8/13/2019 Kant's Anti-realism
1/24
Kant's Moral Anti-Realism
Frederick Rauscher
Journal of the History of Philosophy, Volume 40, Number 4, October
2002, pp. 477-499 (Article)
Published by The Johns Hopkins University Press
DOI: 10.1353/hph.2002.0082
For additional information about this article
Access provided by Boston University Libraries (11 Dec 2013 14:11 GMT)
http://muse.jhu.edu/journals/hph/summary/v040/40.4rauscher.html
http://muse.jhu.edu/journals/hph/summary/v040/40.4rauscher.htmlhttp://muse.jhu.edu/journals/hph/summary/v040/40.4rauscher.html8/13/2019 Kant's Anti-realism
2/24
477K A N T S M O R A L A N T I - R E A L I S M
Journal of the History of Philosophy,vol. 40, no. 3(2002) 47799
[477]
Kants Moral Anti-Realism
F R E D E R I C K R A U S C H E R *
To annihilate the subject of morality in ones own person is to root out the existence of moral-
ity itself from the world, as far as one can. Immanuel Kant writing on suicide, Metaphys-
ics of Morals, 6:423.1
To some Kantians it may seem obvious that Kant was a moral anti-realist since he
appears to have admitted his anti-realism repeatedly by invoking transcendental
idealism in ethics. But to other Kantians it appears just as obvious that Kant was a
moral realist given his claims that we are obligated to morality categorically, that
we must believe that God exists to buttress the moral order of the world, and that
when we think about the world as it is in itself as a noumenal world, we must
employ reason and reasons child, morality. These different claims about Kants
moral theory stem from two sources: disagreements about the proper definition
of moral realism and disagreements about Kants own moral theory. This paper
will first briefly survey different claims about Kants realism or anti-realism and
provide a definition of moral realism; then the bulk of the paper will show that
the metaphysics in Kants moral theory, when properly understood, is anti-realist.
1 . C U R R E N T I N T E R P R E T A T I O N S O F M O R A L R E A L I S M I N K A N T
There is no firm agreement about whether Kant was a moral realist or moral anti-
realist.2 I will review three positions that exemplify different general approaches to
Kants metaethical theory: John Rawlss anti-realist constructivism, Allen Woods
realist focus on the rational will, and Karl Amerikss strong moral realist metaphysics.
* Frederick Rauscheris Assistant Professor of Philosophy at Michigan State University.
1References to Kants works will be to their volume and page number from Kants gesammelte
Schriften, 29vols. (Berlin: Walter de Gruyter [and predecessors], 1902 ), except references to Kritik
der reinen Vernunfttaken from the edition edited by Jens Timmermann (Hamburg: Felix Meiner, 1998),
and identified by page numbers for the first and/or second editions, cited as A and B. These page
numbers are given in the margins of the English translations. I use the following English translations:
Critique of Pure Reason, Paul Guyer and Allen Wood, trans. (New York: Cambridge University Press,
1998) and Practical Philosophy, Mary Gregor, trans. (New York: Cambridge University Press, 1996).2Here I am concerned with interpretations of Kant in particular rather than various contempo-
rary Kantian moral theories. The difference is apparent, for example, in Philip Stratton-Lake, who
recently claimed that there is little doubt that Kants ethics has no place for moral realism and then
proceeded to offer suggestions for Kantian moral realism in contemporary moral theory (Stratton-
Lake, Kant and Contemporary Ethics, Kantian Review2[1998]: 113, 10ff).
8/13/2019 Kant's Anti-realism
3/24
478 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
Rawls takes Kant to offer a constructivist theory in which the categorical im-
perative is, roughly speaking, understood as a procedure for testing maxims. The
result of the procedure will be a set of permissible maxims that form the content
of morality; these are said to be constructed because they do not reflect any prior
moral order. The categorical imperative procedure itself is not the result of con-
struction but rather laid outon the basis [of] the conception of free and equal
persons as reasonable and rational, a conception that is mirrored in the proce-
dureand elicited from our moral experience.3 Thus, practical activity by agents
who view themselves with a resulting collective self-conception provides the basis
for a procedure that in turn provides the content for morality. On this view Kant
is seen as a moral anti-realist because morality is not independent of the practice
and self-conception of certain types of beings.4
Wood focuses on the nature of the rational will as such, not any particular
activities or thoughts of particular beings. He sees Kant as holding that the truth
of moral statements stems from reason itself, identical for Kant with the rational
will. The basis of the independent truth of moral statements, then, is the nature
of reason itself, not the practice or thoughts of particular beings endowed with
reason. Reason itself provides the ground for the derivation of particular moral
principles. Wood holds that since Kant holds that moral truth is irreducible ei-
ther to what people think or to the results of any verification procedures, he is a
moral realist in the most agreed-upon sense that term has in contemporary meta-
physics and metaethics.5 Wood allows for the possibility in principle for us to be
mistaken regarding moral principles because our judgment might err, individu-
ally or collectively, by failing to agree with the idea of the rational will. The real
moral principles are not dependent on our actual beliefs about them.6
3John Rawls, Themes in Kants Moral Philosophy,in Rawls, Collected Papers, Samuel Freeman,
ed. (Cambridge, MA: Harvard University Press, 1999), 497528 , 514 . See also his Kantian
Constructivism in Moral Theoryin the same volume (30358) and Lectures on the History of Moral
Philosophy, Barbara Herman, ed. (Cambridge, MA: Harvard University Press, 2000), 235ff.4Among others who hold a similar view is Christine Korsgaard. Although a constructivist, Korsgaard
is a self-proclaimed Kantian realist. See Korsgaard, Sources of Normativity(New York: Cambridge Uni-
versity Press, 1996), 10812. She distinguishes Kants procedural moral realismfrom metaphysical
realism, which she calls substantive moral realism(op. cit., 35) or objective realismin Creating the
Kingdom of Ends(New York: Cambridge University Press, 1996), 27882. Procedural moral realism is
the view that there are right and wrong methods to find answers to moral questions. Substantive (or
objective) moral realism further holds that there are moral facts or truths that account for the right
answer to those questions.5Allen Wood, Kants Ethical Thought(New York: Cambridge University Press, 1999), 157. The
agreed-upon senseof moral realism Wood cites in his book (374, n. 4) is that of Richard Boyd
discussed below.6Among others who hold a similar view is Dieter Henrich. Henrich focuses on the nature of
moral insightas ultimately, for Kant, based in the fact of reason and not itself a theoretical claim.
The fact of reason is the nature of rational beings as practical agents subject to the categorical impera-
tive (Henrich, Der Begriff der sittlichen Einsicht und Kants Lehre vom Faktum der Vernunft,in
Kant: Zur Deutung seiner Theorie von Erkennen und Handeln, Gerold Prauss, ed. [Kln: Kiepenheuer &
Witsch, 1973], 22354, 2478, translated by Manfred Kuehn as The Concept of Moral Insight and
Kants Doctrine of the Fact of Reason,in The Unity of Reason, Richard Velkley, ed. [Cambridge, MA:
Harvard University Press, 1994], 5587, 83). Although Henrich does not use the terminology of
realismor anti-realism,I believe he would hold that Kant is a moral realist for roughly the same
reasons that Wood does.
8/13/2019 Kant's Anti-realism
4/24
479K A N T S M O R A L A N T I - R E A L I S M
Ameriks takes Kant to insist on a strong metaphysical grounding for morality.
He sees Kant as a realist insisting on an independent standard of morality that can
be known by human reason yet is not created by it:7 In practical philosophy we
move beyond appearances, we have absolute truthof the real standards of moral-
ity in a way very similar to a traditional rationalist.8 Ameriks also cites freedom as
a requirement of morality for Kant.9 Kant insists on a literal notion of
nonempirical agency that would give talk of freedom a real reference and
point.10 Kant requires that we have a real transcendent power of freedom. His
shift in the second Critiqueto a discussion of a fact of reason, Ameriks suggests, is
only an abandonment of any attempt at a theoretical derivation of this freedom
and not a denial of such freedom itself. The strong metaphysical realism required
for morality was reinforced by Kant even at the end of his career when he de-
scribed his ethics as dogmatic and not critical.11 Thus, Kant is a moral realist since
he holds that morality is based on independent standards and requires transcen-
dental freedom.12
Each of these three commentators is merely illustrative of a general view of the
relation of Kants ethics to moral realism. Some of the differences among them
are clearly due to different definitions of moral realism. For example, Christine
Korsgaards Rawlsian constructivist interpretation of Kant is deemed realist by
her but anti-realist by Wood.13 But there are just as clearly quite substantive issues
at stake. The three positions offer different views of the nature of Kants moral
theory. The constructivist position eschews any metaphysics and locates the basis
of moral right in the actual practice of moral agents. Kant is seen as sharply sepa-
rating moral issues from any static basis independent of that practice. Woods
focus on the nature of the rational will, in contrast, requires a conception of rea-son independent of actual practices of moral agents. Kant is seen as resting mo-
rality on the basis of an a priori view of the essence of a certain type of mental
faculty. Ameriksfocus on a strong metaphysical realism goes even further by ar-
guing that this mental faculty must be taken as a non-empirical power of agents in
themselves. Kant is seen as arguing for a two-world transcendental idealism to
defend this positive metaphysical status of agents in themselves as non-empirical.
Even if followers of all three approaches could agree on a definition of moral
realism, then, there could remain disagreement regarding whether Kant himself
was or was not a moral realist.
7Karl Ameriks, On Schneewind and Kants Method in Ethics,Ideas y Valores102(1996): 2853.8 Ibid., 40, 44.9Karl Ameriks, Kant and the Fate of Autonomy(New York: Cambridge University Press,2000a), 69
75; Karl Ameriks, Kants Theory of Mind, 2nd ed. (New York: Oxford University Press, 2000b), 189
220.10Ameriks, 2000a, op. cit., 74.11Ameriks, 2000b, op. cit., 218.12Among others who hold a similar view is Gnther Zller. Zller notes that Kants theoretical
transcendental idealism is balanced by a critically mitigated nonempirical realism on ethical and ethico-
teleological grounds that rehabilitates metaphysics in a moral guise(Zller, Fichtes Transcendental
Philosophy[New York: Cambridge University Press, 1998], 13).13For Korsgaards claim that her position is moral realism, see note 4above. Wood labels her
position moral anti-realism in a review of one of her books (Wood, Review of Creating the Kingdom of
Endsby Christine Korsgaard, Philosophical Review107[1998]: 60711). Since they agree on the con-
tent of her position, their disagreement is only one of the proper definition of moral realism.
8/13/2019 Kant's Anti-realism
5/24
480 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
I will not here evaluate these positions in detail. My own interpretation of the
nature of Kants moral theory given below will make clear my disagreements with
these positions. In brief, the position I develop falls between those of Rawls and Wood.
Like Rawls, I will show that something about the actual experience of moral agents
lies at the basis of moral claims, but unlike Rawls I think this is not a self-conception
drawn from moral practice but an internal experience peculiar to human-like
rational beings. Like Wood, then, I hold that rational nature is involved in the
foundation of morality, but unlike Wood I do not take this basis to be independent
of the actual nature of existing beings. And although my position is distant from the
strong metaphysics invoked by Ameriks, I will show why Kant requires us to have
certain beliefs regarding such a metaphysics. My position is that Kant is a moral anti-
realist because he ultimately holds that the basis of morality is an experience limited
to the minds of human-like beings. That this can be considered moral anti-realism
is shown in a discussion of the definitions of moral realism in my next section.
2 . T H E N A T U R E O F M O R A L R E A L I S M
Moral realism is one of those issues about which philosophers spend much of
their time simply defining the terms of the debate. There is no clear consensus on
what realismactually means, and some have even suggested that the meaning of
realismshifts along with the philosophical tide. According to this last view, some
theories that would have counted as anti-realistic fifty or a hundred years ago
would count as realist today.14
One problem that besets the task of defining moral realism is that various
metaethical theories differ in their interpretation of key terms such as truthand
validity,objectiveand subjective,and obligatoryand permissible.Disagree-ment on these key terms allows for various different theories to claim to present
moral realist positions although they vary widely with one another.
Moral realism also suffers from the connotation of one of the central terms of
the debate: anti-realism.As an anti-, the latter connotes that one is opposed to
some positive claim rather than that one is giving a positive claim of ones own. In
ethics this tendency is exaggerated by the moral import of the terms involved. To
deny moral realism seems to imply a lessening of the claims or values of morality
itself. One writer on the topic, for example, suggests that one might want to be
able to say that the Nazis were really really wrong,but might feel unable to make
such a strong claim without a theory of moral realism behind oneself.15
These two problems combine their effects in several attempts to define moral
realism. Geoffrey Sayre-McCord (who offered the Nazi example I just mentioned)
claims that realism involves embracing just two theses: (1) the claims in ques-
tion, when literally construed, are literally true or false (cognitivism), and (2)
some are literally true.16 Peter Railton holds that The realist about X gives an
account of X discourse that is cognitive, non-error-theoretic, literal, and opinion-
14Peter Railton, Moral Realism: Prospects and Problems,in Moral Knowledge?, Walter Sinnott-
Armstrong and Mark Timmons, eds.(New York: Oxford University Press, 1996), 4981.15Geoffrey Sayre-McCord, Introduction: The Many Moral Realisms,inEssays on Moral Realism,
Geoffrey Sayre-McCord, ed. (Ithaca, NY: Cornell University Press, 1988), 123, 1.16Ibid., 5.
8/13/2019 Kant's Anti-realism
6/24
481K A N T S M O R A L A N T I - R E A L I S M
independent (objective, in our sense).17 These definitions both equate moral
realism with a theory of language, centering on the question of whether our moral
utterances are true without specifying how they would be true.18 David Brink goes
further by adding to his definition a claim that moral facts exist: moral realism is
roughly the view that there are moral facts and true moral claims whose existence
and nature are independent of our beliefs about what is right and wrong.19 This
definition, however, does not specify the criteria for moral facts beyond the crite-
rion that moral facts must be independent of our beliefs; the nature of beliefs is
also not specified. Depending upon how one construes the terms moral facts
and beliefs,one could be labeled either realist or anti-realist.20
Richard Boyd offers three criteria for moral realism: (1) moral statements are
true or false, (2) this truth or falsity is independent of our moral opinions, theo-
ries, etc. . . .,and (3) ordinary moral reasoning is a reliable method for obtain-
ing and improving moral knowledge.21 But the indeterminacy of the list our
moral opinions, theories, etc.in (2) makes it unclear whether moral truth must
be independent of every type of human mental activity, in particular, of the moral
reasoning he mentions in (3). Thus if Rawlss constructivist moral reasoning, which
bases moral truth and falsity upon a procedure stemming from our self-concep-
tions derived from moral practice and not from mere opinion, is excluded from
the etc.in (2), the result is that a constructivist interpretation of Kant such as
that offered by Rawls would count as realist; if that type of constructivist moral
reasoning is included in the etc.in (2), then Rawlss Kant would count as an
anti-realist. The definition is not specific enough to determine this matter.22
The above attempts to characterize moral realism stress epistemological and
linguistic criteria. Panoyot Butchvarov, in contrast, provides a more straightforwardlymetaphysical definition: I shall mean by unqualified realism with respect to x the
view that (1) x exists and has certain properties, a nature, and that its existence and
nature are independent of (2) our awareness of it, of (3) the manner in which we
think of (conceptualize) it, and of (4) the manner in which we speak of it.23
Butchvarovs definition properly, I think, invokes metaphysical criteria for realism.
Its weakness lies in the requirement that the real must have properties and a nature.
Such a criterion would exclude from counting as realist any moral theory that takes
the morally real to be a set of properties supervenient on non-moral entities.
17Railton, op. cit., 57.18A moral realism that draws directly on Wittgensteins theory of language is offered by Savina
Lovibond in Realism and Imagination in Ethics(Oxford: Basil Blackwell, 1983). Because it concerns the
actual form of life of human beings, it resembles Rawlss theory in relating morality to a self-concep-
tion drawn from the actual practice of human beings. Lovibonds theory might then be labeled anti-
realistaccording to some definitions.19David O. Brink, Moral Realism and the Foundations of Ethics (New York: Cambridge University
Press, 1989), 7.20Wood, for example, complains that Rawlss claim to reject moral facts in favor of procedural
construction of moral norms is merely a verbal difference(Wood, 1999, op. cit., 3745, n. 4).21Richard Boyd, How to be a Moral Realist,in Sayre-McCord, op. cit.,181228, 182.22Boyd pursues the former of these alternatives but appears to recognize that the definition does
not dictate such a move (ibid., 199200).23Panoyot Butchvarov, Realism in Ethics,in Midwest Studies in Philosophy Volume XII: Realism and
Antirealism, Peter A. French, Theodore E. Uehling, Jr., and Howard K. Wettstein, eds. (Minneapolis:
University of Minnesota Press, 1988), 395412, 396.
8/13/2019 Kant's Anti-realism
7/24
482 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
The approach to moral realism I want to take does not address the epistemo-
logical question of truth conditions but instead focuses on the metaphysical status
of anti-realism while avoiding the difficulties with Butchvarovs offering. To stress
this metaphysical aspect, I prefer to use the term idealism,a traditional antonym
of realism,instead of anti-realism.This name suggests a set of metaphysical
commitments that narrows the debate dramatically. I will now offer and defend
my definition of moral realism. Later, after showing that Kant is a moral idealist in
this sense, I will return to some of the other definitions of realism to show how
they relate to mine.
Moral realism: The belief that some of the moral characteristics of the world
are independent of the human mind.
Moral idealism: The belief that all of the moral characteristics of the world are
dependent upon the human mind.
By dependentI have in mind ontological dependency such that the charac-
teristics could not exist, or could not exist as they do, independent of the exist-
ence of human minds. When I use the term mindI do not intend to invoke any
Cartesian substance; rather, I intend to be agnostic about the true nature of minds.
At the same time I limit this to the human mind as we generally experience it and
conceive of it. If it were to turn out that the true nature of human minds were that
they are parts of Gods mind, a resulting morality could be realist. A moral charac-
teristic is not the same as a characteristic required for morality. The difference
between a moral characteristic and a characteristic required for morality is that
the latter consists of properties of objects that are not solely moral, the former of
properties that are solely moral. Moral characteristics are to be understood as
involving moral normativity or value. For example, the existence of a human mind
may be a characteristic of entities (humans) required for morality, but because it
can play a role in other areas not pertaining to moral normativity or value, such as
theories of qualia, it is not a moral characteristic. An individuals being good,
however, is a moral characteristic, since this characterization can play no role
except in situations involving moral normativity or value. My definition of this
term moral characteristicis intended to be independent of any particular meta-
physics. As examples of moral characteristics, consider the following: good, evil,
bad, rightness, wrongness, justice, value. As examples of characteristics required
for morality, consider the following: freedom of the will, agent causality, the exist-
ence of minds, etc.
Given these definitions, one can easily see that Rawlss position would count asidealist while Woods and Amerikss positions would count as realist.24 This excur-
sion into the nature of moral realism has at least this result: the anti-realist has to
offer a complete idealist account of morality (or simply reject morality altogether).
24This claim might be contested by someone who would hold that the transcendental ideality of
the self as independent of conditions of space and time and as possessing a set of transcendent powers
or faculties is the true nature of the human mind. In this case, Amerikss type of interpretation would
count as anti-realist. I contend that such a conception is not how we generally experience and con-
ceive of ourselves as required by my definition of moral idealism. This conception would place human
beings in a far different position than we normally allow, and thus might place human beings them-
selves at such a central place ontologically that moral realism would be called for. I will show below
that Kants moral theory does not depend upon this strong metaphysics in this way.
8/13/2019 Kant's Anti-realism
8/24
483K A N T S M O R A L A N T I - R E A L I S M
For moral realism to be true, there must be at least one important moral charac-
teristic of the world that is not dependent for its existence or its truth on the
human mind. Moral idealism, in contrast, entails that every moral characteristic
of the worldmoral propositions, value, obligation, imperatives, etc.depend
for their truth or existence on the human mind.25
The above discussion makes clear the error in equating realism with objectiv-
ity. Objectivity can coincide with idealism provided that the minds in question are
all structured in such a way that they necessarily produce the same ideas. An anal-
ogy can be made with the question of color perception. Simply because our color
qualia may be ideal and phenomenal colors may not inhere in objects indepen-
dent of the human mind, does not mean that the color qualia we perceive are not
objective. As Locke pointed out, the connection appears to us contingent, but
may have some necessity from an unknown source.26 There might be some neces-
sary causal connection between our physical sensations and our mental percep-
tions. We might have a particular mental faculty that provides for the objectivity.
Or God might have set up human minds in such a way that they all must contain
the same ideas. For morality, objectivity can be preserved even when idealism is
accepted, provided that the theory explain how the structure of the human mind
dictates that all humans share the same moral concepts. This last, difficult task is
exactly what I hope to prove Kant was doing.
The next sections of my paper, then, will show that Kant offers an idealist yet
objective moral theory. An idealist moral theory would be one that explains how
every moral characteristic of the world depends upon the human mind. These
moral characteristics come in three major types: the values of ends (value), the
rightness or wrongness of acts (right), and the related properties of moral agents(agency).27 By the value of endsI mean the worth accorded to various states of
affairs, objects, persons, etc., typically but not exclusively chosen as the ends of
moral acts. The states of affairs, objects, persons, etc., themselves would be char-
acteristics required by morality, not moral characteristics. By the rightness or
wrongness of actsI mean the correctness or incorrectness of the behavior of
moral agents according to some moral standard as well as the existence of that
moral standard itself. The behavior of moral agents would be a characteristic re-
quired by morality, not a moral characteristic. By the related properties of moral
agentsI mean the moral characteristics attributed to moral agents solely on the
basis of their performance of moral acts, such as accountable, evil, good, etc. The
moral agents themselves would be characteristics required by morality, not moral
characteristics. To illustrate these points: In a classical hedonistic utilitarian theory,pleasure alone has value; there is one rule of rightness for acts, namely to maxi-
mize pleasure; and agents are judged good or evil based upon their success (or
intended success) in performing those acts. For Kant, value lies in the ends cho-
25One might, of course, adopt a moral realism regarding some moral characteristics and a moral
idealism regarding others. I am not interested in such intermediary positions; rather, my goal is to
show that for Kant every moral characteristic is ideal.26John Locke, An Essay Concerning Human Understanding(1691), Peter H. Nidditch, ed. (Oxford:
Oxford University Press, 1975), IV.iii.11.27I am extremely grateful to Marcia Baron and Hugh Chandler for pointing out problems with
an earlier version of this division.
8/13/2019 Kant's Anti-realism
9/24
484 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
sen by human beings and in those human beings as ends-in-themselves, the right-
ness of acts depends upon their falling under a maxim that can pass the categori-
cal imperative test, and agents are judged morally worthy based upon their freely
chosen intentions in performing those acts. I will look at each of these in turn in
the next three sections. Later sections will consider two general objections to an
idealist reading of Kant concerning the possibility of other moralities and the
nature of objectivity, and discuss why the proper Kantian moral theory is idealist.
3 . I D E A L I S M O F V A L U E I N K A N T
The least controversial claim I am making is that Kants theory of value is idealist.
I will consider the two types of value that appear in Kant: the subjective value of
voluntarily chosen ends and the objective value of objectively required ends.
The value of voluntarily chosen ends is certainly idealist. These ends, such asthe adoption of the hobby of playing a musical instrument or the devotion to
learning about a particular subject or the commitment to a particular communitys
goals, are not forced upon a moral agent. Moral agents are allowed to choose
these ends limited only by the considerations that they not choose immoral ends
and that they choose some ends that develop their talents and aid others in pursu-
ing their ends. Voluntarily chosen ends correspond closely with imperfect duties
to improve ones talents and help others pursue their ends. The particular ends
chosen by any particular agent are voluntary. Their value is entirely dependent
upon the voluntary choice of the agent. Hence, their valuethe moral character-
istic at handis ideal rather than real.
More difficult to prove is that the value of objective ends not subject to the
choice of the agent is idealist. Among these ends are the value of rational beingslives and the respect of their power of choice. Kant described them in the Meta-
physics of Moralsas ends that are also duties(6:382f). One example would be
preserving ones life. Other objective ends are simply the value of humanity itself
as described in the second formulation of the categorical imperative and the hap-
piness of other human beings as described in the Metaphysics of Morals(6:3934).
Human beings must be treated as ends-in-themselves as exemplars of humanity
or rationality.
What is at issue here is not whether these objective ends are in fact ends com-
manded by the categorical imperative. There is no doubt that it is right to per-
form actions to bring about these ends and wrong to avoid these actions. The
adoption of these ends is a duty. What is at issue is whether the value, not the
rightness, of these ends itself is real or ideal. Obligatory ends can be said to derive
their value in one of two ways: either from the voluntary choice of agents or from
the moral law of rights and duties. The former way is clearly ideal, while the latter
way rests on the further question of the ideality of right.
The first option is as follows. The ends themselves are required by duty as
dictated by the categorical imperative. But they do not have any value qua re-
quirement of duty. They only receive value upon the actual volition of an agent to
pursue those ends. The agent, in a sense, is not free to deny the dutifulness of
those ends, but is free to choose not to value those ends. A moral agent cannot
determine that the end is not a requirement of morality to pursue, but can deter-
8/13/2019 Kant's Anti-realism
10/24
485K A N T S M O R A L A N T I - R E A L I S M
mine that she will not pursue it, or that she will pursue the end reluctantly and
thus not value it. For example, one may not deny that it is wrong to lie for ones
own advantage, but one may decide to either lie anyway or to tell the truth grudg-
ingly or for selfish reasons. The value of the end of telling the truth in this first
interpretation thus would come only from the agents voluntary choice. If this is
the proper interpretation of Kants doctrine of obligatory ends, then value is ide-
alist because value is derived from the choice of individuals.
The second interpretation is that the value of these obligatory ends is derived
from their dutifulness. Individual moral agents, of course, still have the choice of
whether to act on the end or not, but they must recognize the value of the end
simultaneous with recognition of the objective rightness of the end. Value is in-
separable from obligatory ends. In this case, the values status as real or ideal
depends upon the status of the obligatory end. The status of the obligatory ends
in turn depends upon the status of their source, which is the categorical impera-
tive itself. Thus the ideality of objective value in this second interpretation would
depend upon the ideality of the right.
I will not here decide which of these two interpretations best fits Kants claims
about ends which are also duties. But either alternative would accord with my
claim that Kant is idealist. Under the first interpretation, the value of ends that
are also duties stems from the voluntary choices of moral agents and is thus en-
tirely dependent upon their minds. Under the second interpretation, the value of
ends that are also duties is derived directly from the categorical imperative, Kants
criterion for moral right; when I show later that Kants theory of right is entirely
mind-dependent, then Kants theory of value will also be shown to be entirely
mind-dependent.This section has shown that Kants theory of value is idealist, at least as a theory
independent of a conception of the right. Value idealism, of course, ought to be
expected from Kant since he is traditionally seen as a deontological theorist who
declares the priorityand even autonomyof the right over any conception of
the good. A more challenging part of proving Kants moral idealism is the topic of
my next, much longer, section.
4 . I D E A L I S M O F R I G H T I N K A N T
The ideality of right is a more difficult matter than the ideality of value since, if
anything is to be considered real for Kantians, it is the categorical imperative.
This section will provide the foundation for my idealist analysis of Kants theory of
right.
To show that Kants theory of right is idealist, I will first discuss the nature of
the categorical imperative and its relation to the moral law in order to show that
Kant differentiates morality as humans experience it from a hypothesized real
morality. The difference between the moral law and the categorical imperative
will be shown to rest on the status of pure practical reason itself and its less pure
realization in human beings. Second, I will evaluate the status of pure practical
reason to reveal that its reality remains subject to some doubt. The grounds for
our belief in pure practical reason will be shown to be only the practical grounds
of our belief in the postulates of practical reason. Third, I will argue that the
8/13/2019 Kant's Anti-realism
11/24
486 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
categorical imperativethe basis of morality for existing humanstherefore is
not derived from the moral lawthe hypothesized real basis for moralitybut by
contrast is itself the foundation of our belief in an objective moral law. Thus,
moral right depends for Kant in its entirety upon actual human consciousness of
the categorical imperative, and is ideal not real.
4.1
To explain the difference between the categorical imperative and the moral law, I
will offer an interpretation of Kants doctrine of the fact of reason.Although
the fact of reason plays an extremely important role for Kant, he never clearly
defined it, forcing commentators to attempt to elucidate it in light of different,
even conflicting, references.28 I will argue here that the fact of reason consists of
the categorical imperative as our consciousness of the moral law. This definitioninvolves Kants distinction between the moral law as such and the categorical im-
perative.
Kant discusses this topic in the Critique of Practical Reason. The moral law holds
of purely rational beings who would use it as the sole principle of action because
they lack any other possible determinants for action. Purely rational beings have
no sensuous impulses, so would always act rationally. Purely rational beings would
conceive of the moral law not as a command but as a description of their prin-
ciple of action. We humans, however, can guide our actions not only using laws of
reason but also by incorporating sensuous needs and desires. Unlike purely ratio-
nal beings, we do not automatically guide our action according to the moral law.
We humans encounter the moral law in the guise of the categorical imperative
(5:32), a command to guide our behavior according to rational form alone: Kantuses the phrase Act only . . .when formulating the categorical imperative to
emphasize its status as a command.
This distinction between the moral law and the categorical imperative implies
that we humans experience the moral law as a categorical imperative: this is the
fact of reason.
Consciousness of this fundamental law may be called a fact of reason because one cannot
reason it out from antecedent data of reason, for example, from consciousness of freedom
(since this is not antecedently given to us) and because it instead forces itself upon us of
itself as a synthetic a priori proposition that is not based on any intuition, either pure or
empirical. (5:31)
The categorical imperative, as a command, presents itself to us as obligatory. We
experience the call of moral obligation in the guise of the categorical imperative.
Accordingly the moral law is for them [finite rational beings] an imperative that com-
mands categorically because the law is unconditional; the relationship of such a will [not
purely rational] to this law is dependence under the name of obligation, which signifies a
necessitation, though only by reason and its objective law, to an action. (5:32)
28See, for example, Lewis White Beck, A Commentary on Kants Critique of Practical Reason(Chi-
cago: University of Chicago Press, 1960), 166f, where he lists Kants possible references for the fact as
either consciousness of the moral law, the moral law itself, or autonomy. Henry Allison reviews Becks
discussion before concluding that the fact of reason is consciousness of standing under the moral law
(Allison, Kants Theory of Freedom[New York: Cambridge University Press, 1990], 2319, esp. 233).
8/13/2019 Kant's Anti-realism
12/24
487K A N T S M O R A L A N T I - R E A L I S M
As rational beings we experience the commanding force of the categorical im-
perative as, in a sense, a call of our rational nature to act according to reason. We
do not experience the moral law as such, that is, as a principle describing the
actions of purely rational beings. Rather, the moral law is for us an imperative
related to our nature as finite, sensuous rational beings.29
The distinction between the moral law and the categorical imperative suggests
a distinction in their sources. The moral law, on the one hand, directly requires
the existence of pure practical reason. The moral law is the principle describing
the behavior of perfectly rational beings; as such, it requires the concept of a
perfecta purely practicalreason in those beings. If the moral law is real, then
pure practical reason is real.
The categorical imperative, on the other hand, is a command experienced by
human beings, whose finite reason is not pure. Our experience of the categorical
imperative can cause us to ask about its ultimate source. One possible source is
pure practical reason; that is, the categorical imperative might be derived from
the moral law as a particular instantiation for finite sensuous and rational beings.
But it is also possible that the experience of the categorical imperative does not
have its source in pure practical reason; it might, for example, merely stem from
empirically conditioned reason (5:15). Because the categorical imperative does
not obviously stem from pure practical reason, Kant is impelled to investigate its
source. The Critique of Practical Reasonis the result. The status of the categorical
imperative will depend upon the status of its source, if any, discovered during the
investigation conducted in the Critique of Practical Reason and elsewhere.
4.2
This point leads to the second major topic of this section: the status of pure prac-
tical reason. Here I will argue that pure practical reason is not known to be real;
rather, its reality is merely adopted for practical purposes.
Consider first the nature of pure reason as such, rather than pure reason in its
practical aspect. Pure reason stands in contrast to empirically conditioned rea-
son. Kant mentions the existence of empirically conditioned reason in the solu-
tion to the Third Antinomy (A549/B577). Pure reason is there contrasted with
empirically conditioned reason as appearances are contrasted with things in them-
selves. Recall that things in themselves (or noumena in the positive sense) are
unknown and unknowable. Pure reason is thus unknown and unknowable.
It may be objected to this argument that when Kant holds that things in them-
selves are unknown and unknowable, he is referring only to their specific nature
not to their mere existence. Kant states this distinction many times (see, for ex-
ample, B306) but I would like to draw particular attention to the passage in the
Preface to the second edition of the Critique of Pure Reasonwhere he notes that
even if we cannot cognize these same objects as things in themselves, we at least must be
able to think them as things in themselves. For otherwise there would follow the absurd
proposition that there is an appearance without anything that appears. (Bxxvivii)
29This way of differentiating the moral law from the categorical imperative expands upon Beck,
op. cit., 121.
8/13/2019 Kant's Anti-realism
13/24
488 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
Since the existence of things in themselves must be admitted, the existence of
reason in itself must also be admitted, since there is an appearance of empirical
reason. And thus, the objection goes, pure reason must be admitted as real.
In response to this objection I would like to point out that all that Kant requires
us to admit is that for every particular empirical thing in appearance, there is some
(not necessarily unique) thing in itself which is its correlate. We can know no more
than its existence. Thus we can claim that for empirical reason, there is some thing
in itself which is its correlate, but we cannot claim to know anything about its nature.
If we used the term reason in itselfor pure reasonfor this thing in itself, we must
avoid reading any positive characterization into the term. This entails that we must
avoid characterizing any principles of pure reason or abilities of pure reason. We
ought even to be skeptical about the individuality of pure reason, because there could
be some single thing in itself which functions as correlate for all mental faculties.
There is a further argument that pure reason is not taken to be real in Kant. In
several places in the first Critique, but in particular in the Third Antinomy, Kant
argues that reason itself must be transcendentally free in order to guarantee the
validity of the moral commands of reason and other a priori propositions (Bxxviii
xxix, A5467/B5745, A803/B831). This transcendental freedom is in fact a cru-
cial part of his concept of pure as opposed to empirical reason. I treat this issue
elsewhere,30 but I will briefly note that I see Kant making the following argument:
In order for a proposition to be a priori valid, it must possess necessity. Any propo-
sition whose generation by a faculty such as reason or the understanding is de-
pendent on some independent cause and does not possess necessity because that
proposition is contingent upon the independent cause for its generation by that
faculty. Therefore, for the proposition to be necessary, the faculty that generatesthat proposition must be free of independent causes. But this is transcendental
freedom. Therefore, since reason generates necessarily valid a priori propositions,
it must be transcendentally free as pure reason.
But now one can see that the status of pure reason is dependent upon the
status of transcendental freedom. And Kant clearly argues that transcendental
freedom is not known to be real. At the very end of his discussion of the solution
to the Third Antinomy, Kant appends a paragraph warning the reader not to
overstate its conclusion (A5578/B5856). Freedom has not been proved actual,
nor even possible, he contends, but only not incompatible with nature.
There is one remaining large obstacle in accepting the arguments I have made
that pure reason is not known to be real. Kant admits that pure reason cannot be
known to be real using any theoretical approach, but its reality can be provedusing a practical approach. The opening paragraph of the Preface to the Critique
of Practical Reason is perhaps the most prominent place where Kant claims that
pure reason, when applied practically, proves its reality and that of its concepts
by what it does(5:3). If pure reasons reality is proved through moral practice,
then perhaps my claim that pure reason is not known to be real is superseded by
a practical awareness that pure reason is real.
30For a detailed defense of this claim and more explanation of the requirement that reason itself
be transcendentally free in order to ground morality, see my Pure Reason and the Moral Law: A
Source of Kants Critical Philosophy,History of Philosophy Quarterly13(1996): 25571.
8/13/2019 Kant's Anti-realism
14/24
489K A N T S M O R A L A N T I - R E A L I S M
But this is a case of putting the boat trailer before the sport utility vehicle (to
update an old clich). Pure reason proves its realityonly for practical purposes.
That is, the reality of pure reason here depends upon its validity in use. Kant does
try to show that moral practice requires an assumption of pure practical reason,
but this conclusion is valid only for moral practice itself.
Consider the postulates of practical reason in the Critique of Practical Reason:
for moral purposes we must believe not only that immortality is real and that God
exists, but also that we are free in the positive sense of the capacity to determine
ones will by the law of an intelligible world(5:132). This positive freedom is
identified as pertaining to the solution to the Third Antinomy (5:133). Kant then
reiterates that this freedom, which must be identified with the freedom of the
faculty of pure practical reason, can be assumed only for practical purposes:
but how freedom is even possible and how this kind of causality has to be representedtheoretically and positively is not thereby seen; that there is such a causality is only postu-
lated by the moral law and for the sake of it. (5:133)
Since freedom is only a postulate, the existence and nature of pure practical rea-
son is also only a postulate. Here the distinction between the moral law as gener-
ated by a pure practical reason and the categorical imperative as our empirical
consciousness of that law helps to make Kants position clear. Our experience of
the categorical imperative begins a process of reflection on the conditions of that
experience. This reflection shows that the validity of this categorical imperative
can be explained by postulating a moral law in a pure practical reason as well as
the three postulates required further to make that moral law coherent. These
postulates thus depend upon the moral law, which in turn depends upon the
practical experience of the categorical imperative. So pure practical reason is to
be treated not as real but only as a postulate which functions to make our moral
lives coherent.
4.3
Since pure practical reason possesses only a postulated status, does this mean that
the categorical imperative is likewise denigrated in its authority? There are two
main reasons to hold that the categorical imperative is not. First, Kant at least
sometimes holds that the categorical imperatives commanding nature trumps
any skeptical challenge in practice, so that the categorical imperative becomes
independent in practice of the existence of pure practical reason. Second, the
categorical imperative is not only not derived from the moral law and pure prac-tical reason but in contrast is itself the foundation of our belief in an objective
moral law in pure practical reason, so that the existence of the latter depends
upon the former.
The first reason I mentionedthat the categorical imperative trumps skepti-
cal challenges in practicedeserves more space than I will give it here;31 here I
will only sketch the idea. Kant often mentions the experience of the categorical
imperative as involving undeniable moral obligation, particularly when giving
31See my Kants Two Priorities of Practical Reason,British Journal for the History of Philosophy6
(1998): 397419.
8/13/2019 Kant's Anti-realism
15/24
490 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
examples of the experience of moral action (for example, 5:30, A475/B503, A803/
B831). The moral agent, when faced with a situation that demands action, is sim-
ply aware of the call of the categorical imperative. The agent at that time and qua
agent cannot doubt the validity of the imperative since it is experienced as an
undeniable command. The possibility for doubting the validity of the categorical
imperative comes only when the moral agent is not faced with a situation that
demands action, and thus when the moral agent in not directly involved in moral
deliberation. At this time, the agent engages in theoretical reflection about moral
deliberation itself, and the experience of the categorical imperative is subject to
philosophical examination. Since it is only in theoretical reflection that the very
idea of pure practical reason arises, in practice the categorical imperative is inde-
pendent of pure practical reason.
I will now turn to the second reason and argue that the difference in certainty
between the categorical imperative and the moral law means that we base our
belief in the moral law solely on the existence of the categorical imperative. The
moral law cannot have any stronger basis for belief. We derive our claim to the
existence of the moral law, and consequently of pure practical reason, from a
need to believe in the validity of the categorical imperative. In short, we create the
conception of pure practical reason to satisfy our idiosyncratic moral experience
of the categorical imperative.
To prove this claim, I will suggest that Kants conception of the relation be-
tween freedom and the moral law in the Critique of Practical Reasonoffers us a way
of understanding how the idea of pure practical reason is dependent upon the
actuality of the categorical imperative. Freedom is, of course, seen by Kant as a
necessary condition of the moral law (5:289). But on the other hand, the morallaws existence apparently depends upon the actuality of freedom (5:29). Rather
than accepting a simple vicious circle of mutual derivation, Kant insists that our
cognition of the unconditionally practical [starts with] the moral law, of which we become
immediately conscious (as soon as we draw up maxims of the will for ourselves) . . . inas-
much as reason presents it as a determining ground not to be outweighed by any sensible
conditions and indeed quite independent of them. (5:2930)
As I have shown above when discussing the fact of reason, our consciousness of
the moral law just is the categorical imperative, and Kant plainly intends that
meaning here when he discusses reason presenting itself as a determining ground
for maxims. Our knowledge of freedom, then, is dependent on our experience of
the categorical imperative.
What is meant here by freedom? In this particular passage, of course, Kant
specifically mentions freedom of the will (Wille). But on the next page, Kant con-
cludes that pure practical reason, in giving this moral law to humans, is here
immediately lawgiving(5:31), that is, pure practical reason generates the moral
law. And, he notes two pages later, this lawgiving of its own on the part of pure
and, as such, practical reason, is freedom in the positive sense(5:33). The legis-
lating power of pure practical reason in generating the moral law requires the
freedom of the faculty of pure practical reason (see also 5:3). Freedom is thus
inextricably linked to the creation of the moral law by pure practical reason. Our
experience of the categorical imperative thus leads us to conclude that we are
8/13/2019 Kant's Anti-realism
16/24
491K A N T S M O R A L A N T I - R E A L I S M
free both in having free wills for choosing independently of inclination and in
having a free faculty of pure practical reason to create the moral law. Our postula-
tion of the moral law as stemming from pure practical reason is thus dependent
upon our experience of the categorical imperative.
Other texts also support this view. In the Canon of Pure ReasonKant argues
that our experience of the necessity of the categorical imperative justifies us in
claiming that God exists, but does not justify us in then deriving the categorical
imperative itself from that concept of God. His reasoning applies just as much to
the question of the existence of pure practical reason itself: For it was these laws
alone whose inner practical necessity led us to the presupposition of a self-suffi-
cient cause or a wise world-regent, in order to give effect to these laws, and hence
we cannot in turn regard these as contingent and derived from a mere will, espe-
cially from a will of which we would have had no concept at all had we not formed
it in accordance with those laws(A8189/B8467). While it is true that in this
passage Kant is arguing specifically against a divine command theory of morality,
one of his reasons applies as well to the question of the relation between the
categorical imperative and pure practical reason. The inner practical necessity
of the categorical imperative leads us to the presuppositionof a pure practical
reason. But then pure practical reason is a faculty of which we would have had no
concept at all had we not formed it in accordance withthe categorical impera-
tive. So although the concept of a pure practical reason is required for a thor-
ough understanding of the validity of the categorical imperative, the latter ought
not be seen as derived from the former, at least in the sense of ontological priority.
One can compare this with the hypothetical use of reason in the Transcen-
dental Dialectic.Reason, in attempting to find the unconditioned and thus tofulfill its needs, postulates broad major premises from which it can then derive
the specific conclusions about propositions it garners from experience (A330/
B386f). While in one sense, of course, the empirical propositions are derived
from the major premises created by reason, in another sense, the major premises
created by reason are themselves dependent upon the empirical proposition, for
without that empirical proposition, reason would not attempt to find its major
premise in a prosyllogism. Exactly in this second sense, the idea of pure practical
reason is dependent upon the categorical imperative, although reason postulates
that pure practical reason exists in order to derive the categorical imperative from
it in the first sense.
It is important that this conclusion not be misconstrued as merely epistemo-
logical. If the dependence of pure practical reason on our experience of the cat-egorical imperative is merely that of a basis for our knowledge rather than a
basis for actual existence, then pure practical reason could be concluded really to
exist as a real basis of our categorical imperative. Kant explicitly denies this alter-
native when discussing the postulates of practical reason (5:132f). He there ar-
gues that practical postulates are not equivalent to theoretical knowledge of the
existence of those objects (God, free will, immortal soul). Rather the concepts are
immanent . . . for practical purposesand as concepts have meaning only in
reference to the moral law(5:133). Regarding God in particular, Kant remarks
in the Metaphysics of Moralsthat any conception of God in morality expresses only
8/13/2019 Kant's Anti-realism
17/24
492 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
the relation of reason to the idea of God which reason makes for itself, and this
does not yet make a duty of religion into a duty to God, as a being existing outside
our idea, since we still abstract from his existence(6:487). Just as the practical
postulate of God abstracts from the existence of God and concerns only the idea
of God in relation to reason, the practical postulation of pure practical reason
merely satisfies reasons need for an explanation of the categorical imperative. It
does not concern epistemological access to anything real but involves only con-
ceptual clarity for reason itself.
Allow me to tie these arguments together. The categorical imperative is dis-
tinct from the moral law in that the former is our experience of the command of
our reason to act only according to certain types of maxims. The moral law itself
is a principle of purely rational action which is derived from the concept of a pure
reason unaffected by inclination. But the existence of pure practical reason is
subject to some doubt, primarily because it requires transcendental freedom that
is subject to some doubt, but also because we do not directly experience pure
reason. The categorical imperative, however, is not subject to any doubt. Based
upon our actual experience of the categorical imperative, we hypothesize the
existence of the moral law in a pure practical reason. But the moral law and pure
practical reason, then, are not independently real but depend for their ontologi-
cal statusa status akin to postulatesupon the categorical imperative.
Moral rightness can now be seen as ideal not real. Moral rightness is a function
of the categorical imperative. Maxims are right to the extent that they conform to
the categorical imperative. But the categorical imperative exists only in human
minds. Any hypothesized existence independent of human minds, particularly
the concept of a pure moral law in a pure practical reason, is itself only a postulatecreated by the human mind for purposes of understanding their experience of
the categorical imperative. Nothing else is required for rightness that is not also
dependent upon the categorical imperative. Therefore, moral rightness is ideal
not real.32
5 . I D E A L I S M O F A G E N C Y I N K A N T
Since I have shown that value and right are ideal in Kants theory, only one type of
moral characteristic remains: agency. Under this category lies moral accountabil-
ity, the moral aspects of freedom, the attribution of good or evil to persons, inten-
tion, etc. I gather these together under the heading agencybecause they all in
some way concern characteristics of moral agents.
32Although Allen Wood offers an interpretation of Kant that insists upon a distinct conception of
reason independent of particular beings, one that he calls realist, there is a suggestion from him that
he would not disagree with the reasoning in this section. In a footnote, he admits that Kants argu-
ment for the basis of morality might not satisfy some extremeskeptics today because, among other
reservations, it draws no distinction between our taking ourselves (from a practical standpoint) to be
responding to moral requirements that are unconditionally obligatory and the actual existence of
such categorical requirements(Wood, 1999, op. cit., 381, n. 30). This claim does not appear compat-
ible with his earlier claim that to ground the moral law on the idea of the will is therefore to distin-
guish moral truth from what any finite rational being (or all such beings) might believe(ibid., 157).
In interpreting Woods position I am taking the latter point to be his considered view since it accords
with the definition of realism he cites.
8/13/2019 Kant's Anti-realism
18/24
493K A N T S M O R A L A N T I - R E A L I S M
Here one must recall the distinction between moral characteristics and char-
acteristics required for morality that I made above. The former are properties
that are solely moral while the latter consists of properties of objects that are not
solely moral. To be solely moral is to have no causal properties except in moral
situations. Many characteristics of moral agents that one might think are moral
characteristics, such as deliberative ability, are instead only characteristics required
for morality because they can possess causal properties even in non-moral situa-
tions. The deliberation that precedes the choice of an action can be used in many
non-moral situations such as scientific investigation, culinary experiences, friend-
ships, and the like. Any related moral characteristics are a subset of the totality of
characteristics associated with these aspects of agency. All moral beings must have
deliberative ability, but not all aspects of deliberation are purely moral character-
istics.
The moral characteristics that must be discussed, then, are specifically moral
aspects of agency. There are several categories of these characteristics: (a) those
concerning an agents ability to make moral decisions and perform right acts,
particularly freedom of choice and development of intentions when performing
acts; (b) those concerning changes to, or the nature of, an agents character, par-
ticularly being good or evil; and (c) those concerning other moral beingsrela-
tionship to the moral agent, particularly attributions of responsibility.
The first type of moral characteristic regarding agency concerns the agents
ability to make the correct moral decisions and act rightly. Kant is well known for
arguing that all and only actions determined by the categorical imperative are
free (4:447; 5:289), so freedom of the will must be considered under this rubric.
Freedom of the will is, for Kant, the freedom to act from the motive of duty ratherthan from the incentives of inclination. Freedom thus encompasses correct moral
decision making and right action to the extent that they are moral characteristics
(deliberation and acting themselves would be characteristics required for moral-
ity). If freedom can be shown ideal, then this type of moral characteristic is ideal.
Given the discussion of the role of freedom in the conception of pure practical
reason above, it is easy to see how freedom of the will is ideal. Freedom is one of
the three postulates Kant identifies in the Critique of Practical Reason(5:132). The
postulates are practical presuppositions made by reason in order to defend the
coherence of the categorical imperative.
The second type of characteristic under consideration here is the agents be-
ing good or evil. Here, it is also relatively easy to see how it is ideal. An agents
being good or evil is distinct from the agent merely performing right or wrongactions; what is required is that the person be motivated to action by a particular
sort of motive. In the Critique of Practical Reason, Kant explains that good or evil
always signifies a reference to the will insofar as it is determined by the law of
reason to make something its objectand refers only to the way of acting, [to]
the maxim of the will, and consequently [to] the acting person himself as a good
or evil human being(5:60). This explanation of good and evil points the way to
two reasons for holding it to be ideal. First, an agents motivation can be deter-
mined by the law of reason only if the agent is free in deciding upon that motive,
and freedom has already been shown to be ideal. Second, an agents maxim can
8/13/2019 Kant's Anti-realism
19/24
494 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
be appraised as permissible or impermissible (certainly part of the determination
of its moral worth) only if the maxim accords with the categorical imperative, and
the categorical imperative has been shown to be ideal.
The third and final type of characteristic related to agency is moral responsi-
bility. Once again, it can be shown that this type of characteristic relies on others
parts of Kants theory already shown to be ideal. In the solution to the Third
Antinomy, Kant links blame of a moral agent to the transcendentally free faculty
of reason: This blame is grounded on the law of reason, which regards reason as
a cause that, regardless of all the empirical conditions just named, could have and
ought to have determined the conduct of the person to be other than it is(A555/
B583). This pure practical reason has already been shown ideal. In the Metaphys-
ics of MoralsKant notes that a deed can be imputed to an agent under two related
conditions: the deed must fall under obligatory laws, and the deed must relate to
the agents freedom of choice (6:223). Both obligatory laws, as the categorical
imperative, and freedom of choice, as the freedom of the will, have been shown to
be ideal.
Thus, all aspects of moral agency that are moral characteristics are also ideal
according to Kant. And since I have already shown in my previous sections that
moral value and moral right are ideal, I conclude that Kants ethics is idealist, or
anti-realist.
6 . T H E P O S S I B I L I T Y O F O T H E R M O R A L I T I E S
If my arguments above are correct, then human morality is limited to beings like
us, that is, beings who share our psychological constitution to the extent that they
experience the categorical imperative. Must it then be possible for other types ofbeings to have moralities different from the human? In other words, would Kant
be committed to moral relativism? This section will show that such a result is not
required for the idealist conclusion above.
The question depends in part on whether the categorical imperativethe
source of all aspects of morality in Kantis intrinsically tied (from broader to
narrower scope) to the possession of consciousness, in which case all conscious
beings must share human morality; or to consciousness along with ability to delib-
erate, in which case all deliberative beings must share human morality; or to con-
sciousness along with ability to deliberate freely, in which case all free deliberative
beings must share human morality; or consciousness along with an ability to de-
liberate freely and rationally, in which case all rational deliberative beings must
share human morality. Kant at a minimum holds to the last of these options since
he asserts that all rational beings are equally obligated to follow the categorical
imperative (for example, 4:389). His argument in the Critique of Practical Reason
that all beings with free wills are subject to the categorical imperative appears to
extend our morality to all beings with any ability to deliberate freely (5:29). Fur-
ther, Kant argues that morality is not possible without freedom, so he appears to
deny that any other type of beingmerely conscious and perhaps deliberative
but not freely deliberativecan have any morality at all.
Given these positions of Kants, it would appear incompatible for him to allow
that there might be other possible moralities. Since in this paper I am only inter-
8/13/2019 Kant's Anti-realism
20/24
495K A N T S M O R A L A N T I - R E A L I S M
preting Kant, I will not attempt to assess the validity of the arguments he gives for
those positions. I will argue, however, that even with the claim that human moral-
ity is the only possible morality, Kant still remains an idealist.
The key to my argument is that the universality of morality among free delib-
erative rational beings does not alter the ontological status of that morality. In
each one of those rational beings, morality rests on the experience of the cat-
egorical imperative. Whether there is only one such being or uncountable tril-
lions, in each of them, morality will rest on each ones experience of the categori-
cal imperative. Morality will still be ideal. A thought is a thought no matter how
many think it.
The argument is not original to me. Kant himself makes a similar point regard-
ing the ideality of space and time:
It is also not necessary for us to limit the kind of intuition in space and time to the sensibil-ity of human beings; it may well be that all finite thinking beings must necessarily agree
with human beings in this regard (though we cannot decide this), yet even given such
universal validity this kind of intuition would not cease to be sensibility, for the very reason
that it is derived (intuitus derivatuvus), not original (intuitus originarius), thus not intellec-
tual intuition. (B72)
Space and time are considered ideal even were it necessary for all finite thinking
beings to share our intuition.33
It is still possible that Kants arguments restricting morality to free rational
deliberative beings like us is faulty. In that case, other moralities are clearly pos-
sible, and the idealism of human morality (shared among all beings like us) would
only be strengthened. I have given a reason for thinking that regardless of whether
or not there are other possible moralities, Kants ethics is still ideal.
7 . O B J E C T I V I T Y , T R U T H , A N D V A L I D I T Y
All of the arguments I have presented indicate that Kant ought to be considered
a moral idealist. In this section I wish to reaffirm that none of those arguments
detracts from the objectivity or validity of morality in a Kantian theory. The claim
that morality cannot be true if there is nothing real to which it is known to corre-
spond will be shown to misunderstand the nature of moral validity for Kant.
Two points need to be made. First, objectivity for Kant stems from sharing a
common cognitive framework. Second, moral validity is derived from the com-
manding nature of the categorical imperative as a part of that framework.
First, then, objectivity must be seen in Kantian terms not as correspondence
with some pre-existing facts but as a result of a shared cognitive framework. Ob-
jectivity is thus limited to the possible experiences we could have in which we
could apply that cognitive framework. It is uncontroversial that Kant holds to the
objectivity of our forms of intuition of space and time and the categories of the
understanding as transformed into principles capable of being applied to our
spatio-temporal experience. I have shown with sufficient support above that our
33Allison (op. cit., 2345) notes the correspondence between space and time on the one hand
and the moral law on the other hand. But Allison does not press the analogy to its logical conclusion
that morality is transcendentally ideal.
8/13/2019 Kant's Anti-realism
21/24
496 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
moral experience is analogously dependent upon the faculty of reasonat least
empirically conditioned reason that contains the categorical imperativeand so
deserves an analogous conception of objectivity. Moral principles are objective in
that they are valid for all human beings in the totality of their experience. Al-
though they are not known to be valid beyond experience, their objectivity is not
lessened.
Moral validity, of course, seems to require more than just the assumed truth of
moral principles; it also requires a corresponding obligation to act on those prin-
ciples. For Kant, moral obligation stems from the force of the categorical impera-
tive. Practical reason is binding upon us as a factthe fact of reason, the categori-
cal imperative as a command. Kant does not pursue obligation beyond the claim
that we humans experience the categorical imperative as commanding obedi-
ence. He defines obligation as the necessity of a free action under a categorical
imperative(6:222). As I have interpreted Kant above, the practical necessity of
actions under the categorical imperative remains the core experience of morality.
Obligation is not lessened by treating morality as ideal rather than real.
Neither objectivity nor obligation are threatened by treating Kant as a moral
idealist. Since my last section also showed that Kantians need not worry about the
possibility of other moralities, I take myself to have defended the idealist interpre-
tation of Kants ethics from these two general objections.
8 . A U T O N O M YA N D K A N T I A N M O R A L I D E A L I S MO R, W H YA L L K A N T I A N S S H O U L D B E M O R A L I D E A L I S T S
The conception of moral idealism I have argued for above is not only the concep-
tion held by Kant but also the most appropriate for any Kantian. Kants notion ofautonomy dictates this result. In this brief section I can only sketch my argument.
Kants notion of autonomy distinguishes Kants cognitivism from other
cognitivist moral theories. A cognitivist holds that moral statements or proposi-
tions can be true or false. Their truth or falsity depends, in all the non-Kantian
cases, on there being some basis for truth external to human beingsown beliefs.
That fact of the matter might be real moral properties of persons and objects in
the world (value realism) or real facts of propositions in an objective, transcen-
dent God or a rationalist reason (rightness realism). In either case, those external
bases for moral propositions ground their truth and falsity independent of any
human beings actual knowledge of their truth or falsity. It might be the case that
humans are wrong about every moral belief they hold, since the truth of moral
propositions is independent of human knowledge.
Kants conception of autonomy precludes such a conception of morality. Hu-
man beings cannot be dependent upon anything distinct from their will for the
moral law that binds them (4:440). Anything independent of humanity is inde-
pendent of the will. Therefore, nothing independent of humanity can ground
morality. This is essential to Kants conception of autonomy.
This conclusion does not entail that morality cannot extend past the set of
human beings. Kant, of course, holds that the categorical imperative is valid for
all rational beings, not just human beings. But this can only mean that the cat-
egorical imperative is valid for all beings relevantly similar to human beings. They
8/13/2019 Kant's Anti-realism
22/24
497K A N T S M O R A L A N T I - R E A L I S M
share with humans the same basis for determining moral validity. It cannot mean
that all rational beings depend upon something furthertranscendent rational-
ity itselfto ground morality.
If the distinctive feature of Kants moral theory is autonomy and avoidance of
heteronomy, and if autonomy requires the dependence of moral principles upon
the human will, and if this dependence on the human will is idealist, then the
distinctive feature of Kants moral theory is its idealism. Kantians ought to em-
brace moral idealism as the distinctive feature of Kants moral theory. Use of the
categorical imperative, universalizability, human dignity, etc., are all undoubtedly
important elements of Kants and Kantian theory, but the distinctive feature of
Kants theory is his idealism based on autonomy.
9 . R E C O N S I D E R I N G D E F I N I T I O N S O F M O R A L R E A L I S M
I have shown that Kants moral theory is, on my definition, an idealist or anti-
realist theory. An obvious objection to my argument is that I have begged the
question by defining moral idealism in such a way that this result obtains. The
objection would claim that under other, better, definitions of moral realism, Kant
turns out to be a realist. In this section I will accept other definitions for the sake
of argument and show how even under those definitions, Kant is a moral anti-
realist. This section will defend my claim that Kant was a moral idealist given the
definitions of realism I considered in my first section.
First, there are the similar definitions of Sayre-McCord and Railton. Sayre-
McCord holds that realism involves embracing just two theses: (1) the claims in
question, when literally construed, are literally true or false (cognitivism), and
(2) some are literally true.Railton holds that The realist about X gives an ac-count of X discourse that is cognitive, non-error-theoretic, literal, and opinion-
independent (objective, in our sense).The crucial term in both of these defini-
tions is literal.34 What does it mean for a moral claim to be literallyconstrued?
Certainly the term at least connotes that actions are themselves right or wrong
and ends are themselves of value and not merely that I am constrained to judge
acts as right or wrong and constrained to value certain ends. When a moral agent
claims It was wrong for Jones to kill Smith since Smith possessed human dignity,
the moral agent would not be construing her terms literally if she meant I am so
constituted that I judge Jonesmurder of Smith to be wrong, and I am so consti-
tuted that I judge Smith to have inherent worth.The latter, however, is the mean-
ing of the terms as I have shown Kant to use them. I have shown that for Kant,
moral claims do not ultimately refer to the act itself, the end itself, or agency
itself; rather, they ultimately refer to the constitution of human rationality as the
source of judgments and beliefs. In contrast, any literal construal of moral claims
would have to refer to acts, ends, and agents themselves. If literal interpretation
of terms is a requirement of moral realism, then Kant is a moral anti-realist.
34Popular use of the term literallyhas reversed its meaning so that it can now denote meta-
phorically.The basketball announcer who says of one player that he literally exploded on the court
uses it in this new sense. I mean the term literallyliterally. I would like to thank Grant Sterling for this
example.
8/13/2019 Kant's Anti-realism
23/24
498 J O U R N A L O F T H E H I S T O R Y O F P H I L O S O P H Y 40 :4 O C T O B E R 2002
Second, Railtons definition includes the further claim that the truth of moral
claims must be opinion independent. Boyds definition includes the similar claim
that the truth or falsity or moral claims is independent of our moral opinions,
theories, etc.Wood, following Boyd, suggested that realism requires that moral
principles be independent of what anyone or everyone thinks. Presumably, by
invoking opinion, Railton and Boyd are excluding from moral realism theories
that depend upon something subjective such as individual or group belief. I have
shown that for Kant, moral claims are in fact dependent on human beliefs that
arise from the structure of specifically human rationality. All aspects of Kantian
morality ultimately rest on our awareness of the categorical imperative. Thus all
aspects of Kantian morality except the categorical imperative itself are clearly anti-
realist, since they are not themselves independent of the categorical imperative
and have no other basis. But is the categorical imperative itself opinion or theory
independent? If so, then Kant would by my criteria be a moral realist.
There are two reasons for holding that the categorical imperative is not inde-
pendent of opinion, theory, or etc.First, the categorical imperative depends
upon the particular nature of human beings. Second, the categorical imperative
is for Kant a subject of belief not knowledge.
The first response notes that the categorical imperative depends upon a par-
ticular mental constitution shared by human beings and perhaps other creatures.
Recall that even the nature of purely rational beings is, for Kant, a concept result-
ing from human reasons need to explain and understand the experience of the
categorical imperative. Kant makes no claim that such purely rational beings ac-
tually exist. When Kant claims that other rational beings would also be subject to
morality, he can thus have in mind only other beings who share the particularmental constitution of human beings, that is, limited rationality. Since the cat-
egorical imperative depends upon that particular mental constitution, Kant can-
not be claiming that the categorical imperative is independent of that particular
mental constitution. While one might not label this mental constitution an opin-
ion or a theory, it is certainly something that falls under the etc.of Boyds defi-
nition. If humans did not have this particular mental constitution, perhaps spe-
cifically only the particular module correlated with the categorical imperative,
then there would be no categorical imperative. Its validity is thus not indepen-
dent of human opinion, theory, etc.
This first response receives further support in relation to the second response,
that is, noting that Kant places reasons claims under belief not knowledge. The
Canon of Pure Reasonof the Critique of Pure Reasoncontrasts opining, believing,and knowing as three stages of conviction (A82031/B84859). Believing occurs
when there is insufficient objective ground for a claim to knowledge, yet suffi-
cient subjective ground to go beyond mere opinion. Specifically, Kant identifies
moral belief with belief in God and a future world (A828/B856). These, he says,
are beliefs that do not enjoy objective grounds (by objectiveKant means related
to an object) but are sufficiently tied to moral principles that every moral actor
must hold them. The subjective basis is their tie to moral principlesand moral
dispositions.Kant takes great pains in this passage to clarify the status of moral
belief as based in the subject with no independent grounds for the belief: regard-
8/13/2019 Kant's Anti-realism
24/24