Kant's Grounding for the Metaphysics of Morals
A Very brief selective summary of sections I and II 
[Copyright  2000 by Geoffrey Sayre-McCord]

First Section

Kant begins the first section by distinguishing between things that are
"good without qualification" or "unconditionally good" and things that are
good, but only qualifiedly or under certain conditions. Although there are
many things that fall into the second category -- everything that is good
only because of its consequences (since its having those consequences is a
condition on its being good) and even all but one thing that is good in
itself. The one thing that is good without qualification, according to
Kant, is (what he calls) a good will. Indeed, he claims, a good will is
the only thing we can even imagine is good without qualification --
everything else being at best good only with qualification. Moreover, he
maintains, the good will itself serves as a condition of the value of
everything else -- something can be good only if it is (in some
appropriate sense) compatible with a good will. In fact, "a good will
seems" he claims, "to constitute the indispensable condition of being even
worthy of happiness." (393-394) If a good will is unconditionally good
then its value, Kant points out, cannot depend upon its having good
effects. For if its value did depend on its having good effects it would
be valuable only on the condition that it had those effects. Take away the
effects and you would take away the source of its value. Since its value
is (by assumption only, so far) unconditional, it must then be valuable
even absent its having any good effects. Its value must be contained
within it. Kant supposes that we all have at least some idea of what he is
referring to in speaking of a good will. Loosely speaking, it is the
determination to do what, in effect, reason requires as right period.

Kant recognizes that the idea that the role of reason is to make possible
a good will rather than to help us satisfy our inclinations or make
ourselves happy may seem highminded nonesense. So first he argues that if
nature's purpose in giving us reason was to help us satisfy our
inclinations or desires or preferences or to make us happy, it would have
made a big mistake. Reason is ill suited to the task. In any case, the
best way to show that the idea is not nonsense is to develop it in a way
that might make it intelligible and Kant sets about doing just that.

To understand what a good will is Kant turns our attention to the
difference between a person (merely) doing her duty and her doing it
because it is her duty. A good will finds its expression only in the
latter case. Clearly, Kant points out, a person is not exercising a good
will when she does what she knows is wrong. But even when she is doing
what she knows to be right she will be exercising a good will only if she
does what she does because it is right and not because, say, she expects
some reward or happens to want to do it. A person exercises a good will
when, but only when, how that person acts is governed by whether so acting
is compatible with her duty. (397-399)

The value of the action a person performs, insofar as it is an expression
of good will, finds its value or worth "not in the purpose that is to be
attained by it" (i.e., not in the consequences it 2 might produce) but in
the "maxim according ot which the action is determined" (i.e., in the
reason the agent had for acting in that way -- the recognition that so
acting was at least compatible with duty). (399-400)

But if the value of an action done from duty is found not in the
consequences it produces but in the respect for duty it expresses then
one's duty must be to express that respect rather than to produce any
particular effects. Doing one's duty because it is one's duty must then be
a matter not of trying to achieve some effect but of conforming one's will
to a principle of duty (law) that commands respect. (400-401)

What sort of principle might this be which can determine the will without
appealing to some expected consequences of acting as it requires? Kant
claims it must be a principle that requires one act so that the reasons
one has for so acting could themselves be principles of the will. Moral
action, it turns out, is in a very deep sense principled action -- action
done on the basis of considerations that could themselves stand as
principles. (401-402) With this view in mind, Kant offers an example of a
person considering whether or not to make a false promise. In asking
oneself whether one may make such a promise one might be asking, Kant
points out, whether so acting will promote one's welfare or one might be
asking whether so acting is compatible with duty. The first question is
often tricky to answer since it turns on so much that might be
undiscoverable. The second question, though, is more readily answered Kant
maintains because it is really the question of whether one could
consistently commend the considerations that would lead one to make a
lying promise as principles that might guide everyone's action. That we
could not commend the considerations as principles of action, Kant holds,
shows that acting on them is immoral, whatever the consequences might be.

Second Section 

The first section was given over to identifying and articulating our
concept of duty, which demands and finds expression in a person's having a
good will (i.e. in her doing her duty because it is her duty rather than
because she expects some advantage or happens to feel like it). This is a
concept with which we are all familiar, Kant maintains. Yet, while we all
have experience of this concept, Kant argues that the concept itself is
not one we get from experience. It is, in his terminology, an a priori,
not an a posteriori, concept. However we come to acquire the concept of
duty, it is not by first experiencing instances of duty and then
extracting our concept from those instances. Central to his argument is
Kant's observation that morality's demands are unconditional, apply to all
rational beings, and allow no exceptions. Whatever experiences we might
have, he holds they couldn't be such as to validate by themselves any such
demands. At the same time, he points out, in order to use examples as
examples of someone acting morally we would already have to have the
concept in order to determine whether the cases were appropriate -- i.e.,
actually cases of someone acting morally. Thus identifying the relevant
cases presupposes our already possessing the concept that was purportedly
derived from the cases. Moreover, he argues, to try to defend one's moral
views by appeal to popular opinion or sentiment or an appeal to experience
only will always result in clouded view of morality. (406-411)

If moral concepts "have their seat and origin completely a priori in
reason," though, they must be understandable wholly without appeal to
aspects of the human condition that can be known only from experience. No
doubt moral concepts can be applied to people and the situations they face
only in light of specific information experience provides, but the
concepts themselves must be comprehensible independent of knowledge
provided by experience. (412) In order, then, to understand moral concepts
conceived as applying to all rational beings regardless of their
particular and contingent circumstances, the place to start (Kant holds)
is with the concept of a rational being. The distinctive and defining
features of rational beings, according to Kant, is their capacity to act
as they do because of their "conception of laws" -- that is, their
capacity not merely to be pushed around by forces but to act as they think
they should (a capacity that involves being governed by their
representation of some option as good or required). In a perfectly
rational being, the representation of something as good or required is by
itself, and without resistance, sufficient for action. For other beings --
such as human beings -- whose rational capacities govern a will that might
be moved by various incentives, temptations, and fears, the representation
of something as good or required is not, by itself, sufficient for action.
In such beings, the determination by their reason that some option is good
or required presents itself as a kind of command -- as the judgment that
they ought to act accordingly (even if they want not to). (412-414)

Importantly, one's judgments that one ought to act in one way or another
fall into two different categories. Sometimes the grounds one has for
judging one ought so to act depend upon certain conditions being
satisfied, so that the imperatives are hypothetical or conditional --
their practical force (i.e. their implications for action) depends upon
the conditions in fact being satisfied. (The relevant conditions, it turns
out, are that one has adopted a particular end or purpose.) Other times,
the grounds one has for judging one ought so to act depend on nothing
contingent, so that the imperatives are categorical -- their practical
force (i.e. their implications for action) is unconditional and so not
dependent on the hypothesis that certain conditions are satisfied. As
various hypothetical imperatives make clear, the judgments at issue here
need not be moral in their ground or upshot. One might, for instance,
judge that one ought not take another bite of the cake (despite the
temptation) on the ground that one has decided to lose weight and
foregoing such pleasures is a necessary step to achieving one's purpose.
One needn't think one has a moral duty to lose weight, nor a moral duty
not to eat given that one has decided to lose weight, even as it is clear
that eating under such circumstances -- unless one abandons the diet --
would be irrational. Hypothetical imperatives in effect represent some
action as good or required on that condition that one has some purpose,
where the purpose is one that is, from the point of view of rationality,
optional. Categorical imperatives, in contrast, in effect represent some
action as good or required unconditionally. (414-415)

Hypothetical imperatives can be divided into those that are 'problematic'
and those that are 'assertoric' depending upon whether the purpose in play
is one that merely might be adopted or is one that has actually been
adopted. Among the assertoric hypothetical imperatives, according to Kant,
is the imperative that one act prudently (since doing so is required in
order to achieve an end he believes we all necessarily have -- the end of
promoting our own welfare). These hypothetical imperatives, whether
problematic or assertoric, differ significantly from the imperatives of
morality precisely in their conditionality, in their force depending on
people having certain purposes. Morality requires that we act in certain
ways not as a means to achieving some further end or purpose but directly
and unconditionally, whether or not acting as it requires happens to
further our ends or not. (415-417)

Against this background Kant turns to the question of how reason can
require anything of us. What makes it true, in the case of hypothetical
imperatives, that given certain ends or purposes it would be irrational
not to act in a certain way? What might make it true, in the case of
categorical imperatives, that it would be irrational not to act in a
certain way regardless of one's ends? (417) 

When it comes to hypothetical imperatives, Kant thinks there is no real
problem in explaining the irrationality. As he sees it, to adopt an end is
to set oneself to be the cause of achieving it, and the idea of one's
being the cause of its achievement contains within it the idea of one's
doing whatever is necessary. So one cannot rationally think of oneself as
the cause and not be thinking of oneself as taking those necessary steps.
To will the end (that is, to set oneself to achieving it) is thus to will
the necessary means. (417) As Kant puts it: "it is one and the same thing
to conceive of something as an effect that is possible in a certain way
through me and to conceive of myself as acting in the same way with regard
to the aforesaid effect." This might be a helpful parallel. To think of
oneself as a bachelor is ipso facto to think of oneself as unmarried. The
idea of one's being a bachelor contains within it the idea of one's being
unmarried. So one cannot rationally think of oneself as a bachelor and not
be thinking of oneself as being unmarried. To believe one is to believe
the other. In neither case, of course, is Kant committed to denying that
people sometimes fail to will -- or believe -- as rationality requires.
Such failures are all too familiar. What he is trying to do is identify
what the failure consists in, and his answer is that (outside of morality)
failures of practical rationality are failures to take what one recognizes
to be the necessary steps to achieving one's ends while not abandoning the
ends. In the face of a hypothetical imperative, there are two ways to
preserve one's practical rationality: (1) take the necessary means to
achieving the ends the give the imperative its practical force or (2)
abandon those ends. Either is a rational response to one's situation. What
is irrational is retaining the end while failing to take the steps one
recognizes as necessary for its achievement. (417-420)

When it comes to categorical imperatives, Kant thinks the problem of
making sense of even its possibility is a genuine challenge. Categorical
imperatives differ from hypothetical imperatives in that, by not
presupposing any particular ends they cannot derive their rational force
from our having adopted those ends. Nor can we hope to discover a
categorical imperative's force by looking to examples because we can never
be sure we have an example on hand and, more important, even if we did
have an example, it would at most reveal the conditional force of the
imperative. Instead of directly defending the force of the categorical
imperative, Kant turns to the problem of identifying its content. What is
it that reason might require unconditionally of all rational agents
without regard to their particular ends or purposes? (420) 

In answering this question, Kant notes that the content of a categorical
imperative must be found solely in the form of the imperative (as an
unconditional law that applies to all) and its demand that the will
conform to it. "Hence," he writes as if the implication were at all clear,
"there is only pone categorical imperative and it is this: Act only
according to that maxim whereby you can at the same time will that it
should become a universal law" [p. 30; italics added]. More colloquially,
the demand is that in whatever you do you should act for reasons that
could serve as acceptable reasons for anyone. To make an exception of
oneself, to act for reasons one could not acknowledge as acceptable were
someone else to act on them, is to act immorally. How exactly this
specific demand is supposed to follow from the form and force of a
categorical imperative is a matter of serious and interesting controversy.
Yet the principle itself has struck many as capturing an important feature
of morality -- that its demands are universal and require, in some
significant sense, that we treat all people the same. To act immorally is,
on this view, to put oneself "above the law" -- not the civil law of some
society or other (which might, after all, be a deeply immoral society) but
the law one's own reason imposes upon one. And the immorally is found not
principally in the acts one performs but, as Socrates emphasized, in the
reasons one has for so acting. (421)

Kant then offers four examples of immoral actions, selected in part
because they highlight the two different ways a person might fail to be
acting on a maxim that would meet the requirements of the categorical
imperative. A maxim might fail by not being universalizable -- by being
such that the very conception of it as being a law governing all is
inconsistent. (We have, as Kant would put it, a perfect duty to refrain
from acting on such maxims.) Or a maxim that might be universalizable
might fail the requirements of the categorical imperative by being such
that a person could not consistently will that the maxim be a universal
law. (We have, as Kant would put it, an imperfect duty to refrain from
acting on such maxims.) In both cases, the failure of the maxim is a
failure of consistency in an important sense. There is no question that an
immoral maxim can itself be perfectly consistent, after all people
actually act on them. What is inconsistent is either (i) the conception of
that maxim as a universal law or (ii) willing that the maxim serve as a
universal law. Thus in testing a maxim (and so evaluating an action that
might be performed on its basis) we can look for two kinds of
inconsistency -- inconsistency in conception and inconsistency in willing.

The first and second examples are supposed to be cases in which the agent
is considering acting on a maxim that cannot consistently be conceived as
a universal law. The third and forth examples are supposed to be cases in
which the agent is considering acting on a maxim that can consistently be
conceived as a universal law but that cannot consistently be willed. The
difference between the first and third, on the one hand, and the second
and the forth on the other, is that the former have to do with one's
treatment of oneself, while the latter have to do with one's treatment of
others. (421-425)

Kant next offers some important observations concerning what could not
serve as the source or foundation of morals. In particular, he argues,
morality cannot legitimately be grounded merely in human nature, even as
it applies to humans and applies in a way that is sensitive to our nature.

The unconditional force of the categorical imperative is explicable, Kant
maintains, only on the supposition that it derives its authority purely
from our capacity, as rational beings, to determine our actions by the
representation of certain options as required by rationally necessary laws
(i.e. hypothetical and categorical imperatives). If those rationally
necessary laws requires some action unconditionally, that action will be
required of all rational beings (human and otherwise), and the actions
prescribed are "objective ends" that are "valid for all rational beings".
They will be the actions that are morally required. If, on the other hand,
the rationally necessary laws require the adoption of some action only
conditionally (say on the condition the agent happens to adopt some end or
purpose) that action's worth is derived from, and contingent on, the agent
having adopted the end or purpose in question. And assuming the action is
not incompatible with a morally required action it will be morally
permissible. (426-428)

How must we conceive of rational beings in order to make sense of their
wills as grounds of unconditional (and conditional) value? Only, Kant
supposes, by thinking of rational beings as ends in themselves (and not
merely more or less useful means to achieving some end a person might
adopt). The value of everything other than rational beings, Kant goes on
to say, is conditional and, in particular, conditional upon the rational
beings happening to value those things. (428)

According to Kant, people inevitably view themselves (insofar as they are
rational) as ends in themselves and as sources of value. Yet the reason
each person has for viewing herself this way is equally a reason for her
to value others similarly. So our viewing ourselves in this way commits us
to viewing others too as ends in themselves. Thus reason requires that you
"act in such a way that you treat hunaity, whether in your own person or
in the person of another, always at the same time as an end and never
simply as a means" (p. 36). This is the second formulation of the
categorical imperative. (429)

Using the same four examples Kant sets out to illustrate the two different
ways in which this formulation of the categorical imperative rules out
certain actions as immoral. The first two examples are of actions that
involve a person treating either himself or another merely as a means. The
second two examples, in contrast, don't involve a person treating someone
as a mere means but do involve a failure actively to embraces others as
ends, as beings with a worth beyond that conferred by their utility.

It is worth noting that the second formulation of the categorical
imperative does not rule out treating people as means -- using them to
help further your own ends. What it rules out is treating them merely as
means, as if their worth depends solely on their serviceability.

The over-all picture is that anything whatsoever a rational being might
decide to do is morally permissible as long as in deciding to do it the
agent is respecting the categorical imperative -- that is, acting for
reasons the agent could will to be universal laws and (Kant thinks
equivalently) treating all rational beings involved as ends and not merely