respect strongly
restricts practices of blaming others in Korsgaard (1996), pp. 71 n. 24, 174.
43 Williams initially aimed this “alienation charge” at utilitarianism “ another alleged member
of the morality system. See his contribution to Smart and Williams (1973), p. 116 ff.
However, in Williams (1981a) it becomes clear that he thinks alienation from one™s own
projects will also be a problem for Kantian morality, see esp. pp. 38“39.
44 See, e.g., Williams (1985), pp. 8“10 and, more recently, Williams (1998).
45 For further discussion, see Louden (1986/1999). See also O™Neill (1996).
46 Williams (1985), p. 181.
116 Robert B. Louden

However, the fear that Kant™s ethics obligates us to do the impossible is
tempered considerably by his candid admission that wide (or “imperfect™)
duties leave “a playroom (latitudo) for free choice in following (complying
with) the law.”47 There are in¬nitely many ways to pursue the general goal
of promoting the happiness of others, and different people will decide to
pursue it in different ways, depending on their own talents and projects.
At the same time, the expansive role of imperfect duties means that they
do have a tendency to crowd out, indeed replace, certain other normative
notions. The most prominent example is supererogation. In his Critique of
Practical Reason, Kant writes: “I do wish that educators would spare their
pupils examples of so-called noble (supermeritorious) actions, with which
our sentimental writings so abound, and would expose them all only to
duty.”48 On Kant™s view, people who focus on actions that are allegedly
beyond duty are liable both to view more mundane obligations as insignif-
icant and beneath their (often self-designated) heroic stature, and to get
caught up in romantic and unattainable images of moral greatness. What
he urges instead is a highly demanding morality that nevertheless constantly
requires agents to exercise their own discretion in determining how best to
pursue morality™s demands within their own particular life-situations.
Self-government, acting under self-willed rational constraint, is indeed
the central feature of human moral experience according to Kant. But does
this mean, as Williams charges, that Kant is pathologically obsessed with
obligation, and with denuding the moral landscape of other important nor-
mative categories? I would say rather that his underlying motive is simply
to convince readers that they are able to bring themselves to act according
to reasons that they have chosen; to show them that they are not sim-
ply creatures of desire. In this basic respect, Kant™s position in ethics ¬ts
squarely into a long and multifaceted philosophical tradition that stresses
the centrality of reason in human life.
What about the other reductionistic tendencies that Williams associates
with the morality system™s obsession with obligation? As concerns con¬‚icts
of obligation, Kant™s notorious claim that “a collision of duties and obligations
is inconceivable” would seem to rule them out entirely.49 He does readily

47 Kant (1797/1996h) Ak. 6: 390; p. 521. Kant™s strong stress on playroom (Spielraum) also
suggests that morality as he understands it is far from being a “system” in which every
choice follows a tightly coordinated organizational scheme. In practice, it is much looser
than this.
48 Kant (1788/1996f) Ak. 5: 155; pp. 263“64. For detailed discussion of this topic, see Baron
49 Kant (1797/1996h), Ak. 6: 224; p. 379. See also Kant (1793/1997c), Ak. 27: 273, 537;
pp. 273, 296“97.
The Critique of the Morality System

acknowledge that there can be competing grounds of obligation (competing
considerations relevant to determining what one™s duty is). However, in such
cases only one of the competing grounds is held to be suf¬cient to actually
put the agent under obligation “ the other one simply is “not a duty.”50
This is not to deny that it may often be dif¬cult, indeed impossible, for
cognitively limited creatures such as human beings to accurately determine
what their moral duty is, nor is it to deny that they will sometimes be faced
with hard choices where they may have to choose the lesser of two evils,
nor is it to deny they may legitimately feel regret after having made a hard
choice, etc. But strictly speaking, in Kant™s rationalist eyes “laws and rules
can never contradict one another.”51 The radical value pluralism assumed
by Williams and many other contemporary thinkers “ a world in which
there exists an irreducible plurality of competing and con¬‚icting values,
with no single scale by which to measure them or rank them in order to
determine our actual duty “ is foreign to Kant™s worldview.
As for the place of emotional response in human moral conduct, Kant,
unlike Williams, certainly would never allow that “emotionally governed
action” constitutes the moral worth of agents (see notes 8 and 9). On the
contrary, action must always be rationally governed if it is to possess moral
worth. Any and all actions determined by inclination “ however amiable
they may be “ lack true moral worth.52 However, this is not at all to say that
emotional response plays no positive role whatsoever in human morality


. 20
( 38 .)