. 21
( 38 .)


on Kant™s view. On the contrary, the presence of certain appropriate emo-
tions constitutes a necessary and important (albeit secondary) feature in the
accurate assessment of human beings™ moral worth. For instance, in The
Metaphysics of Morals he asserts:

it is a duty to sympathize actively in [the fate of others] . . . ; and to this
end it is therefore an indirect duty to cultivate the compassionate natural
(aesthetic) feelings in us, and to make use of them as so many means to
sympathy based on moral principles and the feeling appropriate to them.53

More generally, we have a duty to cultivate those speci¬c feelings that are
appropriate to the performance of each duty “ for example, in the case of the

50 Kant (1797/1996h), Ak. 6: 224; p. 379.
51 Kant (1793/1997c), Ak. 27: 296; p. 296. For a brief discussion of cases in which choosing
the lesser of two evils may be necessary, see Kant (1797/1996h), Ak. 6: 426; p. 550. For
further discussion of Kant and con¬‚icts of obligation, see Donagan (1987) and Hill (1996).
52 See Kant™s (in)famous discussion of the “naturally kind-hearted person,” see Kant
(1797/1996d), Ak. 4: 398; p. 53.
53 Kant (1797/1996h), Ak. 6: 457; p. 575.
118 Robert B. Louden

duty to help others, compassion. And we are also to discipline our emotions
so that we truly enjoy doing what our reason tells us to do:

a heart joyous in the compliance with its duty (not just complacency in the
recognition of it) is the sign of genuineness in virtuous disposition. . . . This
resolve, encouraged by good progress must needs effect a joyous frame of
mind, without which one is never certain of having gained also a love for the
good, i.e., of having incorporated the good into one™s maxim.54

In virtuous agents, the emotions work in harmony with reason: both con-
verge to point agents in the same direction. Reason must ultimately govern
action, but an important part of reason™s task is to cultivate the emotions
so that this humanly-necessary harmony between reason and emotion can
be achieved. This cultivation work in turn focuses on the proper develop-
ment of “moral feeling”; “a motive in which our sensibility concurs with
Practical Necessity. Practical necessity, again, is shorthand for agents™
sense that they must do X. Williams™ position here is that practical necessity
is not unique to ethics; the morality system, by contrast, claims that it is.
On Williams™ view, agents can legitimately conclude that they must do
X (e.g., pursue a career in music, join the underground, help the poor)
based on any number of nonmoral considerations. And even when moral
considerations are included as part of the deliberative picture, what agents
legitimately conclude they must do may not necessarily involve the carrying
out of a moral obligation. How is Kant™s position on this set of issues best
Kant would not deny the common sense view that agents often feel they
“must” do something based on non-moral considerations. Most soldiers
ordered by their commanding of¬cers to stand guard after midnight feel
they must do so; some people feel they must pursue a certain career, and
so on. But for Kant the important question always concerns the rational
assessment of such musts. For Williams, recall (see n. 15), moral obligations
as well as all other forms of practical necessity grow out of preexisting pro-
desires and inclinations. If the desire for X is suf¬ciently strong and is
related to one of the agent™s essential projects (and if the agent deliberates

54 Kant (1793/1996i) Ak. 6: 24 n.; p. 73 n.
55 Kant (1784“85/1997a). Ak. 27: 361; p. 138. Earlier in this lecture “ though in a section
which the translator has taken from the text of Mrongovius, that is, from Kant (1784“
85/1997b) “ Kant notes: “Anyone can see when an action is abhorrent, but only he who
feels this abhorrence has a moral feeling,” Ak 27: 1429; p. 72.
The Critique of the Morality System

correctly, has access to the relevant facts, etc.), then she will conclude that
she must do X. On Williams™ account, this seems to be the end of the story.
In Kantian language, this means that Williams accepts only hypothetical
imperatives, not categorical imperatives. A hypothetical imperative, Kant
notes, is a command to perform an action because it is viewed “as a means
to achieving something else that one wills.”56 In the present context, the
“something else that one wills” would stem from an agent™s fundamental
desires and projects. Agents will feel they must do what is necessary for
realizing their essential projects. Kant, however, famously contrasts hypo-
thetical imperatives to categorical imperatives. A categorical imperative is a
command to perform because it is viewed “as objectively necessary of itself,
without reference to another end” “ that is, regardless of one™s personal
desires and projects.57 And for Kant, only the categorical imperative “may
be called the imperative of morality.”58
This points to a very signi¬cant difference between Williams and Kant
on the topic of practical necessity. The sense of practical necessity that
Williams assumes throughout his writings is for Kant not real practical
necessity. Real practical necessity, in Kant™s sense, exists only when a con-
clusion of practical deliberation convincingly represents an action as in
itself good, regardless of the agent™s desires and projects. But according
to Williams, practical necessity exists only when agents have fundamental
desires and projects and then start deliberating about what they need to do
in order to realize them. For Kant, agents™ desires and projects always need
to be evaluated “ they are not given carte blanche. But for Williams, there
is no court of appeal outside of one™s basic desires and projects. One reasons
instrumentally on how best to achieve them, and that is it.
Thus for Kant, hypothetical imperatives do not present us with cases
of genuine practical necessity, but only with cases of practical contingency.
For if, after re¬‚ecting on the personal desire and project that generated the
particular hypothetical imperative in question we decide that they are, in
the larger scheme of things, not that important after all, then we no longer
need to follow the ought:

the categorical imperative has the tenor of a practical law; all the others can
indeed be called principles of the will but not laws, since what it is necessary
to do merely for achieving a discretionary purpose can be regarded as in

56 Kant (1785/1996d), Ak. 4: 414; p. 67. See also the opening footnote to Williams (1981a),
where he states that he agrees with the substance of Philippa Foot™s position in Foot (1978).
57 Kant (1785/1996d), Ak. 4: 414; p. 67.
58 Kant (1785/1996d), Ak. 4: 416; p. 69. See also Ak. 4: 425; p. 76.
120 Robert B. Louden

itself contingent and we can always be released from the precept if we give
up the purpose; on the contrary, the unconditional command leaves the will
no discretion with respect to the opposite, so that it alone brings with it
that necessity which we require of law.59

Brie¬‚y, on Kant™s view we are to assess the conclusions of all of our practical
deliberations (including those stemming from our desires and projects)
by means of the concepts of universality and necessity. Those desires and
projects that upon re¬‚ection do not carry a ¬rm sense of universality and
necessity with them, if and when they con¬‚ict with other practical concerns
that do convincingly convey a sense of universality and necessity, must take
a back seat to the latter. For example: suppose the project of mastering
Corelli™s violin sonatas is extremely important to me. When I deliberate
about how this project is to be realized, I conclude correctly that I will
need to devote a great deal of practice time to the violin. However, if in the
middle of an intense practice session (when I am ¬nally ¬guring out how
to pull off some of the tricky double-stops with ¬nesse), I suddenly see a
neighborhood child crying alone in agony on the sidewalk because he has
just slashed his foot on a broken window pane, then I must put down the
¬ddle and go help the child. The latter ought is clearly more pressing in
the larger scheme of things.
Admittedly, in real life it is often dif¬cult to correctly determine which
projects truly are most important. Kant is perhaps a bit too sanguine (not
to mention over the top, in his Star Trek conviction that true moral oughts
necessarily reach beyond the human species to all rational beings) in assum-
ing that it is always possible for agents to know which ones do as a matter
of fact involve absolute necessity and universality.60 But he at least puts
forward a fundamental criterion (one which he ¬rmly believes “common
human reason also agrees completely with”) by means of which we are to
test and evaluate the conclusions of our practical deliberations.61 Williams,
as far as I can see, offers nothing in its place. Agents are charged with deter-
mining which projects are “essential” to their own lives, but they are given

59 Kant (1785/1996d), Ak. 4: 420; p. 72.
60 “Everyone must grant that a law, if it is to hold morally, that is, as a ground of obligation,
must carry with it absolute necessity; that, for example, the command ˜thou shalt not lie™
does not hold only for human beings, as if other rational beings did not have to heed it, and
so with all other moral laws properly so called; that, therefore, the ground of obligation here
must not be sought in the nature of the human being or in the circumstances of the world
in which he is placed, but a priori simply in concepts of pure reason” (Kant (1785/1996d),
Ak. 4: 389; pp. 44“45). See also Kant (1785/1996d), Ak. Ak. 4: 408; pp. 62“63.
61 Kant (1785/1996d), Ak. 4: 402; p. 57. For further discussion, see Louden (1992), pp. 116“
The Critique of the Morality System

no conceptual tools by means of which to ascertain what should really count
as essential and why.
Ethical Practical Necessity. The third philosophical mistake of the moral-
ity system concerns the issue of whether or not conclusions of practical
necessity, even when they are grounded in ethical reasons, necessarily sig-
nal a moral obligation. On Williams™ view, they don™t. Sometimes, what
we legitimately conclude we must do, ethically speaking, has nothing to do
with moral obligation.
As noted earlier in our discussion of Kant and obligation, Kant is
committed to the general view that, because human beings are less-than-
perfectly-rational beings whose psychological make-ups often hinder them
from ful¬lling moral projects, morality confronts them as an ought which
they must bring themselves to strive to carry out “ not as an automatic is
(or even a will be) that necessarily comes about. Again though, this does not
mean that human moral agents always experience moral oughts as unwel-
come demands. Quite the contrary. Virtuous agents, in most cases, will in
fact want to do what they ought to do, because they have educated their emo-
tions to work in harmony with practical reason. But even here, in the ideal
case, reason will be in charge: desire is guided by reason. However, it does
follow from Kant™s basic “morality-confronts-human-beings-as-an-ought”
starting point that on his view all cases of practical necessity grounded in
ethical reasons (which, as we saw earlier, are also the only legitimate cases of
practical necessity, strictly speaking) will signal a moral obligation of some
sort. For in all such cases, there will be a particularly strong ought (one indi-
cating universality and strict necessity) leading to the practical conclusion
about what to do.
But the debate does not quite stop here. The issue is complicated on
Kant™s side due to the fact that he recognizes a plurality of different kinds
of obligation. Agents are often subject to con¬‚icting grounds of obligation,
and in such cases they need to try and determine which particular ground
should prevail. Some of the most dif¬cult cases will involve imperfect obli-
gations, obligations which, as we saw, leave a playroom for free choice in
determining how best to ful¬ll them. We have an imperfect duty to promote
the happiness of others, but which others shall we help? Family relatives?
People in our local community? Strangers in other parts of the world? In
cases involving competing grounds of imperfect obligation, the ought that
prevails will be an imperfect rather than a perfect one. And in such cases,
what Williams calls a case of ethical practical necessity that does not sig-
nal a moral obligation would simply be, in Kant™s language, one ground of
imperfect duty prevailing over another.
122 Robert B. Louden

However, the example given earlier (see Ethical Practical Necessity)
involved an apparent con¬‚ict between a perfect duty (keeping one™s promise
to a friend) and an imperfect one (trying to help others, in this case, part-
time university faculty). Williams, recall, holds that in this particular case
we should break the promise and attend the demonstration in support of
health insurance bene¬ts for part-time faculty. Kant, by contrast, holds
that “imperfect duties always succumb to perfect ones.”62 A perfect duty,
he notes at one point in the Groundwork, “admits no exception in favor of
inclination”;63 whereas every imperfect duty gives us some leeway in deter-
mining how best to ful¬ll it. Kant™s conviction that perfect duties always
trump imperfect ones give his position an absolutist color that contempo-
rary nonconsequentialists are not always eager to defend, but it does square
well with (parts of) traditional moral codes “ for example, St. Paul™s maxim
that we may not “do evil that good may come” (steal from one person in
order to help needy others, kill one person in order to save ¬ve, or “ less
dramatically but much more problematically “ break an appointment with
one person in order to help others).64 At any rate, regardless of what kind
of duties are at stake, Kant does hold that ethical practical necessity signals
Inclination, the Voluntary, and Purity. As regards inclination, much of
what was said earlier concerning the proper role of emotions in ethics is
relevant here as well. Again, it is certainly not Kant™s view that obligation

62 Kant (1793/1997c), Ak. 27: 537; p. 296.
63 Kant (1785/1996d), Ak. 4: 421 n.; p. 73 n.
64 Romans 3: 8. See also Donagan (1977), pp. 149“57, who invokes Paul™s maxim to defend
Kant™s absolutism. Kant™s own defense of the priority of perfect duties is more abstract.
Maxims opposed to perfect duties, he holds, cannot even be conceived as universal laws of
nature without contradiction; whereas maxims opposed to imperfect duties cannot be so
willed. E.g., a world in which the maxim of promise-breaking is universalized is conceptually
incoherent, for the practice of promising would then no longer exist. But a world in which
the maxim of not helping others is universalized is conceptually coherent, though Kant
claims we could not coherently wish to live in such a world “ we might need others™ help
sometime, and it would thus be irrational to choose to live in a world where it would never be
given. This distinction between a contradiction in conception vs. willing, Kant holds, is the
ground for the priority of perfect over imperfect duties (see, e.g., Kant (1785/1996d), Ak. 4:
424; p. 75). I am a contemporary non-consequentialist who denies the absolutist conception
of perfect duties. That is, on my view, keeping an appointment is not always what matters
most, morally speaking. Donagan™s use of the Pauline principle helps make sense of the
traditional view that negative duties (e.g., not to harm) are more stringent than positive
duties (e.g., to aid). But I don™t think it helps support Kant™s more controversial claim that
perfect duties (regardless of how trivial they seem) must always outweigh imperfect duties.
I am also sceptical of Kant™s blanket assertion that every maxim opposed to each and every
perfect duty necessarily involves a contradiction in conception. On my view, this is an open
question “ each such maxim needs to be analyzed. Or, to use Paul™s language: it is not clear
that every single instance of breaking an appointment constitutes “doing evil” to someone.
The Critique of the Morality System

and inclination are always necessarily opposed to one another “ that, as
Schiller famously objected, it is only when we act “with aversion” that we
can be sure to have acted from the motive of duty.65 Rather, virtuous agents
will educate their emotions so that they truly enjoy pursuing morality™s
demands whenever possible, and so that they exhibit the appropriate moral
feelings in the right contexts (e.g., compassion when helping others). Kant
would thus not necessarily oppose weaker versions of Williams™ thesis con-
cerning the relationship between inclination and obligation “ for example,
he could accept the claim that in human experience obligation sometimes
or even frequently presupposes inclinations and projects that are essential
to the agent. But again, even in those cases where an inclination to perform
the obligation is already present, reason must still be in charge. It is only the
stronger, neo-Humean version of Williams™ claim (viz., obligation neces-
sarily always presupposes inclination) to which Kant is ¬rmly opposed. To
assert that moral agents are incapable of bringing themselves to do some-
thing unless they already feel like doing it is to assume much too paltry a
picture of practical reason. The human being is not merely “a plaything of


. 21
( 38 .)