UNIVERSAL LAWS AND ENDS-IN-THEMSELVES
Kant's Groundwork is the most read and surely the most exasperating
of his works on practical philosophy. Both its structure and its arguments
remain obscure and controversial. A quick list of unsettled questions
reminds one how much is in doubt. The list might include the following:
Why does Kant shift the framework of his discussion three times in a short
work? Does he establish that there is a supreme principle of morality? Does
he show that the Categorical Imperative is that supreme principle? Does he
show that human beings are free agents for whom such principles of morali
ty are important? What is the relationship between the various apparently
distinct formulations of the Categorical Imperative? To what extent are any
(or all) of them action-guiding?
This paper concentrates on the last two of these questions. It is mainly
about the equivalence of the various formulations of the Categorical Im
perative: it also sketches ways in which the Categorical Imperative can guide
action. I shall comment only on three significantly different formulations,
the Formula of Universal Law (FUL), The Formula of the End-in-Itself
(FEI) and the (more briefly) on the Formula of the Kingdom of Ends
(FKE).1
By way of reminder the three formulations may be stated:
FUL: Act only on that maxim through which you can at the same time
will that it should become a universal law. (421)2
FEI: Act in such a way that you always treat humanity whether in your
own person or in the person of any other never simply as a means
but always at the same time as an ends. (429)
FKE is not initially stated as a single second-order practical principle. Kant
writes that "morality consists in the relation of all action to the making of
laws whereby alone a kingdom of ends is possible" (434), where a
"kingdom of ends" is characterised as "a systematic union of rational be
ings under common objective laws" (433). A later version runs: "All max
ims as proceeding from our own making of law ought to harmonize with a
possible kingdom of ends as a kingdom of nature." (436)
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
342 ONORA O'NEILL
Since these are the versions of Categorical Imperative for which Kant
himself claims equivalence it seems reasonable to restrict an initial discus
sion to them. He asserts that these three formulations of the Categorical Im
perative are
at bottom merely so many formulations of precisely the same law, one of them
by itself containing a combination of the other two. (436)
This is a puzzling claim. For he promptly interprets FUL as specifying the
form which maxims of duty must have and FEI as determining the matter or
end that they must have, while asserting that FKE provides a complete
determination of all morally worthy maxims. How can all three formulae be
"so many formulations" of the same law if the first two are essentially in
complete and complementary, while the third combines the two incomplete
formulae and is itself complete? How can he say this and then go on rather
dismissively to assert that the significant different is "subjectively rather
than objectively practical." (436) and to suggest that it is just that FUL is
best followed in moral judgement "in accordance with the strict method"
(436) while FEI "is useful when we want also to secure acceptance of the
moral law"? (436) Surely Kant cannot have it both ways. If the three for
mulations are at bottom the same, then the first two are also complete, and
contain all that the third contains, and any differences are indeed merely
subjective; if the first two are incomplete and specify distinct aspects of the
third, then none of them is at bottom the same as any other, and the dif
ference between them is by no means merely subjective.
A surprising amount hinges on the resolution of this dilemma. If the
claim of equivalence cannot be sustained, the argument of Groundwork,
and more generally of Kant's ethics, is deeply unsatisfactory. Most of the
arguments or argument sketches that he provides for the supreme principle
of morality lead us to (at least towards) FUL; yet much that he and many of
his admirers (and even of his critics) find attractive and significant in
guiding moral reflection and action derives from FEI. It is the ideal of
treating persons as ends and avoiding using them as means, not the ideal of
acting on universalizable principles, that has become part of our culture. If
the formulations are not equivalent, then the attractive idea of treating
others as ends and never as means may not be groundable by Kantian
arguments, while the charges of rigourism and formalism which are peren
nially levelled against FUL may lead us to conclude that even if Kantian
arguments show that this is the supreme principle of morality, still we have
not discovered a principle that can help us lead our lives.
A preliminary consideration of the three formulations suggests that
they must be distinct, for two reasons. First, they rely on distinct sets of
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
UNIVERSAL LAWS AND ENDS-IN-THEMSELVES 343
concepts. FUL invokes the notions of action on a maxim and of universal
law; FEI those of action, persons, means and ends and of humanity; FKE
those of action, law, and kingdom of ends. Second, FUL apparently pro
poses a single test of the morality of actions?that they be done on univer
salizable maxims?while FEI apparently makes two demands?that others
not be treated as mere means and they be treated as ends-in-themselves. If
these initial impressions are confirmed Kant's practical philosophy is deeply
flawed. We can make sense of the structure of Groundwork only if there is
some reading of the formulations under which the claimed equivalences
hold. The most demanding task for such a reading would be to connect
FUL and FEI. If these two can be shown "at bottom the same", then so
plausibly can FKE. Conversely any otherwise plausible reading of FUL and
FKE which sustains their equivalence gains some support from the fact that
Kant claims that they are equivalent and that it is vital to his argument that
they be so.
1. Agents and Maxims: The Common Content of the Formulae
The three formulations are all offered as tests that agents can apply to
proposals for action. The Categorical Imperative is nowhere proposed as a
principle that will by itself generate or entail a universal moral code. It is not
a moral algorithm (unlike the Principle of Utility) but (supposedly) a
criterion of moral action for agents who act freely, so may start with
various possible proposals for action. The common assumption of the three
principles is that there is some way by which agents can filter these initial
proposals to check whether they are morally acceptable. Each formulation
of the Categorical Imperative is offered as an answer to the agent's question
"What ought I do?",3 on the assumption that agents will have certain ten
tative plans, proposals and policies which they can consider, revise or re
ject?or endorse and pursue.
A first account of the difference between the three formulations might
stress the differing perspectives from which this agent's question is taken
up. FUL addresses the question from the perspective of agents who
acknowledge that others too are agents, and enjoins them to shun principles
that could not be adopted by others i.e. that could not be universal laws.
FEI addresses the agent's question from the perspective of agents who
acknowledge that action affects others, and enjoins them to avoid dam
aging others' capacities to act. To settle whether or not FUL and FEI
are equivalent, the answers the agent's question receives when explicated
from these two perspectives must be compared. First, however, the two ver
sions of the agent's question must be explicated.
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
344 ONORA O'NEILL
Kant sees action as undertaken on certain principles, which he speaks
of as determinations of the will or as agents' maxims. The interpretation of
the notion of a maxim has been the scene of much argument. For present
purposes five points are needed. I shall state them briefly without textual
comment; these are points on which there is some agreement.4
First, a maxim is a subjective practical principle in the sense that it is a
principle of action of a subject or agent at some time. This is no more than a
restatement of the point that Kant's ethics presupposes agents with prin
ciples or policies of action, which are then to be tested, rejected or accepted,
rather than offering a practical algorithm that prescribes a correct act for
each situation.
Second, we can speak of the maxim of a given act. From this it follows
that not every principle which an act exemplifies is its maxim, nor even
every principle which embodies a description under which the agent acts.
Maxims are not to be equated simply with intentions, which may be multi
ple, some of them profound and others superficial. Rather a maxim is the
underlying or fundamental principle of an action in the sense that any other
principles to which the act conforms are selected and explicable because that
is what it takes to act on a certain maxim in that situation (as perceived by
the agent). The maxim of an act is the principle that governs the selection of
ancillary principles of action that express or implement the maxim in a way
that is adjusted to the agent's (perceived) circumstances. (The maxim is the
maxima propositio or highest principle of some piece of practical
reasoning). This point is crucial for Kant: if acts could have multiple max
ims no test of the moral character of maxims could guide action. One of the
commonest lines of criticism of Kant on this issue works by assuming that
maxims are to be identified with agents' intentions, or perhaps simply with
practical principles which an act exemplifies, infers that acts do not have
unique maxims and concludes that the Categorical Imperative whatever its
demands may be cannot in principle guide action.
Third, it does not follow from the claim that maxims are underlying
principles that they must be the policies of a life time or even of a prolonged
stretch of life, although they may be just that. Scrooge made miserliness his
maxim, but was not doomed to perpetual miserliness. It is a corollary of
taking the freedom of agents seriously that maxims are not unchangeable
dispositions. Even uncharacteristic action is done on some maxim.
However, many agents will in fact hold some maxims for long periods, us
ing them repeatedly to guide their action in varying situations. The specific
acts which express, say, loyalty in friendship or commitment to the profit
motive will, of course, vary enormously depending on the contexts in which
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
UNIVERSAL LAWS AND ENDS-IN-THEMSELVES 345
agents find themselves. (If this account of what a maxim is correct, the long
litany of claims about Kant's demand for rigid and insensitive uniformities
of conduct may be misplaced).
Fourth, the maxim of an act may be a principle which embodies no
description under which an agent acts. Kant takes it that agents' self
consciousness is fallible: we are opaque to ourselves (as also to others) and
may be unsure which principle governs our actions in any situation. We may
hope that we are fundamentally honest, but be well aware that situations we
have faced have been ones in which, as luck would have it, honesty was the
best policy, so that we were never put to the test. All that we can do to try to
ensure that we are honest on principle rather than by luck is to align our
outward action with those that would express a maxim of honesty in ways
appropriate to the each situation we face. It remains possible that some new
situation will disclose to us how limited and fragile our honesty is, leading
us to doubt whether the stretch of life that conformed so well to the out
ward demands of honesty was actually governed by a maxim of honesty.
Fifth, since the implementation of maxims will differ according to cir
cumstances, a test on maxims is not and cannot be enough to determine the
Tightness or wrongness of particular acts (their "legality"); it can only
reveal the moral quality or worth of maxims (so is in Kant's terms a test of
"morality"). Kant defines duty not (as would be common today) as out
ward performance of a certain sort, but as action that embodies a good will
i.e., action on a maxim of a certain sort. (397) However, although moral
worth is more fundamental than Tightness in Kant's theory, Tightness and
wrongness are more easily ascertainable. This is simply a corollary of the
opacity of our self knowledge. If we are unsure what the maxim of a given
act is, we cannot be sure whether it is morally worthy. Despite their best ef
forts at principled and self-conscious action, agents are prey to self
deception and selective perception. This is not rare or exotic but com
monplace?we are repeatedly tempted to ascribe maxims that place acts and
agents in a more flattering or a more lurid light. By contrast, it would be
relatively easy // we had a test to identify morally worthy maxims to deter
mine whether an agent who acted on such a maxim would have acted in a
specific way in those circumstances. In his most pessimistic moments Kant
doubts whether we can ever know that a morally worthy action has been
done, which would mean that we can never judge the morality of acts. This
pessimism need not stop us from judging whether acts that conform to such
maxims have been done?provided that we have a criterion for identifying
morally worthy maxims.5 The various formulations of the Categorical Im
perative are supposed to provide this criterion.
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
346 ONORA O'NEILL
This preliminary account of Kant's theory of action provides the com
mon context for each formulation of the Categorical Imperative. It enables
us to distinguish two views of what it might be for FUL and FEI to be "at
bottom the same." The two formulations might be equivalent in that both
classify maxims, and derivatively the acts that conform to or violate those
maxims, in the same ways: they might be simply extensionally equivalent.
Any maxim which would be rejected as morally unworthy by FUL would
also be rejected as morally unworthy by FEI; the same maxims would be
identified as maxims of duty by both tests; the same acts would be classified
as right or wrong according to their conformity or nonconformity to those
maxims of duty. Alternatively FUL and FEI might be intensionally
equivalent, if it could be shown not merely that they in fact yield the same
results, but that this follows from the nature of the formulations. If FUL
and FEI can be shown intensionally equivalent, then extensional
equivalence is also shown; but merely extensional equivalence would not
guarantee that the formulations are "at bottom" the same. A merely exten
sional equivalence would have practical use, for it would show that either
formulation could be used to identify maxims of duty. However, if we want
insight into why these formulations are both versions of the supreme princi
ple of morality we will need to be shown not merely that they yield the same
results, but why they do so.
2. The Formula of Universal Law
FUL states that we should act only on those maxims through which we
can will at the same time that they be universal laws. This is often
misconstrued as a claim that morally worthy maxims must be ones that we
are willing, i.e., want to see universally adopted (Cf. Golden Rules, Univer
sal Prescriptivism). This may not be such deep died heteronomy as a
Utilitarian pursuit of maximal satisfaction of desires; but it is heteronomy
nonetheless, and Kant rejects it decisively (e.g., 430n.).
Kant's understanding of FUL is uncompromisingly rationalist. He asks
whether we can without contradiction will (not want) a maxim (underlying
principle) to hold as a universal law. His explication of this idea reveals that
FUL (like FEI) has two components. He writes
We must be able to will that a maxim of our action should become a universal
law?this is the general canon for all moral judgement of action. Some actions
are so constituted that their maxim cannot even be conceived as a universal law
of nature without contradiction, let alone willed as what ought to become one.
In the case of others we do not find this inner impossibility, but it is still im
possible to will that that their maxim should be raised to the universality of a law
of nature because such a will would contradict itself. (424)
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
UNIVERSAL LAWS AND ENDS-IN-THEMSELVES 347
These two distinct aspects of FUL are to serve as the criteria for max
ims of strict (or perfect) and of wide (or imperfect) duties. Kant brings
duties of justice and of respect for self and others under the first heading,
and duties of beneficence and self-development under the second.6
Maxims which violate strict duties are said to yield contradictions in
conception if we try to universalize them: the very attempt to think of the
maxim as universally adopted breaks down owing to some incoherence in
the way the world would have to be if it was universally acted on. For exam
ple, the means required for all to adopt and act on the maxim might be in
compatible with the results of all adopting and acting on the maxim. A max
im of deceit can readily be seen as one that we cannot even conceive as
universally adopted. The project of deceit requires a world with sufficient
trust for deceivers to get others to believe them; the results of universal
deception would be a world in which such trust is lacking, and the deceiver's
project is impossible. Of course, this is not to say that there is some con
tradiction in the actual thinking of each deceiver. Far from it. Deceivers
simply aim to use the trust that others have created to get their own decep
tion believed. They rely, as Kant puts it, on substituting generality (and
their own exemption) for universality. (424) The Categorical Imperative is a
thought experiment that we can reject: that is why Kant must show that it is
the supreme principle of morality, and why we can defy it. It is only when
we take up the thought experiment and try to universalize a non
universalizable maxim that a contradiction shows up in our thinking.
There is a fair amount of agreement on this type of account of con
tradictions in conception, but little on whether it provides a plausible
criterion for the range of duties conventionally classified as perfect duties.
Some commentators argue that contradictions in conception emerge only
from attempts to universalize maxims whose universalization would destroy
a practice on which any action on the maxim depends (as universalizing
false promising undercuts the practice of promising on which each proposed
false promise depends). They suggest that other maxims, for example those
of coercive and of brute violence, which aim to destroy or damage respec
tively agents' plans and their bodies, can sail past the contradiction in con
ception test. They suggest that there is no contradiction in universal coer
cion or in universal killing or murder or assault. However, both instrumen
tal and brute violence undercut the agency of those whom they victimise. It
is not merely that victims do not in fact will the maxims of their destroyers
and coercers: they are deliberately made unable to do so, or unable to do so
for some period of time. A test that demands action only on maxims that all
can adopt will require that action not be based on maxims of victimising.
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
348 ONORA O'NEILL
It is unclear how the contradiction in conception test would deal with
self-inflicted violence, such as suicide or self-mutilation, or with violence to
willing victims, such as assisted suicide and sadism towards masochists.
Kant thought such maxims violations of duty; but the Categorical Im
perative may not show this. These cases are at least complex, and need to be
discussed in their own right, rather then with the aim of "rescuing" or con
demning the way Kant articulated them. I shall not do so here because the
sorts of violence that most concern us are brute and coercive violence, and
here the implications of FUL are definite. There is a palpable contradiction
in the thinking of an agent who adopts a maxim of murder or assault, or of
duress and intimidation, which aims to destroy or undercut at least some
other's agency, yet (tries to) will the same maxim as a universal law. Agents
cannot coherently (nor honestly) assume that the agency of those whom
they plan to destroy or damage can already be discounted. It is only after a
killing that its victims are no longer agents; before the killing they are agents
and must fall within the scope of FUL: victims even of minor coercive
violence are evidently agents before and after the violence, which cannot be
willed as a universal law because it aims to undercut agency, at least for
some time.7
Maxims which violate wide or imperfect duties are said to generate con
tradictions in the will when we try to universalize them. A contradiction in
the will is not a contradiction in thinking, but a contradiction between the
thought experiment of universalizing a maxim and the background condi
tions of the lives of specifically finite rational agents. Kant speaks of beings
such as ourselves as finite rational beings not only because their rationality
is limited, but because they are finite in many ways. They have limited
capacities to act which can be destroyed or undercut in many ways. Self
sufficiency is an incoherent goal for finite rational beings; at most they can
coherently aim to minimize their dependence on others. They cannot
universalize maxims either of refusing to accept any help or of refusing to
offer any help, since help may be needed for the survival of their agency.
The thought experiment of willing a world of principled nonbeneficience is
not one that finite rational beings can make consistent with an awareness of
the limitations of their own agency, on which all their plans for action (in
cluding the futile?or perhaps self-deceiving?plan of self-sufficiency) are
premised. A duty of beneficience grounded in this way is only an imperfect
duty: it demands only the rejection of a maxim of refusing (to give or
receive) any help, and not the adoption of a maxim of providing or accep
ting all help (which would in any case be impossible). Which particular
forms of help should be offered or accepted by finite rational beings must
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
UNIVERSAL LAWS AND ENDS-IN-THEMSELVES 349
vary. The types of helping and being helped that are vital to sustain agency
will vary in different situations and with different sorts of finitude.
These considerations show, if they are plausible, only that there is a
reading of FUL which escapes the common charge that the formula iden
tifies no maxims of duty, and the rather less frequent claim that there is no
difference between the two aspects of the formula.8 It is a further matter to
show that Kant provides an account of practical reasoning that includes
procedures of deliberation that lead from maxims to particular decisions.
Although it is slightly tangential to the main point of this paper I shall
sketch an account of deliberation to which I believe Kant is committed;
without this it is hard to set out the various moral distinctions which agents
can draw using the Categorical Imperative.
In the first place the Categorical Imperative allows us to distinguish
maxims of different sorts. Maxims that are not universalizable are contrary
to duty; to act out of such maxims is morally unworthy. Maxims that are
universalizable are not contrary to duty; to act out of them would not be
morally unworthy. Where a maxim is universalizable but the maxim of re
jecting it is not, the first maxim is one of duty and to act out of it would not
merely not be morally unworthy but morally worthy.
Most practical reasoning is not a matter of determining the moral
status of maxims. We usually already have learnt or worked out the moral
standing of many common maxims of duty and of many '' cautionary"
maxims whose adoption would be contrary to duty. These standard maxims
are the principles which we take care to inculcate and identify before we
ever meet life's problems. They, rather than the second-order Categorical
Imperative, comprise the almanac with which we commonly set sail on the
sea of life. We have good reason to check the almanac we inherit; but for
tunately it does not need to be recalculated before every move.9
The almanac can be used to guide deliberation about specific proposed
acts. Kant distinguishes duty from conformity to duty, and much practical
reasoning remains to be done after an agent knows and seeks to observe (or
flout) the standard maxims of duty. For example, a certain proposed act
may be one that could not (in the actual circumstances) be done by
somebody whose maxim was a maxim of duty; such an act will be forbid
den. Another act may be one that whose omission would (in the actual cir
cumstances) be incompatible with acting on a maxim of duty; such an act
will be obligatory. Presumably most acts are neither forbidden nor required
in most circumstances. If Kantian reasoning does not classify maxims ex
haustively into maxims of duty and maxims that are contrary to duty, and
Kantian deliberation reveals that the acts that express or that violate a max
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
350 ONORA O'NEILL
im in particular circumstances vary, then there may be some acts which are
forbidden or required in certain circumstances, and others that are forbid
den or required in all circumstances. Kantian reasoning does not even aim
to provide an algorithm for action; nor does it automatically generate an
* 'overload* ' of obligations. On the other hand, the common claim that FUL
is without practical import is apparently rnistaken. FUL provides a way to
discriminate maxims of duty, and the pattern of deliberation just sketched
can link those maxims to particular contexts of action and decision.
3. The Formula of the End in Itself:
Humanity and Rational Nature
FEI states that we should treat humanity, in ourselves and in others,
"never simply as a means but always at the same time as an ends". There
are so many apparent discrepancies between this formula and FUL, that it is
worth beginning by reemphasizing a basic similarity. FEI too answers an
agent's question. It purports to tell agents what they ought to do. Although
the term maxim does not occur explicitly in the formulation of FEI, I shall
take it that we must read it as a claim about the maxims that ought to guide
action. The notion of a maxim plays the central role already explicated in
Kant's theory of action; without it we can neither distinguish the types of
actions that are to be prescribed or proscribed by duty, nor consequently
work out which particular acts may be forbidden or obligatory in specific
situations.
The most striking discrepancy between FUL and FEI is that FEI refers
explicitly to humanity. This might suggest that the two formulations cannot
be equivalent. Is not FUL formulated for rational beings as such, and FEI a
much more restricted formulation that is relevant only to human beings?
Does not this show that FEI can at best be a special case of FUL? If this is
the whole story we shall not be able to make much sense of Kant's subse
quent claim that the formulae are "at bottom the same."
To see whether it is the whole story we need to consider the relationship
between claims about rational beings as such and claims about human be
ings in Groundwork. Kant insists in the Preface on the need for "Pure
moral philosophy completely cleansed of everything that can only be em
pirical and appropriate to anthropology." (389) As in other works, he pro
ceeds on the assumption that there may be many species of rational beings,
but that we are acquainted only with our own. Hence his illustrations of the
Categorical Imperative are constrained to use instances drawn from human
affairs, but they are intended to illustrate a theory that is not restricted to
human beings. Given the structure of the work it cannot be shown in
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
UNIVERSAL LAWS AND ENDS-IN-THEMSELVES 351
Chapters I and II that these are genuine illustrations, for it is not until the
latter parts of Chapter III?after 450?that Kant gives reasons for thinking
that human beings are indeed free and rational beings.
So long as the agent's question is taken to ask "What ought I do if my
maxims are to be such any other free and rational agent can adopt them?"
this causes no problems. FUL can be stated without assuming even that
there are any other agents: the boundaries of the class of free and rational
beings can be left indeterminate. But if the agent's question is to be asked in
a form that emphasizes the agency of whomever may be affected,
something must be said about the scope of the formula. We can only answer
"What ought I do if my maxims are to leave intact the agency of those
whom my action may affect?" if we take some view of who those other agents
are. Yet the structure of Groundwork means that Kant is no position in
chapter II assert that or whether there are any free and rational agents. In
particular he is no position to assert human freedom or rationality. (He is,
of course, well aware that it is part of our common understanding of
morality that we are free and rational beings?but only the considerations
of Chapter III could vindicate that assumption).
Kant's move in this predicament is appropriate and quite explicit. He
argues hypothetically, using an assumption that can only be vindicated at a
later stage of the argument. He invites us to suppose that there are free and
rational beings, for example ourselves:
Suppose, however, that there is something whose existence has in itself an ab
solute value, something which as an end itself could be a ground of determinate laws_(428)
He continues, in a mode he uses rarely but (I believe) quite deliberately, by
signalling that what he says is (at this point in the argument) mere
assertion:10
Now I say that man, and in general every rational being, exists as an end in
himself, not merely as a means for arbitrary use for this or that will.... (428)
He tells us quite explicitly that the principle rational nature exists as an end
in itself is "put forward here as a postulate" and that the "grounds for it
will be found in the final chapter" (429, inc. note) At the moment we need
not concern outselves with the plausibility of the supposition. What we are
given is a provisional means by which we can refer to a specific class of free
and rational beings who can be thought of as on the receiving end of pro
posals for action. The actions of rational beings of a specific sort do not af
fect all other rational beings without restriction. They affect a restricted
This content downloaded from 128.226.136.66 on Tue, 21 Jan 2020 00:10:19 UTC
All use subject to https://about.jstor.org/terms
352 ONORA O'NEILL
class of rational beings who are, to put the matter vaguely, part of the same
world.
Hence for human beings the Categorical Imperative can be formulated
as a principle constraining the maxims to be adopted by those whose action
affects humanity. However, the term "humanity" is no more than a place
holder. Since it is "rational nature in general" of whom Kant postulates
that it exists as an end-in-itself, FEI could also be formulated, for example,
as the requirement to treat martianness or rational animality or rational ex
traterrestriality never as mere means but always as an end-in-itself.11

Question

Ask a Philosopher · Accepted Answer

I prithee, let us delve into the depths of Kant's Groundwork, that most enigmatic of works on practical philosophy. The structure, the arguments, they do confound and provoke debate. What be the ultimate principle of morality, and doth the Categorical Imperative reign supreme? Dost it truly guide the actions of free agents, those creatures of moral importance? The relationship betwixt the various formulations of the Categorical Imperative doth perplex the mind. How doth one discern the action-guiding nature of these distinct formulations? Let us focus on the Formula of Universal Law, the Formula of the End-in-Itself, and briefly touch upon the Formula of the Kingdom of Ends.

The Formula of Universal Law doth bid us act on maxims universalizable, that we may will them to become universal laws. A test of morality, it distinguishes strict from wide duties. Actions that violate strict duties entail contradictions in conception, for they undermine the very foundation of their own existence. Deceit, coercion, violence, they are but examples of maxims that cannot withstand the test of universalizability. The thought experiment of universalizing such maxims doth reveal the incoherence of their essence.

But lo, the Formula of the End-in-Itself doth beckon us to treat humanity with reverence, as ends and not as mere means. A principle that encompasses rational nature in general, it postulates that rational beings exist as ends in themselves. Humanity, in ourselves and in others, must be respected and cherished. The actions we take must uphold the dignity and autonomy of all rational beings, for they are not mere instruments for our whims, but ends unto themselves.

In this intricate web of moral philosophy, the distinction between these formulations may seem vast. Yet, at their core, they both offer answers to the agent's question: what ought I do? The test of universalizability and the principle of treating rational beings as ends in themselves guide us in navigating the complexities of moral decision-making. And in this realm of ethical inquiry, where reason and duty intersect, we seek to unravel the mysteries of the Categorical Imperative and its implications for our actions in this world of moral quandaries.