Вы находитесь на странице: 1из 221

Freedom and Anthropology in Kant’s Moral Philosophy

This book is the first comprehensive account of Kant’s theory of free-

dom and his moral anthropology. The point of departure is the appar-
ent conflict between three claims to which Kant is committed: that
human beings are transcendentally free, that moral anthropology
studies the empirical influences on human beings, and that anthro-
pology is morally relevant. Frierson shows why this conflict is only
apparent. He draws on Kant’s transcendental idealism and his theory
of the will and describes how empirical influences can affect the em-
pirical expression of one’s will in a way that is morally significant but
still consistent with Kant’s concept of freedom.
As the first work on Kant to integrate his anthropology with his
philosophy as a whole, this book will be an unusually important source
of study for all Kant scholars and advanced students of Kant.

Patrick R. Frierson is an assistant professor of philosophy at Whitman

College, Washington.
Freedom and Anthropology
in Kant’s Moral Philosophy

Whitman College
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge cb2 2ru, UK
Published in the United States of America by Cambridge University Press, New York
Information on this title: www.cambridge.org/9780521824002

© Patrick R. Frierson 2003

This publication is in copyright. Subject to statutory exception and to the provision of

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2003

isbn-13 978-0-511-17009-6 eBook (MyiLibrary)

isbn-10 0-511-17009-2 eBook (MyiLibrary)

isbn-13 978-0-521-82400-2 hardback

isbn-10 0-521-82400-1 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
for J.C.

Preface page ix

Introduction: A Problem with Kant’s Moral Anthropology 1

i. the problem
1 The Asymmetry in Kant’s Conception of Freedom 13
2 Anthropology as an Empirical Science 31
3 The Moral Importance of Kant’s “Pragmatic” Anthropology 48
4 Moral Anthropology in Contemporary Neokantian Ethics 68

ii. the solution

5 Transcendental Idealism, Radical Evil, and Moral
Anthropology 95
6 Moral Influence on Others 136
Epilogue: Incorporating Moral Anthropology and
Defending Kantian Moral Philosophy 163

Notes 167
References 200
Index of Kant’s Works 205
Name Index 207
Subject Index 209


From my earliest exposure to Kant’s moral theory, I was drawn to his

emphasis on the centrality of freedom but bothered by the apparent ab-
stractness of the moral law. Thus when I first approached the Anthropology
many years ago, I did so with excitement. I hoped that Kant would in-
corporate all the rich details of human life that I found lacking in his
Grounding, and that he would show how these details fit with the dis-
tinctive focus on freedom that attracted me to his moral theory. At first,
the Anthropology seemed more amusing than philosophically satisfying.
However, as I came to appreciate the details of Kant’s Anthropology and
as more neokantians incorporated anthropological insights into moral
theory, I saw that Kantians could provide as rich and concrete a moral
theory as anyone.
During graduate school, especially as a result of interaction with neo-
Aristotelian ethical theories and more recent neokantian accounts, I be-
came interested in Kant’s treatment of the cultivation of moral character.
In parts of Kant’s anthropology he seemed to suggest that there could be
empirical influences on moral development, and these showed that Kant
could provide the sort of nuanced theory of human nature that often
made Aristotle attractive. But I was still puzzled about the compatibility
of these new (for me) aspects of Kant’s account with the treatment of
freedom that initially attracted me to Kant.
This puzzle led me to write my dissertation on Schleiermacher’s cri-
tique of Kant’s Anthropology, in which Schleiermacher argues that tran-
scendental freedom is incompatible with a robust anthropology. And
Schleiermacher’s critique finally led to this book, which is my attempt to
show how Kant’s moral theory can incorporate the anthropology that I
x Preface

have grown to appreciate into the theory of freedom that first drew me
to Kant.

Without the generous and perceptive criticism of Karl Ameriks, this book
would not exist. As my advisor at the University of Notre Dame, he pushed
me to refine and expand ideas that eventually made their way into this
book. Since that time, he has continued to offer suggestions for which I
am extremely grateful.
G. Felicitas Munzel, Philip Quinn, and Robert Solomon also read var-
ious drafts and provided extensive and invaluable comments. Robert
Louden was extremely generous, not only to give me an early draft of
his book, but to discuss my comments on his work in detail at an earlier
stage of this project. Natalie Brender also gave very helpful comments
on an early draft of this book. I also thank Eric Newman and my anony-
mous reviewers at Cambridge University Press. Travis Exstrom at Whitman
College provided help with the index of Kant’s works and the name index.
Samuel Fleischacker deserves a special thanks for introducing me first
to philosophy, and then to Kant, while I was an undergraduate at Williams
Finally, this work is built on the support and sacrifice of family. I am
grateful to my parents for their encouragement and patience. And I thank
my wife, Katheryn, whose calls for clarity and relevance in my work cer-
tainly improved this book and made me both a better philosopher and a
better person.

A Problem with Kant’s Moral Anthropology

1. Kant’s Anthropology and Schleiermacher’s Objection

In 1798, Immanuel Kant published his Anthropology from a Pragmatic Point
of View. In this work, he discusses “what man makes, can, or should make
of himself” (7:119). The book offers detailed, even if incomplete, ac-
counts of human capacities and character, and these accounts help flesh
out Kant’s Critical philosophy with empirical information about human
beings. This 1798 Anthropology was not Kant’s first foray into anthropology.
Starting in 1772, Kant offered yearly lectures on anthropology that paral-
lel the published work. In addition, anthropological insights are scattered
throughout Kant’s other publications. The essays on history (primarily
from 1784 to 1786), the third Critique (1790), Religion within the Bound-
aries of Mere Reason (1793), and the Metaphysics of Morals (1797) all present
various aspects of Kant’s anthropology. But the Anthropology of 1798 is the
most detailed, systematic, and public treatment of anthropological issues
in Kant’s corpus.
Within a year and a half of its publication, at least eleven reviews of
Kant’s Anthropology appeared.1 Among these was an important review
by a young Friedrich Schleiermacher, published in the Romantic jour-
nal Athaeneaum. In his review, Schleiermacher sarcastically suggests that
Kant’s Anthropology must have been intended as a “negation of all an-
thropology” (Schleiermacher 1984: v.1, p. 366, cf. Schleiermacher 1998)
because it blatantly conflicts with the rest of Kant’s philosophy. The re-
view criticizes the Anthropology for disorganization and triviality and ac-
cuses Kant of failing to combine systematicity and popularity in it.2 But
the most philosophically important objection comes in the form of a

2 Introduction

challenge to Kant to choose between his anthropology and his theory of

freedom, insisting that Kant cannot have both.
The conflict is between two claims, that “nature is choice” and that
“choice is nature” (Schleiermacher 1984: v.1, p. 366). Kant’s anthropol-
ogy, according to Schleiermacher, must affirm that “choice is nature.”
That is, human beings and all their choices must be considered objects in
nature if they are to be studied by anthropology. But Kant must also affirm
that “nature is choice” – that is, that an individual’s human nature is due
to that individual’s choice. The “nature” that anthropology studies cannot
be merely the result of natural causes. Schleiermacher gives two reasons
for insisting that Kant must reconcile his anthropology with this strong
claim about freedom. First, he points out that the conception of freedom
developed in Kant’s Critical philosophy commits Kant to the view that
freedom grounds human choices, and thereby human “nature.” Second,
he suggests that any anthropology must have some account of choice un-
derlying nature to make sense of the epistemic norms implicit in scientific
Kant could have offered an anthropology that would not conflict with
his Critical philosophy. The problem is the particular sort of anthropology
he presents, one that is both empirical and morally relevant. The dilemma
can be stated in terms of a conflict among three claims to which Kant
seems committed:

r Human beings are transcendentally free, in the sense that empirical

influences can have no effect on the moral status of a human be-
ing and in the sense that choice is fundamentally prior to natural
r Moral anthropology is an empirical science that studies empirical in-
fluences on human beings.
r Moral anthropology is morally relevant, in that it describes influences
on moral development.

Schleiermacher suggests that Kant is committed to all three of these

claims and that the claims are inconsistent with one another. Any two
of them could be held consistently, but all three cannot. Unless Kant is
willing to sacrifice the conception of freedom on which his moral philos-
ophy depends, his Anthropology can be nothing more than a “negation of
all anthropology.”4
A Problem with Kant’s Moral Anthropology 3

2. The Practical Problem of Moral Anthropology

At this point, it is important to distinguish Schleiermacher’s objection
to Kant’s anthropology from a familiar objection to Kant’s metaphysics.
From Kant’s earliest critics, such as Rehberg, Fichte, and Hegel,5 to more
recent commentators, Kant’s account of freedom has been criticized as
an incoherent form of compatibilism. This criticism takes different forms,
but the basic point is that one cannot claim both that one’s actions are
causally determined in a series of natural events and that one is free in the
sense that one’s actions are ultimately caused by some more fundamental
freedom. There is simply no room for both natural and free causes, when
freedom is understood in Kant’s anti- determinist sense. Responses to this
objection have been almost as varied as the formulations of the objection
The problem that Schleiermacher raises is not primarily this metaphys-
ical one. Schleiermacher is not claiming merely that studying human be-
ings as natural objects is impossible because they are metaphysically free.
His objection also, and more fundamentally, involves a problem from
the standpoint of practical reason. This problem arises in the context of
moral anthropology.7 The practical problem is how to account for moral
judgments that make use of anthropological insights regarding helps and
hindrances for moral development.
Anthropological insight into empirical influences on moral choice is
not merely scientific knowledge of human beings. For one thing, it is
knowledge that is specifically articulated for practical use. In his anthro-
pology, for example, Kant argues that politeness promotes virtues.8 This
observation leads to a duty to promote politeness. The anthropological
perspective enters into the practical one. Of course, scientific perspectives
enter into practical deliberation all the time. When I serve tea to guests,
I make use of my knowledge that tea will contribute to their happiness. A
murderer who decides to pull a trigger makes use of his or her knowledge
that the gun will fire and kill the victim. But moral anthropology enters
into deliberation in a more problematic way. Specifically, a human agent
must be considered at once as both empirically influenced and morally
responsible. One must consider people as capable of influence by polite
society, or one does not have any responsibility to promote polite soci-
ety. But one must also think of this influence as bearing on the moral
status of those people, because that is the particular sort of influence
that makes politeness so important.9 But then people must be consid-
ered from a practical perspective and thus as free from any empirical
4 Introduction

influence. The conflict between freedom and moral anthropology arises

as a practical problem even if metaphysical issues surrounding freedom
can be resolved.

3. Kant’s Susceptibility to the Problem

Ultimately, this book defends Kant’s account of freedom and anthropol-
ogy. But the first half of the book shows just how closely Schleiermacher’s
objection makes contact with Kant. One easy way to defend Kant would be
to deny one of the three claims that constitute Schleiermacher’s dilemma.
If Kant does not affirm transcendental freedom, or does not hold that
anthropology is empirical, or restricts anthropology to nonmoral con-
texts, then he is easily saved. But Kant cannot be saved that easily. He
does affirm all three of the claims that form the dilemma. The project of
saving Kant thus involves showing that there is a way that they can all be
held consistently.

In Chapter 1, I show that Kant has a strong, noncombatibilist concep-

tion of human freedom. I focus on one crucial feature of Kant’s ac-
count that makes his anthropological work difficult. That feature is an
asymmetry in the causal relation between the noumenal free self and its
phenomenal appearance in the world, an asymmetry that arises whether
one holds a two-object, two-aspect, or two-perspective account of Kant’s
Kant resolves the third antinomy of the first Critique by suggesting
that although there cannot be a free cause in nature, there can be a free
ground of effects that are in nature (see A537f./B565f.). This ground has
a relation to its effects in the world similar to that of a natural cause,
though it is not spatiotemporal. In the first Critique, Kant does not show
that there is a free ground of empirical effects, only that for all we know
there can be such a ground. What is important, however, is that Kant
specifies the metaphysical place that such a ground would occupy. A free
ground of effects in the world would have to lie outside of nature in the
sense that it would not be susceptible to being an effect of natural causes.
This is precisely what it means for such a cause to be free. Theoretical
reason provides a basis for saying that if there are free causes, they must
not be influenced by other causes in the empirical world. In that sense,
the relationship between freedom and nature is asymmetrical.
In the second Critique, Kant argues that human beings actually are
free agents. Human beings fill the spot left open but empty by the first
A Problem with Kant’s Moral Anthropology 5

Critique. And in the second Critique, Kant again emphasizes the asymmetry
between the free cause – the agent – and the empirical world in which
one’s agency is effective. Not only for theoretical reasons (the nature
of causation) but for moral ones (the conditions of possibility of moral
responsibility), the free agent must affect but not be affected by the world.

In Chapter 2, I show that Kant’s anthropology is empirical. This is the

most consistent claim that Kant makes about his work in anthropology,
persisting through all his lectures on it and enduring in his published
work from his earliest works to his published Anthropology. In the Ground-
work, he distinguishes pure morality, which is the rational part of ethics,
from “practical anthropology,” which is the “empirical part” (4:388). In
the Metaphysics of Morals, he explains that in anthropology “we shall often
have to take as our object the particular nature of human beings, which
is cognized only by experience” (6:216–17).
In the Anthropology itself Kant makes clear that anthropological in-
vestigation is a matter of empirical observation, not a priori theorizing.
Although he dismisses observations that are not put to use as “specula-
tive theorizing” that “is a sheer waste of time,” proper anthropological
investigation also consists in “observations,” but only when one “distin-
guishes between those observations which have been found to hinder
and those which have been found to promote” the faculty under inves-
tigation (7:119). Both the fruitless and the proper sorts of anthropology
are empirical. The difference is that proper anthropology puts empirical
observation to use. In Chapter 2, I articulate what it means for anthro-
pology to make claims that are at once universal, related to the free self,
and empirical.

In Chapter 3, I describe how Kant’s empirical anthropology is a moral

anthropology. That is, anthropology takes moral choice as one of its ob-
jects. Based on Kant’s distinction between moral, pragmatic, and tech-
nical considerations in the Groundwork, one might think that the title of
Kant’s published work, Anthropology from a Pragmatic Point of View (empha-
sis added), speaks against its inclusion of moral considerations. However,
the term pragmatic is used in several different ways in Kant’s work, and
I show that the primary sense of pragmatic in the Anthropology contrasts
not with moral but with merely theoretical or physiological concerns.
Kant’s anthropology is meant to be useful, rather than mere knowledge
of human beings. And one of the uses of anthropology is to cultivate and
encourage good moral choices.
6 Introduction

Thus one finds in the Anthropology examples of empirical helps and

hindrances to having a good will. Kant’s discussions of politeness, of the
passions, and of character all present these as helps or hindrances to
morality itself, so I briefly discuss each of these examples. The chapter
concludes by pointing out the systematic place of moral anthropology in
Kant’s anthropology as a whole. The first three chapters thus show that
Kant’s moral anthropology makes him susceptible – at least at first sight –
to Schleiermacher’s criticism.

4. Freedom and Anthropology in Contemporary Moral Theories

In recent years, neokantian moral theorists have begun to pay more at-
tention to Kant’s moral anthropology. Onora O’Neill, in a chapter of
Constructions of Reason called “Action, Anthropology, and Autonomy,” ac-
counts for a “gap between Kant’s practical philosophy and contempo-
rary would-be Kantian writing on ethics” in part by pointing out that
“modern protagonists of ‘Kantian’ ethics are mainly interested in rights,
which for Kant are one element in a broader picture” (O’Neill 1989:
66). Anthropology helps to flesh out this broader picture. Allen Wood
explicitly articulates his conception of Kant’s ethical thought in contrast
to approaches that are open to “common charges that Kantian ethics is
unconcerned with the empirical realities of psychology, society, and his-
tory, that it sees no value in the affective side of our nature, and that it is
individualistic” (Wood 2000: xiv). This new approach to Kant involves an
extensive treatment of Kant’s anthropology (see pp. 193–320). Robert
Louden’s Kant’s Impure Ethics is devoted entirely to drawing attention to
“the second part of Kant’s ethics, a part that . . . unfortunately remains a
well-kept secret [and that] Kant referred to . . . as ‘moral anthropology’ ”
(Louden 2000: vii). And G. Felicitas Munzel’s Kant’s Conception of Moral
Character (1999) seeks to integrate Kant’s anthropology into his moral
philosophy through a study of the notion of “character.”
Even neokantians who do not discuss Kant’s anthropology as a whole
often turn to specific aspects of moral anthropology to flesh out their
Kantian moral theories. Kant’s remark that people are “not to shun sick-
rooms or debtors’ prisons” (6:457) has become a popular text to point
out Kant’s awareness of the importance of cultivating sympathy for the
moral life.10 Barbara Herman has drawn attention to the role that com-
munity and education can play in promoting moral behavior (Herman
1993, esp. pp. 82–3), and she has drawn attention to the importance
of “character” for Kant (Herman 1996). Nancy Sherman, in a series of
A Problem with Kant’s Moral Anthropology 7

books and articles, has shown that Kant’s account of emotions can con-
tribute to a richer Kantian ethical theory (Sherman 1990, 1995, 1997,
and 1998). All of these developments spring from Kant’s anthropological
The attention to Kant’s anthropology in contemporary ethics is not
merely an attempt to be historically accurate. Neokantian moral theo-
rists find in Kant’s anthropology a richness of detail and attention to
human particularity that should be an important part of any moral the-
ory. The recent rise of neo-Aristotelian, Humean, and anti-theoretical
approaches to moral theory has presented serious challenges for Kantian
moral theories.11 These apparent alternatives to Kant tend to focus on
character rather than action, virtues rather than rights or duties, and take
into account a wide range of features of human psychology that Kantians
have sometimes ignored. They thus present sensitive accounts of moral
development and the role of emotions in moral motivation, and they can
seem to provide a very nuanced account of ethical life. The focus on for-
mulaic applications of the categorical imperative, and a general empha-
sis on the Groundwork in Kantian moral theory, has made some Kantians
particularly susceptible to challenges from these alternative accounts of
Kant’s anthropology provides effective responses to many of these ob-
jections. His moral anthropology includes extensive discussions of the
importance of community and education for moral development. He
discusses and differentiates different sorts of emotions and various roles
that these can play in moral life. His moral anthropology focuses on culti-
vating a virtuous character, rather than on merely doing good deeds. And
throughout his anthropological writings, Kant discusses character, dispo-
sition, and virtue.12 Moreover, his anthropology provides detailed, even
if scattered, accounts of the particulars of human life. He analyzes the
psychology that underlies sexual temptation, gives a sophisticated treat-
ment of the role of politeness in modern life, and even provides advice
on conducting an excellent dinner party. Even when these descriptions
of human life fall short of what one might hope for, they go far beyond
the abstraction of the categorical imperative.

Thus neokantians have been right to look to Kant’s anthropology for

an ethical theory that can hold its own against recent virtue-based and
anti-theoretical approaches to ethics. But there has been insufficient at-
tention to the problems that Kant’s anthropology presents for his over-
all moral theory. In Chapter 4, I take up three of the more prominent
8 Introduction

current neokantians who draw extensively from Kant’s anthropology –

Nancy Sherman, Robert Louden, and G. Felicitas Munzel. All three ex-
plicitly articulate their accounts as theories that can meet some of the
challenges recently raised against Kant, especially by those sympathetic
to Aristotle. Because of their use of Kant’s anthropology, these contem-
porary neokantian moral theories are susceptible, in varying degrees, to
Schleiermacher’s objections to Kant.
Unfortunately, no one has yet offered a sufficient integration of Kant’s
moral anthropology with his conception of freedom. Thus contemporary
accounts often fall short of seeing the full significance of Kant’s moral
anthropology. In some cases, they simply fail to recognize all the ways in
which Kant’s anthropology affects his ethics. Nancy Sherman, for exam-
ple, allows for important anthropological influences but ultimately does
not give anthropology the range of moral significance that Kant allows.
In other cases, neokantians fail to save Kant’s theory of freedom. The
result is a moral theory that is so tied to anthropology that it loses its dis-
tinctive Kantian emphasis on freedom. At times, Louden and Munzel go
in this direction. Given the increasing emphasis on moral anthropology
as an important part of a contemporary Kantian ethics, there is a need
to articulate an answer to Schleiermacher’s challenge that can justify the
integration of anthropological insights into a genuinely Kantian moral

5. Solving Schleiermacher’s Dilemma

The second half of this book, especially Chapters 5 and 6, offers the
needed solution to Schleiermacher’s dilemma. In Chapter 5, I show that
Kant has the resources to distinguish between the empirical will, which
can be affected by empirical influences, and the free will, which cannot.
The connection between these is such that the empirical will is morally
relevant as the expression of the moral status of the free will. In the simplest
case, an action in the world such as making a false promise for personal
gain expresses an evil will. But the situation is complicated by the pres-
ence in human beings of what Kant calls radical evil. Radical evil involves
both choosing badly and making choices that reinforce one’s tendency
to choose badly. This evil forces Kant to reconceive of the nature of the
human good will and its expression in the world. According to Kant’s ac-
count in Religion within the Boundaries of Mere Reason, the best that humans
can hope for is a “revolution” against evil, the expression of which is a
constant struggle against evil in one’s nature. Although the revolution
A Problem with Kant’s Moral Anthropology 9

itself is not something temporal, the struggle that expresses a revolution

in the will is temporally extended. To combat radical evil, one must not
only choose rightly but also act to counteract one’s tendency to choose
poorly in the future. In cool hours of moral self-control, one who is in
revolution will act to effect a good empirical will not only in the present
but in the future as well. In this context, moral anthropology is crucial.
Moral anthropology explains the means for effectively correcting and im-
proving one’s empirical will. And the effort to correct and improve one’s
empirical will is part of the struggle against evil that expresses the will
in revolution. Promoting a good character through methods explained
in moral anthropology is an expression of one’s free will, so the asymme-
try between nature and freedom in Kant’s philosophy is preserved. But
because it is an expression of one’s moral status, anthropology has moral
significance. This solution to Schleiermacher’s dilemma is worked out in
detail in Chapter 5.

In Chapter 6, I take up an important remaining problem with the ac-

count offered in Chapter 5. If considerations from moral anthropology
are relevant because they enable one to express a revolution against evil,
it is not clear how interpersonal moral influence can be morally significant.
I argue that Kant is not as committed to the possibility of interpersonal
moral influence as some have suggested. He does not think that one
should seek to effect moral revolutions in others. Nonetheless, acting to
improve the character of others is morally significant for several reasons,
which I explore in Chapter 6. Most important, because one’s own empir-
ical will is connected to the wills of others, acting to improve the wills of
others expresses one’s own struggle against evil. I even argue that there is
some room for Kant to allow that the actions of one agent can genuinely
affect the moral status of another. Still, one can never know how this
occurs and should not consider it a reason for acting to promote moral
development in others.

In the brief Epilogue, I reflect on where the debate between Kant and
Schleiermacher, and Kant’s many other critics, stands given the account
of Kant’s moral theory offered in this book. Although this book does not
show that Kant’s moral theory is the only reasonable option, it does show
that one of the most important objections to that moral theory fails. Kant
can integrate moral anthropology into his ethics without sacrificing the
account of freedom that lies at the core of his philosophy.
part i


The Asymmetry in Kant’s Conception of Freedom

In the Introduction, I raised a problem with Kant’s moral anthropology.

For this problem to make any contact with Kant’s philosophy, Kant’s
theory of freedom must preclude the influence of empirical causes
on moral choice. In this chapter, I argue that Kant does articulate
just such a conception of human agency. In section 1, I briefly sketch
Kant’s mature argument for freedom and draw attention to the role of
moral responsibility in this argument. In section 2, I argue that what-
ever one believes about the relationship between freedom and nature,
or noumena and phenomena, or practical and theoretical perspectives,
it is essential to recognize that for Kant, freedom is the ground of na-
ture and not vice versa. Although Kant cannot explain how freedom
grounds nature, he can say something about what this relation means,
and why one must assume it. To avoid entering into recent debates
about whether Kant offers primarily a two-object or a two-perspective
account of the self, I defend the asymmetry of the relation between na-
ture and freedom on both the two-object and two-perspective interpre-
tations of the relation. In section 3, in order to clarify this asymmetry,
I point out how the priority of the practical precludes any traditional
determinist account of freedom. Finally, in the last section I touch on
the nature of our knowledge or lack thereof of the free self. The na-
ture of knowledge of the free self is important to clarify the sort of re-
sources available to Kant to explain the relation between freedom and

14 The Problem

1. Kant’s Argument for Freedom

Kant’s argument that persons are free has two stages. In the Critique of
Pure Reason, he argues that the possibility of freedom is not precluded
by the nature of our experience of the world, the requirements of scien-
tific knowledge, or any justifiable metaphysical theories. This argument
depends on Kant’s transcendental idealism. Kant claims that experience
is structured according to two forms of intuition – space and time – and
twelve categories, one of which is cause and effect. Because these intu-
itions and categories provide the structure of our experience, one knows
a priori that any experience of ours will be spatiotemporal and will fit
into a series of natural causes that are objects of possible experience.
However, because these intuitions and categories structure our experience,
one cannot apply them in order to know “things in themselves.”1 In the
context of arguing for this transcendental idealism Kant presents sev-
eral antinomies, the third of which highlights the conflict between the
claims that there is freedom and that there is no freedom.2 The anti-
nomy arises because while understanding demands that every object of
experience be explained in terms of a (temporally)3 prior cause, reason
demands that this cause be sufficient, such that it leaves nothing to be
explained. Only a cause that is itself not in the world can serve this pur-
pose, because the world is structured by spatiotemporal causation such
that every cause has a prior cause. If a cause c1 of an effect e1 has an-
other prior cause c2 , then c1 does not fully explain e1 , because even after
granting c1 , there is something left to explain (namely, what caused c1 ).
To satisfy the demands of reason, one would have to find some cause that
does not require (or does not allow) any further explanation. Ultimately,
Kant argues that the resolution of the antinomy depends on the fact that
one can distinguish things as they are in themselves from things as they
Because experience is structured according to the laws of causality,
nothing can be experienced except as determined by prior causes. Thus
nothing can be experienced as free. But it is thinkable that free things in
themselves provide a ground for the series of appearances that is ordered
according to natural laws, and these free things neither need nor allow
further explanation in terms of further causes. Because these things in
themselves are not structured according to the schematized categories of
human understanding, they cannot be thought of as possible objects of
our experience. The way in which they “ground” appearances is at best
only analogous to the way appearances ground one another.5 But Kant
Asymmetry in Kant’s Conception of Freedom 15

at least shows that nothing that one knows about natural causation rules
out some sort of freedom. As he explains,

It should be noted that we have not been trying to establish the reality of
freedom. . ., [and] we have not even tried to prove the possibility of freedom . . . .
[To show] that . . . nature at least does not conflict with causality through freedom –
that was the sole thing we could accomplish, and it alone was our sole concern.

In the Critique of Pure Reason, Kant’s goal is simply to open a space for
In the second stage of the argument, Kant argues for the reality of
freedom.6 In the second Critique, Kant claims it is a “fact of reason” that
people have moral obligations. Kant uses this fact “of which we become
immediately conscious” to support the reality of freedom (5:29, cf. 5:6,
31–2, 42, 47, 55, 91, and 104–5; and Rawls 2000: 253–72; Beck 1960:
166–75; Allison 1990: 230–49; and Ameriks 1981). Moreover, this moral
obligation does not depend upon any particular details about oneself.
Rather, the good concerns something that anyone ought to do in the same
situation. The moral law that underlies this obligation binds universally.
And this universality is not merely contingent. Obligation is not due to
some feature that agents happen to have in common such that if particular
persons came to lack this shared characteristic they could be excused from
morality. Moral obligation is both universal and necessary. From these
general features of moral obligation, Kant argues in the Groundwork that
the moral law that underlies obligation must have nothing to do with any
particular or contingent content or context of choices but with the form of
morally acceptable choice.7 Because whether or not a choice is permitted
does not depend on the particular empirical influences in the context of
which the choice is made, the way in which one is obligated to act is not
determined by natural causes in the world. And yet, to be obligated, it must
be possible to act in the way required. But then one must be free in the
sense that one can act in a way that is not causally determined by empirical
influences alone. Because the moral law requires action on a basis distinct
from natural causes, and because transcendental idealism allows freedom
beyond natural causes, it must be possible to act on such a basis whenever
the moral law is applicable. When morality is at stake, then, one is free
in the sense that one’s actions need not have their ultimate ground in
natural causes but can be chosen freely in accordance with a universal
moral law.8
16 The Problem

2. The Priority of Freedom

Although every person is free from determination by natural causes, a
person’s actions, as appearances in the world, still are explicable in terms
of those causes. In this section I briefly consider the issue of how a free and
morally responsible person is related to that person’s actions as objects
of experience. In “Kant’s Compatibilism” (Wood 1984), Allen Wood has
dealt with several of the metaphysical issues that arise in this context. My
aim here is to show merely that freedom is in an important sense prior
to actions that can be empirically described. Exactly how this priority is
described and defended depends on whether one takes Kant’s distinction
between things in themselves and appearances to describe merely two
different ways of thinking about things (two perspectives) or two different
sorts of things that might interact (two objects). In this section, I do not
arbitrate between these two interpretations of Kant. Instead, I explain
how the priority of freedom can be explained and justified on either

A. Asymmetry on a Two-Object Account of Kant’s Transcendental Idealism

Richard Aquila offers the best defense of a two-object account of Kant’s
transcendental idealism. Aquila claims to “elaborate a defense of the two-
object approach” to Kant’s transcendental idealism (Aquila 1979: 295)
and explicitly applies this approach to the nature of the free self. Aquila
focuses on the apparent problem that in the case of the person, “a number
of passages concerning these phenomena do seem to require the con-
ception of a single entity which may be regarded both phenomenally and
as a thing in itself ” (Aquila 1979: 305). Aquila deals with these passages
by articulating a two-object intepretation of the notion of a “phenomenal
person.” He explains,

It is compatible with the two-object interpretation to maintain that a “phenomenal

person” is to be identified not with a particular phenomenal object, but rather
with a noumenal entity regarded as “ground” of at least some of the behav-
ior exhibited by some phenomenal object. . . . [A] phenomenal person is a cer-
tain sort of composite : a composite of a noumenal subject and a phenomenal
object. . . . All propositions about phenomenal persons (qua persons) are an-
alyzable into propositions about both a noumenal subject and a phenomenal
object. (306)

This account is a two-object account because there is a real distinction

between the noumenal subject and the phenomenal object of which that
subject is the ground.
Asymmetry in Kant’s Conception of Freedom 17

The “grounding” relationship between the subject and the phenome-

nal object is akin to a causal relationship, and this relationship is asym-
metrical. The free subject is always the ground of the phenomenal object,
while the phenomenal object never exerts causal influence on the free
subject. Both the metaphysical reflections of the first Critique and the prac-
tical arguments of the second Critique defend this asymmetry. In the first
Critique, Kant explains how it is possible to talk about relations between
the phenomenal and the noumenal. He admits that one cannot make
sense of the relation between the free subject and the phenomenal ap-
pearance of that subject using schematized categories of the understand-
ing. But one can think about this relation in terms of unschematized
concepts. Kant is not totally consistent about whether these concepts are
unschematized categories or merely logical forms of judgment. He often
says that the free self is the ground (Grund) of its appearance (see, e.g.,
A545–6/B573–4), making use of the ground-consequent relation from
the table of judgments. But at other times he claims that the free self
is the cause (Ursache or Kausalität) of its effects in appearance (see, e.g.,
A539/B567), making use of the category of cause and effect. Kant is best
understood as holding that categories, but not schematized categories, can
be used to think of things in themselves. When used to think about a
free self, “cause” must be taken in a sense that does not involve intuitions
of space and time. Kant’s use of the term Grund can then be seen as a
way to draw attention to the fact that things in themselves do not cause
appearances in the way that we ordinarily understand causation – that
is, spatiotemporally. On any interpretation of this relation, it is crucial
that things in themselves ground appearances, and not vice versa. In the
personal case, the free self in itself is the ground of the appearance of
that self, hence of one’s actions and thereby of changes in the world.
The relationship is asymmetrical in that the morally responsible agent
determines the way the world is, whereas changes in the world do not
determine the status of a moral agent. And that means that the ultimate
ground of one’s actions is oneself, not prior empirical causes.9
The theoretical need for this asymmetry is found in Kant’s resolution
of the third antinomy. This resolution involves positing the possibility10
of freedom as a ground of appearances that is not within the series of
appearances themselves. As Kant explains,
While for every effect in the [field of ] appearance a connection with its cause in
accordance with the laws of empirical causality is indeed required, this empirical
causality, without the least violation of its connection with natural causes, is itself
an effect of a causality that is not empirical but intelligible. (A544/B572)
18 The Problem

Whereas there can be no freedom in the series of natural causes, freedom

can characterize a purely intelligible cause of the empirical series. This
need not require that there in fact be such a free intelligible cause. But it
establishes that if there is such a cause, it is prior to the appearances that
it causes, prior just by virtue of the fact that the intelligible causes the
empirical and not vice versa.
To this theoretical basis for asymmetry, Kant adds a practical one. Prac-
tical reason establishes not just a space for freedom but for the reality of
freedom. But the practical argument for freedom also requires an asym-
metry between freedom and nature, because Kant’s conception of moral
responsibility precludes the freedom of a “marionette or an automaton,”
in which one is “free” but at a deeper level determined by natural causes
(5:101, cf. 28:267). For practical reasons, the noumenal self must be
truly free. It must not be itself affected by the series of appearances that
it grounds.

B. Asymmetry on a Two-Perspective Account of Kant’s Transcendental Idealism

When one interprets Kant according to a two-perspective theory, the
picture gets more complicated. On this account, there is neither a real
distinction nor causal interaction between things in themselves and their
appearances. Instead, the free self and the self as appearance merely refer
to two different perspectives one can take on the self. Christine Korsgaard
explains the distinction in terms of two different “standpoints”:

The deliberating agent, employing reason practically, views the world as it were
from a noumenal standpoint. . . . The theorizing spectator, on the other hand,
views the world as phenomena, mechanistic and fully determined. The interests
of morality demand a different conceptual organization of the world than those
of theoretical explanation (MM 217; 221; 225). Both interests are rational and
legitimate. (Korsgaard 1996a: 173)

People have different reasons to give accounts of actions. Depending on

the “interests” that motivate the account, one assumes either a practical
standpoint according to which one is the ultimate free cause of the ac-
tion or a theoretical standpoint within which one can trace the natural
causes that give rise to the action. Korsgaard is right to point out that
both are rational and legitimate. But she then goes on to say that “it
is important that neither standpoint is privileged over the other – each
has its own territory” (173). Korsgaard quickly catches herself, recog-
nizing that this symmetry is contrary to Kant’s own account, and adds,
“Or, if either is privileged, it is the practical, because, according to Kant,
Asymmetry in Kant’s Conception of Freedom 19

‘every interest is ultimately practical’ (C2 121)” (173). Korsgaard’s con-

cession to Kant is important here, but she does not sufficiently work out
a Kantian justification of the asymmetry.11 As a result, it is all too easy to
think that it really does not matter whether one perspective is privileged
over another. They just have separate spheres. And, it might seem, there
is no third “metaphysical” perspective that gets at what we “really” are.
For Kant, however, it is crucial to establish the priority of the practical.12
Kant does not establish this priority through any third perspective
that goes beyond the theoretical or the practical. Instead, he argues that
from within each of these perspectives, the status of the other perspective
is raised. In the third antinomy of the first Critique, for example, the
demands of reason within the theoretical perspective raise the problem of
the need for an unconditioned ground of the series of appearances.
In the course of discussing the relationship between the theoretical and
practical perspectives, John Rawls explains the interests of the theoretical

Theoretical reason has two legitimate interests: one is the positive interest in
regulating the understanding and unifying into the highest possible systematic
unity the low-level empirical knowledge it provides; the other is the negative
interest in restricting speculative folly. (Rawls 2000: 324)

Rawls goes so far as to say that “the unity of reason is established . . .

by the full satisfaction of the legitimate claims of theoretical and practical
reason as articulated in the form and structure of the two points of view”
(324). In his description of this unity of reason, Rawls focuses on the fact
that the theoretical and practical perspectives do not make conflicting
claims. Theoretical reason “abdicates” a “space” within which practical
reason can make practically justified claims (324).
The interest of theoretical reason in the systematic unity of knowledge
leads to a need on the part of theoretical reason for some fundamental
ground. In the context of a two-perspective account of Kant’s idealism,
this cannot be some fundamental metaphysical “cause” of appearances
but must be a more fundamental perspective on the same world that the-
oretical reason investigates. Now Kant insists that one cannot give any
theoretical argument to show that there is such an ultimate perspec-
tive. The demand of theoretical reason for completeness within empirical
investigation goes beyond the conditions of empirical knowledge. But
reason’s theoretical demands are real and legitimate, and the Critical
solution is to hold open the possibility, and even the hope, of a new per-
spective that is not itself restricted by the forms of intuition. By positing
20 The Problem

the priority of this new perspective to its own, theoretical reason can hold
open the possibility of satisfying its own demands. Thus Kant suggests
in the first Critique that “the architectonic interest of reason” provides
a speculative recommendation for the thesis of freedom, because with-
out it, “a complete edifice of [empirical] knowledge is . . . altogether
impossible” (A474–5/B502–3). And in the second Critique, Kant reminds
his readers that the transcendental freedom for which he argues on the
basis of the fact of the moral law is the unconditioned ground that “spec-
ulative reason needed” (5:3, see also 5:7, 48–9, 107, 120). This implies
that the structure of the two points of view involves an asymmetry between
points of view.
In “What Is Orientation in Thinking” – published just after the Ground-
work and before the second edition of the first Critique – Kant offers an
extended account of this sort of “need which reason imposes on itself ”
(8:136). Kant suggests that the “need of reason can be regarded as twofold
in character: firstly, it has a theoretical use, and secondly, a practical use”
(8:139). In this essay, Kant is particularly interested in the need of reason
to posit God, but the need of theoretical reason to posit God has a parallel
in reason’s need for a unconditioned free cause in the first Critique. Kant
suggests that the need of reason to posit something unconditioned as a
means of “orientating ourselves in thought” can give a “subjective ground
for presupposing and accepting something which [theoretical] reason
cannot presume to know on objective grounds . . . purely by means of
the need of reason itself ” (8:137). Kant is careful, later in the essay, to
point out that this subjective ground is merely conditional and thus does
not provide the kind of justified belief that is provided by the postulates
of practical reason in the second Critique. That is, whereas theoretical
reason requires positing God or freedom only “if we wish to pass judgment
on the first causes of all contingent things,” practical reason compels us
“to assume that God exists . . . because we must pass judgment ” in practical
matters (8:139). As a result, insofar as freedom is established as a condi-
tion of moral responsibility, it justifies a “conviction of truth [that] is not
inferior in any degree to [empirical] knowledge . . . even if it is totally
different from it in kind” (8:141). Because of its “need” for completeness
in empirical investigation, theoretical reason itself points to a perspective
that it cannot justify or characterize. Moreover, for this other perspective
to satisfy the need of theoretical reason, it must provide an explanation
that conditions but is not conditioned by what is described empirically.
In the second Critique, Kant nicely summarizes both the range and lim-
its of this demand on the part of theoretical reason. He there explains
Asymmetry in Kant’s Conception of Freedom 21

what could be said on the basis of theoretical reason alone, in the first
Critique, about the relationship between the theoretical and practical

We could defend the thought of a freely acting cause. . . . On the one side, in the
explanation of events in the world and so too of the actions of a rational being, I
grant the mechanism of natural necessity . . ., while on the other side I keep open
for speculative reason the place which for it is vacant, namely the intelligible. . . .

Theoretical reason has a need for and holds open a vacant space for
freedom. And this vacant space, as described by theoretical reason,
must be a perspective that is “causally” prior to the theoretical space
the need of which it satisfies. Whatever perspective fills that vacant
space must be considered, even by theoretical reason, to provide the
ultimate explanation that empirical investigation, because the forms
of intuition condition it, can never provide. Theoretical reason holds
open the possibility13 that the ultimate explanation of actions is beyond
the conditions of possibility of experience because only on this sort of
foundation can it hope for that “complete edifice of knowledge” which
is the regulative ideal of empirical investigation. What theoretical reason
can not do, however, is “realize this thought” – that is, show that there
actually is another legitimate perspective on the actions of a rational
being. “Pure practical reason now fills this vacant place” (5:49).14
At the abstract level of theoretical reason’s need, the perspective that
theoretical reason privileges is any perspective from which freedom can
be justified, not particularly the practical one. There is a space left open,
from within the perspective of theoretical reason, for a perspective that
satisfies theoretical reason’s interest in systematic completeness. But this
new perspective can play this role only by being more fundamental than
the theoretical perspective itself. Once the practical perspective provides
a basis for freedom, it satisfies the demand raised within the theoretical
perspective and is considered, from that theoretical perspective, to have
From the practical standpoint, one also considers the status of the em-
pirical. An agent is interested in evaluating actions or volitions that occur
in the realm of experience. This is true for both moral evaluation and
moral deliberation. When one evaluates oneself or another, one evalu-
ates an action or a volition that is experienced. Only insofar as a volition
enters the realm of experience16 can it be an object of evaluation, but if
a volition is considered as experienced, it must be considered according
22 The Problem

to the perspective of empirical investigation. This does not mean that the
empirical perspective is the proper perspective for moral evaluation; it
clearly is not. But when we evaluate something, the perspective of empir-
ical investigation is at hand. We can privilege the moral perspective for
the purpose of evaluation only because we consider the moral perspective
more fundamental than the empirical one. The self considered as free is
the ultimate ground of what is observable, and not vice versa.
In the case of moral deliberation, the relationship between the em-
pirical and practical perspectives also arises. When one deliberates, one
deliberates about what to do. The decision that I make is a decision that
is a possible object of empirical investigation. What is more, and this is
crucial, I care about the decision insofar as it is an object of possible
empirical investigation. My choice has weight for me because it will have
effects in the world. This does not mean that all I care about is results. It
is crucial for Kant that morality not be based on consequences. But what I
care about is an actual decision, a decision that enters the series of causes
in experience. As Kant explains in the Critique of Practical Reason, “this
[moral] law is to furnish the sensible world . . . with the form of . . . a super-
sensible nature “(5:43, emphasis added). What I deliberate about from
the practical perspective is what I will choose to do, where this choice and
the action that proceeds from it are possible objects of empirical investi-
gation. Thus it makes all the difference in the world, from the practical
perspective, that this choice can be thought of from an empirical perspec-
tive. But it also makes all the difference in the world that the practical
perspective, according to which I am free, is more fundamental than this
empirical perspective. In deliberating I take my choice understood from
a practical perspective to be the real ground of that choice understood
from an empirical perspective, and not vice versa. Thus from both the
theoretical and the practical perspectives, the practical is considered more
fundamental than the theoretical.

Even if one does not accept in general that the theoretical standpoint
must be prior to the practical, the practical perspective has priority over
the theoretical in the arena of moral anthropology. As we will see, moral
anthropology involves empirical insights about human beings that are
put to practical use in cultivating moral character. The person putting the
insights to use must be considered from a practical perspective because he
or she is acting, and the one in whom moral character is cultivated must
be considered from a practical perspective because what is cultivated is
moral character. And within the practical perspective itself, the practical
Asymmetry in Kant’s Conception of Freedom 23

certainly has priority over the empirical. Thus even if the priority of the
practical is not maintained in general, it must be affirmed in precisely the
context in which it poses problems for Kant’s anthropology.

Whether one holds a two-object or a two-perspective view of the relation

between the free self and the naturally determined self, these two objects
or perspectives are not symmetrical. The free self (or the self considered
as free) is the ground of the empirical (or empirically considered) one
and not vice versa. Even if the empirical self is also phenomenally deter-
mined according to natural causation, nothing empirical determines the
fundamental nature – in particular, the moral status – of the free self. Kant
rightly describes the relation between the free self and its appearance,
or between two perspectives on the self, as one of ground and conse-
quence, and even of cause and effect.17 And so he rightly suggests that
when one empirically investigates oneself or others, one merely observes
what people look like, not what they really are.

3. Incompatibilist Freedom
The belief in the priority of freedom precludes any straightforward
compatibilism. Like compatibilists, Kant holds that human actions are
both causally predetermined and free. But unlike most compatibilists –
so-called “soft determinists” – Kant insists that the free cause of one’s
actions cannot be predetermined. Whereas soft determinists interpret
“freedom” in such a way that a free cause can be determined by prior nat-
ural causes, Kant affirms that people are free causes of their actions, and
that those actions are determined by an infinite series of natural causes,
but also that a free cause cannot itself be causally determined. In this
sense, Allen Wood’s description of Kant’s position as a “compatibilism
between compatibilism and incompatibilism” is apt (Wood 1984). Kant
is adamant about the insufficiency of standard compatibilist accounts of
freedom as grounds for moral imputation.18 He is able to avoid these
standard accounts and the problems with them because of his belief in
the priority of freedom over natural causation.
In the Critique of Practical Reason, Kant clearly expresses his dissatisfac-
tion with “soft determinist” accounts of freedom. If such an account were
true, Kant says,

A human being would be a marionette or an automaton . . . built and wound up

by the supreme artist; . . . the consciousness of his own spontaneity . . . would be
mere delusion. (5:101)
24 The Problem

Freedom that has its ultimate cause in anything else, whether this is a
natural world or divine Creator, is not real freedom but “mere delusion.”19
And this soft determinist freedom is simply not enough to provide a
condition of the possibility of moral responsibility, as Kant makes clear
elsewhere in the second Critique:
It is a wretched subterfuge to . . . say . . . that the kind of determining grounds of
[man’s] causality in accordance with natural law agrees with a comparative concept
of freedom . . ., e.g., that which a projectile accomplishes when it is in free motion,
in which case one uses the word “freedom” because while it is in flight it is not
impelled from without. (5:96)

A few sentences later, Kant explains why he believes that this compatibilist
solution is unacceptable:
They therefore leave no transcendental freedom, which must be thought as inde-
pendence from everything empirical and so from nature generally . . .; without
this freedom . . ., which alone is practical a priori, no moral law is possible and
no imputation in accordance with it [is possible]. (5:97)

Kant clearly believes that a compatibilist account of freedom does not

leave room for the freedom necessary for moral responsibility. If one is
to be held morally responsible for an action, it must be possible to be
free of empirical determination, because this sort of autonomy is morally
required. If forces of nature ultimately cause all action, the best that
anyone can hope for is action in conformity with the demands of the
moral law. One could never act from the moral law itself, because the
moral law is not a force of nature. And acting from the moral law is
precisely what the moral law demands.
But at the same time that Kant holds as a condition of possibility of
moral responsibility a sort of freedom that is not compatible with soft
determinism, he is committed to the view that human actions can be
described in terms of natural causes. Whereas most reconciliations of
freedom and natural determination insist that the ultimate explanation
of an action is according to natural causation – such that one’s “freedom”
is just a proximate but predetermined cause of an action – Kant insists
that the fundamental level of explanation is freedom. Natural causes just
express that freedom. The difference between Kant’s “compatibilism”
and the accounts that he attacks is “the difference between the laws of
a nature to which the will is subject and of a nature which is subject to a
will ” (5:44). That is, the primacy of freedom over nature distinguishes
Kant from his compatibilist predecessors. Freedom is not a means by
which natural causes affect the world. Nature with its causes is instead
Asymmetry in Kant’s Conception of Freedom 25

the means by which freedom effects change in the world. There are, of
course, numerous problems with this view.20 What is crucial here, though,
is just that this description allows Kant to hold to a peculiar kind of com-
patibilism without a soft determinist conception of freedom. Freedom
and natural causation are compatible only because free choices, totally
undetermined by anything prior to them, are themselves the grounds of
a series of actions in the natural world.
There is one problem with Kant’s account that I take up directly. Kant’s
remarks about the cultivation of character and the role that empirical
influences can have in shaping and enabling freedom might lead one
to think that his conception of freedom cannot be as absolute as this
account makes it out to be. This problem is the subject of the rest of this
book. Before turning there, however, I must make one more provisional
comment on Kant’s account of freedom.

4. The Inscrutability of Freedom

Crucial to Kant’s account of freedom is the notion that one can distin-
guish between a free self (or practical perspective on the self) and a
causally determined self (or empirical perspective on the self). As we
have seen in sections 2 and 3, at least some detail with respect to the
relationship between these selves or ways of thinking of the self is im-
portant for differentiating Kant’s theory from soft determinist versions
of compatibilism. But Kant is also wary of giving too much detail about
the nature of freedom and its relation to the empirical self. When he
introduces the distinction between intelligible and empirical character
in the first Critique, for example, he insists,

This intelligible character can never, indeed, be immediately known, for nothing
can be perceived except in so far as it appears. It would have to be thought in
accordance with the empirical character. (A539–40/B567–8)

This theme, that the intelligible character can be known only through
its appearance, or empirical character, is reiterated throughout Kant’s
moral philosophy.
Still, Kant does insist that one can intelligibly say at least some
things about freedom. The most important of these is that freedom is

In our judgments in regard to the causality of free actions, we can get as far as the
intelligible cause, but not beyond it. We can know that it is free. (A557/B585)
26 The Problem

In making the claim that there actually is an intelligible cause that is

free, Kant goes beyond21 the strict speculative limits of the first Critique
and acknowledges as much. His argument for the reality of freedom is
based, even in a work that is supposed to limit itself to theoretical reason,
on the fact that we impute responsibility to human agents, the fact of
practical reason. Kant’s fuller articulation of the way in which practical
reason establishes the reality of freedom comes in the Critique of Practical

We could [in the first Critique] defend the thought of a freely acting cause . . .
by showing that it is not self-contradictory to regard all its actions as physi-
cally conditioned insofar as they are appearances and yet also to regard their
causality as physically unconditioned insofar as the acting being is a being of the
understanding; . . . I do not cognize at all the object to which such causality [i.e.
freedom] is attributed . . . but I nevertheless remove the obstacle inasmuch as
on the one side, . . . I grant the mechanism of natural necessity . . ., while on the
other side I keep open for speculative reason the place which for it is vacant,
namely the intelligible. . . . Pure practical reason now fills this vacant place with a
determinate law of causality in an intelligible world (with freedom), namely the
moral law. (5:49)

While speculative reason can establish a “vacant space” for freedom, prac-
tical reason establishes its reality.
Although both of these passages suggest room for some knowledge of
the free self, Kant is careful, in both contexts, to limit such knowledge
to the mere fact that one is free without any understanding of how one
is free. Hence he says in the first Critique that we can get as far as the
intelligible cause – that is, its reality – but not beyond it to understanding
its nature. He goes on,

But to explain why in the given circumstances the intelligible character should
give just these appearances and this empirical character transcends all the powers
of our reason, indeed all its rights of questioning. (A557/B585)

Likewise in the second Critique, Kant insists that there is an important

sense in which one cannot even extend the category of causation to the
free self:

Even the concept of causality, which has application and so too significance strictly
speaking only in reference to appearances, in order to connect them into expe-
riences (as the Critique of Pure Reason proves) is not enlarged in such a way as to
extend its use beyond the boundaries mentioned. For, if reason sought to do this
it would have to try to show how the logical relation of ground and consequence
could be used synthetically with a kind of intuition different from the sensible,
that is, how a causa noumenon is possible; this it cannot do. (5:49)
Asymmetry in Kant’s Conception of Freedom 27

This inscrutability of freedom is reiterated later in the Critique of Practical

Reason (see 5:72), throughout Religion within the Boundaries of Mere Reason
(6:22, 43–4, 49fn, 50, 59, 62), and in Kant’s lectures in both metaphysics
and ethics (27:505, 29:1020–2).
Despite this inscrutability, however, Kant claims, “Freedom is regarded
as a kind of causality . . . with respect to actions possible through it as ap-
pearances in the sensible world” (5:67). This application of “a kind of
causality” to freedom seems to contradict Kant’s insistence that the cate-
gories cannot apply to the free self. The apparent contradiction can be
resolved if we get clearer on exactly why and in what sense Kant prohibits
the application of the categories beyond experience.
Kant is concerned with three related problems that could arise if one
applies the categories to the free self. First, his restriction of the categories
to empirical objects is part of a general limitation of theoretical knowledge
to objects of possible experience. Because the free self is not an object
of possible experience, one cannot have theoretical knowledge of it. The
categories as concepts through which one theoretically knows objects do
not apply to the free self. Hence, after establishing the reality of the free
self, Kant warns,

I do not now claim to know theoretically by this concept the constitution of a being
insofar as it has a pure will; it is enough for me to . . . consider myself authorized
to make no other use of it than with regard to the moral law which determines
its reality, that is, only a practical use. (5:55–6)

Insofar as the categories apply to the free self, they apply only for practical
use. One cannot have any theoretically warranted knowledge about the
causality of a free being, nor can one put the knowledge that one is free
to use in any theoretical justifications of further claims.
Second, one of the important uses of the categories is to establish the
real possibility of an object. To show that something is a real possibility,
one must show that it is a possible object of experience, and this is done by
showing that the categories and forms of intuition apply to it. In the first
Critique, Kant considers the possibility of something that cannot be an ob-
ject of possible experience for human beings because it does not conform
to the form of human intuition. About such an “object,” Kant says,

If we suppose an object of a non-sensible intuition to be given, we can indeed

represent it through all the predicates which are implied in the presupposition
that it has none of the characteristics proper to sensible intuition; that it is not
extended or in space, that its duration is not in time, that no change (succession
of determinations in time) is to be met with in it, etc. But there is no proper
28 The Problem

knowledge if I thus merely indicate what the intuition of an object is not, without
being able to say what it is that is contained in the intuition. For I have not then
shown that the object which I am thinking through my pure concept is even so
much as possible. . . . To such a something not a single one of all the categories
could be applied. (B149)

Here Kant considers precisely the situation in which he finds himself in

his resolution to the third antinomy, where he supposes something that is
free and hence not sensible. In this passage, he suggests first that one is en-
titled to make various negative claims about such a free subject. That enti-
tlement is important for his future accounts of human freedom because it
allows him to deny to the free will any properties that depend on space and
time. Here Kant also denies that the categories can apply to such an ob-
ject. What is important to recognize, however, is that the reason Kant gives
for this restriction is that “save in so far as empirical intuition provides
the instance to which I apply [the category], I do not know whether there
can be anything that corresponds to such a form of thought” (B149). The
problem with applying the categories is that their application depends
upon the reality of the object to which they are applied. In the case of
freedom, one has no theoretical knowledge of its reality, and hence one
cannot put the categories to any theoretical use with respect to the free
self. However, once one establishes the reality of freedom on practical
grounds, one applies the categories – the only way in which one can
think of a thing – in one’s practical use of the concept of freedom. In this
sense, one thinks of freedom as the ground or cause of one’s actions.
Third, Kant is suspicious of the application of the categories to free-
dom because knowledge of objects is always knowledge of objects under
schematized categories, and these cannot be applied in any sense to a free

We have seen that concepts are altogether impossible, and can have no mean-
ing, if no object is given for them, or at least for the elements of which
they are composed. . . . We have already proved that the only manner in
which objects can be given to us is by modification of our sensibility. . . .
The conditions of sensibility [thus] constitute the universal condition un-
der which alone the category can be applied to any object. This formal and
pure condition of sensibility to which the employment of the concept of
understanding is restricted, we shall entitle the schema of the concept. (A139–40/

In the first Critique, Kant emphasizes that the categories can be applied
in a way that is intelligible for theoretical reason only when they are
schematized according to the intuitions of time and, for external objects,
Asymmetry in Kant’s Conception of Freedom 29

space. Whenever one seeks to picture what it means for something to be

the cause of some other thing, one can imagine this category only under
the conditions of time. But time does not apply to the free agent, which
is not a part of experience. Hence the categories, insofar as they can be
imagined by human beings, cannot apply to free agents.22
Nonetheless, one can apply the categories from the standpoint of prac-
tical reason, where they take on purely practical significance. Thus one
says that the free agent “causes” some appearance, not in the sense that
a free act precedes and is necessarily succeeded by that appearance but
in the sense that moral responsibility for the appearance is imputed to
the agent. Likewise, a free agent is a “substance” not in the sense that it
persists through time but in the sense that the appearances of the past,
present, and future are imputed to a single agent. In this context, it is
appropriate to think of the categories as applying to free agents in a way
that is analogous to the way schematized categories apply to objects in the
world. One cannot strictly explain the nature of that agent, but one can un-
derstand the relationship between it and its appearance by analogy with
relationships in experience. Kant explains this in the Critique of Judgment:

Analogy . . . is the identity of the relation between bases and consequences (causes
and effects) insofar as it is present despite what difference in kind there is between
the things themselves. . . . Though I can conceive of the causality of the supreme
world cause when I compare its purposive products in the world with the works
of art of man, by analogy with an understanding, I cannot by analogy infer that it
has these [same] properties; for in this case the principle that authorizes such an
inference is just what is lacking, i.e., we do not have paritas rationis for including
the supreme being in one and the same general kind as man (as regards their
respective causalities). The causality of world beings (which includes the causality
through understanding) is always conditioned by the sensible, [and so] cannot be
transferred to a being that has no generic concept in common with them except
that of a thing as such. (5:464fn)

Here Kant explicitly allows that the relation of causality can be applied
even to the relationship between something noumenal (the supreme
world cause) and the empirical world. However, this relation is applied
without giving knowledge of what is noumenal. It merely forms the basis
of an analogy. Similarly, Kant can allow, once the reality of the free self
has been established on practical grounds, the use of the relation of
causality to make analogical statements about freedom, but these must
be understood as merely analogical and put to purely practical use.
Kant’s restriction of the scope of knowledge of freedom has two impor-
tant implications for his account of the relationship between freedom and
30 The Problem

anthropology. First, that account need not take the form of an explanation
of freedom. Kant has theoretical reasons for claiming that it is impossible
for human beings to understand the nature of freedom and how a free
self underlies an empirical one. His account must show that moral anthro-
pology is consistent with freedom, but it need not explain how freedom
itself is possible. Second, his response cannot make use of claims about
freedom in theoretical arguments, because these claims have no theo-
retical warrant. Kant cannot apply the categories in their schematized
form to free agents. For that matter, Kant cannot apply to free agents any
concepts that make use of the intuitions of space and time. If he does
make statements that seem to apply the schematized categories, his argu-
ment must work when these are interpreted analogically.23 Moreover, he
cannot make use of the categories for any theoretical knowledge of the
self. Kant is perfectly free to discuss freedom in terms of the categories
because we cannot think of it in any other way, but such discussion must
be purely practical. It must be limited to pointing out the conditions of
possibility of moral obligation.

5. Conclusion
In this chapter, I presented Kant’s arguments for human freedom, argu-
ments based on the fact of moral responsibility. I have shown that the
freedom established on the basis of these arguments must be considered
prior to one’s appearances in the world, whether one understands this
relationship to hold between two objects or between two perspectives.
I have further shown why Kant takes this asymmetry to both imply and
allow for a stronger conception of freedom than is available from soft de-
terminist accounts. This conception of freedom raises several problems,
but the one on which I focus is that it seems to preclude any robust moral
anthropology. If the asymmetry between nature and freedom is absolute,
such that freedom can influence nature but never vice versa, as I have
indicated here, then there seems to be no room for observations about
empirical influences on the free will. The relation between freedom and
nature allows influence in only one direction. If moral anthropology re-
quires mutual influence, there seems to be no way out of the dilemma.
In the next two chapters, I take up the question of just what sorts of influ-
ences anthropology does investigate. I show that the influences are both
empirical and morally relevant, and thus that there is at least a prima
facie problem with Kant’s mature moral theory. Chapters 5 and 6 will
then show how Kant can solve that problem.

Anthropology as an Empirical Science

In the previous chapter I showed that Kant’s theory of freedom implies

an asymmetry according to which freedom is prior to nature. On one
account, this means that the free self can influence but cannot be influ-
enced by the empirical world. To interpret in another way, one might say
that the empirical perspective on the self cannot play a fundamental role
within the practical perspective, whereas the practical perspective is nec-
essary in order to complete the empirical one. Whichever interpretation
one prefers, one would not expect to find Kant outlining empirical helps
and hindrances to moral progress. Such empirical influences on moral-
ity represent a reversal of the required asymmetry. However, throughout
Kant’s mature moral philosophy, one finds references to apparent in-
fluences on morality that seem to be empirical. Among the helps and
hindrances to which Kant refers are beliefs such as the practical postu-
lates (5:122–35, 452–3, 472), belief in grace (6:44–78), and even belief
that virtue is rewarded in one’s life (6:216, 474); feelings such as sympathy
(6:457), respect (5:71–89, 4:400, 6:39), and love (6:399, 8:337–8); social
institutions such as polite society (6:473–4, 7: 151f., 25:502–5, 1455), a
peaceful republic (8:375), and churches (6:93–202); the beautiful or sub-
lime (5:268–9, 299, 354–6); and moral education (5:151f., 6:474f., 9:480–
99). A great deal could be said about each of these influences. Some are
more clearly empirical than others are. Some are more clearly influences
on morality than others are. All of them potentially raise problems for
Kant’s theory of freedom.
The problem outlined in the Introduction, and my eventual solution,
applies to most of the helps and hindrances discussed in Kant’s moral phi-
losophy. But the explicit focus of my attention is the way in which Kant’s
32 The Problem

anthropology presents empirical helps and hindrances to having a good

will. In this chapter and the next, I show the extent to which Kant’s an-
thropology satisfies the description of anthropology that Schleiermacher
criticizes. That is, Kant’s anthropology is at once empirical and morally
relevant. In this chapter, I show that anthropology is empirical. In the
next, I show that it is morally relevant.
One of the difficulties in offering an analysis of Kant’s anthropology
is that there remains disagreement about where that anthropology is to
be found. There can be no doubt that the published Anthropology from
a Pragmatic Point of View and the lectures on anthropology should be
considered anthropology for Kant, although even in these cases there
is disagreement about the relation of this anthropology to the moral an-
thropology referred to in the Groundwork and Metaphysics of Morals.1 Many
also include as “anthropological” works such as Religion within the Bound-
aries of Mere Reason, parts of the Critique of Judgment, and many of Kant’s
shorter essays, especially those dealing with history (cf. Anderson-Gold
2001, Munzel 1999, and Louden 2000). My account focuses on the pub-
lished Anthropology and related lectures because it is this subject matter
more than any other that bothered Schleiermacher, and this is the most
uncontroversially anthropological of Kant’s works. This focus should not
be interpreted as a denial of the anthropological nature of Kant’s other
works, however, and in the course of my discussion I occasionally point
out connections between his explicitly anthropological works and his
other writings.

There are two important dimensions to the question of whether anthro-

pology is empirical, which I take up in turn. First, one can ask whether
anthropology is empirical in its method. When one undertakes the study
of anthropology, does one investigate empirically? Second, one can ask
whether the subject matter of anthropology is empirical. Is the anthro-
pos that is studied by anthropology the human as a free, noumenal
agent, or humans as observable in the world? And insofar as anthro-
pology finds influences on human beings, are these purely empirical
The status of both the method and the subject matter is important
in the context of Kant‘s account of freedom, and the two issues are in-
timately related. Insofar as moral anthropology is empirical in method,
it suggests that there is a perspective within which human beings must
be considered objects of empirical study but at the same time as morally
responsible. Moreover, if it can be shown that anthropology is empirical
Anthropology as an Empirical Science 33

in its method, this proof will provide an important piece of evidence that
the subject matter is empirical. Whether or not the subject matter is em-
pirical is important both to highlight the problem of moral anthropology
for Kant and to move toward a solution to this problem. The problem
becomes clearer because if empirical anthropology finds empirical helps
and hindrances, and if these are also moral helps and hindrances (which
will be discussed in the next chapter), then there is a good prima facie
case that this sort of anthropology is incompatible with Kant’s account
of freedom. At the same time, the fact that Kant’s anthropology is pri-
marily related to the self as observed, rather than to the free noumenal
self, may provide a way to escape this apparent incompatibility. There
is no problem with empirical influences’ affecting the observed self. To
show how this limitation helps solve the dilemma, it will be necessary to
explain how the observed and free selves (or the self considered as free
and as observed) relate to each other. I will take up this task in Chapter 5.
At this point it is enough to mention briefly that by drawing attention to
the observed self, anthropology may begin to point toward Kant’s overall
account of anthropology and freedom.

1. Kant’s Conception of His Empirical Anthropology

That Kant’s anthropology is empirical is the most consistent claim that
Kant makes about his work in anthropology, persisting through all his
lectures on it and enduring in his published work from his earliest
works to his published Anthropology. As far back as 1755, in his Allgemeine
Naturgeschichte, Kant says, “consciousness and the senses should teach us
what man is” (1:366). In the Groundwork, he distinguishes pure morality,
which is the rational part of ethics, from “practical anthropology,” which
is the “empirical part” (4:388). In the Metaphysics of Morals, Kant explains
that in anthropology, to which a metaphysics of morals can be applied,
“we shall often have to take as our object the particular nature of human
beings, which is cognized only by experience” (6:216–17, emphasis added).
And in the Anthropology itself Kant makes clear that anthropological
investigation is a matter of empirical observation, not a priori theorizing.
Observations about the causes of memory that are not put to any use are
“speculative theorizing” that “is a sheer waste of time.” However, proper
anthropological investigation also consists in “observations,” but only
when one “distinguishes between those observations which have been
found to hinder and those which have been found to promote” the fac-
ulty under investigation (7:119). Both the fruitless and the proper sorts of
34 The Problem

anthropology are empirical. The difference is that proper anthropology

puts empirical observation to use.
The empirical nature of anthropology is also evident in the difficulties
that Kant raises for the prospect of anthropology: (1) one “who notices
that he is being observed . . . will not be able to behave as he really is”;
(2) “if he wants to observe himself . . . he does not observe himself when
his impulses are in action, and when he observes himself, the impulses
are at peace,” so he cannot truly observe his own impulsive behavior;
and (3) “conditions of time and place, when lasting, result in habits
which, it is said, constitute second nature, which makes man’s judgment
of himself more difficult” (7:121, cf. 4:471, 15:660 [Reflexion 1482], and
25:1212). All three difficulties pose a problem for anthropology only be-
cause anthropology is a science that depends upon accurate observations.2
Anthropology is dependent upon good observation precisely because it
is an empirical science. Showing in exactly what sense anthropology is
empirical, however, requires a closer look at Kant’s explanation of the
methods of anthropology. Before turning to this closer look, however,
we must confront one seeming difficulty with the view that all of Kant’s
anthropology is empirical.

2. A Problem: How Can Anthropology Be Empirical and Universal?

Kant’s anthropology is universal. Kant says that anthropology must be-
gin with “universal knowledge” and that the subject of anthropology is
human beings as such. In fact, one of the difficulties for anthropology is
precisely that one might mistakenly include contingent details reflecting
particular times and places as part of the scientific core of anthropology.
This universality does not refer to all anthropological knowledge. Some
of the most well-known and most embarrassing parts of Kant’s anthropol-
ogy are his reflections on women and various races.3 These are meant only
to refer to the gender or race under discussion. Kant also distinguishes
different characteristics of people with different ages or temperaments,
or at different stages in the cultural development of the species. But most
of his anthropological observations are meant as universal claims about
all human beings.
This poses an apparent problem because of Kant’s general account of
knowledge. There, Kant connects universality with necessity and argues
that the only possible basis for universal claims is a priori reasoning. In
the Critique of Pure Reason, he says, “Experience tells us, indeed, what is,
but not that it must necessarily be so, and not otherwise. It therefore
Anthropology as an Empirical Science 35

gives us no true universality” (A1–2, cf. B3–4, A91/B124, A196/B241).

Whereas in the Anthropology Kant seems to want to extend his claims to
all people without exception, his general claims about universality would
make this extension seem “merely fictitious” (A196/B241). One seems
forced to choose between anthropology’s being universal and its being
empirical, when Kant clearly states that it is both.
This tension has not been discussed with reference to the Anthropology
in particular,4 but it has provoked discussion in the context of Kant’s claim
in the Religion that humans are radically evil. In his discussion of radical
evil in Kant’s Theory of Freedom, Henry Allison insists that the proper
Kantian grounds for believing in radical evil in human nature must
be a priori synthetic, although Kant himself does not give the requisite
argument.5 Although this is not specifically an argument about anthro-
pology per se, the discussion of radical evil is often taken rightly as an ex-
ample of Kantian anthropology, and Kant himself claims that it “can only
be proved [by] anthropological research” (6:25). Moreover, Allison’s pri-
mary reason for insisting that radical evil requires a priori justification also
applies to many of the claims of Kant’s Anthropology. Specifically, Allison
points out that Kant regards radical evil as “ ‘rooted in humanity itself’
and, therefore, universal” and asks “what grounds . . . does Kant offer for
this claim?” (Allison 1990: 154). He then rightly points out that Kant
simply “seems to treat it as an unproblematic empirical generalization”
(see 6:32–4) but objects that “the most that this evidence can show is that
evil is widespread, not that there is a universal propensity to it” (Allison
1990: 154). This is exactly the sort of objection that one seems able to
raise against the Anthropology. Moreover, if there is any case in which the
universality of an anthropological claim is crucial, it is the case of radi-
cal evil.6 Radical evil thus provides a useful test case for the relationship
between being empirically based and holding universally.
There are two stages to Allison’s argument that Kant’s ascription to
humans of a universal propensity to radical evil must be a priori. First, he
argues against Kant’s empirical defense of universal radical evil. Second,
he offers his own a priori argument. I take these up in turn.
Allison first objects to Kant’s empirical argument on two grounds:

[1] the most that this [empirical] evidence can show is that evil is widespread, not
that there is a universal propensity to it. [2] Moreover, since Kant insists that this
propensity concerns only the ultimate subjective ground of one’s maxims and is
perfectly compatible with a virtuous empirical character, it is difficult to see what
could conceivably falsify his claim. Consequently, it is difficult to take seriously
the suggestion that it is an empirical generalization. (Allison 1990: 154–5)
36 The Problem

The second problem – that because Kant could not give empirical con-
ditions that would falsify his claim, he cannot mean it seriously – is not
relevant to Kant. It is true that there is no empirical means by which one
can falsify Kant’s claim, but this does not mean that it cannot be empiri-
cally established. The claim cannot be falsified because a good empirical
will is consistent with either a good or an evil intelligible one. One might
look good to all appearances – even to oneself – and yet have hidden
motives that make one evil. But while a good empirical will is consistent
with really being either good or evil, an evil empirical will is consistent
only with genuine evil.7 If one can show the presence of evil in a person’s
empirical will, one can be sure that the person is evil. And showing evil
in the empirical will is something that can be accomplished empirically.
Kant goes so far as to say that “we call a man evil . . . because [his] actions
are of such a nature that we may infer from them the presence in him
of evil maxims” (5:20).8 Although one cannot falsify the claim that one
is evil by showing that one is good, one could undermine support for that
claim by showing that one’s will is empirically good. Unless Kant can show
that one has an evil empirical will, he cannot confidently assert anything
about one’s intelligible will. In this sense he has an empirical negative
test for his claim, though failing that test does not prove that his claim is
false, only that it might be false.
Still, even if one can demonstrate an evil intelligible will from an evil
empirical one, Allison seems right that one could never show on the basis
of empirical criteria that evil or the propensity thereto is a fact of human
nature, universally shared by all. Here, as we have already pointed out
from the first Critique, Kant must agree with Allison at least in a sense. His
claims that human beings are radically evil or have a propensity thereto
cannot have strict universality if they are empirical. But Kant also might
add that there is another sense of “universal” according to which his
claims can hold universally. Moreover, it is crucially important that these
claims not have strict universality. In the first Critique, Kant consistently
connects “strict universality” with “necessity,” going so far as to insist that
these are “inseparable from each other” (B4). For a predicate to hold of
human nature according to strict universality, it must be not only true of all
people but necessarily true of all. But for Kant it is crucial that the propen-
sity to evil not be universal in that sense. It must not be strictly universal
that one have a propensity to evil precisely because this propensity is the
result of one’s choice. Kant is adamant on this point: “this propensity
must itself be considered as morally evil . . . [and hence] as something
Anthropology as an Empirical Science 37

that can be imputed to man” (5:32). Allen Wood points out, in contrast
to Allison,
Since this propensity is one for which we ourselves are culpable, it could not have
derived its existence merely from the finitude of human nature. The moral im-
perfections of our volition are limitations we impose upon ourselves, limitations
for which we ourselves are responsible, and which must be inferred from the use
of our freedom, and not merely from the concept of finite rational volition in
general. (Wood 1970: 223–4)

Allison is surely correct that no empirical investigation can yield a judg-

ment about radical evil that would have strict universality. But Kant’s judg-
ments about radical evil impute that evil to free human beings as their own
fault, and hence as the result of free choice. Because necessity accompanies
strict universality and undermines moral accountability, to argue for strict
universality would undermine the moral imputability of the propensity
to radical evil.
Before moving on to discuss the sense in which Kant’s claim is uni-
versal, it is important to address briefly the second stage of Allison’s ar-
gument, in which Allison claims to offer an a priori argument showing
the propensity for radical evil. Kant would not want an a priori argument
for the propensity of radical evil, because such an argument would imply
that radical evil is a necessary propensity in human beings rather than
a moral fault imputable to free agents. But Allison does not just claim
that such a proof is needed; he claims to give such a proof.9 His deduc-
tion rests on the claim that unless it is possible for human beings to have
propensities to good, it is necessary that they have propensities to evil.
Although some have denied this claim (see Wood 1970: 222–3), I agree
with Allison’s general thrust here, and it is not necessary to contest this
point to show where he goes wrong. Instead, I argue that for Kant it is
possible that human beings have what Allison calls a propensity to good,
and thus not necessary that they have a propensity to evil. Allison identifies
a propensity to good with “a spontaneous preference for the impersonal
requirements of morality over one’s own needs as a rational animal with
a built-in desire for happiness” (Allison 1990: 155). He then claims that if
someone had such a propensity, “the moral incentive would, as a matter
of course, always outweigh the incentive of self-love” (155). This account
of one with a propensity to good is just a description of one who always
does what morality requires. To take one’s moral interest to outweigh all
interests of self-love is what Kant requires for one to be morally good. The
moral law precisely demands that one choose it over self-love, and it is
38 The Problem

through choice that the moral law becomes an incentive that outweighs
self-love.10 Allison claims, however, that if this were the case,
For an agent blessed with such a propensity, there would be no temptation to
adopt maxims that run counter to the law and, therefore, no thought of the law
as constraining. Within the Kantian framework, this means that the law would
not take the form of an imperative and moral requirements would not be viewed
as duties. (Allison 1990: 155)

Here Allison goes too far. The fact that one takes the demands of the
moral law to outweigh incentives of self-love does not in any way remove
the incentives of self-love. And as long as self-love is present, it is a temp-
tation. The fact that one resists that temptation, that one chooses what
is right over what is in one’s interest, does not mean that there is no
temptation to face. And this would not be any different if one always
chose the moral law over self-love. Allison’s a priori argument fails, as
Kant would hope. Still, Allison’s contention that empirical arguments
cannot establish universality is strong. If there is no a priori argument
for radical evil, it seems that Kant should not have claimed that it is

3. Anthropological Universality as Empirical Universality

In fact, however, although Kant must certainly deny that his ascription
of the propensity to evil to the human race is strictly universal, strict uni-
versality is not the only kind of universality there is. In the first Critique,
Kant repeatedly refers to a kind of universality that experience can pro-
vide. Although “experience never confers on its judgments true or strict”
universality, it can confer on them “assumed and comparative universality,
through induction” (B3). Later, he says that “empirical rules . . . can ac-
quire through induction only comparative universality, that is, extensive
applicability” (B124/A91). According to this kind of universality,
We can properly only say, therefore, that, so far as we have hitherto observed,
there is no exception to this or that rule. . . . Empirical universality is only an
arbitrary extension of a validity holding in most cases to one which holds in all,
for instance, in the proposition “all bodies are heavy.” (B 3–4)

The first Critique downplays the role of this kind of empirical universal
judgment in its search for a priori conditions of experience. Still, Kant’s
insistence that this universality is merely “extensive applicability” is only
a contrast to that strict universality according to which “no exception
is allowed as possible” (B4, emphasis added). As his example of heavy
Anthropology as an Empirical Science 39

bodies makes clear, these judgments need not be taken as merely claims
that something is widespread. They hold without exception, but with the
possibility of exception.11 In the Critique of Judgment, Kant even explicitly
distinguishes these two kinds of universality: “Here it is understood that
the universality [Allgemeinheit] is only comparative, so that the rules are
only general [generale] (as all empirical rules are), not universal [universale]”
(5:213). Allgemein is the term generally translated as “universal” in Kant’s
writings and in particular is the term used to describe the “universality” of
radical evil in the Religion (6:29). In the Anthropology as well, Kant is careful
to use the term Generalkenntnis to describe the “universal” knowledge
involved in anthropology (7:120). In both cases, the knowledge at issue is
universal only in the comparative sense. As Kant repeatedly insists, “there
is no cause for exempting anyone from it” (6:25), but this universality is
merely contingent (6:29).

4. The Primary Source of Anthropology: Inner Sense

In support of universal empirical claims, in both the Religion and the
Anthropology, Kant offers several supporting lines of investigation.12
Through arguing in all of these ways, Kant hopes to establish claims
that go legitimately beyond empirical claims about merely widespread
human phenomena and that are therefore generalizations that apply to
all human beings.13 As we will see, this generality need not imply that
Kant takes these claims to be infallible, but if they are correct then they
apply to everyone.
At the beginning of his Anthropology, Kant mentions three different
sorts of observation that can play a role in anthropological investigation.
He says,

[1] Travel is among the means of enlarging the scope of anthropology even if
such knowledge is only acquired by reading books of travel. One must, however,
have gained his knowledge of man through [2] interaction with one’s fellowmen
at home if one wishes to know what to look for abroad in order to increase one’s
range of knowledge of man. Without such a program (which presupposes [3] the
knowledge of man), the anthropology of the citizen of the world will remain very
limited. (7:120)

Kant here distinguishes three levels of observation. Travel or travel books,

through which one is exposed to a wide range of peoples, are the final
stage of anthropology. Prior to that, Kant claims that one must study one’s
fellow human beings locally. Even prior to this investigation of others,
40 The Problem

one must acquire a certain “knowledge of man,” a general knowledge

that investigation of others refines.
Kant makes the source of this general knowledge clearer later in the
Anthropology, insisting on the importance of “knowledge of man through
inner experience, according to which he judges others” (7:143, emphasis
added). This emphasis on inner experience as a basis for anthropol-
ogy fits well with Kant’s account of the difficulties one encounters in
attempting to do anthropological research. Of three difficulties that he
mentions, two arise when one is “observing oneself” or seeking to issue a
“judgment about oneself” (7:121).14 Later, Kant points out that “on the
whole, knowledge of man through inner experience, according to which
he judges others, is of great importance” (7:143). In an early draft of the
Anthropology, Kant goes so far as to say that “anthropology is supplied by
the inner sense with content” (7:398).
In the lectures on anthropology, this emphasis on inner sense is even
clearer.15 There Kant describes anthropology as “the knowledge of people
as objects of inner sense” (25:473). In one lecture, Kant begins with a pas-
sage very similar to his treatment of travel in the published Anthropology:

One who is well-traveled, and has come to know many people . . . cannot say
that he knows the person, for he has only known the condition, which is very
changeable. If, on the other hand, I know humanity, it must fit all kinds of people.
Anthropology is thus not a description of people, but of the nature of people.

Kant immediately follows this up with an explanation of the source of the

knowledge of humanity that provides the basis for further anthropologi-
cal investigation:

The knowledge of humanity is at the same time my knowledge. . . . Hence we

must study ourselves and because we want to apply this to others, we must study
humanity. (25:471)

Although anthropological knowledge is knowledge of all people, hence

of others as well as oneself, it properly begins with introspection. Only
with respect to oneself does one have an inner experience, by means of
which motives, inclinations, and desires can be observed. Hence one
is equipped to apply anthropology to others only once one has self-
Appealing to inner sense is important because it provides a means
by which judgments can be made that are both empirical and universal.
Claims based on inner sense are empirical, for Kant, because observations
Anthropology as an Empirical Science 41

of oneself are empirical observations. This insistence on the empirical

nature of self-knowledge is one of the most important and controversial
claims in his Critique of Pure Reason. In the second edition in particular,
Kant goes out of his way to insist that “we know our own subject only as
appearance, not as it is in itself” (B156, cf. B69). In the preface to that
edition, he explains that “through inner experience I am conscious of my ex-
istence in time” (Bxl, fn). But existence in time is necessarily the existence
of an appearance only, because things in themselves are not dependent
upon the intuitions of space or time. In the “Refutation of Idealism”
and the B edition of the “Paralogisms,” Kant further reiterates that self-
knowledge through inner experience is merely empirical knowledge of
oneself (see B275–9, B422, and B429f.).
Although judgments based on inner sense are empirical, however, they
can assume a certain kind of universality. Inner sense is generically hu-
man, and when one observes the structure of one’s cognition, feelings,
and desires, one observes a structure that is likely to be shared by others.
This is not to say that the mere form of inner sense implies the details
of one’s inner experience. If one’s reflections were determined in that
way, they would be necessary, synthetic a priori judgments, whereas they
are in fact contingent, empirical, and still “universal.” But unless one has
a reason to doubt that others have in common a structure in the con-
text of which they experience themselves, one can explore that structure
within oneself as a means of discerning general truths about humanity.
In particular, the investigation of inner sense gives one the means of un-
derstanding the motives, desires, and other aspects of inner experience
that underlie the actions of others.
The problem, of course, is that one does not observe a generic human
inner sense but only one’s own. Thus there is a danger of mistaking
contingent facts about one’s particular inner experience for universal
facts about humans in general. Kant is well aware of this difficulty. In
the Anthropology, two of his three basic warnings about anthropological
investigation relate directly to the problem of extending particular facts
about oneself to all people. The first of these points out limits to the
depth of self-observation. The second points out that “conditions of time
and place, when lasting, result in habits which, it is said, constitute second
nature, which makes man’s judgment of himself more difficult” (7:121).
This is a “difficulty inherent in human nature itself,” and it seems to
threaten the possibility of any universal anthropology.
Before looking at how Kant mitigates against these difficulties, it is im-
portant to remember that for anthropology to be universal, it need not
42 The Problem

establish anthropological claims on grounds that cannot be doubted, but

only on grounds that one reasonably claims to hold for all people. Good
judgment is needed to distinguish those aspects of one’s inner experi-
ence that can rightly be applied to others and those that are peculiar to
oneself. One can always make mistakes in one’s judgments about what is
truly universal. Luckily, there are some aids to correcting these sorts of

5. Secondary Sources of Anthropology: Corrections to Inner Sense

First, there is one’s experience of others. Self-observation is only the
first stage in anthropology. The next sphere of investigation is the cir-
cle of one’s acquaintances. Although Kant suggests that self-observation
is the primary method of anthropology, he is clear that anthropology
also involves interaction with and observation of others. Self-centered
observation provides a basis for further observations, but it is not in itself
sufficient because of the problem of identifying particular characteristics
of one’s second nature with universal human nature. Thus Kant writes
that “one must have gained his knowledge of man through interaction
with one’s fellow man at home” (7:120). And in his lectures, after asking,
“Where will we then get to know the world, without traveling around it?”
he answers,

1) The observation of people that are around us, and a sharp reflection, can
replace for us extensive experience and widely surpass whatever an unthinking
traveler gets. . . .
2) Civil association. What is essential here is the attention to the human dis-
positions that often appear in many forms. (25:734)

It is essential to recognize that when Kant points out the value of observ-
ing others for anthropology, he insists that a “sharp reflection” accom-
pany this observation. Anthropological knowledge from self-observation
comes first, especially insofar as it leads to understanding the way in which
inner dispositions are related to external forms in which they appear. But
it should be supplemented with observations of one’s fellow citizens.
Even the observation of one’s fellows, however, may not sufficiently
broaden one’s horizons. At least two potential sources of error remain.
First, the observation of one’s fellows at home may be insufficient to dis-
tinguish human nature from a second nature that is common to a group.
The scope of investigation calls for expansion to ensure that men and
women around the world share the characteristics one identifies in one’s
Anthropology as an Empirical Science 43

fellow men at home. Second, one’s observations of others ultimately de-

pend on interpretations of their actions and expressions, interpretations
that are rooted in one’s observations of oneself. Thus one is in danger of
imposing on others what properly belongs only to oneself. One might, for
instance, interpret expressions of extreme feeling as a sign of loss of self-
control even when they are not, simply because if one were to express such
feeling oneself it would involve such a loss. Or one might naı̈vely suppose
that another’s smiles and affection are genuine acts of kindness, sim-
ply because one lacks the experience of certain emotions or of being
duplicitous. To further refine the judgment necessary for anthropology,
one must look beyond one’s fellow citizens to observations that are both
broader and deeper.
One can broaden one’s horizons through travel or reading books of
travel. Throughout his lectures Kant insists that travel is neither a nec-
essary nor a sufficient condition of good anthropology, but he does rec-
ognize it as a useful means for expanding and refining one’s knowledge.
Thus he allows that “travel is among the means of enlarging the scope
of anthropology, even if such travel is only acquired by reading books of
travel” (7:120). In addition, one can both broaden and deepen one’s obser-
vations through literature. In answering the question of the sources of
anthropology in his lectures, Kant adds to observations of others at home
and abroad a third source of anthropological insight: “3) Plays, novels,
histories, and various biographies” (25:734). This source is in fact Kant’s
favorite and most consistent external source of anthropology. In part this
is because he lacked the resources to travel, but Kant also values literature
because a well-written book incorporates both the author’s observations
of others and his or her inner reflection on those observations, such that
one gets an account rich in anthropological insight. Thus Kant says in a
lecture in 1775,

From where does anthropology spring? Through collections of many observations

of people [in] various authors that have had a sharp knowledge of man, e.g.
Shakespeare’s theatrical works, the English Spectator, and Montaigne’s essays
together with his life. (25:472)

In his published Anthropology, Kant defends the use of literature for an-
thropology, saying,

Such characters as are created, for instance, by Richardson and Molière, still had
to be derived in their basic traits from observance of the actual doings of man.
Exaggerated as these traits may be in degree, they must still conform to human
nature. (7:121)
44 The Problem

Books and plays present the world interpreted according to the inner
experiences of their authors. Thus one is exposed not only to a broader
extent of experience (in Kant’s case, being able to explore parts of the
world he would never be able to visit) but also to an enriching intensive
perspective on human affairs. When one reads Richardson, Shakespeare,
or Montaigne, one comes to see more clearly what aspects of one’s own in-
ner and outer experience are truly universal. This refinement is especially
important for those whose judgment is immature, who are prone to mis-
take particular details of themselves for universal phenomena. By being
exposed to good literature, one comes to see both the variety of human
experience and the universal aspects of human nature. Even for someone
(Kant, perhaps?) with a relatively cultivated anthropological judgment,
reading Jane Austen, Charlotte and Emily Brontë, and George Eliot can
help refine that judgment. One may find that what seemed universal is
more limited than one imagined. If one goes on to read the Mahabharata
and the Tale of Genji and encounters literature from people all over the
world, one may again rethink one’s judgments. This broadening of one’s
horizons need not lead one to renounce all universal anthropological
claims, but it may require adjusting one’s views about which claims are
truly universal.16

What, then, are the sources of anthropology? The primary source is one’s
own inner experience, on the basis of which one can make reflective
judgments about the actions of others. Aids to the cultivation of this
judgment include not only direct engagement with and observations of
others but also the rich anthropological data provided by books and plays.
All of these aids to anthropological judgment, it is worth pointing out, are
empirical. One does not give a priori arguments based on the concept of a
human being or the necessary structure of experience. One’s arguments
are based on experience, enriched by a reflective judgment that is itself
shaped by experience. So anthropology is deeply empirical.
But then just how universal is anthropology? It is properly universal,
although not in the strict sense, and not infallibly. When Kant says, for
instance, “what the affect of anger does not accomplish quickly will not
be accomplished at all” (7:252), he takes this to describe a universal fact
about the feeling that he characterizes as the affect of anger. Human
beings, he supposes, are not capable of sustaining this particular sort of
feeling for long periods of time. If he is correct, then he is making a
true and universal, but not a necessary, claim. Is it possible that he is
wrong? Of course. In fact, Kant changes several of his anthropological
Anthropology as an Empirical Science 45

views over the course of his lectures.17 And he emphatically denounces

various politicians of his time for holding false anthropological beliefs.
In Perpetual Peace, he criticizes “moralizing politicians,” who

try to cover up political principles which are contrary to right, under the pretext
that human nature is incapable of attaining the good which reason prescribes as
an idea. . . . They may boast to know men (which is certainly to be expected, since
they have to do with so many of them), although they certainly do not know
man and what can be made of him, for this requires a higher anthropological
vantage-point. (8:373)

Although Kant does not detail the higher anthropological vantage point
to which these politicians should turn to refine their false anthropolog-
ical claims,18 he does make clear that interaction with and observations
of many people do not guarantee that one’s anthropological judgments
will be true. People can always make mistakes. And given that certain er-
rors, especially errors about human capacities, can be useful for justifying
wickedness, one might even expect mistakes to turn up in anthropology.
But in addition to pointing out that these anthropological facts give no
excuse for error, which Kant does in Perpetual Peace and throughout his
moral philosophy, it is important to fight bad anthropology with good
anthropology. And that is what Kant does.19
The fact that a claim holds for all people does not mean that one
affirms that claim with absolute certainty, nor that it holds necessarily.
The danger with empirical investigations of any sort, regardless of how
universal they are, is that one might be wrong. Even to claim that some
particular person has some characteristic at some particular time might
be wrong. So it surely might be wrong that every person has a character-
istic at every time. But one still can make reasoned judgments of these
sorts. One may conclude, on the basis of a refined judgment about what
sorts of things are due to one’s nature, that such-and-such a characteris-
tic is universal. One’s judgment can be refined by more experience, and
as it is one may change some of one’s anthropological claims. None of
this means that one is incapable of ever being right. And until one has a
reason to doubt one’s judgment, one does best to trust it.
This is especially true given the scope of the application of anthro-
pology. One is most likely to use anthropology in the context of those
with whom one is most familiar, and hence those who have primarily
formed the basis of one’s anthropological claims. Errors are less likely
to have serious practical implications than if one’s anthropology were
applied more broadly. This assumes that one gives close attention in
46 The Problem

one’s anthropological investigations to all of those with whom one has

contact. If one systematically excludes from anthropological investigation
a class of people with whom one regularly deals, one is not only in dan-
ger of being in error, but these errors are likely to have serious practical

6. The Empirical Objects of Anthropological Study

Thus far I have shown only that the discipline of anthropology is empir-
ical in method, not that it studies empirical helps and hindrances. It is
necessary to point out that anthropology is not only an empirical science
but also a science with an empirical subject matter. One might think that
there is no reason an empirical science would have to highlight empirical
helps and hindrances. Perhaps it could be an anthropological fact about
human beings that they must believe the postulates of practical reason to
sustain commitment to obeying the moral law, but the postulates could
be considered an influence of reason itself rather than a purely empirical
influence.21 Or one might argue that it is an empirical fact about human
beings that one’s happiness is promoted by acting on the basis of the
principle that to achieve an end one must pursue the means to that end.
But this principle could be a purely rational influence on happiness.
For Kant, however, the fact that a science is empirical implies that its
subject matter is also empirical. One of the most important implications
of the first Critique is the general point that experience is always only
of appearance and never of things in themselves, and Kant specifically
applies this to knowledge of the self through inner sense (see B155).
Thus all empirical knowledge can take only objects of experience as its
objects. The dialectic in the first Critique shows that even the use of pure
theoretical reason, which is based on the a priori conditions of the possi-
bility of experience, errs in seeking to discern the existence or properties
of objects that transcend experience. To attempt any such extension of
knowledge on empirical grounds would be an even greater error.22

7. Conclusion
Anthropology, for Kant, is an empirical study of human beings. It makes
claims that are contingent but universal, in the sense that they apply to
all human beings without exception. Moreover, anthropological claims
study relations among empirical objects, because this is the proper subject
matter of an empirical discipline. Thus insofar as anthropology studies
Anthropology as an Empirical Science 47

influences on human beings and human faculties, it studies only empiri-

cal influences. In that context, and given the asymmetry between nature
and freedom discussed in Chapter 1, one might expect Kant to restrict
the scope of anthropology. In particular, because there can be no em-
pirical influences on one’s free and morally responsible self, one might
think that anthropology would be limited to spheres that do not involve
specifically moral development. But as we will see in the next chapter,
Kant does not limit his anthropology in this way.

The Moral Importance of Kant’s

“Pragmatic” Anthropology

In the preceding chapter, we saw that anthropology for Kant is an empir-

ical science with empirical subject matter. Not only is its method empir-
ical, but the helps and hindrances that anthropology finds are empirical
helps and hindrances. This anthropology will seem to raise problems for
Kant’s account of freedom if it is put to moral use, such that empiri-
cal helps and hindrances aid moral development. If anthropology were
merely an empirical description of human beings or merely pointed out
empirical influences on faculties like memory, it would not even appear
to conflict with the asymmetrical account of the relation between nature
and freedom outlined in Chapter 1. If Schleiermacher’s criticism is to
make contact with Kant’s anthropology, that anthropology must be not
only empirical but also morally relevant. Thus this chapter takes up the
issue of the extent to which Kant’s “pragmatic” anthropology is a moral
Before showing that Kant’s anthropology is moral, it is important to
clear up a potential misconception about what moral anthropology in-
volves. In particular, moral anthropology in Kant’s mature moral theory
is not the kind of anthropology needed to apply the categorical imperative
to concrete human situations.1 That is, moral anthropology does not have
the kind of role indicated by Kant’s claim in the Groundwork that moral
“laws require a power of judgment sharpened by experience . . . in order
to distinguish in what cases they are applicable” (4:389). If this were the
only role that anthropology played in acting well, anthropology would not
raise any prima facie problems for Kant’s account of freedom, because
such an anthropology need not consider moral agency as itself susceptible
to empirical influence in order to apply the moral law to human beings.
Kant’s “Pragmatic” Anthropology 49

To apply the general moral requirement to promote the happiness of oth-

ers, one might need to know something specific about what makes people
happy. And to know that false promises are immoral, one must know that
human beings are capable of basing trust on promises, incapable of read-
ing one another’s minds, and so on. But none of these anthropological
claims need have any bearing on the transcendental freedom of human
Kant ascribes a different role to moral anthropology in his mature
moral theory. Even in the Groundwork, he describes moral anthropol-
ogy as important to “gain for [the moral law] access to the human
will” (4:389).2 In the Metaphysics of Morals, this role is more clearly de-
fined: “moral anthropology . . . would deal only with the subjective con-
ditions in human nature that help or hinder in fulfilling the laws of a
metaphysics of morals. It would deal with the development, spreading,
and strengthening of moral principles” (6:217). And in his lectures on
ethics, Kant explains the distinction between practical philosophy and
anthropology by pointing out that “consideration of rules [in practical
philosophy] is useless if one cannot make men ready to follow them
[which depends on anthropological insight], so these two sciences are
connected” (27:244).
In Kant’s mature moral philosophy, applying moral principles is part
of the metaphysics of morals. In his Metaphysics of Morals, Kant even goes
so far as to offer various “casuistical questions,” including, for example,
“Can one at least justify . . . a use of wine bordering on intoxication, since
it enlivens the company’s conversation and in doing so makes them speak
more freely?” (6:428). Anthropology is very important for these sorts of
applications of the moral law, but anthropology that merely serves to apply
the moral law is not “moral anthropology.” Moral anthropology deals
specifically with subjective conditions that spread and strengthen, rather
than merely specify, moral principles. That is, moral anthropology studies
the influences that help or hinder one in adopting and acting on moral
principles. In a certain sense, of course, this anthropology still involves the
application of the moral law to particular human situations, but its scope
is greatly restricted. The application with which moral anthropology is
concerned is the working out of the imperfect duty to develop one’s
moral perfection.

This narrower task of laying out helps and hindrances to moral choice is
particularly problematic from the standpoint of Kant’s theory of freedom.
And this narrower role is prominent in Kant’s Anthropology from a Pragmatic
50 The Problem

Point of View. Unfortunately, whereas Kant is very explicit about the fact
that his Anthropology is empirical (as we saw in the previous chapter),
he does not as clearly indicate that it is moral. In fact, there are sug-
gestions that might lead one to think that Kant deliberately avoids call-
ing his anthropology moral. Reinhard Brandt points out that whenever
Kant discusses anthropology in his purely moral works (whether lectures
or published work), he designates it “moral” or “practical” anthropol-
ogy (Brandt 1997). Yet in the published Anthropology and throughout
his lectures on anthropology, he never calls the subject “moral” and
only rarely refers to it as “practical.”3 His use of the term “pragmatic”
(pragmatisch) to describe anthropology seems to distance Kant’s specif-
ically anthropological work from the anthropology to which he refers
in his moral philosophy. The title of Kant’s published work, Anthropol-
ogy from a Pragmatic Point of View (emphasis added), seems to reinforce
In this context, one would expect Kant’s anthropology to focus on
that sphere of practical reasoning which he elsewhere calls “pragmatic.”
And one might expect this sphere to be distinct from the realm of the
moral. But Kant’s use of the term pragmatic varies throughout his work.
This term has at least three different meanings in Kant’s moral phi-
losophy, and aspects of the Anthropology reflect each of these different
senses. “Pragmatic” reasoning can involve (1) one’s happiness, (2) the
whole sphere of the practical, and/or (3) the use of others to achieve one’s
ends.5 Of these, only the first excludes moral elements from its consid-
eration, while the second necessarily includes them. The third involves
morality, though more obliquely. In the rest of this chapter I first show
that Kant uses the term pragmatic in all of these senses, and hence that
the term need not be taken to exclude morality. Then I point out three
examples of moral anthropology in Anthropology from a Pragmatic Point of
View and the related lectures. Finally, I conclude with a brief remark on
the systematic place of moral anthropology in Kant’s anthropology as a

1. Three Conceptions of Pragmatic in Kant’s Philosophy

A. Pragmatic as Relating to Happiness
The term pragmatic can refer to the sphere of practical reasoning related
to the promotion of happiness. In this context, pragmatic reasoning
is distinguished from technical and moral types of reasoning. In the
Kant’s “Pragmatic” Anthropology 51

Groundwork, Kant distinguishes three types of imperatives according to

corresponding spheres of practical consideration:

The first kind of imperatives might also be called technical (belonging to art), the
second kind pragmatic (belonging to welfare), the third kind moral (belonging
to free conduct as such, i.e., to morals). (4:416–17)6

In this context, Kant contrasts pragmatic and moral imperatives. Whereas

moral imperatives are concerned with the necessary laws governing an
autonomous being, pragmatic ones are merely “counsels of prudence”
that help one promote happiness. This is reiterated in the Metaphysics
of Morals, where Kant distinguishes “a human being’s duty himself to
develop and increase his natural perfection, that is for a pragmatic purpose”
with “a human being’s duty to himself to increase his moral perfection,
that is, for a moral purpose only” (6:444, 446). In fact, this distinction
between technical, pragmatic, and moral spheres is a persistent theme in
Kant’s practical philosophy (cf. 5:12n, 27:255–6).
If the Anthropology were presented from a “pragmatic point of view”
in this sense, it would primarily offer investigations of human beings
that are conducive to achieving happiness. One would expect to find
investigations of what sorts of things make people happy, what the most
reliable means of achieving those things are, and how to improve one’s
future capacities for happiness. One does find these sorts of concerns
throughout the Anthropology. To name just one example, Kant’s discussion
of the “highest physical good” offers suggestions about the importance
of meaningful work to a happy life: “The greatest sensuous pleasure . . .
is found under healthy conditions in resting after work” (7:276).7
If this were the only or primary consideration of anthropology, it would
not raise any problems for Kant’s theory of freedom. There might still
seem to be a problem, in that human beings would be considered objects
of empirical investigation and at the same time free in the sense that
they can act on counsels of prudence. But this weak sort of prudential
freedom is not, on Kant’s understanding, incompatible with ultimate
determination by empirical influences. In particular, while it is crucial
for Kant’s moral theory that one always be absolutely free to do what is
morally required, it is not essential that one always be capable of doing
what is in one’s best interest. The freedom involved in prudence could
be a soft determinist freedom that would eliminate the problems raised
in the Introduction.8
52 The Problem

B. Pragmatic as Practical
A second use of pragmatic contrasts the term not with technical and moral
but with physiological or theoretical reasoning. In this context, pragmatic
considerations include not only those related to happiness but moral
and technical ones as well. Anything that is practical is pragmatic. In
their introduction to Kant’s lectures on anthropology, Brandt and Stark
explain this sense of pragmatic:

The overarching concept “pragmatic” ends up being for Kant an approximate

frame for the triad of technical skill (for any ends), pragmatic prudence (for
our actual end in association with others) and wisdom (for moral ends). . . . The
anthropological doctrine of prudence . . . includes the whole territory of possible
praxis. Pragmatic is on the one hand a general concept, whose opposite concept is
“speculative”; on the other, it is one of three specific concepts, between “technical”
and “moral.” (25:xvi–xvii.)9

Brandt and Stark defend this interpretation of pragmatic by drawing on

Kant’s letter to Marcus Herz, in which Kant contrasts his anthropology
with that of Ernst Platner:

This winter I am giving, for the second time, a lecture course on anthropology,
a subject that I now intend to make into a proper academic discipline. But my
plan is quite unique. I intend to disclose the bases of all sciences, the principles
of morality, of skill, of human intercourse, of the method of molding and gov-
erning men, and thus of everything that pertains to the practical. (X:146, emphasis

Kant wrote this letter in 1773, when he first began teaching his anthro-
pology lectures. Later, in a Reflexion written sometime between 1776 and
1789, Kant reiterates this general notion of pragmatic: “The historical
kind of teaching is pragmatic, when it . . . is not merely for the school,
but also for the world or ethics” (Refl. 3376, 16:804, emphasis added, cf.
The broad focus on all that is practical characterizes Kant’s lectures
in anthropology. Thus in a set of lecture notes taken by Kant’s stu-
dent Pillau in 1777–8, he outlines the “Uses of Anthropology,” of which
one is that “it gives us contact with the subjective principles of all the
sciences. And these subjective principles have a great influence . . . in
morals” (25:734–5).11 In the published Anthropology, Kant is not as ex-
plicit about including moral purposes in his general description of an-
thropology, but his claim that “pragmatic knowledge of man aims at
what man makes, can, or should make of himself as a freely acting
Kant’s “Pragmatic” Anthropology 53

being” (7:119) suggests a broad notion of pragmatic as practical. Aim-

ing at what man “should make” of himself in particular suggests a moral
Further, Kant’s use of the term pragmatic in the published Anthropology
(1798) is specifically designed to contrast his anthropology with “physi-
ological” studies of human beings. Kant explains, “A systematic doctrine
containing our knowledge of man (anthropology) can either be given
from a physiological or a pragmatic point of view, [where] physiological
knowledge of man aims at the investigation of what Nature makes of man”
(7:119). This contrast between a purely physiological anthropology and
his own is exactly the contrast that Kant sought to make, twenty years ear-
lier, between his own pragmatic anthropology and Platner’s. This similar-
ity suggests another reason to interpret the “pragmatic” nature of Kant’s
Anthropology in a broad sense, meaning “everything that pertains to the

C. Pragmatic as Pertaining to the Use of Others

A third use of pragmatic, related to but different from the previous two,
considers pragmatic reasoning to involve the use of others for one’s ends.
This is related to the connection between prudence and pragmatic rea-
soning in the Groundwork and to Kant’s suggestion that pragmatic an-
thropology teaches prudence, or Klugheit (cf. Sturm 1997 and Munzel
1999:162). In the lectures, for instance, Kant says that “we must study
mankind, but not psychologically or speculatively, but rather pragmati-
cally, for all pragmatic instruction is instruction in prudence [Klugheits
Lehren]” (25:471), and that lectures that promote “prudence [are] prag-
matic” (25:1436).12
Kant explains in an important footnote in the Groundwork what he
means by prudence : “The word ‘prudence’ . . . can mean worldly wis-
dom, . . . the skill of someone in influencing others so as to use them
for his own purposes” (4:416n.).13 This sense of prudence as worldly
wisdom is the primary sense of prudence in Kant’s discussions of his
pragmatic anthropology. In his lectures, for instance, immediately af-
ter explaining that anthropology is connected with prudence, Kant says
that “the reason that morals and preaching that are full of admoni-
tions . . . have little effect is the lack of knowledge of man” (25:471–2,
cf. 27:358). Kant’s lectures on anthropology are meant to supply the
knowledge of man necessary so that his students, who will go on to be
doctors, lawyers, teachers, or priests, can affect their patients, judges, stu-
dents, and parishioners. In his later lectures, Kant again emphasizes the
54 The Problem

topic of influence on others in connection with the pragmatic nature of


We must trouble ourselves to form the way of thinking and the capacities of those
people with whom we have to do, so that we are not too hard nor too offensive.
So we are taught anthropology, which shows us how we can use people to our
ends. (25:1436)

In the published Anthropology the same theme emerges. Kant says in his
discussion of the different characters of various national groups:

We are interested only in [what] would permit judgment about what each has to
know about the other, and how each could use the other to its own advantage.

The emphasis in these passages is on the ability to use others for one’s
own ends.
Although this use of others is occasionally connected with promot-
ing what is “to one’s own advantage,” prudence need not be limited to
ends of self-interest. It is involved in any context in which one must learn
how to influence others to promote one’s ends, whether those be par-
ticular ends, one’s happiness, or moral ends. Kant even specifically says,
precisely in the context of pointing out that anthropology helps one ma-
nipulate others, that “morals must be bound together with knowledge of
man” (25:472, emphasis added). Thus Kant explicitly discusses the uses
to which anthropology can be put, and he specifically mentions moral
cultivation. Kant says,

Anthropology is . . . indispensable and manages great uses.

1. In pedagogy.
2. With respect to the influence that we can have on others. Especially
commanders, who with a proper knowledge of man can get total oppo-
sites to work [together], which otherwise can be set right only through
3. With respect to the influences on morals and religion, that through
this knowledge one can give these duties the power of inclinations. (25:

Each of these three uses of anthropology can concern moral develop-

ment. Education, which Kant mentions first, is emphasized throughout
his anthropology. Insofar as one learns how to direct others, not only in
action but also in thought and choice, one can effectively educate those
under one’s care. In several sets of lectures Kant discusses education in
Kant’s “Pragmatic” Anthropology 55

depth and often ends his course on anthropology with applications of his
insights to the education of others (see, e.g., 25:722–8). In all of these
discussions, one of the most important kinds of education for Kant is
moral. In the Anthropology, Kant’s discussion of education includes man’s
need to be “moralized,” to be “educated to the good” (7:324, 325). In his
lectures, he claims that “respect for the worth of humanity in one’s own
person is the final level of education” (25:727). And in his lectures on ed-
ucation, Kant discusses at length “moral culture [moralische Cultur]” and
a “practical education [praktischen Erziehung]” that includes “morality”
(9:480, 486).
The second important use of anthropology relates to a passage we
looked at earlier, where Kant insisted that learning to direct others to
one’s ends can help one avoid being “too hard or too offensive.” Here
Kant examines the most extreme case, insisting that the lessons of anthro-
pology are an important alternative to violence in getting others to do
one’s will. Kant does not suggest means by which one can or should force
or deceive another into promoting one’s ends. Rather, by learning about
others, one can affect them without force, deceit, or even offense, by play-
ing off of their own natural inclinations and, in some cases, weaknesses.
This is perfectly consistent with the categorical imperative, whereby one
can never use another merely as a means. The commander who uses an-
thropological prudence not only acts in an acceptable way; she does posi-
tive good insofar as she promotes social cooperation by getting opposites
to work together. This social cooperation is a crucial part of a moral com-
munity, for it establishes the necessary condition of stability under which
such a community can flourish. In the Religion and his writings on history,
Kant points out the need for political stability as a precondition of moral
stability – that is, for a political commonwealth prior to a moral com-
monwealth (cf. Wood 1999, Louden 2000, and Anderson-Gold 2001).
Peaceful social cooperation is thus an important part of promoting moral
While the first two uses can apply to moral and nonmoral concerns,
Kant’s third use for pragmatic anthropology connects it directly with
his explanations of moral anthropology in the Metaphysics of Morals. By
teaching how to get others to adopt one’s ends, one gains specific insight
into how someone can be influenced to adopt obedience to the moral law
as an end. The unconditional value of the moral law may be an end that
one seeks to influence another to adopt in the case of moral education
and may even be an end that one wants to influence oneself to adopt, as we
will see more clearly in Chapter 5. Not only does this “pragmatic” nature
56 The Problem

of anthropology not exclude consideration of how to promote morality,

but Kant’s remarks on the uses of anthropology suggest that increasing
the force of duty is one of the most important roles of anthropological

This brief discussion of the various senses in which Kant uses the term
pragmatic shows that his description of the Anthropology as being given from
a “pragmatic point of view” does not preclude important moral dimen-
sions to that work. Moreover, Kant’s own claims about his anthropology,
how it differs from Platner’s as well as the uses that he mentions to his
students, suggest that Kant meant to include moral anthropology in his
anthropology from a pragmatic point of view. Of course, neither the possi-
bility that anthropology includes moral anthropology nor Kant’s intention
to include moral anthropology shows that there actually is anything moral
about Kant’s specifically anthropological work. And if Schleiermacher
is right about the conflict between Kant’s Critical philosophy and his
anthropology, one might even think that Kant cannot, or should not,
succeed in actually carrying out a moral anthropology. Moreover, much
of Kant’s anthropology is clearly not moral anthropology in the narrow
sense. Almost all of his anthropological observations are relevant to ap-
plying the moral law, but only some relate directly to the task of showing
what influences strengthen or weaken the force of the moral law on the
will. This is not surprising, because there is no reading of pragmatic that
identifies it with moral concerns. But Kant’s pragmatic anthropology does
include moral anthropology, and the most effective means of illustrating
that is to offer examples. To that task we now turn.

2. Three Examples of Kant’s Moral Anthropology

In order to show that Kant’s anthropology includes empirical influences
that promote a good will, I examine three particular cases in which an-
thropological insights are morally relevant in the sense of giving sub-
jective conditions that hinder or help people in fulfilling the laws of a
metaphysics of morals. Exploring the full richness of these examples is
not necessary to establish the presence of moral anthropology in his work,
so I focus in each case on showing that the example is moral, in the sense
that it reflects a help or a hindrance to moral maturity. When relevant, I
will show how the Anthropology enriches accounts in Kant’s more specif-
ically moral works in just the ways that one would expect from moral
anthropology. I also briefly show that each example is practical, in the
Kant’s “Pragmatic” Anthropology 57

sense that it is to be put to use rather than merely a matter of theoretical

knowledge, and empirical, in the sense discussed in the previous chapter.
In the concluding section of this chapter, I briefly point out how the
structure of the Anthropology and Kant’s corresponding lectures shows a
deep concern on Kant’s part with the moral dimension of his work that
comes out in the examples.

A. Politeness
Kant’s discussion of politeness occurs primarily in a section of the An-
thropology called “on the admissible moral perception” (7:151–3) and in
Kant’s discussion of “sociability” in the Metaphysics of Morals (6:473–4).14
In Kant’s lectures on anthropology, politeness is included within a more
general discussion of the distinction between illusion and deceit (see, e.g.,
25:502–5, 1455). Kant’s account of politeness is not the most prominent
example of moral anthropology, but it is one of the most clear. Kant insists
that because of certain facts about human beings, we are susceptible to
influence through politeness, and this influence should be used for the
strengthening of moral principles in oneself and others. Politeness is an
empirical aid to the development of a good will, and it is explained in de-
tail only in Kant’s anthropology. Thus it highlights Kant’s attention to the
influence of empirical helps and hindrances on subjective conditions of
morality, as well as his recognition that anthropology is the proper place
for such attention.
Kant’s account of politeness is clearly moral anthropology, in the sense
that it discusses an aid to moral cultivation. One should be polite be-
cause politeness fosters virtue in oneself and others. In the Metaphysics of
Morals, Kant explains that although polite graces “are, indeed, only . . . a
beautiful illusion resembling virtue . . ., yet they promote the feeling for
virtue itself” (6:473). The Metaphysics of Morals focuses on the claim that
“it is a duty . . . to associate the graces with virtue” because these graces
promote virtue itself (6:473). The Metaphysics of Morals does not discuss,
however, how politeness promotes virtue. And this is proper, because such
a discussion of the empirical aids to virtue is the proper subject of moral
anthropology. In the Anthropology, Kant reiterates the moral purpose of
politeness, saying that “when men play [virtuous] roles, virtues are gradu-
ally established, whose appearance had up until now only been affected”
(7:151, cf. 7:152–3). But he also goes further, pointing out how such
politeness works to cultivate morality.
Politeness cultivates morality by “deceiving the deceiver in our-
selves” (7:151), which occurs in two main ways. First, one who is polite
58 The Problem

must often refrain from directly satisfying illegitimate desires, and this
self-restraint both reveals and cultivates a capacity for self-control that
may be ignored or undeveloped otherwise. As Kant explains, “This be-
trays a self-mastery and is the beginning of conquering oneself. It is a step
towards virtue or at least towards a capacity thereto” (25:930). Second,
politeness presents a “beautiful illusion resembling virtue” (6:473), such
that “one who loves the illusion of the good eventually is won over to
actually loving the good” (25:931).15 Kant’s account of politeness in his
Anthropology is fundamentally moral anthropology, because it studies an
aid to the cultivation of virtue, and it provides needed anthropological
background to flesh out Kant’s reference to sociability in the Metaphysics
of Morals.
Kant’s account of politeness is also not merely speculative theorizing;
it is meant to be put to use. In the Anthropology itself, this is clearest in
the language that Kant uses to describe its effects. In pointing out that
through politeness “the virtues are gradually established,” Kant adds that
this is “a good thing . . . in this world” (7:151). Throughout he speaks of
being polite in concrete terms, directing his readers’ attention to what
they can actually do in response to his anthropological insights. And in
case one is unmotivated by the good effects of politeness, the discussion
in the Metaphysics of Morals points out that one has a duty to put into
practice the insights of the Anthropology: “It is a duty to oneself as well as
to others not to isolate oneself but to use one’s moral perfections in social
intercourse . . . and so to associate the graces with virtue” (6:473). Unlike
Platner’s physiological anthropology, Kant’s discussion of politeness aims
at what one can and should do; it is practical.
Finally, Kant’s discussion of politeness in the Anthropology makes clear
that politeness is an empirical influence. In the published Anthropology,
Kant repeatedly emphasizes the fact that politeness is merely an appear-
ance (Schein). Virtue involves an “appearance of attachment” (7:151) to
others, an “appearance of affability” (7:152), and “good and honorable
formal behavior [which] is an external appearance” (7:152). It is pre-
cisely “the appearance of good in others [which] must have some value for
us, because in the long run something serious can come of [it]” (7:153).
Politeness fosters virtue precisely as an illusion, an external appearance
that can affect behavior even when one knows that it is mere appearance.
Hence politeness is a paradigm case of an empirical influence on moral-
ity, and it is discussed in detail only in Kant’s anthropology. Thus it high-
lights not only Kant’s attention to empirical aids but also his recognition
that anthropology is the proper science to which such attention belongs.
Kant’s “Pragmatic” Anthropology 59

B. Affects and Passions

Politeness is an important empirical help for developing a good will, but
Kant’s Anthropology also discusses two important empirical hindrances: pas-
sions and affects.16 Passions and affects are not significant for morality
alone. In both his published Anthropology and his lectures, Kant points
out that passions and affects hinder one’s pursuit of happiness (25:592).
But they also hinder virtue, and thus they are properly a part of moral
anthropology. The moral importance of affects and passions is clearest
in Kant’s lectures in ethics, where Kant insists that “in duty to ourselves,
and for the dignity of mankind, the demand upon a man is that we have
no passions or affects at all” (27:368). In the Metaphysics of Morals, Kant
reiterates this theme, saying that affects involve a “lack of virtue” and pas-
sions are “properly evil” (6:408). The key to their hindering effect, Kant
points out, is the way in which they inhibit self-mastery. Kant’s Anthropology
reiterates this, saying that “to be subject to affects and passions is . . .
an illness of mind because both . . . exclude the sovereignty of reason”
(7:251). But Kant’s anthropology goes further than simply pointing out
that affects and passions are morally dangerous.17 Anthropology explains
the differences between passions and affects and explains how each hinders
The primary purposes of the descriptions of affects and passions in the
Anthropology and related lectures are, first, to make clear in what way each
is an “illness of mind,” and, second, to classify various different affects and
passions that hinder morality (and the pursuit of happiness) in different
ways. In this context, Kant also clarifies the claim in the Metaphysics of
Morals that affects are merely a “lack of virtue,” whereas passions are
“properly evil,” and he presents at least some suggestions regarding the
“distinct . . . method[s] of prevention as well as . . . cure that the physician
of souls would have to employ” to treat affects and passions (7:251).
With regard to the first purpose, that of clarifying what affects and
passions are, Kant explains,

The inclination that can hardly, or not at all, be controlled by reason is passion. On
the other hand, affect is the feeling of a pleasure or displeasure at a particular
moment, which does not give rise to reflection (namely the process of reason
whether one should submit to it or reject it). (7:251)

Here Kant makes clear that his condemnation of affects and passions
is not a condemnation of all emotions, nor even of all intense emo-
tions. What distinguishes affects and passions from ordinary emotions
is the “want of reflection” (7:254). This want of reflection, and “not the
60 The Problem

intensity of a certain feeling,” qualifies a feeling as an affect (7:254). In

this passage, Kant also distinguishes affects and passions based on the
different faculties of soul that they affect. Passions are uncontrolled in-
clinations, related to the faculty of desire, while affects are uncontrolled
feelings, related to the distinct faculty of pleasure and displeasure. The
Anthropology clearly distinguishes these faculties, and this difference plays
an important role in understanding how passions can be evil while affects
are merely a lack of virtue.
In particular, the faculty of desire is the seat of choices (see 7:251 and
6:212–13). One who is under the influence of a passion makes one incli-
nation the overriding inclination in life. This choice of one inclination
over all others does not involve the comparison of that inclination with
others, nor the consideration of the demands of reason. But it does not
consider those other interests because it deliberately rejects them. Thus
Kant claims that “passion is an illness that abhors all medication” (7:266).
Because it is deliberate, it is “properly evil” (6:408).
By contrast, affects, which merely infect one’s faculty of pleasure and
displeasure, bypass normal processes of choice. Kant offers as an example
one who is paralyzed by seeing a child fall into the water, who is “so
shocked that one thereby cannot do anything” (25:592). Because affect
bypasses rather than corrupts choice, it “produces a momentary loss of . . .
self-control” (7:267) and reflects merely a “lack of virtue” (6:408). As we
will see later, however, even this “lack of virtue” is a hindrance to the
cultivation of a good character, because it involves a rest in a struggle
against evil that should be incessant.
These general points about affects and passions help classify and char-
acterize two important hindrances to the good will. Kant also gives details
about specific affects, such as anguish, rage, and shock; and specific pas-
sions, such as ambition, thirst for revenge, and lust for power. The details
of these accounts are unnecessary here. The primary purpose of these
discussions is to draw attention to the symptoms of particular cases so that
one can be more aware in oneself and others of the presence of these
hindrances to virtue.
In addition to classifying and characterizing passions and affects, Kant
offers at least some advice on how to treat and prevent these hindrances.
Thus this discussion is properly practical. With respect to passions, Kant
points out that “most of them are incurable because the sick person does
not want to be cured” (7:266). Thus the most important measures with
respect to passions are preventative, and Kant gives specific suggestions
for preventing various particular passions. For instance, he explains that
Kant’s “Pragmatic” Anthropology 61

“in nature there is a propensity for frugality in elders. But reason requires
that this inclination be resisted and that one saves only for a purpose”
(25:520). By being aware of the passions that are most dangerous for
one, one can take preventative measures. In addition, there are at least
some things that one can do to lead another person out of passion despite
himself. In this context, for example, Kant reiterates the importance of
politeness as an illusion that can free one from passions: “The passion
of love is much moderated through [politeness], when one . . . conceals
the red-hot inclination that otherwise would be difficult to suppress”
With respect to affects, Kant suggests some preventative measures,
such as avoiding “romances and maudlin plays” (5:273), and suggests
that refining feeling through culture can help prevent and treat affects
(25:622–3). Because affects are only “precipitate” (7:252; 6:407) and
“what the affect . . . does not accomplish quickly will not be accomplished
at all” (7:252), affects provide much more room for self-treatment. One
can and probably will seek out refining influences of culture and avoid
situations that prompt affects.
Finally, affects and passions are empirical influences. Kant describes
affects as “surprise through sensation” (7:252) and offers an empirical
description of desire that culminates in his claim that “a subject’s sensu-
ous desire that has become customary . . . is called inclination . . . [and]
inclination that hinders the use of reason . . . is passion” (7:265). In both
cases, Kant describes empirical processes that characterize passions and
affects. And this is proper, not only because they are included in his
Anthropology, which is empirical in general, but because each is a disorder
of a different faculty of soul, and these faculties are themselves outlined
in Kant’s empirical description of the human being.
Whereas politeness is an empirical aid to virtue, affects and pas-
sions are empirical hindrances to virtue. And Kant describes and char-
acterizes them in his Anthropology, providing a needed anthropological
background to the brief treatment in the Metaphysics of Morals. More-
over, he suggests practical measures to take in the light of his account,
confirming that his anthropology is not only moral, but also properly

C. Character as a Subjective Condition of Moral Agency

The most important aid to morality discussed in the Anthropology is charac-
ter. In the Anthropology Kant discusses several different senses of character,
only some of which serve to characterize human beings in general. The
62 The Problem

discussion is divided into sections on the character of the person, of the

sexes, of nations, of races, and of the species. Of these, the character
of the person is the most morally significant. The character of the per-
son is explicitly identified with “moral character” (7:285) and described
as “above all price” (7:292), linking this discussion with specifically nor-
mative concerns. In the other sections, the term character often merely
refers to characteristics.18 Thus my discussion of character focuses on the
character of the person.
Character is clearly relevant to morality. The repeated reference to
character as “moral” (moralische) in Kant’s introductory section on char-
acter and the claim that “character has an inner value and is above all
price” (7:292) might even suggest that character is identical to a good
will. But Kant makes clear that character and virtue, though related, are
not identical. Kant’s only example of character in his discussion is Sulla,
“a person of evil [bösem] character . . . [who] is, nevertheless, also an object
of admiration, because we generally admire strength of soul” (7:293).
The strength of soul that is possessed even by the evil Sulla is not
identical to a good will, but it is an invaluable aid to a good will. In his
lectures on anthropology, Kant defines “character” as “the use of our
choice [Willkühr] to act according to rules and principles” (25:630, cf.
25:1384). In the Anthropology, he expands on this:

Simply to have a character relates to that property of the will by which the subject
has tied himself to certain practical principles. . . . Although these principles may
sometimes indeed be false or defective, nevertheless the formal element of will
as such, which is determined to act according to firm principles (not shifting
hither and yon like a swarm of gnats), has something precious and admirable to
it, which is also something rare. (7:292)

When one remains firmly committed to defective principles, as in the

case of Sulla, one is not morally good. But in order to be morally good,
one must remain firmly committed to good principles.
Thus having a character, a commitment to principled action, is a nec-
essary but not a sufficient condition for complete virtue. As Kant explains,
“before a good or evil character is built for a human being, a character
altogether must be built, with which he first has a character in general,
i.e., he first must get into the habit of acting from principles” (25:630–1).
Hence Kant says later that “such people [without character] are like soft
wax19 ; every instant they take up another rule. Towards evil they are quite
steerable but not to good, for to become good already requires prin-
ciples” (25:631). People without character are steerable to evil but not
Kant’s “Pragmatic” Anthropology 63

good because good action is necessarily action according to principle. To

be good, for Kant, just is to act according to maxims that conform to the
categorical imperative because they conform to the categorical imperative.
But this implies an overall commitment to a principle (the categorical im-
perative) as a guide to action. Thus, for Kant, “character rises to the level
of morality insofar as it is determined as either good or evil” (25:652),
such that “good character would be the good will” (25:648).
The close connection between character and morality might lead one
to forget that the study of character is part of an empirical anthropology.
In that context, it is important to note that Kant’s account of character
is not an account of what constitutes a good or evil will. That subject is
proper to moral philosophy as such. Character is a subjective condition of
a will choosing to be either good or evil. Hence Kant’s discussion focuses
exclusively on character as such, not on how character becomes good or
Kant’s account of the origin and development of character further
emphasizes its empirical basis. Although Kant claims that “character
comes not from nature, but rather must be acquired” (25:1172), he
adds that the “seed” of character comes from “nature” (25:1172). This
is fleshed out in terms of an innate and natural propensity for char-
acter (25:823, 1176). Kant also suggests that some human beings are
better prepared for character than others by virtue of another natu-
ral characteristic – temperament. Kant says, “not every temperament
is inclined to adopt a character, e.g. the melancholy one adopts a
character first, the sanguine one not so easily” (25:1388), and of one
with a phlegmatic temperament, Kant claims, “without being brilliant,
he will still proceed from principles and not from instinct” (7:290).
Kant’s discussion of temperaments is “psychological” but still empir-
ical, and temperaments can even “be influenced . . . by the physical
condition of a person” (7:286). Thus insofar as they play a role in
the formation of character, they show the extent to which character is
Kant discusses further aids to the cultivation of character that are not
natural endowments but are nonetheless empirical influences. Among
these, education is the most important.20 “The acquisition of good charac-
ter with people happens through education” (25:1172). Throughout his
writings Kant’s accounts of character are accompanied by specific peda-
gogical recommendations. He suggests that because “imitation . . . greatly
hinders character, in education one must never refer one’s children to
the neighbor’s children, . . . but rather build their character directly,
64 The Problem

[using] principles of good and bad to inspire righteousness and nobility”

(25:635, cf. 7:325; 5:154; 25:599, 722ff., 1386). Politeness also plays an
important role in the cultivation of character by combating passions and
promoting self-control. In his writings on history and politics, Kant gives
further explanations for how natural inclinations can give rise to vari-
ous social institutions – including stable and just political regimes, peace,
and even progress in the arts and sciences – that may have beneficial ef-
fects on character.21 Thus “experience and history” provide reasons that
“we should not despair about our species’ progress towards the better”
(7:329). The increasing presence of stable political structures and advanc-
ing culture, like the presence of stable norms of polite society, can help the
cultivation of constancy in principled action. These external influences,
like the more direct influence of education, support the establishment of
None of these external supports for character absolve people from
responsibility for developing character themselves, however. Thus in ad-
dition to the practical suggestions for promoting character in others,
Kant’s discussion of character also provides a practical program for de-
veloping character in oneself. He suggests five basic principles to aid the
cultivation of character.

a) Not to speak an untruth intentionally . . .

b) not to dissemble . . .
c) not to break one’s legitimate promise . . .
d) not to join the company of evil-minded people . . .
e) not to pay attention to slander. . . . (7:294, cf. 25:1387–8, 1392)

These are all practical principles that support and constitute the develop-
ment of character as such. The pursuit of these methods for developing
character depends on one’s already having at least some level of char-
acter. Unless one can act on the basis of principles, one will be unable
even to follow the principles for developing character. But keeping these
principles even sporadically can have some beneficial effect. The more
one avoids duplicity, bad company, and slander, the easier it will be for
one to stick to principles in the future. Insofar as one has some minimal
level of constancy, these principles can reinforce one’s character. They are
important aids to self-improvement, even if they are not sufficient. And
their inclusion here is enough to show that character as an empirical aid
to virtue is part of a practical anthropology, rather than mere speculative
Kant’s “Pragmatic” Anthropology 65

3. Conclusion: The Systematic Place of Moral Concerns

in Kant’s Anthropology
These three examples of the moral anthropology within Kant’s “prag-
matic” anthropology show that his anthropology is not concerned merely
with techniques for achieving happiness or worldly success. And this, of
course, is what one would expect, given that Kant considered his course
in anthropology a pre- or at least co-requisite of his course in ethics.22
But although there are several areas in which Kant’s anthropology is
morally relevant, one still might think that these are only occasional and
peripheral.23 And Kant’s anthropology certainly contains much that is not
particularly moral, including a long section on cognition that has little
to do with subjective conditions of moral choice.24
Nonetheless, there is a clear structure to Kant’s anthropology that high-
lights the importance of moral anthropology. Throughout his lectures
and in his published anthropology, Kant begins his account with a de-
scription of the cognitive faculty, then discusses pleasure and displeasure,
and ends the first part of his work with an account of the faculty of desire,
the faculty most directly related to morality. Consistently, Kant ends this
first part of his anthropology with discussions of passions and affects.25
This discussion is especially prominent in the published Anthropology, in
which Kant’s account of the faculty of desire is taken up almost entirely
with the passions and affects. Not only does this prominence reflect an
extremely narrow range for a discussion that is supposed to be about
the entire faculty of desire, but it does not fit Kant’s systematic division
of faculties into cognition, feeling, and desire, because affects are mat-
ters of feeling. Kant seems more concerned with placing a discussion of
passions and affects prior to his account of character than with holding
rigorously to an account divided by faculties of the soul. In all of his
lectures in anthropology as well, the discussion of passions and affects
precedes an account of character that makes up the second major part
of his anthropology.
The importance of this structure should not be missed. Kant’s ac-
counts of passions and affects and character are the two aspects of his
anthropology that are most directly related to moral anthropology, and
they mark the climax of nearly all his lectures as well as his published
Anthropology. Moreover, they are closely related, such that Kant uses his
account of passions and affects to introduce and provide a transition
into his account of character. In his lectures, this is especially dramatic.
Kant ends the first part of his anthropology in his 1775–6 lectures with
66 The Problem

“general remarks on the passions and affects” (25:613). He concludes this


All these [passions and affects] are set against the steadfastness, wherein one does
not deviate from his principle and persists firmly in his decision. It is important
to have firmness in one’s decisions and not to depart from one’s principle, and
better to endure disadvantage than to let go one’s principle. When then a human
being knows something for sure, he carries it out. But whoever is not steadfast
therein often seizes a principle that he is certain that nothing will come of, because
he knows that already he has frequently broken principles. Thereby the human
being is a cream puff in his own eyes; he believes himself capable of nothing
more, from which springs hopelessness. That is a comfortless condition, when
one always postpones hope. So it is with late conversions. So it is with other
things, that one wants to break the habit of, e.g., oversleeping, for it always says,
just one more time but then no more, and so one philosophizes oneself further
away from carrying out [one’s resolution]. In such a condition one never has
hope to become better; this is an important point in morals. One must therefore
seek to hold himself ever so promptly to this word as to others. From that springs
a firm confidence in ourselves. (25:624)

This passage is a perfect conclusion to a long discussion of passions and

affects, in that it highlights the importance of holding firm to principles
and the hopelessness of a life lacking in self-mastery, in which one cannot
follow through on one’s own resolution. But it is also a perfect introduc-
tion to the second part of the anthropology, dealing with character.
After a long discussion of human capacities, Kant always discusses the
danger posed by affects and passions, a danger whereby all of one’s capac-
ities can prove fruitless through one’s loss of self-mastery. He then turns
to a discussion of character that stresses the importance of constancy.
Kant’s anthropology, which begins with discussions of cognitive capabil-
ities, ends with the importance of developing character in oneself and
in the species as a whole. And although this character is not necessarily
morally good, it is a form of self-mastery that is a subjective condition of
It is especially significant that in these sections on character (and only
in these), Kant describes his account of the character of the person in
language that makes clear its deep connection to pragmatic anthropol-
ogy as a whole. Recall that in his preface Kant distinguished pragmatic
from mere physiological anthropology by saying that whereas physiolog-
ical anthropology investigates “what Nature makes of man,” pragmatic
anthropology deals with “what man makes . . . of himself” (7:119). It is
only in his discussion of character that Kant points out explicitly that that
Kant’s “Pragmatic” Anthropology 67

purpose of pragmatic anthropology is fully realized: “Here it does not

depend on what Nature makes of man, but what man makes of himself”
Kant’s account of character finally ends in a discussion of the human
species in which he explains how the race can “advance constantly from
the evil to the good” (7:333). As he summarizes it,
The sum total of findings generated by pragmatic anthropology as to the classi-
fication of man and the characterization of his development is as follows: man
is destined by his reason to live in a society of other people and in this society
he has to cultivate himself, civilize himself, and moralize himself by the arts and
sciences. . . . (7:324, emphasis added)

Pragmatic anthropology deals with cultivating all human capacities, to be

sure, but it reaches its conclusion only in the civilization, and eventually
moralization, of humanity itself.
This means that although it is empirical, anthropology is deeply con-
cerned with helps and hindrances to moral development. Politeness and
character mark two important helps, and Kant’s discussion of character
points to a wide variety of other helps, such as education, republican
government, world peace, and the development of the arts and sciences.
Passions and affects are two significant hindrances, for which Kant pro-
poses at least some treatments. And all of these influences are empirical
influences, the proper objects of an empirical science of anthropology.
Schleiermacher thus seems to have accurately characterized Kant’s an-
thropology, and he has drawn attention to a real tension between this
anthropology and Kant’s conception of transcendental freedom.

In the next chapter, I draw from important work in contemporary

neokantian ethics to articulate several ways in which Kant might seem to
escape Schleiermacher’s dilemma. While all are promising for explaining
certain sorts of empirical helps and hindrances, all fall short in one of
two ways, either by failing to sufficiently account for an empirical moral
anthropology in the robust sense outlined in this chapter, or by sacrific-
ing Kant’s theory of freedom. In Chapter 5, I offer my own resolution
of the tension, drawn from Kant’s Religion and later writings in moral

Moral Anthropology in Contemporary

Neokantian Ethics

Neokantian moral theorists have recently begun to pay more attention to

Kant’s moral anthropology. Robert Louden devotes the longest chapter of
Kant’s Impure Ethics (Louden 2000, cf. Louden 1986) to the treatment of
the role of anthropology in the “second part” of Kant’s moral theory. And
G. Felicitas Munzel devotes considerable portions of her Kant’s Concep-
tion of Moral Character (Munzel 1999) to pointing out the insights that can
be gleaned from the anthropology for Kant’s moral theory. Aside from
these direct treatments of anthropology, many of the most influential con-
temporary appropriations of Kant’s moral theory have shown increasing
attention to various little-known gems of anthropological reflection in
Kant. Nancy Sherman and others have pointed out the important role of
the emotions in Kant’s moral philosophy, drawing in large part from the
Anthropology and the most anthropological sections of Kant’s other writ-
ings. Allen Wood and Sharon Anderson-Gold, in their attention to the
social dimension of ethical life (Wood 1999; Anderson-Gold 2001), draw
from anthropological considerations in Kant’s Religion and historical es-
says. Paul Guyer’s rich accounts of the experience of freedom also draw
extensively on anthropological considerations (Guyer 1993). And this
just scratches the surface of the explosion of attention to anthropology
in contemporary Kantian moral philosophy.
In part, this attention is a response to challenges from Kant’s critics,
especially those sympathetic to contemporary virtue ethics, who often ac-
cuse Kantian moral theory of paying insufficient attention to character,
virtue, and the richness of human life. In part, it is the result of the re-
cent publication in Germany of Kant’s Lectures on Anthropology (vol. 25
of the Academy Edition of Kant’s works). And in part it is simply a
Moral Anthropology in Neokantian Ethics 69

long-overdue recognition of the importance that Kant himself gave to an-

thropology. These developments are welcome. They flesh out a moral the-
ory that though provocative and appealing seemed to have lost touch with
human life.
But this enrichment of moral theory through anthropology has led
some Kantians to neglect the central importance of freedom in Kant’s
moral theory. Even those who continue to attend to Kant’s theory of
freedom while seeking to incorporate anthropological insights fail to
explain adequately how one can coherently integrate moral anthropology
into Kant’s theory of freedom. As we have seen in the past three chapters,
Kant’s theory of freedom does not fit well with his empirical and morally
relevant anthropology. As neokantians incorporate anthropology into
moral theory, they often fail to give moral anthropology the significance
it deserves, or they sacrifice Kant’s commitment to freedom.
In this chapter, I examine four roles for moral anthropology that seem
to offer promising possibilities for integrating empirical influences with
Kant’s theory of freedom. I point out the place of these possibilities in the
work of some of the most innovative recent work in Kantian ethics today,
especially that of Barbara Herman, Nancy Sherman, Robert Louden, and
G. Felicitas Munzel.1 I argue that none of these possibilities offers an ad-
equate account of how a genuine moral anthropology can be integrated
with Kantian freedom.
In section 1, I consider the possibility that morally beneficial empirical
influences play an epistemic role in helping human beings determine the
content, relevance, or proper application of the moral law in concrete
situations. In section 2, I consider the possibility that these influences
play a necessary or useful instrumental role in carrying out the demands of
morality. That is, they provide resources that one can use to do more good
deeds. In section 3, I consider the possibility that helpful “influences” are
actually constitutive of the good will, such that to have character, or to be
polite, or to be part of a good community, is part of what it is to be good.
In sections 4 and 5, I consider the suggestion that empirical influences
are a necessary or helpful “propaedeutic” to morality.
The first possibility plays an important role in the work of both Nancy
Sherman and Barbara Herman (Herman 1993 and 1998; Sherman 1990,
1995, 1997a, 1997b, and 1998). Nancy Sherman is also the primary source
for the second and third options.
The final role is discussed by Robert Louden, who raises it in the
context of an explicit treatment of the relationship between anthropo-
logical influences and Kant’s theory of freedom (Louden 2000). This
70 The Problem

final option – of influences as propaedeutics – is also central to recent

accounts offered by G. Felicitas Munzel (Munzel 1999) and Allen Wood
(Wood 1999). Wood deals extensively even if obliquely with the issue
of the relationship between history, political and social structures, and
morality. He suggests that empirical influences can play roles as neces-
sary preconditions of moral choice, and he seems to think that at least
some empirical influences are necessary preparations for choosing rightly.
But Wood does not offer as much detail as Louden, because his book is
focused more generally on Kant’s ethical theory. Moreover, on the issue
of helps and hindrances there does not seem to be any significant dif-
ference between their views.2 Thus although I focus on Louden, most of
what I say here could apply to Wood as well. In section 5, I take up Mun-
zel’s propaedeutic account. At times, Munzel suggests a conception of
“propaedeutic” that leaves room for “influences” that are not necessarily
“prior” to choice, but she does not sufficiently develop this conception.

1. Empirical Influences as Epistemic Aids

The first role that one might ascribe to empirical influences is an epis-
temic one. According to this account, certain influences, such as polite
society, a well-functioning political system, a good moral education, or
well-cultivated emotions, aid in discerning morally relevant features of
a situation. One who lives in a society with rigidly enforced gender or
class roles, for instance, might not recognize the harm done to women
or to members of a lower class as morally relevant in deliberation. These
empirical influences would thus hinder one from being morally good.
On the other hand, emotional sensitivity to the pain of others or a moral
education that teaches one to recognize morally relevant features of a
situation can be invaluable aids to choosing rightly. By revealing or con-
cealing morally relevant features of one’s choices, empirical influences
can – or so it seems – affect whether one turns out to be good or evil. But
because these influences affect merely one’s cognitive faculties, one still
freely chooses whether to act on good or evil maxims. Assigning this role
to moral anthropology thus seems to reconcile empirical influence with
transcendental freedom.

Nancy Sherman assigns this sort of role to the emotions, the particu-
lar sphere of moral anthropology on which she focuses.3 Sherman not
only suggests that emotions can be part of a structure that supports
the good will, but she lays out, clearly and in several essays, “what this
Moral Anthropology in Neokantian Ethics 71

supportive structure could be like.”4 The first “way in which emotions

can serve supportive roles” is “as modes of attention that help us track what
is morally salient in our circumstances, and thus locate possible moments
for morally permissible and required actions” (1997a:145; 1997b:273).
In this respect emotions “serve instrumental (epistemic) roles indispens-
able for acting from moral interest” (1997a:158; 1997b:278).
In this explanation of the role of emotions, Sherman draws heavily
from Barbara Herman’s work on moral judgment. Herman points out
the importance of “rules of moral salience” for moral evaluation and
deliberation. These rules of moral salience “structure an agent’s per-
ception of the world so that what he perceives is a world with moral
features” (Herman 1993:77). Only in this context can an agent properly
test maxims against the moral law, because such a test requires that one’s
maxims pick out those features of a situation that are morally relevant.
Like Sherman, Herman closely connects these rules of moral salience
with empirical aids. She says that these rules are “acquired as elements
in a moral education” (Herman 1993:77) and that “they are acquired
in childhood as part of socialization” (78). External influences thus de-
termine, at least in part, the structure of our moral sensitivity. And this
structure affects not only evaluation of situations but also practical de-
liberation. As Herman explains, “issues of excuse and justification [that
depend on rules of moral salience] do not enter a moral thought only
after an action has been done; they are part of the way a normal moral
agent explains the permissibility of his actions to himself” (Herman
Like Sherman, Herman suggests the possibility that “affective capac-
ities of response” may be necessary to properly perceive moral features
of a situation, and if these are necessary, then “we will have found
a Kantian argument for the development of the affective capacities”
(Herman 1993:82). Because Herman does not show that affects are nec-
essary, Sherman’s arguments that emotions actually are necessary provide
a crucial supplement to Herman’s account. On the other hand, Herman
makes clear that the epistemic instrumental role that Sherman assigns
to the emotions is a role that can and probably must be filled by a wide
variety of abilities that can be encouraged by empirical influences.
Herman is also much clearer than Sherman about precisely what sort
of help is provided by emotions or other influences that contribute to
rules of moral salience. And it is clear in Herman’s discussion that these
sorts of aids are not aids to actually having a good will. On Herman’s
account, there are two critical ways in which rules of moral salience relate
72 The Problem

to the good will. First, an agent must have some rules of moral salience
in order to be capable of choosing to be good or evil. Empirical helps are
thus necessary for being a moral agent at all. As Herman explains, “the
Kantian moral agent must have a characteristic way of seeing if he is to
judge [morally] at all. To be a moral agent one must be trained to perceive
situations in terms of their morally significant features (as described by
the RMS [Rules of Moral Salience])” (Herman 1993:83). For Kant, the
moral status of an action depends on the maxim that underlies that action.
Beings that “act” merely from impulse, without any maxims at all, cannot
be morally accountable at all. Likewise, those who act on maxims that
incorporate no moral features at all cannot be morally accountable. One
can imagine, for example, a “person” with a certain sort of cognitive
failing such that he does not recognize basic rules of moral salience.
Such a person might “kill” another but recognize in the action nothing
more than the lifting of an arm and the moving of a finger. The person
does not even perceive – or does not see as relevant – that the hand
at the end of the arm is holding a gun and that the finger pulls the
trigger and that the gun is pointed at another person. A person for whom
every action is of this sort might act, say, for the pleasure of experiencing
new bodily positions, without any awareness of the mayhem caused by
these new positions. For such a person, moral issues simply do not arise.
Awareness is so constrained that it never captures anything of any moral
salience. Because of this peculiar cognitive failure, such a person would
be incapable of choosing either good or evil. That person would not be a
moral agent at all.
A second role of rules of moral salience is to allow a moral agent – and
Herman discusses only the case of morally good agents – to apply moral
principles more easily to particular situations. In this case, better rules of
moral salience allow one who is committed to acting only on good maxims
to choose maxims that are not only consistent with the moral law but that
also “more accurately reflect moral facts that can be obscured from plain
sight” (Herman 1993:88). A person without good rules of moral salience
would be someone who “is attentive to duty, but not very perceptive, and
so [who] does not see that his circumstances fall under a principle of duty.
(He fails to recognize some situation as one calling for help.)” (Herman
1993:81). In this case, as Herman is careful to point out, the moral sta-
tus of the agent is not at stake. More or better rules of moral salience do
not affect whether or not the agent has a good will. Because on Kant’s
account “there is no way to judge actions apart from the way they are
willed,” “morally defective RMS [rules of moral salience] may not yield
Moral Anthropology in Neokantian Ethics 73

morally defective actions” (Herman 1993:91). “It can be permissible for

agents with mistaken RMS to act in ways that would be judged impermis-
sible if their RMS were correct” (Herman 1993:89). For example, one
may be morally good, act in accordance with the moral law, and refer to
women as “girls,” or to African-Americans as “niggers,” because one fails
to recognize that referring to them in these ways is demeaning. Failing to
recognize this, one may deeply harm women or African-Americans while
acting on a maxim that is consistent with or even based on respect for the
dignity and happiness of women and African-Americans. One just fails
to consider important dimensions of their situation.5 Because Kantian
moral evaluation of agents is based on the status of their maxims, better
rules of moral salience do not affect the status of a person’s will. Instead,
as Herman points out, “the scope of beneficent actions (how much good
is done) will be greater for persons who can more readily perceive the
distress of others” (Herman 1993:81). Insofar as one has better epistemic
access to relevant features of a situation, one is able to do more good, given
a good will. But epistemic access is not relevant to the status of the will
On Herman’s account, then, empirical influences7 play an important
epistemic role for the good will, but this in no case affects whether a given
moral agent has a good will or not. Empirical influences may be necessary
for someone to become a moral agent at all, or they can enable someone
who already chooses rightly to perform more and better actions. But they
cannot help a morally responsible person to choose rightly. They cannot
help or hinder having a good will.
Nancy Sherman draws from Herman’s account but adds one other
way in which empirical influences – in particular, emotions – can serve as
epistemic helps toward moral improvement. She suggests that they can
provide better epistemic access not only to morally salient features of a
situation but also to self-knowledge:

The notion that emotions are modes of attending to moral salience may also
have application to the duty of self-knowledge. . . . Emotions can turn inward to
alert us of our own inner states, to help us to read, as he says, our own heart.
As such, they may be thought of as mediums by which we can become more
aware of our thoughts and motives. (Sherman 1997a:146; this passage is not in

This additional epistemic role of emotions is similar to Kant’s accounts

of the importance of politeness,8 and Kant does argue that this kind
of self-knowledge aids moral development. But it is important to notice
74 The Problem

that Sherman does not give us any account of how this aid would work,
nor even how it could be possible, given Kant’s account of freedom. In
particular, if one comes to think – based on insights into oneself provided
by the emotions – that one is in fact committed to the moral law, there is
no obvious reason why this awareness should affect one’s commitment.
One might even think that such knowledge would lead to an unhealthy
complacency in one’s moral status.9 Moreover, if one is committed to
some evil maxim, there is no obvious reason why becoming aware of that
commitment will lead to any kind of moral improvement. One might
simply regard it with indifference as a moral fact about oneself, or one
might be led to a moral despair that would undermine one’s commitment
to pursue the good.10

Nothing is this section should be taken to deny the important epistemic

role of empirical influences in general or of emotions in particular. Sher-
man and Herman enrich contemporary Kantian ethics by pointing out
the crucial roles of empirical influences as preconditions of moral agency
and as aids to the performance of beneficent actions. But neither of these
roles explains the distinctive role of empirical influences in moral an-
thropology. Neither explains how a moral agent can become morally better
through empirical influence.

2. Empirical Influences as Instruments

Rather than ascribing a purely epistemic role to moral influences, one
might think that they are necessary or useful means for acting well in
the world. A stable political order, for instance, provides a framework
within which one can make and honor promises, help others achieve
their morally worthy ends, and practice the virtues of self-cultivation. A
good education and certain basic material goods may be necessary or
at least very useful to make it possible for one to act in a variety of
morally praiseworthy ways. Certain emotions might be needed to min-
ister to a friend in need, to honor one’s duties to one’s family, or to
effectively serve one’s nation. On the other hand, empirical influences
that incapacitate or weaken one physically, mentally, and emotionally
can hinder one’s performance of one’s duties. One might argue that
in none of these cases do empirical influences help or hinder in a way
that violates the priority of transcendental freedom. They merely extend
or limit one’s abilities in ways that affect the extent to which one can
perform the deeds that are morally required. Thus this instrumental
Moral Anthropology in Neokantian Ethics 75

role for empirical influences seems to integrate moral anthropology and


In her account of the ways in which emotions support virtue, Sherman

details the way in which emotions in particular can affect moral life as

Not only do we read circumstances through the emotions, but we respond through
them. Manifest affect is a vehicle or mode for conveying moral interest. Through
emotions, we communicate or signal moral interest to others in ways tailored to
particular circumstances and needs. (1997a:147; 1997b: 273)11

Emotions are often means by which one can carry out duties to others.
When friends suffer, one’s efforts to provide for their happiness and one’s
obligations to them as friends require that one do more than merely
say compassionate things to them, spend time with them, and so on.
One ought to have and express certain emotions such as sympathy and
Sherman is certainly correct to point out the important role that emo-
tions can play in carrying out one’s duties, but there is an important
danger in taking this role too far. Given that emotions are not totally
within one’s control, Kant cannot allow them to have a necessary role in
the fulfillment of duties. Sherman rightly points out that Kant never says
that emotions are totally beyond one’s control.12 But for emotions to be
necessary for carrying out one’s duty, it is not enough that they be partially
within one’s control. They would have to be totally within one’s control.
Otherwise, one would be required to do something – feel sympathy, for
example – that in certain cases would be impossible for one. And this
would violate Kant’s fundamental principle that ought implies can.
Sherman at times seems to recognize that emotions themselves cannot
be necessary helps to carrying out one’s duty. As she explains,

We cannot ensure success in actually expressing the emotions that we think are ap-
propriate. For the Kantian, the most we can demand is good willing, that is,
sincere effort in certain relevant endeavors, constrained by the principles practi-
cal agency itself generates. If an agent tries to be sympathetic because she realizes
it is important for morality, but simply cannot make it happen – if she makes the
attempt part of her sincere maxim of what is relevant to helping another, but
can’t bring it off – is she subject to moral self-criticism? Presumably, a Kantian
would argue “no,” especially if the agent is one who, on the whole, takes seriously
a project of developing sympathetic feelings, exhorts herself to do better when
she fails, cultivates the right outward gestures, and monitors her behavior to see
how she is doing. A moral defect in character, on this Kantian view, shows up not
76 The Problem

so much in the finished product as in whether or not there is adequate striving

of a will toward the right ends. (1997a:184)13

Emotions may serve important roles in more effectively pursuing virtuous

ends, and as a result the cultivation of emotions may be a requirement to
satisfy one’s duties of self-perfection and beneficence, but the emotions
themselves cannot be required. More generally, anything that is useful
for carrying out one’s duty can be the necessary object of an imperfect
duty. If some empirical influences help one more effectively promote the
happiness of others, for instance, then one has a derivative obligation to
seek those influences.
Unfortunately, whether one emphasizes the importance of emotions
themselves or merely the cultivation of emotions, this role does not
capture the distinctive role of emotions in moral anthropology. Sherman
certainly highlights an important role that emotions play in moral life,
but this role is not relevant to whether or not one actually has a good
will. Rather, emotions, like much in the empirical world, can serve a will
once that will is already good. For carrying out its purposes, emotions are
important, and one therefore has a duty to cultivate them, just as phys-
ical strength can be important and one therefore has a duty to develop
physical strength. But one also has a duty to develop and control one’s
emotions in a variety of ways because such control can actually help one
acquire and sustain a will that can make proper use of them. Sherman
fails to show how emotions – or other empirical influences – can serve
this role.14 Empirical influences that serve as instruments to acting well
are important, but they do not have the special moral importance of the
helps and hindrances that are central to moral anthropology.

3. Empirical Influences as Constitutive

One might argue, however, that some empirical influences are important
not merely as means by which a will does what is good but also as constitutive
parts of a good will. Some “instruments” might be so enmeshed with acting
well, for instance, that to act well necessarily includes acting with those
instruments. Being a part of polite society may play such an important
role in acting well that part of what it is to have a good will is to act
politely and be part of a polite society. A just government, or at least
some minimal level of justice, may be so intrinsic to respecting the rights
of others that part of what it is to have a good will is to be a citizen under
such a government. Perhaps most plausibly, one might think that certain
Moral Anthropology in Neokantian Ethics 77

emotions are just so important for virtue that part of what it is to have a
good will is to have those emotions.

Nancy Sherman suggests this last view as an important way in which emo-
tions fit into one’s moral status. Sherman explains that “in the light of
this last claim [that emotions are “ways in which we act and convey moral
interest”] especially, I would argue that emotions are constitutive of virtue
in human embodiment” (1997a:158; 1997b:278). The notion that emo-
tions can be constitutive of a good will draws on Kant’s account of im-
perfect duties in the Doctrine of Virtue as well as his description of virtue
as a “hierarchically structured composite” in the Religion. In this section,
I examine Sherman’s suggestions about the way in which empirical aids
can be constitutive of the good will in the context of both the Doctrine of
Virtue and the Religion.
In the Doctrine of Virtue, Kant insists that one has at least two obligatory
ends, one’s own perfection and the happiness of others (see 6:385–6).
One’s own perfection includes anything that helps the good will either
in an epistemic sense as described in section 1 of this chapter or in an
instrumental sense as described in section 2. Insofar as emotions serve
either of these roles, they are necessary ends of a good will, and having
them may seem to be constitutive of a good will. Moreover, if emotions
are required for the moral awareness needed to be capable of choosing
either good or evil, then they are constitutive of a good will in that they
are constitutive of any morally responsible will at all.
As in the preceding two sections, I agree with Sherman that emotions
as well as other empirical aids are constitutive of a good will at least in
the sense that a will that fails to cultivate them fails to fulfill its imperfect
duties. Moreover, I grant that there may be some minimum level of emo-
tional sensitivity or other empirical assistance that is necessary if one is
to be morally responsible at all. However, Sherman sometimes takes this
insight too far. She suggests not only that some level of moral awareness is
necessary to be morally responsible and that a deeper level is a required
end for a good agent but also that some deep level of moral awareness is a
necessary attribute of a good agent. But all that a Kantian – or anyone who
takes seriously that ought implies can – should require is that one must
pursue, to an extent that may be different for different people, emotional
Moreover, Kant specifically distinguishes in the Doctrine of Virtue be-
tween instrumental aids for the good will – which would include not only
certain emotions, moral attentiveness, and so on, but also such things
78 The Problem

as physical strength and wealth – and specifically moral helps and hin-
drances. A person’s duty to perfect himself includes “a duty to cultivate
his natural powers (powers of spirit, mind, and body), as means to all sorts
of possible ends” (6:444). Some of the natural powers that one cultivates,
such as sensitivity to the pain of others, may serve primarily moral ends,
but none are uniquely moral.15 In contrast to these “natural powers,”
however, Kant points out that the duty to perfect oneself also requires
that one “increase his moral perfection” (6:446). Moral anthropology –
as opposed to anthropology that is primarily technical or pragmatic –
points out empirical helps and hindrances to this sort of perfection.
In this context, Sherman, Guyer (Guyer 1993, Chapters 9 and 10), and
others are correct to point out that the pursuit of various empirical aids
is constitutive of having a good will. And it is true that these empirical
aids are “helps” to having a good will in the sense that they are objects of
an imperfect duty to cultivate one’s natural powers. But these accounts
fail to capture the distinctive sense in which specifically moral helps and
hindrances affect the status of one’s will. It is constitutive of having a good
will that one interact with polite society or develop a sense of good taste
because these empirical influences can aid in the cultivation of a good
will. But to claim that they aid in this cultivation only in the sense that
their pursuit is constitutive of having a good will is to argue in a circle. We
can say that the pursuit of emotional sensitivity is constitutive of having a
good will16 because it is an important means for promoting the happiness
of others, which is itself a necessary end of the good will. And a good will
necessarily wills the necessary or helpful means to its necessary ends. But
some empirical influences serve not only as means to general ends that
a good will ought to have but as means by which one comes to have a
good will itself. These influences are constitutive of the good will, and
in that sense their pursuit is required, but this cannot be the end of the
story. They are means to having a good will prior to being constitutive of
such a will, and are constitutive only because they are a means. Saying that
they are constitutive does not solve the problem of moral anthropology.
It only pushes it to a new level.

4. Empirical Helps as Necessary Propaedeutics

One final possible role for empirical influences is as necessary propaedeu-
tics. This view is in some ways an expansion of the epistemic option
discussed in section 1. There we saw that certain influences could be
considered necessary means for formulating proper maxims. In that way,
Moral Anthropology in Neokantian Ethics 79

they seemed to provide important preparatory steps to being morally

good. But the fact that these were merely epistemic aids limited the sig-
nificance of their role. They could serve as basic requirements for some-
one to be morally responsible at all, such that to lack such aids would
preclude choosing good or evil. Or these epistemic aids could improve
the effectiveness of people already committed to choosing well. Thinking
of empirical aids as necessary propaedeutics broadens their scope by al-
lowing that even those who properly formulate maxims may need some
help in resisting temptations to act wrongly. This need would be due to
their human limitations and thus would not – or so some say – interfere
with freedom. On this account, empirical influences prepare free and
responsible agents to choose well. They are not sufficient conditions for
morally good choice, so there is still an important role for freedom, but
they are necessary propaedeutics.

Although various versions of the “propaedeutic” role have been offered in

recent years, this section focuses on Robert Louden’s use of the account.17
Louden repeatedly emphasizes the role of empirical helps as “neces-
sary preparations” for morality. To offer just a few examples, he claims
that “human morality does presuppose a sufficiently developed, inter-
connected web of cultural institutions as a necessary condition for its own
presence” (Louden 2000:21), that “moralization . . . necessarily presupposes
the preparatory steps of culture and civilization” (41), that “art serves
as an important and necessary preparation for morality in the lives of hu-
man beings” (109) and “the aesthetic is indispensible in cultivating hu-
man moral virtue” (116), that “history . . . ‘sets the stage’ for morality;
it presents us with ‘quasi-moral “veneers” that are not themselves wholly
moral’ but are necessary preconditions for morality” (149), that “the state
plays a necessary preparatory role in the development of morality” (169), and
that “political and legal progress are both necessary presuppositions for this
deeper moral progress, as are cultural and scientific advances, growth in
foreign trade; indeed as we have seen, even war itself” (160).18
In some of these discussions, it is not entirely clear to what the term
morality refers. It could be taken to mean simply that certain preconditions
are necessary before one can be capable of choosing either good or
evil. This claim would be similar to the important preparatory role that
Herman at times sees for rules of moral salience. In these cases, Louden
would be saying that some people cannot be held morally responsible
at all because they lack those aids that enable moral choice. The term
morality could also refer merely to the moral appearances. Certain empirical
80 The Problem

aids might be necessary for good wills to manifest themselves in certain

ways, or for more actions that conform to morality to be performed in
spite of people’s evil wills. “Moral” progress even might be taken to refer
to something that is merely a matter of the human species as a whole and
not also related to the moral status of individuals.
Louden clearly has in mind, however, a notion of morality and moral
progress that is individual, even if also present at the level of the species,
and that already assumes that agents who need these preparations are
morally responsible. Empirical influences enable already morally respon-
sible wills to choose in conformity with the moral law. He makes clear
that morality at least includes individual moral choice when he says, for
example, that

There is no guarantee that people who have been exposed to these preparatory
steps will be morally good, but human beings who lack all contact with them
cannot possibly be morally good. (Louden 2000:53)

Moralization is a matter of individuals being morally good. Moreover, this

moral goodness is not merely a matter of appearance. Louden makes this
clear in an extended discussion of how Kant can justifiably talk about true
moral progress throughout history. Louden points out that moral progress
involves a “transformation of the way of thinking” that takes place at the
level of an unobservable “radical innerness” (Louden 2000:151). He goes
on to suggest,

We must try the best we can to decipher the inner from the outer. And since
certain manifestations of the latter (e.g., law, art, culture, religious community)
are themselves necessary preparatory steps for the former, it is correct to say
that, from a human perspective, clear signs of outer progress also afford us clues
regarding inner progress. (Louden 2000:152)

We will come back to this passage again when I discuss a central problem
with Louden’s account, but for now it is important to note that various
empirical influences such as law, art, and culture are necessary prepara-
tions for an inner, moral conversion.
Louden ends his book with an assessment of Kant’s ethics after its
“impure” part has been reintegrated. He ends this assessment with the
question of the “internal coherence” of Kant’s ethics and specifically
questions whether “Kant’s commitment to transcendental freedom” is
consistent with his impure ethics (Louden 2000:180). Louden does
not offer a detailed response there. Instead he says, “I have demon-
strated the coherence of Kant’s concept of empirical ethics in this study”
Moral Anthropology in Neokantian Ethics 81

(Louden 2000:180). Has he? Does Louden’s account of empirical aids

as necessary preparations for having a good will really cohere with Kant’s
doctrine of transcendental freedom? After raising a basic objection to
Louden’s view, I consider the two most explicit responses that Louden
seems to offer in his book and show that neither of these resolves the
basic incoherence between his account of necessary preparations and
Kant’s account of transcendental freedom. I then show that the problem
of incoherence arises not merely when empirical aids are necessary prepa-
rations, but on any account according to which they are ultimately prior
in a practical sense to the good will.19
To put it simply, Kant cannot ascribe moral responsibility to any agent
that lacks what is necessary to act in accord with the moral law. As a
result, if Louden is correct that empirical aids – such as rational and cos-
mopolitan churches, a stable political order, and aesthetic sensitivity – are
necessary for one to be virtuous, then unless these aids are present, one is
not morally responsible. Freedom means – if it means anything for Kant –
that one can choose to obey the moral law regardless of empirical condi-
tions. There are no excuses. So empirical influences cannot be necessary
preparations for morality unless they are necessary for moral responsibil-
ity itself, and neither Kant nor Louden claims that all empirical aids are
necessary in that sense. There is a straightforward contradiction between
the claim that an agent is morally responsible and thus transcendentally
free and the claim that certain empirical aids are in addition necessary
for that agent to be morally good.
Throughout his book, Louden at times seems to tackle this contra-
diction, but in almost every case he actually addresses a narrower and
less significant problem. For example, at the end of his chapter “Edu-
cation,” Louden considers several objections to the claims that “culture
and education . . . are necessary . . . for human moralization” (Louden
2000:53) and that there must therefore be progress in human moral-
ity insofar as earlier generations must work to establish the culture and
education that are necessary preconditions of the morality of later gen-
erations. In this context, the contradiction between freedom and nec-
essary conditions arises in a particularly poignant way. It looks as if ei-
ther earlier generations – who establish the conditions of possibility of
morality – are not transcendentally free, or the “conditions of possibility”
are not really necessary conditions.20 Louden considers several objec-
tions that seem to reflect concern about this basic contradiction between
transcendental freedom and the necessary social conditions of moraliza-
tion. First, “Kant’s philosophy of education and history contradicts the
82 The Problem

individualist strand of his ethical theory which holds that each moral
agent has absolute moral worth” (Louden 2000:55). Second, there is “a
charge of moral unfairness” because past generations lacked even the
possibility to be good (Louden 2000:55). Third, there is the charge that
Louden explicitly calls “Education vs. Transcendental Freedom,” which raises
the question “how exactly is the crossing-over from nature to freedom
to be achieved, and how can two such qualitatively different realms in-
teract with one another?” (Louden 2000:56), and finally there is the
problem of moral luck, that “later generations are eventually forced into
a moral condition for which they themselves deserve no credit” (Louden

Unfortunately, Louden’s solutions to these problems avoid rather than

solve the real issue of tension between freedom and empirical influence.
In response to the first problem, Louden only makes the tension more
severe. He describes this problem in such a way that the conflict between
“progress” and “dignity” is a matter of a supposed conflict between valu-
ing people based on what they accomplish and valuing them based on
their intrinsic moral worth. Louden thus reconstrues the problem as a
conflict between Kant’s “teleological view concerning the true worth of
the species” and the view that a good will would have dignity “even if
[that] morally good will were not able to accomplish anything” (Louden
2000:57). His solution to this tension is to argue that insofar as the cul-
tural and political development of the species is itself a precondition of
human moralization, this conflict is only apparent. The dignity of human
beings presupposes the development of the species. Without the latter,
the former does not exist (Louden 2000:58).22 What Louden appears to
be saying is that humans at later stages of development are not valued
merely on the basis of the increase in good consequences, but that they
actually come for the first time at the end of history to have good wills.
Thus the two forms of evaluation do not conflict. But this response just
makes the tension between transcendental freedom and teleological his-
tory all the more problematic. It looks as if progress, which establishes
the necessary preconditions of morality, really does conflict with holding
past generations to have even the minimum dignity that allows us to hold
them morally responsible for their actions. And Kant, who insists that all
human beings are morally responsible, cannot think that.23 Moreover,
given that many of the supposed necessary conditions for moralization –
conditions like world peace and high-quality moral education – are still
not present, not even people today can be held morally responsible.
Moral Anthropology in Neokantian Ethics 83

With respect to the problem of unfairness, Louden suggests two re-

sponses. First, “once we recognize that complete perfection is an unattain-
able ideal, it becomes clear that no human being or group participates
fully in this destiny: we can only approximate it” (Louden 2000:58). Ac-
cording to this response, it looks as if there is no moral unfairness, because
all people and groups fail to achieve perfection, and all can in some way
approximate it. The problem, of course, is that insofar as this is a claim
about what is possible, Louden solves one problem – unfairness – by in-
troducing the unfortunate consequence that no one is ever even capable of
being morally good. But given that one is morally responsible on Kant’s
account only if one is capable of acting morally, no one is ever morally
Louden’s second response is to treat the issue of unfairness as one
of progressive happiness, where the misery of early generations makes
possible the happiness of later ones. It is not clear that Kant holds such a
view, but if he does, then Louden’s response is correct. Louden explains
that this unhappiness “is a matter of theodicy [in that it] makes possible
the progress of the species” (Louden 2000:58), where the progress in
arts, sciences, and (for Louden) morality is the final end of the species.
While this solution would be legitimate in the context of happiness, it
does not shed any light on the conflict between freedom and empirical
Louden’s response to the third problem, the most directly relevant
formulation of the tension, is the most troubling. Louden bites the

Kant does believe that efficacious moral education is education that somehow cuts
through the surface causal network in order to effect the grounding of character.
How this process works is something human beings cannot fully understand; we
cannot know intelligible character, nor can we ever know with certainty that our
attempts to shape and influence it are effective. But we can assume that such
efforts may succeed, and, indeed, this assumption is a necessary presupposition
of any program of moral education. (Louden 2000:59)

This is an absolutely amazing claim, and if true it would mean that Kant
rejects the asymmetrical relationship between the intelligible and the
empirical character that we saw in Chapter 1. If it is true that empiri-
cal influences can “cut through the surface network” to actually influ-
ence intelligible character, then Kant must give up the asymmetry in
his theoretical account of the relationship between appearances and
things in themselves as well as the practical insistence that freedom
84 The Problem

precludes determination by anything tainted with empirical influence.

Unless it is really a “necessary presupposition of any program of moral
education” – and I argue in the next two chapters that it is not – this
should be our last resort. And if it is a necessary presupposition, we
should just admit that Kant’s transcendental freedom is sacrificed at
the altar of impure ethics and not pretend that the conflict has been
In Louden’s response to the last problem, moral luck, he returns to
an emphasis on freedom and moral responsibility. He admits that “luck
plays a role in many areas of human life” but goes on to claim that “the
normative status of a person’s moral character is a function of how well
this person ‘has tied himself to certain practical principles that he has un-
alterably prescribed for himself through his own reason’ [7:292]. Here
the question is what we do to ourselves, not what nature and the environ-
ment have done to us” (Louden 2000:60). This is an accurate account
of what is required for moral approbation on Kant’s account. But it does
not fit well with Louden’s conception of aids as necessary preparations
for morality nor with his extreme claims that education can somehow cut
through the causal network to effect changes at the intelligible level. The
latter claim suggests that one’s moral education can – at least in part –
determine whether or not one is morally good. In this context, whether
one is lucky enough to have good moral education has a lot to do with
one’s moral status. It is not only a matter of what one does to oneself but
is also a matter of “what nature and the environment have done to us.”
And even to say that moral education is merely a necessary preparation
means that those who are unlucky enough to be denied moral education
or other empirical aids will not be morally good. Again, what nature or
the environment does to us would have a lot to do with our moral sta-
tus, a conclusion that would conflict with Kant’s conception of human

One final proposal that Louden occasionally offers to reconcile transcen-

dental freedom with empirical propaedeutics is somewhat more promis-
ing. He says,

Culture (in particular the arts and sciences) and education are, along with law,
politics, and religion, all necessary but not sufficient conditions for human mor-
alization. There is no guarantee that people who have been exposed to all these
preparatory steps will be morally good, but human beings who lack all contact
with them cannot possibly be good. (Louden 2000:53, emphasis added, see too
pp. 149–50, 159)
Moral Anthropology in Neokantian Ethics 85

Louden suggests that because empirical influences are not sufficient con-
ditions for morality, they do not preclude transcendental freedom. Re-
gardless of how favored persons are in terms of empirical aids to morality,
they still will not be morally good unless they choose to be morally good.
In fact, Louden points out, Kant thought that in the case of peace as
an empirical aid, “peace can (and according to Kant will ) be attained
without a fundamental change in moral character occurring” (Louden
This reconciliation of necessary preconditions with transcendental
freedom does successfully solve the contradiction, but only in those cases
in which all the necessary preconditions are present. By insisting that no con-
ditions are sufficient, Louden can preserve transcendental freedom and
hence moral responsibility for all those people who may live at that won-
derful time in which all necessary preconditions are present. But those liv-
ing without those advantages are not transcendentally free and hence not
morally responsible. And perpetual peace – to take just one empirical pre-
condition – is not yet available to people today. This would imply the un-
tenable conclusion that no one now or at any point in the past can be held
morally responsible, because good willing is not even an option for us.

So far I have argued that Louden is mistaken to take empirical aids to

be necessary preconditions of morality. There may be some preconditions
to moral responsibility, but these must be minimal enough to allow us to
hold most human beings today and in the past morally responsible. Kant
clearly thought that the basic cognitive and motivational requirements
for being morally responsible at all are not very rare. He insists that

The moral law commands compliance from everyone, and in fact the most exact
compliance. Appraising what is to be done in accordance with it must, therefore,
not be so difficult that the most common and unpracticed understanding should
not know how to go about it, even without worldly prudence. (5:36)

By everyone in this passage, Kant presumably can exclude most nonhuman

animals, some children too young to be held morally responsible, and
perhaps even people with certain severe forms of insanity or disability.
But he cannot exclude everyone who lacks all the advantages that Louden
suggests are necessary. To make all empirical aids necessary is simply to go
too far.

But even the claim that empirical aids are propaedeutics of any kind fails to
adequately address the problem of transcendental freedom. First, to say
86 The Problem

that empirical influences are preconditions that “somehow cut through

the surface causal network” to exercise an effect on one’s intelligible
character does not explain how this influence can be consistent with
our freedom. Louden insists that “how this process works is something
human beings cannot fully understand” (Louden 2000:59), an appropri-
ate Kantian move given the limits on knowledge of the intelligible. But
Louden must at least show that this influence does not contradict Kant’s in-
sistence that the will is free. Kant himself, in resolving the antinomies, did
not simply assert that certain things are beyond human understanding.
He showed how two apparently contradictory theoretical claims could be
made consistent. Louden seems to recognize the apparent contradiction
between two practically grounded claims – the transcendental freedom of
the will and the potential for that free will to be empirically influenced –
but he just dodges the central issue of how they can be reconciled. The
claim that empirical aids are “propaedeutics” does not resolve the prima
facie contradiction in Kant’s ethics.
The suggestion that there is an important practical sense in which
empirical aids, as “propaedeutics,” are prior to moral choice raises serious
problems for making sense of Kant’s descriptions of the relation between
the free will and its manifestation in the world. These problems arise
in a particularly striking way in Louden’s most detailed account of the
relationship between inner moral transformation and external historical
progress (Louden 2000:143–52). There, Louden explains that

Moral character, at least on Kant’s view, is not about external events in time
and space: indeed, it is not even about an internal process taking place in time
and space. The latter is still only empirical character; and morality . . . is about
intelligible character. . . . Human beings are cognitively limited creatures. Given
these cognitive limitations, we must make do with what we have; we must try as best
we can to decipher the inner from the outer. And since certain manifestations of
the latter (e.g., law, art, culture, religious community) are themselves necessary
preparatory steps for the former, it is correct to say that, from a human perspective,
clear signs of outer progress also afford us clues regarding inner progress. . . .
Human beings have no choice but to try to read inner moral character from its
outer, empirical manifestations. (Louden 2000:152)

Louden seems not to appreciate the conflict between thinking of em-

pirical aids as “manifestations” of inner moral character and thinking of
them as “necessary preparations” for moral character.
For Kant, looking to someone’s “deeds” – which include not only
purely external processes but also internal temporal processes – is the best
way that one can “try and read” that person’s moral character precisely
Moral Anthropology in Neokantian Ethics 87

because those visible processes are always the expression of moral charac-
ter. Insofar as empirical influences are preparations for morality, however,
they fail as expressions. A precondition of a good will, especially if it is
a necessary precondition, cannot rightly be considered an expression of a
good will.

In sum, Louden’s account of the role of helps and hindrances fails on

three counts. First, insofar as he suggests that empirical aids are necessary
preconditions, he undermines Kant’s most basic convictions about the
nature of freedom. Second, insofar as he merely points out that empirical
aids are “preconditions,” he fails to enrich our understanding of their
role sufficiently to show that there is no conflict between these aids and
transcendental freedom. Third, by giving practical priority to these helps
and hindrances, he undermines the relation wherein what is empirical
manifests one’s moral status.

5. Munzel’s Alternative Account of Helps as Propaedeutics

Like Louden, G. Felicitas Munzel offers an account of empirical influ-
ences as propaedeutics. And like Louden’s, her account is explicitly of-
fered in the context of integrating Kant’s anthropology into his moral
theory in general and his account of freedom in particular. At times,
Munzel’s account of the propaedeutic function of helps and hindrances
seems virtually identical to the account offered by Louden. She portrays
the republican constitution, for example, as an “indispensable vehicle
of moral cultivation” (Munzel 1999:259, emphasis added) and follows
this up with an account of how this necessity does not interfere with
freedom because it is not sufficient for establishing morality. She de-
scribes discipline in education as an “indispensable first step” (Munzel
1999:280, emphasis added). She names at least some empirical condi-
tions “necessary (albeit not sufficient) conditions for producing char-
acter” (Munzel 1999:340). And, she claims, “the pedagogical process
complete with propaedeutic functions which prepare and develop these
aptitudes, even initially awakening them to a state of moral responsive-
ness, is deemed by Kant to be essential ” (Munzel 1999:340–1). All of this
might lead one to think that Munzel, like Louden, offers an account of
empirical aids as necessary but not sufficient propaedeutics to having a
good will.
But in other places Munzel’s account is more nuanced. She describes
the cultivation of natural aptitudes not as necessary, but as a “central factor”
88 The Problem

for developing a good character. In the course of outlining hindrances

to morality, she suggests that “life in even the so-called civilized state can
present . . . nearly insurmountable obstacles” (Munzel 1999:177, empha-
sis added). In the presence of numerous hindrances to morality, “condi-
tions in the state are not conducive for the successful cultivation of inner
peace within . . . individuals” (Munzel 1999:179, emphasis added). Apti-
tudes for good “serve to promote the observance of the law” (Munzel 1999:
110, emphasis added). “Discipline and cultivation prepare the conditions
most conducive to the formation of character in its strict sense” (Munzel
1999:332, emphasis added). In these cases, Munzel backs off from strong
language of necessity, pointing out that helps merely help, rather than
enable, and hindrances merely hinder, rather than prevent, developing a
good will.
If this were the only difference, Munzel’s account would be better
than Louden’s in that it would not be susceptible to objections against
the notion of a necessary propaedeutic. But her account would still seem to
imply that natural influences “cut through the surface causal network”
to influence one’s free will, and – like Louden – she does not explain
how this can be reconciled with Kant’s asymmetry between freedom and
nature. But Munzel provides several important illustrations of the re-
lationship between the good will and empirical aids that suggest a dif-
ferent conception of the relation between the two. I will not discuss all
the details of her account here but will focus on the two most fruitful
analogies she offers for understanding how empirical influences relate
to the free choice that ultimately establishes a morally good character in a
The first analogy is medical. Munzel suggests that empirical aids can be
viewed as treatments for the symptoms of a lack of morally good character.
This analogy is particularly significant in the context of the helpful dis-
cussion of the relationship between empirical and intelligible character,
a distinction that is important for my own account of the relationship be-
tween empirical aids and moral formation. Empirical character, Munzel
points out, can be thought of as the effect of intelligible character. In-
sofar as one’s intelligible character is evil or unformed, one will have
various symptoms at the empirical level, symptoms that can be treated by
empirical means. Munzel says,
Reflection on Kant’s own metaphor likening the virtuous comportment of mind
to a moral state of health leads one to recognize that it is not at all unusual
in cases of illness for treatment to begin with the mitigation of the symp-
toms, for fever and swelling to be reduced (for example), before surgery is
Moral Anthropology in Neokantian Ethics 89

undertaken to correct the actual source of these effects or symptoms. By ex-

tension, it is no more necessary to assume that the external discipline of the
inclinations executed under just, civil government has become Kant’s basis for
the formation of good moral character, than one would assume that the exter-
nal application of ice pack and anti-inflammatory medications would correct
or even begin to address the underlying cause in a diseased organ. (Munzel
This analogy is markedly different from Louden’s account of the role of
propaedeutics in that it need not imply an actual influence of empirical
aids on moral choice at the level of free choice. The “diseased organ” –
the evil intelligible character – does not seem to be affected at all by the
empirical influences, which merely treat the “symptoms”: the empirical
character that is manifested in various evil deeds.
Unfortunately, it is not at all clear in what sense this treatment of symp-
toms is morally significant. It is, of course, important that the number of
evil deeds in the world be reduced. Because one ought to be concerned
about the happiness of others, one must seek to reduce those deeds, per-
formed by oneself or by others, that interfere with this happiness. But if
treating symptoms is merely treating symptoms, if it does not at all relate
to the diseased organ, then these empirical influences seem to have no
moral significance at all.
Munzel seems to recognize this problem, and she later recasts treating
the symptoms in a more significant light. She continues to insist that
treatment of symptoms is never the basis for moral improvement (see,
e.g., Munzel 1999:322), but she suggests that

The lawless exercise of human nature appears feverish and swollen and calls
for discipline as the propaedeutic first step in bringing about a state of moral
health by alleviating the symptoms. Only then are the human capacities fit for,
able to undergo, their positive cultivation, which is the next step. (Munzel 1999:

If alleviating the symptoms is a first step, such that a person is ready

to be healed only after such alleviation, then helps and hindrances
are morally significant. Unfortunately, this moral significance is bought
with a price. In particular, an illustration that at first seemed to pro-
vide an important alternative to Louden’s account of necessary but
not sufficient conditions collapses back into that account. This anal-
ogy gives a different interpretation of the way empirical influences are
necessary conditions – an interpretation that takes much more seri-
ously Kant’s distinction between intelligible and empirical character –
but it is susceptible to the same objections.
90 The Problem

Munzel’s second illustration avoids this problem. She describes the

analogy in terms of an oath of office. She says,
The relation of the formation of character . . . to discipline and cultivation [i.e.,
empirical influences] may be understood in terms of an analogy with the relation
of preparatory training to taking an oath of office. In this case, of course, the
office is that of the vocation of our humanity. No amount of schooling equipping
us (in our subjective capacities) to carry out the office well, even if essential for
conducting ourselves well in office, can necessitate our taking the oath of office, our
committing ourselves consciously and freely to assuming personal responsibility
for fulfilling it. This choice to make it our own vocation, to make the promise (in
effect) to ourselves and to others that we will duly perform it, can only be our
own. It is not a degree ceremoniously conferred upon us that we may passively
receive at the conclusion of our studies; we cannot be made or pronounced by
another to be a person of good moral character. (Munzel 1999:331–2)

So far, this description of the relationship between the training and the
oath seems to imply that the training is a necessary but not sufficient
condition of taking the oath. Munzel’s emphasis is on the fact that no
amount of training can be sufficient. But she goes on to insist that training
is not necessary either:
In principle, consistent with Kant’s description of the adoption of this resolve as
a conversion or transformation in conduct of thought and with his fundamental
premise of freedom, there is in fact nothing to prevent someone from taking
this particular oath of office in the absence of preparatory schooling and then
acquiring the latter as a kind of on-the-job training. Pragmatically speaking, to do
so would be to take the hard road . . ., but it would not be per se impossible. (Munzel

Munzel’s attention to Kant’s “fundamental premise of freedom” as a con-

straint on her own account of propaedeutics leads to an important in-
sight. It must be possible to be morally good – to take the “oath” – prior
to any empirical aids. This passage also suggests what is certainly correct,
that these empirical aids are still important subsequent helps, as “on-the-
job training.” Unfortunately, Munzel does not adequately distinguish this
training from the sorts of roles in Sherman’s accounts of empirical influ-
ences. Moreover, her account is still permeated with temporal language
that confuses the sort of priority that the oath taking must have. As a re-
sult, she still seems to think of the sort of relationship between empirical
influence and moral choice that she describes here as the exceptional
case, the “hard road.”

All of the authors discussed in this chapter offer profitable contribu-

tions to Kantian ethics, but none offers a satisfying account of how moral
Moral Anthropology in Neokantian Ethics 91

anthropology can be integrated with Kant’s views about human freedom.

Sherman and Herman present important analyses of various roles that
helps and hindrances – especially emotions – can play in moral life, but
none of these roles fits the unique importance of empirical influences in
moral anthropology. In particular, none shows how the presence or ab-
sence of helps and hindrances can make a difference for the moral status
of someone already morally responsible. Louden attempts to address this
issue head-on, but in the process he seems to sacrifice transcendental
freedom to preserve the importance of impure ethics. Munzel offers the
most promising approach, but she fails to clearly articulate in detail the
relationship between one’s moral commitment and empirical influences.
In the next chapter, I offer an account of the relationship between the
good will and empirical influences that shows how these influences can
have genuine moral significance without undermining the fundamental
priority of freedom in Kant’s moral theory.
part ii


Transcendental Idealism, Radical Evil,

and Moral Anthropology

Over the past several chapters, I have shown that there seems to be a deep
problem in Kant’s moral philosophy, one that Schleiermacher recognized
and to which he drew attention. This problem is not limited to Kant but
is shared by any moral theory that takes seriously a strong, nondetermin-
ist conception of human freedom as a condition of moral responsibility
while at the same time recognizing the obvious importance of helps and
hindrances to moral development. Essentially, Schleiermacher points out
an apparent conflict between three fundamental assertions: that the will
is transcendentally free, that anthropology is empirical, and that anthro-
pology studies morally significant helps and hindrances. Because Kant so
forcefully articulates the nature of freedom, the problem arises for him
in a particularly stark way. The fact that Kant’s recognition of helps and
hindrances has been ignored for so long only heightens the sense today
that this recognition is incompatible with the rest of his theory. Moreover,
as we saw in the previous chapter, recent attempts to make sense of these
helps and hindrances either deny their full moral significance or sacrifice
Kant’s strong conception of freedom.
In this chapter, I offer a Kantian solution to Schleiermacher’s dilemma.
In section 1, I show that Kant has the resources to distinguish between the
empirical will, which can be affected by empirical influences, and the free
will, which cannot.1 In section 2, I explain that changes in the empirical
will are morally relevant as expressions of the moral status of the free will.
In section 3, I argue that the Kantian notion of “radical evil” requires
rethinking the nature of the good human will, and grace makes such a
rethinking possible. In section 4, I show that proper expressions of a good
human will involve attention to considerations of moral anthropology.
96 The Solution

Thus moral anthropology is morally relevant because it describes aspects

of human life that are important expressions of a good human will.

1. Kant’s Crucial Distinction

Unfortunately, Kant does not offer an extended, systematic treatment of
the relationship between freedom and moral anthropology. However, his
mature moral theory does provide the resources for solving the problem
raised by Schleiermacher, and Kant provides scattered but important re-
marks that directly address it. The most direct treatment is a footnote in
the Critique of Judgment. There Kant says,

One of the various supposed contradictions in this complete distinction of natural

causality from the causality through freedom is given in the following objection
to it. It is held that when I talk about nature putting obstacles [Hindernissen] in the
way of the causality governed by laws of freedom (moral laws), or about nature
furthering it, I do after all grant that nature influences freedom. (5:196)

Here Kant sketches a criticism similar to the one that Schleiermacher

raises several years later.2 The helps and hindrances that are a particular
problem for freedom are those that are part of nature, which for Kant is
just the empirical world. The problem of nature’s influencing freedom
is the problem of empirical helps and hindrances’ influencing moral
choice. To that problem, Kant offers the following solution:

This is a misinterpretation, which is easily avoided merely by understanding what

I have said. The resistance or furtherance is not between nature and freedom, but
between nature as appearance and the effects of freedom as appearances in the
world of sense; and even the causality of freedom (of pure and practical reason)
is the causality of a natural cause (the subject, regarded as a human being and
hence as an appearance) subject to the [laws of] nature.3 It is this causality’s
determination whose basis is contained, in a way not otherwise explicable, in the
intelligible that is thought under freedom ( just as in the case of the intelligible
that is the supersensible substrate of nature). (5:196)

Kant’s explanation of the relationship between freedom and empirical in-

fluences depends on his distinction between phenomena and noumena,
and specifically on his claim that each person exists both as a phenom-
enal series (nature) and a noumenal (free) ground of that series.4 The
solution to Schleiermacher’s dilemma is that moral responsibility and
hence freedom apply to oneself as a free ground of appearances, while
helps and hindrances affect only one’s “natural” appearance.5
Radical Evil and Moral Anthropology 97

This solution, however, is no more than a preliminary gesture. In par-

ticular, it does not explain why, if empirical influences cannot affect the
free self, they are morally relevant. One can be good or evil only insofar
as one is free. A series of appearances is relevant to morality only insofar
as that series expresses one’s free will. If empirical influences change the
way one appears without changing the free will that grounds this appear-
ance, then the appearances are not pure expressions of the free self – or so
it would seem. Unless empirical influences can change the moral status
of the free will, any effect these influences have on a series of appearances
seems to be morally secondary. If empirical influences actually are rele-
vant to one’s moral status, as Kant suggests in his anthropology, then they
must have some important relation to oneself insofar as one is free. They
cannot just influence how one appears.
Thus this important footnote does not completely reconcile transcen-
dental freedom with moral anthropology. It rightly points to the relation-
ship between the free will and the appearance of that will as the core of
the solution, but Kant still needs an account of how appearances relate
to the free will. Given the inscrutability of freedom, this account cannot
be a metaphysical one. But it also need not be metaphysical. All that Kant
needs to show is why one should be morally concerned with those appear-
ances that can be affected by empirical influences. In particular, he needs
to show why one should be morally concerned about the sorts of changes
involved in moral anthropology.

2. The Moral Importance of Appearances

A. Connecting the Good Will to Its Appearance in the World
A good will is a will that acts in accordance with the moral law. In the
Groundwork, Kant focuses his attention on the concept of “duty, which
includes that of a good will, though with certain subjective restrictions
and hindrances” (4:397). For such a will, action in accordance with the
moral law is action in accordance with a categorical imperative. From
this categorical imperative are derived “all imperatives of duty” (4:421, cf.
4:422–3, 4:429–30, and 6:205–491). For the purposes of understanding
the relationship between a good will and its expression, the details of
these derivations are less important than the nature of the imperatives
that are derived. Action for the sake of one’s duty characterizes the good
human will, but the imperatives of duty are themselves directed toward
appearances. In that sense, action in appearance that conforms to certain
standards characterizes the good free will.
98 The Solution

For example, in the Groundwork Kant explains that one ought never act
on a maxim of the form “When I believe myself to be in need of money,
I will borrow money and promise to pay it back, although I know that I
can never do so” (4:422). One is morally good only by refraining from
false promises out of respect for the moral law. This respect for the moral
law is the deliberate and constant commitment to act in such a way that
one’s maxims can conform to the moral law.
But although neither one’s maxim nor one’s commitment to the moral
law is itself an appearance, both are directed toward appearances. A maxim is
a rule for action that specifies a certain action or type of action as a means
to an end or type of end in a certain situation or type of situation. For ex-
ample, take the maxim “When I believe myself to be in need of money, I
will borrow money and promise to pay it back, although I know that I can
never do so” (4:422). Here the action is promising, the end is acquiring
money, and the situation is having certain beliefs about one’s needs and
inability to repay. Every aspect of this maxim is tied to appearances. The
action is an action that will take place in appearances. A purely noume-
nal “act” is not likely to convince anyone to give money, because others
cannot see or hear or experience this act. A promise is something that
can be heard or otherwise perceived. Likewise, the acquisition of money,
the desire to acquire money, the inclinations that are satisfied by hav-
ing money, the pleasure of their satisfaction, and the beliefs about one’s
needs and abilities are all appearances. Although inclinations, desires,
pleasures, and beliefs may appear only in inner sense and might not be
observable to anyone other than the one who has the desire, inclination,
pleasure, or belief, they are nevertheless experiences.
Now the moral status of an agent is not purely a matter of the action,
the end, and the situation that together form the content of the agent’s
maxims. Rather, Kant insists that the form of the maxim determines its
moral worth. One might perform the same action, or act for the same
end, or act in the same situation, without moral fault. Fault enters because
of the relation between the action, end, and situation.6 And one might
think that even if every aspect of the content of a maxim is appearance
directed, the form of the maxim is not.
In fact, however, concern with the form of a maxim is concern about
the structure of one’s actions in the world. The conformity of a particular
maxim to the moral law is not reducible to the content of the maxim,
but neither is it separable from that content. To have a good will is to
choose to act on certain sorts of maxims rather than on others. And the
sorts of maxims on which one acts describe actions in the world for ends
Radical Evil and Moral Anthropology 99

in the world in situations that arise in the world. Morally upright agents
are deeply concerned with appearances because effecting certain sorts of
appearances is the way in which one expresses one’s moral goodness.
It is important to note here that the term expression is used in the
sense in which one expresses oneself through the clothes one wears, or
the music to which one listens. This sort of self-expression is not the
revealing of an already determined nature. One becomes the self one is
through one’s self-expression.7 In the moral case this is true as well. When
one expresses one’s goodness by refraining from false promising out of
a sense of duty, one does not merely show a goodness that was “already
there.” Instead, having a good will is a matter of choosing in certain
ways. Refraining from false promises, when this flows from respect for
the moral law, is a choice of the morally praiseworthy sort. By choosing in
this way, one expresses one’s moral status in the sense that one chooses
to be a morally good person.
This emphasis on expressing one’s moral goodness may seem to shift
the focus of moral attention from the moral law to oneself. Given that Kant
says that one must act purely “from respect for [the moral] law” (4:400),
this emphasis on oneself may seem inappropriate. In fact, however, acting
from respect for the moral law involves performing deeds in the world
(acting) in order to meet the demands of morality. And one always focuses
on making one’s own actions, not just actions in general, conform to the
moral law. In that sense, a person who acts so as to express his or her
moral goodness is one who acts out of respect for the moral law.
This model of the relationship between the good free will and its ap-
pearance helps to make sense of why one cares morally about one’s ap-
pearance in the world. As the expression of one’s noumenal will, one’s
appearance in the world is morally relevant in the strongest possible
sense. It is by being a certain sort of appearance that one shows oneself
to be a certain sort of free agent. Not every feature of one’s appearance is
relevant to the kind of free agent that one is. One’s physical appearance
and natural temperament, for example, do not express one’s freedom.
But all voluntary actions that are morally relevant8 express one’s free will,
even though these actions are also the effects of empirical causes.
This account still has two remaining problems, to which I will now
turn. First, it is not clear how the sorts of concerns that are central to
moral anthropology are relevant to expressing moral goodness. I ex-
plain this problem and its solution in more detail in sections 3 and
4. Second, and of more immediate concern, Kant often suggests that
there is an insuperable barrier to inferring anything about one’s ultimate
100 The Solution

moral status from one’s appearance in the world. This barrier might seem
to undermine any connection at all between the will and its expression.
Before turning to the specific issues raised by moral anthropology, I de-
vote the rest of this section to the general problem of the inscrutability
of one’s moral status.

B. The Inscrutability of One’s Moral Status

In the midst of a discussion of freedom in the Groundwork, Kant claims that
“we shall never be able to comprehend how freedom is possible” (4:456).
This concern about inscrutability arises within both Kant’s theoretical and
his practical philosophy. The theoretical limitations to knowledge about
the relationship between the free self and its appearance were discussed
in detail in Chapter 1. There we saw that although Kant can appropriately
describe the free self as the cause or ground of its appearances, he cannot
go further in describing the nature of this causal relation. The category
of causation is invoked without a schematization. Thus it gives no true
“knowledge” – by which Kant means theoretical knowledge – of the self.
This theoretical limitation, however, need not interfere with practical
knowledge of a nontheoretical kind about the relationship between the
appearance of a self and the moral status of that self. However, it is crucial
that any causal language about the relationship between the free will and
its expression be purely practical.
A further challenge comes from Kant’s practical philosophy. Given
that the only “knowledge” of freedom that is possible is practical, it is
crucial to know what limits Kant sets on knowledge of freedom from the
standpoint of morality. We will find that one can know enough about
the relationship between one’s appearance and one’s moral status to
justify ascribing moral weight to influences on one’s appearance. In Kant’s
practical philosophy, there is no general claim that one cannot know
anything about one’s free self from the appearances of that self. But Kant
sometimes seems to suggest that one cannot know the most important
thing – one’s moral status – on the basis of appearances. In this section,
I show that while Kant does insist that one can never know for certain
that one is morally good, there is quite a bit that one can be justified in
believing about one’s moral status from how one appears.
The first Critique shows that one cannot have theoretical knowledge
of the noumenal self, but Kant asserts on the basis of practical reason
that one is free. Although this postulate is called mere “belief” or “faith”
(Glaube), it is absolutely certain from the standpoint of practical reason.
Thus it could be called “knowledge” in a practical sense, to distinguish it
Radical Evil and Moral Anthropology 101

from mere beliefs not held with practical certainty. The fact that one can
know (in this sense) that one is free raises the question of how much more
one can say about this free self on the basis of practical reason. One might
think that one could at least know whether or not one is morally good,
but Kant denies this. However, one can be certain that one is radically
evil, and one can have reasons that justify a rational hope that one is good.
Kant denies knowledge of one’s moral goodness for two reasons. The
first is an epistemological point about moral awareness: We cannot know
our own maxims, nor our volitional dispositions, with certainty. The sec-
ond is more specifically moral: If one could be certain of having a good
will, that certainty might lead to a moral complacency that would under-
mine such a will. Whereas a free will is necessary to make sense of moral
action, knowledge of one’s good will actually undermines moral action.
This practical limitation on self-knowledge will also turn out to be cru-
cial for making sense of Kant’s account of helps and hindrances. Neither
epistemic nor moral considerations, however, undermine knowledge of
one’s own radical evil or of the fact that one can have some reasons for
some confidence in one’s goodness. They limit only the extent to which
one can be sure that one is good.

Kant insists that we lack epistemic access to those aspects of ourselves

that would allow for certainty regarding our moral goodness. The most
famous declaration to this effect is in the Groundwork:

In fact there is absolutely no possibility by means of experience to make out

with complete certainty a single case in which a maxim of an action that may in
other respects conform to duty has rested solely on moral grounds and on the
representation of one’s duty. (4:407)

For Kant, neither the consequences of one’s actions nor the actions them-
selves, either of which could be empirically verified, determine one’s
moral status. Instead, what matters are the reasons for acting, the maxims
according to which one chooses. And there are no actions for which one
cannot have a bad reason. As Kant makes clear in the Religion, even when
one is evil, one’s “actions can still turn out to be as much in conformity
to the law as if they had originated from true principles,” such that “the
empirical character is then good but the intelligible character still evil”
Moreover, although observations of one’s inner life can reveal
thoughts, feelings, and even patterns of deliberation, there is no guaran-
tee that they will reveal the actual reasons for action. This is especially true,
102 The Solution

as Kant suggests in the rest of the passage from the Groundwork, because
people have a very strong ulterior motive to present their actions as con-
forming to morality.

We like to flatter ourselves with the false claim to a more noble motive; but in fact
we can never, even by the strictest examination, completely plumb the depths of
the secret incentives of our actions. For when moral value is being considered,
the concern is not with the actions, which are seen, but rather with their inner
principles, which are not seen. (4:407)

Human beings are adept at self-deception, especially when that deception

can soften the bite of moral self-reproach.9 This deception takes many
forms, and one of the most important is the presentation of one’s motives
as more noble than they truly are. One is frequently able to mislead even
oneself about the true maxims of one’s actions, covering one’s wickedness
with a whitewash of illusory good intentions.10 Self-examination can never
yield a final verdict on the moral status of one’s will because there may
always be hidden or ignored motives at play.
In the Religion, Kant explains a further difficulty with discerning moral

A human being’s inner experience of himself does not allow him so to fathom the
depths of his heart as to be able to attain, through self-observation, an entirely
reliable cognition of the basis of the maxims which he professes, and of their
purity and stability. (6:63, emphasis added)

To be truly good, at the level of the free will, one’s maxims must be not only
pure but also stable. Even if self-examination could somehow overcome
self-deception and complacency to reveal the purity of one’s maxims,
no action can provide reliable assurance that one’s will is stably oriented
toward the good. The moral status of one’s will thus remains in doubt.

This inscrutability does not distress Kant, however. On the contrary, he

notes an important practical benefit in remaining somewhat ignorant
of one’s moral status. Kant cautions that “it seems never advisable to be
encouraged to such a state of confidence [in one’s moral goodness] but
much more beneficial (for morality) to ‘work out one’s salvation with
fear and trembling’ ” (6:68). Good actions and good intentions offer
agents some reasonable hope that they have a good will. Kant says only
that one always lacks an “entirely reliable” cognition of the bases of one’s
maxims. If one constantly acts in apparent accordance with and out of
apparent respect for the moral law, one can reasonably hope that one is
genuinely good. One might be deceiving oneself, but this hope is enough
Radical Evil and Moral Anthropology 103

to ground a connection in deliberation between moral status and actions

in the world.
But the ultimate inscrutability of moral status erodes overconfident
complacency in one’s good will. The limit to self-knowledge actually pro-
motes the maintenance of a good will. This consideration alone could
not of course prove that one is epistemically limited, but it can help ex-
plain the importance for Kant of reminding his readers of the fact. And
in this context, it is clear that inscrutability does not undermine the im-
portance of appearances in moral action. One still seeks to effect the
sorts of appearances that express a morally good will. But because no
appearance is an incontrovertible expression of a good will, one never
rests complacently in one’s moral goodness. Moreover, as we will see in
the next section, there is at least one sense in which one’s moral status
is not inscrutable. And it turns out that what one can know about one’s
moral status provides a basis for being morally concerned about the sorts
of considerations that arise in moral anthropology.
Thus appearances are morally important as expressions of one’s free will.
One evaluates and deliberates about one’s appearance in the world. Even
though one can never be sure that any particular action is an expression
of a good will, the possibility of justified hope is sufficient to motivate
one to act well. This account of the moral importance of appearances
has the advantage of maintaining Kant’s asymmetry between nature and
freedom. Because appearances matter only as expressions of a free will but
do not effect changes in that free will, the relationship is asymmetrical.
Moreover, because this expression relation is purely a matter of moral
responsibility, in that an expression is an expression of a moral status;
it does not transgress the theoretical limitations on knowledge. Finally,
because the expression itself is empirical, the account offered here shows
how at least some empirical facts can be morally relevant. But I still have
not shown how this account applies to moral anthropology in particular.
The next two sections accomplish that task.

3. Toward the Moral Importance

of Moral Anthropology: Radical Evil
Kant’s account of the connection of the good will with appearance-
directed maxims provides a reason for concern about appearances, and
even the limitations on knowledge of one’s moral status do not under-
mine this concern. But this concern seems to be justified only in those
cases when one’s appearance reflects a maxim that does not conform
to the moral law, or a lack of respect for the moral law in the choice of
104 The Solution

one’s maxim. The account offered so far shows only that one should care
about acting, in the realm of appearance, in accordance with the moral
law. But the sorts of considerations that are part of moral anthropology
are not themselves directly required by the moral law.11 In the next sec-
tion, I explain why one has an obligation not only to act in the realm of
appearance out of respect for the moral law but also to take into account
moral anthropology to promote aids and avoid obstacles to morality. For
that purpose, however, one must first take a closer look at what can be
known about one’s moral condition. Kant’s most extensive discussion of
this is in his Religion within the Boundaries of Mere Reason.12

A. The Proof of Radical Evil

In section 2, we saw that Kant imposes stringent limits on what one can
know about one’s moral condition. Kant’s remarks on the inscrutability of
one’s moral status seem to be very general. It seems as though one of the
most important facts about the relationship between a free will and the
series of appearances that is its expression is that one can never be sure of
anything about the free will from the details of its appearance. But Kant
is careful to make a narrower claim. Kant’s discussions of inscrutability
focus on the impossibility of knowing that one is morally good.13 Thus in
the Groundwork he specifically argues that one might always have a maxim
of self-love that underlies an action apparently done from duty alone. In
the Religion, his claim is that inner experience does not allow one to know
the “purity and stability” of one’s maxims. One cannot know that one’s
maxims are good. And the state of confidence against which he warns is a
state of confidence in one’s moral goodness. Human self-deception tries
to make one seem better than one truly is. Kant seeks to silence those who
assert their moral virtue, not to deny justification for any moral claims at
all about oneself.
In fact, in his Religion, Kant makes a rather startling claim, made even
more startling by the fact that his support for it is empirical.14 He claims
that human beings are “radically evil.” In the context of this discussion,
Kant emphasizes the difficulty of judging moral status from the mere
appearance of the will, but he goes on to explain how one can show
moral evil from appearances. His reasoning is complicated, so it is worth
looking at closely. Kant begins by explaining,

We call a human being evil, however, not because he performs actions that are
evil (contrary to law), but because these are so constituted that they allow the
inference of evil maxims in him. (6:20)
Radical Evil and Moral Anthropology 105

This is the essence of the empirical argument. From certain kinds of

actions, one can infer the presence of evil maxims. But then Kant makes
a remark that suggests the impossibility of such an inference.

Now through experience we can indeed notice unlawful actions, and also notice
(at least within ourselves) that they are consciously contrary to the law. But we
cannot observe maxims, we cannot do so unproblematically even within ourselves;
hence the judgment that an agent is an evil human being cannot reliably be based
on experience. (6:20)

Based on this passage, one might think that Kant will give an a priori
argument for the evil in human beings, or just give up his assertion that
one can know that they are radically evil. But Kant does neither. Rather,
he goes on later in the section to insist that “according to the cognition
we have of the human being through experience . . ., we may presuppose evil
as subjectively necessary in every human being” and that “the existence
of this propensity to evil in human nature can be established through expe-
riential demonstrations” (6:32, 6:35, emphases added). And immediately
following his apparent claim that one cannot know reliably from experi-
ence that a person is evil, Kant suggests that there might be an empirical
basis after all. He says,

In order, then, to call a human being evil, it must be possible to infer a priori from
a number of consciously evil actions, or even from a single one, an underlying
evil maxim. (6:20)

An a priori inference is necessary in order to “establish” the propensity

to evil in human beings “through experiential demonstrations of the
actual resistance in time of the human power of choice against the law”
(6:35). Thus when Kant says, “[T]he judgment that an agent is an evil
human being cannot be reliably based on experience” (6:20), he cannot
mean that experience is irrelevant to such a judgment. Rather, he must
mean that such a judgment cannot be based on experience alone, because
an a priori argument is necessary in order to show a connection between
actions and maxims.
Kant must show a priori that there is a connection between actions and
maxims such that one can conclude that a person has evil maxims on the
basis of his or her actions. But this a priori argument alone will not prove
that actual human beings are radically evil. To do that, Kant will have to
make the further empirical argument that human beings actually perform
the actions that justify inferring that they are evil. The a priori argument
showing how actions can provide a basis for judgment about maxims is
106 The Solution

an important preliminary step before the empirical argument can get

started. This inference from actions to maxims is based on the fact that
certain actions are possible only because the moral law is overridden by
other considerations. As Kant explains,

If the law fails nevertheless to determine somebody’s free power of choice with
respect to an action relating to it, an incentive opposed to it must have influence
on the power of choice of the human being in question; . . . by hypothesis, this
can only happen because this human being incorporates this incentive (and
consequently also the deviation from the moral law) into his maxim (in which
case he is an evil human being). (6:24)

From the failure to act in certain ways, one can infer the status of a per-
son’s disposition.
One might argue that what remains unknowable is precisely whether
“the law fails to determine someone’s free power of choice,” and thus that
while a person’s disposition could be inferred from knowledge of that, it
cannot be inferred from actions (or internal motivational states) alone.
In fact, however, some actions are contrary to duty and therefore cannot
be done from duty. In the Groundwork, Kant briefly mentions this when
he introduces his candidates for dutiful action:

I here omit all actions already recognized as contrary to duty, even though they
may be useful for this or that end; for in the case of these the question does not
arise at all as to whether they might be done from duty, since they even conflict
with duty. (4:397, emphasis added)

Some actions are recognized as in themselves contrary to duty, such that

there is no maxim according to which they are justified.15 In the Meta-
physics of Morals Kant takes this even further, articulating a whole doctrine
of right based on the principle that “any action is right if it can coexist
with everyone’s freedom in accordance with a universal law” (6:231). The
doctrine of right deals with the moral status of actions, regardless of the
maxim according to which they are performed. For Kant, governments
can and should define those actions that are wrong. And although the
state cannot punish all moral evil, those crimes that it can punish are
included within the scope of moral evil. As Kant explains, “that I make it
my maxim to act rightly is a demand that ethics makes on me” (6:231).
Every act contrary to right implies a choice that is contrary to the moral
law. Thus there is a whole set of actions, spelled out in the Doctrine of Right,
the performance of which implies an evil will.
Moreover, there is some reason to trust self-consciousness of evil even
for those actions that could be consistent with some morally acceptable
Radical Evil and Moral Anthropology 107

maxim. Kant’s explanations of the difficulty of determining the true

maxim of one’s conduct are directed against those who consider their
conduct to be good because he recognizes the tendency of human be-
ings to deceive themselves about their own goodness. However, there
does not seem to be a comparable tendency to deceive oneself to one’s
own detriment.16 If, on reflection, one recognizes deliberate violations
of the moral law out of self-love, these are unlikely to be rooted in respect
for the moral law at some deeper level. The awareness of action done
with evil intention is good evidence of an evil disposition. Even if it is
conceivable that one’s actual maxim is not the evil maxim of which one
is conscious, it is likely to be a deeper, equally immoral maxim. And even
if one’s true underlying maxim is consistent with the moral law, one who
cannot – even on careful introspection – recognize the consistency of
one’s maxims with the moral law is unlikely to have a stable commitment
to acting only on those maxims that are morally permissible.

In that context, Kant offers his empirical argument for evil. He has shown
that at least for actions contrary to right, it is possible to infer with certainty
the state of one’s maxims from appearances, and thus to the disposition
of one’s free choice. More generally, for maxims that seem contrary to the
moral law, it is reasonable to conclude from evil appearances that one has
an evil disposition. To show that humans are evil, then, Kant must simply
show that humans perform the sorts of actions from which the presence
of evil can be inferred.17 And that is not particularly hard to do. Kant pref-
aces his litany of examples with the unfortunate truth that “we can spare
ourselves the formal proof that there must be such a corrupt propensity
rooted in the human being, in view of the multitude of woeful exam-
ples that the experience of human deeds parades before us” (6:32–3).
This is followed by examples of unjust cruelty, betrayal, ingratitude, and
finally war. The first and last are contrary to right, while the others are
examples of which his readers cannot deny they have at times been guilty.
All that it takes, for any individual, to be assured of one’s own evil is to
reflect upon one’s actions. If one has ever knowingly violated the rights
of others, one is evil. Kant’s argument for rigorism establishes that even a
single evil action is enough to show that one has a will uncommitted to the
moral law, which is to say an evil will.18 Even if one has never violated strict
right, if one has ever acted from self-love in contradiction to or disregard
of the moral law, one cannot reasonably deny that one is evil. Thus Kant
explains that although “this quality [evil] may [not] be inferred from the
concept of . . . a human being, . . . according to the cognition we have of
108 The Solution

the human being through experience, he cannot be judged otherwise”


B. The Nature of Radical Evil

Once Kant sketches his argument that human beings are radically evil,
he goes on to characterize the nature of that evil. And whereas the
fact of radical evil is established on empirical grounds, the nature of
radical evil must be explained a priori. Kant explains, “even though
the existence of this propensity to evil in human nature can be estab-
lished through experiential demonstrations . . ., these demonstrations
still do not teach us the real nature of that propensity . . .; that nature . . .
must be cognized a priori from the concept of evil” (6:35). On Kant’s
account, the evil in human nature is the subordination of the moral law
to maxims of self-love.19

If [a person] took [incentives of his sensuous nature] into his maxim as of themselves
sufficient for the determination of his power of choice, without minding the moral
law . . ., he would then become morally evil. (6:36)


The statement, “The human being is evil,” cannot mean anything else than that
he is conscious of the moral law and yet has incorporated into his maxim the
(occasional) deviation from it. (6:32)

When a person allows him- or herself to choose in such a way that the
moral law can be overridden by sensuous incentives, that person is evil.
In characterizing this evil, however, Kant also makes use of the im-
portant concept of a propensity to evil, a propensity that is a universal
consequence of the radical evil in human nature.20 Kant explains his
notion of a propensity as follows:

Propensity is actually only the predisposition to desire an enjoyment which, when

the subject has experienced it, arouses inclination to it. Thus all savages have a
propensity for intoxicants; for although many of them have no acquiantance at
all with intoxication, and hence absolutely no desire for the things that produce
it, let them try these things but once, and there is aroused in them an almost
inextinguishable desire for them. (6:29)

A propensity is distinguished from an inclination in that a propensity is

merely a potential inclination. A propensity is merely potential, however,
in an unusual sense. The propensity for intoxicants that Kant ascribes to
“savages” implies not merely that when they “try these things” they can
come to have an inclination. Rather, a propensity is potential in that it
Radical Evil and Moral Anthropology 109

depends upon certain conditions, but once those conditions are satisfied,
“there is aroused in them a . . . desire.” Someone who does not desire in-
toxicants even after trying them lacks a propensity to them. Propensities
cannot be ignored, although of course the desires to which they irre-
sistibly give rise can be resisted.
These aspects of Kant’s discussion are general, and it is difficult to apply
this conception of propensity to the propensity for moral evil. After all,
moral evil is not something that depends upon one’s inclinations but
is just a matter of the way that one considers one’s inclination in one’s
choices. So if propensities are just proto-inclinations, how can there be a
propensity for evil itself ? Fortunately, Kant explains the special way that
the notion of a propensity should be applied to moral evil, and in the
process he drops all language of inclination. He says,
A propensity can indeed be innate yet may be represented as not being such:
it can rather also be thought of (if it is good) as acquired, or (if evil) as brought
by the human being upon himself. – Here, however, we are only talking of a
propensity to genuine evil, i.e. moral evil, which, since it is only possible as the
determination of a free power of choice and this power for its part can be judged
good or evil only on the basis of its maxims, must reside in the subjective ground
of the possibility of the deviation of the maxims from the moral law. (6:29)

One with an ordinary propensity has a potential inclination (say, to drink

intoxicants); one with a propensity to evil has a potential choice, that is, a
potential “deviation of maxims from the moral law.” Thus, whereas an ordi-
nary propensity can be overridden by the will because it affects only one’s
desires, the propensity to evil affects the will itself. Moreover, whereas or-
dinary propensities may or may not be due to oneself, the propensity to
evil is “brought by the human being upon himself.”21
This “possibility of deviation” has in common with other propensities,
however, that it is not a mere possibility. The propensity to evil is not simply
what makes it possible for human beings to choose evil. All that is neces-
sary for it to be possible to be evil is that human beings have inclinations
that can conflict with the moral law. A propensity is a potential that de-
pends only upon circumstance for its actualization.22 Kant does not make
clear exactly what these circumstances are. They might involve merely the
opportunity for evil, or inclinations of a certain strength, or certain kinds
of social pressures. Whatever the conditions, the point is that a human
being will in fact choose evil when they are present. Thus just as one
who does not desire intoxicants even when he experiences them must be
said to lack the propensity to intoxicants, one who does not do evil given
appropriate conditions lacks the propensity to evil. The propensity to evil
110 The Solution

in human beings is not identical to choosing evil, but it is the condition

that implies that one will choose evil.23
This propensity need not imply that one always make choices that are
contrary to the moral law. To choose evil is to make choices in such a way
that one allows the moral law to be overridden given a sufficient sensuous
incentive. Thus to have a propensity to evil is to have a breaking point,
a degree of temptation to which one will submit. Kant illustrates this at
the end of a section explaining that “the human being is by nature evil.”
There he says,
A member of the English Parliament exclaimed in the heat of debate: “Every man
has his price, for which he sells himself.” If this is true (and everyone can decide
for himself), if nowhere is a virtue which no level of temptation can overthrow,
if whether good or evil wins us over only depends on which bids the most and
affords the promptest pay-off, then what the Apostle says might indeed hold true
of human beings universally, “There is no distinction here, they are all under
sin – there is none righteous (in the spirit of the law), no, not one.” (6:39)

That all are under sin is precisely the claim that Kant puts forward in
this section, and his basis for this claim is an empirical argument that
everyone can decide about for him- or herself. What this passage adds is
a clarification of the nature of that universal evil. The universal propen-
sity to evil implies that everyone has a price, that none is in principle
committed to the moral law above all sensuous inclination.
This propensity to evil is due to the choice of the person affected by it
because the person brings it upon him- or herself. Thus it expresses an
already present24 moral evil. Kant explains,
This propensity itself is morally evil, since it must ultimately be sought in a free
power of choice, and hence is imputable. (6:37, cf. 6:32)

The propensity to moral evil is an expression of an evil will, but because

it is a propensity, it is also a potential that will be actualized under certain
conditions. And even if these conditions are never present, the propensity
marks a willingness to transgress that is itself evil. The radical evil of
human beings thus includes choosing both maxims that are themselves
contrary to the moral law and other maxims that constitute or bring about
a propensity to and persistence in evil choices. One who is evil – and all
humans are – chooses evil and also ensures that evil will continue in one’s
choices throughout life.

This propensity to evil, constituted and reinforced by an evil will

that refuses to submit to the moral law, employs three degrees of
self-manipulation to ensure the continued dominance of sensuous
Radical Evil and Moral Anthropology 111

incentives over the will. Kant outlines these as the “frailty . . ., impurity . . .,
[and] depravity of human nature” (6:29). All three degrees of evil are
forms of a propensity to evil that is itself the consequence of evil self-
manipulation. Frailty, the first degree of evil, is when

I incorporate the good (the law) into the maxim of my power of choice; but this
good, which is an irresistible incentive objectively or ideally, is subjectively the
weaker (in comparison with inclination) whenever the maxim is to be followed.

This frailty is closely connected to the problem of a lack of character,

as discussed in Chapter 3. In one’s cool, quiet moments, one articulates
principles, but one is incapable of sticking to them. As Kant indicates in
his anthropology, this lack of character precludes one’s having a good
will, precisely because a good will depends on one’s consistently acting
according to the moral law. Thus one way in which one’s evil will seeks to
undermine even the possibility of future reform is by establishing a nature
so frail that it is incapable of consistent action.25 This frailty ensures that
ever-shifting sensuous incentives continue to get the upper hand over the
principle of the moral law.
A second means of ensuring the persistence of evil is impurity, accord-
ing to which

Although the maxim is good with respect to its object . . . and perhaps even
powerful enough in practice, it is not purely moral, i.e. it . . . often (and perhaps
always) needs other incentives besides [the moral law] in order to determine the
power of choice for what duty requires. (6:30)

In this case, one ensures that even apparently morally praiseworthy ac-
tions will in fact be corrupted by sensuous motives. One of the most insidi-
ous forms of self-deception is the tendency to interpret as morally upright
actions that have underlying sensuous motives, and Kant connects self-
deception of this sort with the impurity of the will.26 In a description of
human depravity that parallels his account of the three propensities to
evil, Kant explains that one origin of impurity is

The frailty of human nature . . . coupled with its dishonesty in not screening
incentives . . . in accordance with the moral guide, and hence . . . in seeing only
to the conformity of those incentives with the law, not to whether they have been
derived from the latter itself. (6:37)

Adding dishonesty to frailty helps produce impurity of will. One begins

by ignoring certain incentives, focusing on the moral law as an incentive
for action. In this way, one can act from self-love, along with the motive of
duty, without even recognizing it. This is precisely the sort of deception
112 The Solution

that makes knowledge of one’s moral status so difficult, and here Kant
points out that it also makes becoming truly good more difficult. Because
one is likely to interpret even actions that are bad (in that they are not due
to the moral law as a sufficient motive) as morally good, one’s resolution
in remaining true to the demands of morality is weakened. One is likely to
be complacent in impurity through self-imposed ignorance. Eventually,
at the worst extent of impurity, one does not act from the incentive of
duty at all, but only such that one’s actions are “consistent” with the moral
law. One’s evil will maintains a foothold by eroding the resolution needed
to have a truly good will.
The final degree of the propensity to evil is the most extreme. The
depravity of the human heart is “the propensity of the power of choice to
maxims that subordinate the incentives of the moral law to others (not
moral ones)” (6:30). The previous two degrees of the propensity to evil
really represent the attempts of an evil will to undermine the possibility of
good. The final degree of the propensity to evil is the direct promotion
of evil. Insofar as an agent seeks to bring upon himself the propensity
for evil in this sense, he directly promotes his future performance of evil
actions. At this level too, self-deception is important. Kant explains,

The third [degree of the propensity to evil can be judged] as deliberate guilt
(dolus), and is characterized by a certain perfidy on the part of the human heart
(dolus malus) in deceiving itself [emphasis added] as regards its own good or evil
disposition. (6:38)

The final degree of the propensity to evil is pervaded by self-deception.

This is a benefit from the standpoint of happiness in that it allows for a
self-contentment that would be otherwise impossible. But it is extremely
dangerous from the standpoint of morality, for it contributes to the deep-
ening grip of evil on a person. Kant insists that

This dishonesty . . . hinders the establishment in us of a genuine moral disposition. . . .

It rests on the radical evil of human nature which (inasmuch as it puts out of
tune the moral ability to judge what to think of a human being and renders any
imputability entirely uncertain, whether internal or external) constitutes the foul
stain of our species – and so long as we do not remove it, it hinders the germ of good
from developing as it otherwise would. (6:38, emphases added)

Self-deception is a hindrance of the most drastic kind, and it is therefore

an important part of the radical evil in human nature that promotes ever
deeper and more persistent evil.
For Kant, then, evil is present in human nature as a failure to subor-
dinate sensuous inclination to duty and also as a propensity to continue
Radical Evil and Moral Anthropology 113

this failure, a propensity characterized by the cultivation – in large part

through self-deception – of frailty, impurity, and ultimately wickedness.
Radical evil in human nature is a fact.

C. The Problem of Radical Evil

The fact of radical evil poses an important problem, which Kant spends
his Religion within the Boundaries of Mere Reason solving. Put simply, the
problem is that evil seems inextirpable. As Kant says,

This evil is radical, since it corrupts the grounds of all maxims; as a natural
propensity, it is also not to be extirpated through human forces, for this could
happen only through good maxims – something that cannot take place if the
subjective supreme ground of all maxims is presupposed to be corrupted. (6:37,
cf. 6:45)

People are radically evil, such that they not only act on evil maxims but
also establish in themselves a propensity to evil which ensures that future
choices will be made against the moral law. In that context, it seems
impossible to choose to reject evil and pursue good.
This ineradicability of evil does not imply any kind of moral de-
terminism. Because the evil that is ineradicable is freely chosen, that
evil is not determined prior to choice. If one chose to be good, one
could be. The problem with radical evil is that one in fact does not
choose to be good. Thus, the ineradicability of evil does not in any
way decrease one’s culpability for that evil. Nonetheless, there is an
important practical problem for an agent who recognizes the presence
of radical evil. One is always obligated to do what is morally right. But if
one is already committed to evil, then there seems to be a relevant sense
in which one can no longer choose to obey the moral law. This is true for
two very different reasons.
First, given that one has subordinated the moral law to sensuous incli-
nations, it can never be the case that one completely prioritizes morality
over inclination. One’s overall moral status depends on one’s life as a
whole.27 To be morally good, one cannot ever compromise morality. But
if one has already compromised morality, even if one always does one’s
duty from now on, one is nonetheless a person who, given the right cir-
cumstances (which may include temporal conditions), violates the moral
law. That is to say, one is nonetheless evil. In Kant’s terms,

However steadfastly a human being may have persevered in such a [good] dispo-
sition in a life conduct conformable to it, he nevertheless started from evil, and this
is a debt which is impossible for him to wipe out. (6:72)28
114 The Solution

The first reason that one cannot be good is that one has a stain of past
evil that cannot be erased.
The second reason is that one’s evil is not merely past evil deeds but
includes a propensity to evil. The fundamental maxim governing one’s life
is a commitment to prefer inclination over morality. And this fundamental
maxim provides no ground for its own overturning. One will not reject
evil because the basis of one’s decisions is evil. And although obstacles
to choosing rightly do not erode one’s responsibility for one’s evil, they
seem to undermine the real possibility of moral reform. But Kant insists
that what is required by morality is that one do what is right, which includes
improving the basis of one’s choices. As Kant explains, “In spite of the
fall, the command that we ought to become better human beings still
resounds unabated in our souls; consequently, we must also be capable
of it” (6:45). Throughout Kant’s moral philosophy, he takes for granted
that human beings have moral obligations. In the second Critique, he
draws attention to a “fact of reason,” according to which one recognizes
that one is bound by duty.29 On the basis of this “fact,” Kant argues that
one is free, that God must exist, and that one is immortal.30 Just as those
arguments begin with the assumption of moral responsibility, Kant here
simply takes for granted that one who is radically evil is obligated. In this
case, however, because of the problem of one’s propensity to evil, one is
obligated not only to do good but even to become better, because this
is a necessary condition of choosing rightly. Because one still ought to
become better, one must still be capable of moral reform. But that means
that the radical evil that seemed ineradicable must, somehow, be possible
to overcome.

D. The Solution to the Problem of Radical Evil: Grace

Kant’s attempt to show how radical evil can be overcome involves two
central concepts. The one that is crucial for understanding the place
of anthropology in his moral theory is a new conception of the human
good will. I discuss this concept in detail in the next section. Before
turning to that, however, I must briefly mention another part of Kant’s
solution to the problem of radical evil: grace.31 My goal in this sec-
tion is not to give a full treatment of Kant’s account of grace, nor to
enter into the Christian background of this account, but only to show
how Kant’s conception of grace makes possible a new conception of the
good will, one that is consistent with his claims about radical evil. Thus I
do not address the many nuances and problems with Kant’s account of
Radical Evil and Moral Anthropology 115

Kant’s account of radical evil involves the claim that evil cannot be
extirpated from one’s nature “through human forces” (6:37). Kant em-
phasizes both the depth of radical evil and the fact that the root of evil
lies in one’s own will, but he still insists that we must be capable, in some
sense, of moral improvement. Thus there must be some nonhuman force
through which evil can be overcome: “Some supernatural cooperation is
also needed to [a person’s] becoming good or better” (6:44).33 To un-
derstand the roles that grace plays, it is important to remember the two
very different problems posed by radical evil. First, the stain of radical evil
cannot be removed regardless of the extent to which one improves in the
future because one has done wrong in the past. Second, radical evil seems
to undermine the possibility of transformation by deliberately hindering
one’s own moral development, promoting a “propensity to evil.” Kant ad-
dresses the first problem with a conception of atoning grace, whereby one
is justified before God, “after the fact,” so to speak. The second problem
is addressed by a conception of sanctifying grace according to which God
actually facilitates moral transformation.34
Both sorts of grace reflect attempts by Kant to articulate the possibility
of a will that is fundamentally good – hence good in the eyes of God –
but the temporal expression of which reflects radical evil. Kant wants to
show how such a will is possible without weakening the requirements of
the moral law. The way that grace functions depends, in part, on the type
of grace involved. With respect to atoning grace, Kant’s emphasis is on
providing a supplement for actual (past) violations of the moral law. In
the case of sanctifying grace, Kant is concerned with a supplement that
can counteract one’s propensity to evil. I first focus on atoning grace, and
then I briefly discuss sanctifying grace.
In a series of ethics lectures given just a few years before writing the
Religion, Kant explains the role of atoning grace in his moral theory:

The holy law necessarily entails that punishments should be appropriate to ac-
tions. But is man, then, to be left without help, seeing that he is frail, after all, in
regard to morality? He cannot, indeed, hope for any remission of punishment
for his crimes from a benevolent ruler, since in that case the divine will would
not be holy; but . . . if, for our part, we do everything we can, we may hope for
a supplementation, such that we may stand before God’s justice and be found
adequate to the holy laws. . . . In that case, then, instead of a lenient justice, we
have a supplementation of justice. (27:331)

Kant’s challenge is to reconcile three fundamental commitments: (1)

One can rationally hope to be fundamentally morally good. (2) One is
radically evil, both in the sense that one has fallen short of strict obedience
116 The Solution

to the moral law and in the sense that one deliberately undermines one’s
capacity for good. (3) The moral law is rigorous; there is no middle ground
between moral good and evil because the moral law insists on unswerving
So how does grace allow Kant to reconcile these commitments? Ap-
parently the notion of a “supplementation of justice” is distinguished
from “lenient justice” in that while either would preserve moral hope (1)
and radical evil (2), “lenient justice” would do so at the cost of rigorism
(3). What is not clear is how grace preserves all three claims. It looks
as if any role that grace can play will compromise one of the claims it
is meant to save. Perhaps grace can save the rigor of the moral law at
least in one sense, in that there might not be any degrees of goodness:
One either receives grace and is thus good, or one does not. But if the
moral law demands perfect conformity with morality, then any grace that
makes one good despite one’s imperfection is for all apparent practical
purposes indistinguishable from lenient justice. Perhaps Kant thinks that
the concept of grace provides a better basis on which to argue for a con-
ception of the good will that does not allow excuses in the moment of
deliberation. Exactly what this involves will become clearer in the next
For now it is enough to note that for Kant, it is crucial that in delib-
eration one measure one’s proposed maxims against the moral law in all
its rigor. To avoid despair, one must hope that one can be good despite
past evil deeds. And so some concept is necessary to allow this hope. But
whereas a lenient justice would taint one’s present deliberations by ex-
cusing one from strict obedience to the law, atoning grace should allow
one to avoid despair without providing any excuse from the full demands
of morality in deliberation. Ultimately, the issue of whether grace is more
effective in this regard than a nuanced version of lenient justice would
go beyond the limited needs of this discussion. But this is clearly a role
that Kant wants grace to play.
More important for the present purposes is the fact that one can invoke
the concept of grace at all only if it can be given some coherent sense.
To preserve his commitments to moral hope, radical evil, and rigorism,
Kant needs a concept that is distinct from “lenient justice.” He gives
this concept a name: “grace” or the “supplementation of justice.” But
it is not clear that Kant shows that anything corresponds, or even could
correspond, to the name. In general, Kant’s strategy at this juncture is to
point out that the mechanisms of grace are inscrutable to human reason.
In the middle of the passage quoted above from the lectures on ethics,
Radical Evil and Moral Anthropology 117

Kant says, “How God brings about this supplementation, and what sort
of means he employs for it, we know not, nor do we have any need to
know; but we can hope for it” (27:331, cf. 28.2.2: 1120–1, 1225, 1319). In
the Religion, Kant says that of “supernatural assistance . . . we can have not
the least cognition” (6:191).36 And in the Conflict of the Faculties, written
after the Religion,37 Kant repeats this point:

If worst comes to worst reason is entitled [befugt] to adopt on faith a supernatural

supplement to fill what is lacking for his justification (though [reason is] not to specify
in what this consists). . . . We need not be able to understand and state exactly what
the means of this replenishment is (for in the final analysis this is transcendent
and, despite all that God Himself might tell us about it, inconceivable to us); even to
lay claim to this knowledge would, in fact, be presumptuous. (7:43–4, emphases

The mechanism of grace cannot be understood. To avoid despair, one

must believe that there is some way to be morally good, but practical rea-
son will not condone any lenience in the interpretation of the moral law.
Thus there must be some “inconceivable” supplement for our failings,
some supplement that reason does not fully specify.

The status of grace in Kant’s practical philosophy can be understood

by analogy with the status of the unconditioned in Kant’s theoretical
philosophy. In Chapter 1, we saw that for Kant, speculative reason has
a “need” that the concept of freedom satisfies. This need is due to the
demand on the part of speculative reason for completeness, a demand
that requires some unconditioned ground for empirical knowledge. We
can see Kant’s writings on Religion as suggesting that practical reason has
a similar “need” for a concept of grace. This need is due to a demand
on the part of practical reason, a demand for uniting three fundamental
claims about human nature, the reasonableness of moral hope, the fact
of radical evil, and rigorism.
Like the demand for an unconditioned ground, this need provides a
merely conditional warrant for the belief in grace. Neither practical nor
theoretical reason strictly proves the reality of grace. In this sense, the
grounds for grace are less solid than even the grounds for the postulates
of pure practical reason.38 In particular, practical reason does not have
to remain committed to moral hope – the conviction that we are still free
to be good. Although one must be transcendentally free to be morally
responsible, practical reason by itself need not claim that one who does
in fact choose contrary to the moral law can nonetheless be morally good.
118 The Solution

But given the role that practical reason must play in our temporally situ-
ated deliberation, we have a subjective need to believe in something like
By itself, this need is not enough to justify even a conditional warrant
for grace. Recall that the speculative need for some unconditioned does
not specify freedom or even God as the content of that unconditioned.
Until the second Critique argues on nontheoretical grounds for freedom and
God, there is merely a need for something unconditioned, not a need
for freedom or God per se. Likewise, the need of practical reason, in
the context of radical evil, is just for some concept that will unify moral
hope, radical evil, and rigorism. This need is not specifically a need for
Moreover, Kant gives no details about the perspective from within
which one can justify grace. In the realm of theoretical reason, it is not
enough to show that theoretical reason has a “need” for some uncon-
ditioned ground of appearances. Kant also shows, in the first Critique,
that such a ground would not conflict with the conditions of possibility
of experience, as long as the ground is not itself an object of possible
experience and not justified on theoretical grounds. Only once he artic-
ulates his transcendental idealism can Kant justify freedom on practical
grounds or believe in it on the grounds of a subjective need of theoretical
reason.39 In the case of grace, Kant must show that some concept of grace
can be consistent with the constraints of practical and theoretical reason.
He does not necessarily have to explain how it can be consistent, but he
must say at least enough to convince us that there is no contradiction with
the requirements of theoretical or practical reason.
In the Religion, Kant gives some detail about the way in which grace
can fill the need of practical reason, and he offers some argument that
it is not ruled out by practical or theoretical reason. With respect to
atoning grace, Kant’s account involves God’s superior knowledge of what
is nontemporal and a rational reconstruction of the Christian doctrine
of substitutionary atonement. The substitutionary doctrine is ingenious,
in that Kant essentially posits that one can suffer on one’s own behalf.
The temporal appearance that expresses a good will involves suffering,
and this suffering wipes out the stain of one’s past wickedness. Philip
Quinn has worked out Kant’s doctrine of atonement in detail in terms of
a distinction between “pre-revolutionary and post-revolutionary persons”
(Quinn 1986:450). I will not go over his arguments here. Quinn rightly
notes that while Kant’s rational reconstruction improves on some aspects
of the traditional Christian doctrine of Christ’s vicarious atonement for
Radical Evil and Moral Anthropology 119

sins, it has many problems of its own (Quinn 1986). Whether or not
Kant can solve these problems, his account of atoning grace does at
least provide some sense of how grace could be consistent with the de-
mands of theoretical and practical reason. Grace in this context is re-
ally a matter of God’s recognizing an intelligible truth – one’s good-
ness – in the face of the complicated appearance of one’s whole life. As
long as Kant does not claim any theoretical warrant or empirical inter-
pretation of this grace, there do not seem to be theoretical reasons to
reject it.
Practically, it raises more problems. In particular, it is not entirely clear
how this account of atonement provides a real alternative to “lenient
justice” to save moral hope, radical evil, and rigorism. Nothing in Kant’s
account of the moral law suggests that suffering can somehow make up
for violations of the moral law. And it is crucial that this be the case, lest
people excuse themselves from obedience only to follow up vice with
further (monkish) vices of self-flagellation. This account of atonement
seems to amount to lenience in the demands of morality, albeit a very
complicated lenience. Kant seems to respond to this objection when he

It can . . . be asked whether this deduction of the idea of a justification of a human

being who is indeed guilty . . . has any practical use at all, and what such use
could be. It is hard to see what positive use can be made of it for religion and
for the conduct of life. . . . The investigation is only an answer to a speculative
question, but one that cannot therefore be passed over in silence, since reason
could then be accused of being absolutely incapable of reconciling the human
being’s hope of absolution from his guilt with divine justice, and this accusa-
tion might be disadvantageous to reason in many respects, most of all morally.

Just as freedom meets a need of theoretical reason but is not itself to be

put to use theoretically, atoning grace meets a need of practical reason but
can be dangerous when put to use practically. Thus it is totally different
from “lenient justice,” which would be a modification of the demands of
practical reason itself.
Unfortunately, Kant’s claim that the investigation is “speculative” is not
helpful here. It is just not clear what kind of warrant the argument for
grace is supposed to have, if it is not practical. The argument is not a
theoretical one, so there must be some other sense – as a condition of
possibility of hope, perhaps – within which grace makes sense. But the re-
lationship between this sort of reason and practical reason is not specified,
and it is unclear what the relationship would be. This lack of specificity
120 The Solution

is markedly different, for example, from the detailed description of the

status of the postulates in the second Critique.

Kant’s account of sanctifying grace raises even more problems. Sancti-

fying grace is necessary not to make up for past misdeeds but to make
possible future good deeds, given that the basis of one’s choices is corrupt.
Kant does not discuss this grace in detail prior to the Religion, and his dis-
cussion in the Religion emphasizes the problems that arise if one tries
to account for sanctifying grace. Kant raises an apparent antinomy. The
first prong of the antinomy is that one “cannot but regard [grace] as
only conditional, that is, consider the improvement of his life conduct,
as much as lies in his power, as having to come first, before he gives even
the least credit to the hope that the favor from on high will redound to
his good” (6:117). Grace must be conditional because grace is a benefit,
which cannot rightly accrue to someone worthy only of punishment. The
second prong of the antinomy is that because of radical evil, it seems
impossible to first turn to the good and only later receive grace, because
one’s will itself is corrupt.

Faith in a merit which is not his own, but through which he is reconciled with God,
would therefore have to precede any striving for good works, and this contradicts
the previous proposition. (6:117)

Kant insists that the antinomy cannot be resolved at a theoretical level.

This conflict cannot be mediated through insight into the causal determination
of the freedom of a human being, i.e. into the causes that make a human being
become good or bad: in other words, it cannot be resolved theoretically, for this
question totally surpasses the speculative capacity of our reason. (6:117–18)

From a practical standpoint, there is an answer to the question of priority.

When the question is not whether one’s action or God’s cooperation
actually comes first, but which one should effect first, the answer is easy.

Practically, however, where the question is . . . whence . . . we are to make our start,
whether from faith in what God has done for our sake, or from what we ought to
do in order to become worthy of it (whatever this may be), there is no hesitation
in deciding for the second alternative. (6:117–18)

The second prong of the antinomy, that one must become worthy of
grace before one can receive it, is true in a practical sense. Whether or not
there is an actual priority (temporal or otherwise) of merit over grace,
when deciding what to do, one must think not about first getting grace but
rather first becoming worthy of it.
Radical Evil and Moral Anthropology 121

Again, “we can admit an effect of grace as something incomprehensi-

ble, but cannot incorporate it . . . for either theoretical or practical use”
(6:53). But as in the case of atoning grace, Kant does not specify what al-
ternative there is to theoretical or practical uses for the concept. And just
as Kant could not actually provide a philosophical justification of free-
dom without the second Critique, despite the need of theoretical reason
for some unconditioned, he cannot provide a philosophical justification
of grace without some further Critique that gives a sense to the “grace”
that is supposed to meet this need of practical reason.40 More troubling
is that the arguments that Kant gives for this concept seem to imply that it
should have some practical use, but Kant cannot allow such use without
compromising his doctrines of radical evil and moral rigorism. In any
event, a fully satisfying Kantian moral theory would have to either show
how moral hope, radical evil, and/or rigorism can be given up with-
out compromising Kant’s moral philosophy or provide some framework
within which the need of practical reason to unify these commitments
could be met. I will not attempt either of these tasks here.
Kant’s account of grace fills an important need in his practical philos-
ophy. There are several problems with his account of grace as it stands.
First, Kant does not seem justified in saying that there is need of practical
reason for grace in particular. Something is needed if moral hope, radical
evil, and rigorism are to be maintained, but it is not clear that “grace” is
the only option. Second, Kant’s description of grace does not show how it
is relevantly different from the “lenient justice” he seeks to avoid. Third,
Kant does not adequately show that grace is consistent with practical and
theoretical reason. Especially with respect to sanctifying grace, there is
an apparent contradiction with practical reason that is bypassed rather
than resolved. Fourth, Kant gives insufficient details about the perspec-
tive within which grace would be justified. He provides a subjective need,
but not a real justification, for the concept of grace. Finally, the details
that Kant does provide raise problems of their own.
Dealing with all of these problems is beyond the scope of this book.41
What is important for our purposes is the role that grace plays in Kant’s
practical philosophy. In section 3, I have discussed Kant’s arguments for
radical evil. Radical evil poses serious problems for giving an account of a
good human will. But Kant claims that grace, by reconciling moral hope,
radical evil, and rigorism, makes room for a conception of the human
good will that is good but not identical to a will that would express itself in
a life of perfect conformity to the moral law. Moreover, it does this without
weakening the force of the moral law itself and without admitting degrees
122 The Solution

of goodness. In the next section, we turn to Kant’s detailed description

of this new conception of the human good will.

4. Integrating Moral Anthropology: A New Conception

of the Human Good Will
Solving the problem of radical evil presents for Kant the challenge of
articulating a new conception of what a human good will looks like in the
world. Without awareness of the radical evil in human nature, one might
think that a good will would express itself in a will that always chooses
in accordance with the categorical imperative. The presence of radical
evil makes such a will an impossibility for human beings because one
can never completely escape the stain of past misdeeds and the enduring
presence of a propensity to evil. But the possibility of atoning and sancti-
fying grace frees Kant to develop a new account of the good will. It must
be possible to act and choose so that one is a proper recipient of grace,
such that one’s free will can be considered good. There must be a sort of
good will that is consistent with the reality of radical evil.
Kant articulates this conception of a good will as a “new man,” a will
that engages in a “revolution” against its radical evil. As Kant explains,

That a human being should become not merely legally good, but morally good
(pleasing to God), i.e. virtuous according to the intelligible character . . . – that,
so long as the foundation of the maxims of the human being remains impure,
cannot be effected through gradual reform but must rather be effected through
a revolution in the disposition of the human being (a transition to the maxim of
holiness of disposition). And so a “new man” can come about only through a kind
of rebirth, as it were a new creation . . ., and a change of heart. (6:47)

Despite the Biblical language, Kant’s point is philosophical. Because of

the nature of evil, a change in moral status cannot be gradual but must
be an absolute transformation from good to evil. Evil is the willingness to
subordinate the moral law to sensuous inclinations, even only occasionally.
One does not become less evil by subordinating the moral law less often,
or to stronger inclinations only. Only a complete shift, such that the moral
law assumes absolute priority, constitutes genuine moral improvement
on Kant’s account. Thus until and unless one makes this change, one
remains evil. And this change cannot be made on the basis of the radical
evil that one, as human, always already adopts. It instead depends on a
revolution against one’s evil maxims and toward the good, a revolution
made possible, though still not comprehensible, through grace.
Radical Evil and Moral Anthropology 123

As Kant makes clear in his discussion of the revolution, this description

applies only to a person as free. He even uses the term virtus noumenon
to describe the virtue that this revolution reflects (6:47). This discussion
raises some important problems for Kant’s claim that one cannot have
any knowledge of the noumenal realm, as well as for his insistence in the
first Critique that noumena cannot be understood temporally. Kant does
not directly address either of these problems, but one can make the most
sense of the account by understanding the term revolution to refer not to
something temporal but simply to a good will. Thus we can refer to this
new man as a will “in revolution” to capture the fact that this revolution
is not a temporal act but a noumenal reality.42 The will in revolution is
a will that is fundamentally good, but good as a reaction against its own
radical evil, rather than as an absence of any evil.
This good will can be described by the term revolution to capture two
important points. First, the will in revolution is a good will that is consis-
tent with radical evil expressed in evil deeds and tendencies to perform
such deeds. Second, this will is not temporal in the sense that it does not
in itself involve gradual reform. This latter point is the main emphasis
in Kant’s discussion of the revolution. He opposes any notion of gradual-
ism which might suggest that one can be “partly good,” because goodness
must be an absolute commitment to the moral law. Thus it is consistency
with radical evil and resistance to gradualism that Kant captures with his
notion of “revolution.” Rather than introduce a temporal notion, the
whole point of the term revolution is to shift away from thinking of this
change temporally. Whether there is a word that is better suited to such
a nontemporal situation is less important than clarifying the necessity
and possibility of the revolution and the nature of its appearance in the
Instead of discussing the extent to which a “revolution” can be as-
cribed to a noumenal will, Kant focuses on the problem of explaining
what this revolution would look like in appearance. This is important for
reasons that have been discussed in section 2. Appearances are crucial
for moral evaluation, because one evaluates agents on the basis of how
their appearance measures up against a moral standard. To judge whether
someone is good, one must have some sense of what the appearance of
a good will is. More important, one makes moral decisions based upon
effecting certain appearances. To understand what is required for one to
receive grace by a noumenal revolution against evil, one must understand
what such a revolution would look like in the world of appearances. Only
then can it be understood how one can choose to pursue a pattern of
124 The Solution

behavior that corresponds to such an appearance of revolution. And it

will turn out for Kant that this pattern of behavior includes pursuing helps
and avoiding hindrances. Thus the problem of helps and hindrances is
ultimately solved by Kant’s account of the appearance of the revolution
that is necessary for a human will to be a good will.

The first step in Kant’s articulation of the appearance of the good will
is to apply the notion of atemporal revolution to the temporal series
that constitutes the human being as appearance. In this context, Kant

A revolution is necessary in the mode of thought but a gradual reformation in

the mode of sense . . ., and [both] must therefore be possible also to a human
being. That is: if by a single and unalterable decision a human being reverses the
supreme ground of his maxims by which he was an evil human being (and thereby
puts on a “new man”), he is to this extent, by principle and mode of thought,43
a subject receptive to the good; but he is a good human being only in incessant
laboring and becoming; i.e. he can hope – in view of the purity of the principle
which he has adopted as the supreme maxim of his power of choice, and in view
of the stability of this principle – to find himself upon the good (though narrow)
path of constant progress from bad to better. (6:47–8)

The “single and unalterable decision” is the revolution that makes one
morally good at the level of one’s free self. But this decision must be
expressed as progress in the appearance of that self. The revolution is
directly relevant to morality, but it can never be known directly through
experience. For God, who intuits one’s free will directly, “this [revolution]
is the same as actually being a good person (pleasing to him)” (6:48). One
can never know that one effects a revolution in one’s mode of thought;
one can only perceive signs of that revolution through its expression in
moral progress. As Kant explains,

For the judgment of human beings . . . who can assess themselves and the strength
of their maxims only by the upper hand they gain over the senses in time, the
change is to be regarded only as . . . a gradual reformation of the propensity to
evil. (6:48)

Just as one cannot know that one’s fundamental maxim is good, one
cannot know that one is in revolution against evil. But just as one can
reasonably hope for and pursue goodness, one can reasonably hope
for and pursue a moral revolution. And the basis for hope, as well as
the immediate object of one’s pursuit, will be the expression of moral
Radical Evil and Moral Anthropology 125

In that context, Kant considers several possibilities for the temporal

expression of a nontemporal free agent undergoing a “revolution.” The
first crucial requirement for a series of appearances to express a will in
revolution is that the entire series express the free agent. In that context,
Kant insists that the appearance that is relevant in evaluating whether or
not one is good in the sense of revolting against evil be one’s entire life.
Kant says,

Reason, when it is a question of the law of our intelligible existence (the moral
law), recognizes no distinction of time and asks only whether the event belongs to
me as a deed and, if it does, then always connects the same feeling with it morally,
whether it was done just now or long ago. (5:99)

When evaluating whether or not one truly revolts against evil, one
must not simply look for some particular moment or phase of life but
at one’s life as a whole. In this context, Kant considers several pos-
sibilities for how a good life would look. The most obvious option –
a life free from any evil – is ruled out already by the presence of radical
evil. Kant’s problem at this point is to consider what the most plausible
remaining option is.
Kant argues that a will in revolution must express itself in a complete
life by contrasting morality with happiness. With respect to happiness,

Griefs once endured, when we feel safe from them, leave no painful reminis-
cences behind but rather a feeling of gladness that makes the enjoyment of the
supervening good fortune all the sweeter. For pleasure and pain (since they be-
long to the senses) are both included in the temporal series, and disappear with
it; they do not constitute a totality with the present enjoyment of life but are
rather displaced by it as it succeeds them. (6:70)

Insofar as one experiences proper pleasures and no pains at any moment,

one can be described as happy, whatever pains one experienced in the
past. In contrast, however, moral life is a nontemporal unity.

The moral subjective principle of the disposition by which our life is to be judged
is (as transcending senses) not of the kind that its existence can be thought as
divisible into temporal segments but rather only as an absolute unity. (6:70)

Kant goes on to discuss, and to reject, one particular account of how a

moral revolution might appear in the world. He sums this option up with
the slogan “all’s well that ends well.”
According to this view, what matters in life is the moral condition
at which one arrives at the end of one’s life. Regardless of how wicked
someone may be in youth, acting from duty alone in the minutes before
126 The Solution

death shows that such a person is a morally good person, that is, one who
has undergone the moral revolution. In discussing this view, Kant focuses
on the danger that it poses for morality:

Give [a human being] encouragement (as with the proverb “All is well that ends
well”) and from early on he will make his plans accordingly, with a view not
to forfeit too much of life’s pleasures unnecessarily and, by life’s end, to settle
accounts with speed and to his advantage. (6: 77–8, cf. 6:70)

The thought that one might be morally evil brings with it “moral suffer-
ings, the reproaches of one’s conscience” (6:78 fn.), not to mention fears
of suffering after death. Still, human beings have wants and needs that
tempt them to violate the moral law. If one comes to believe that one can
be morally good, can be spared from the sufferings (moral and otherwise)
associated with having an evil will, and at the same time can spend all but
the last few seconds of one’s life pursuing sensuous inclinations in disre-
gard of morality, one seems able to have one’s cake and eat it too. One
can actually be morally good while acting for most of one’s life in a way
that disregards morality.
For Kant, of course, such a possibility is a moral disaster. A view of moral
goodness that justifies evil choices erodes the foundation of morality. It is
no longer morally required, apparently, that every agent choose well. As
long as an agent has a couple of well-timed good choices, the force of all
other moral obligations erodes. And this can lead to moral complacency,
which the doctrine of grace and the new conception of a good will are
supposed to avoid. If it is impossible to be morally good, one can hardly
make a serious effort at it – this is the problem of radical evil. But if it is too
easy to be morally good, one will be almost as likely to ignore the moral
law. When Kant contrasts the application of “all’s well that ends well” to
happiness with its application to morality, he highlights this problem. Not
only does this account fail to take into account one’s whole life, but it also
has dangerous practical consequences for moral life.
Still, Kant does not completely reject the possibility of conversion at the
end of one’s life. He claims, “the common saying, ‘all’s well that ends well,’
can indeed be applied to moral cases, but only if by the ‘good ending’ we
understand that a human being becomes a genuinely good human being”
(6:70). This view fits well with his description of the sensible appearance
of one’s noumenal revolution as “progress.” One measures progress, at
least in part, by the end point at which one arrives. However, Kant is quick
to point out a problem with resting one’s entire self-evaluation on a last-
minute conversion. If one “becomes a genuinely good human being” at
Radical Evil and Moral Anthropology 127

the end of life, things will have turned out well. But, Kant asks rhetorically,
“where is he to recognize himself to be such [a good human being], since
he can draw this conclusion only from the constancy of his consequent
good conduct, and, at the end of life, there is no time left for this?” (6:70).
The inscrutability of one’s true disposition undermines shortcut attempts
at moral reform. Because one has no direct access to the fundamental
maxims of one’s actions, one can judge one’s disposition only on the
basis of consistent struggle against evil in intentions and actions. If even
the best appearances can be misleading, apparent dispositions at the end
of life, without any corresponding consistency over time, are particularly
suspect. And this suspicion will be especially pronounced when one has
lived a life of wickedness.
Of course, one can never rule out completely the possibility that a
conversion at the end of life is genuine. Kant even remarks that it is par-
ticularly important to be careful how one interacts with those on their
deathbeds. There is always hope for true reform, hope that can be un-
dermined by bad clergy offering false comforts. It is important to hold
out the need for genuine moral reform, for actually doing what is right and
correcting past wrongs, even to those near death (see 6:78 fn).
There is another sort of moral conversion that Kant does not discuss
separately but that might seem more plausible as an account of the ap-
pearance of a good will. Rather than a conversion at the end of life, or a
conversion in disposition only, one might imagine that a good will would
appear as a person who, though wicked for an early portion of his or her
life, at some point changes radically and from then on makes all choices
in accordance with the moral law. Such a person would not base hope
merely on an impression of a different resolution but would have a long
record of deeds. On this account, one would always be justified in seeking
to be good, because one can effect at any time the radical transforma-
tion that characterizes the good will. One could counteract any danger
of postponing that transformation on the grounds that only when one
turns to the good with enough time to exhibit that goodness as a stable
aspect of one’s character does one truly express a will in revolution. This
provision would eliminate many of the practical problems with deathbed
At times, Kant seems to endorse a position similar to this one, although
he ultimately rejects it in favor of an alternative account. His endorsement
comes in the form of an emphasis on temporal conversion, on an estab-
lishment of good character at a certain time. In the Anthropology, Kant de-
scribes the establishment of character as a revolution. Because character
128 The Solution

is a precondition of having a will that is thoroughly good, this description

seems to have some relation to the revolution described in the Religion.44
At times, Kant even uses similar language to describe the two revolutions.
Thus in the Anthropology he says,

The establishment of character is similar to a kind of rebirth, a certain solemn

resolution which the person himself makes. . . . This stability and persistence in
principles can generally not be effected . . . by degrees, but it can only be done
by an explosion which suddenly occurs. . . . Perhaps there will be only a few who
have attempted this revolution. . . . (7:294)

As in the Religion, the revolution that establishes character is distinguished

from anything that happens gradually. The language of “rebirth” fits
well with Kant’s description of the moral revolution as giving rise to a
“new man.” But the revolution described here is very clearly a revolution
that occurs at a particular time. Kant says, “this resolution and the moment
[Zeitpunkt] at which the transformation took place remain unforgettable”
(7:294, emphasis added). And his claim that few attempt this revolution
continues with the following qualification: “Perhaps there will be few
who have attempted this revolution before their thirtieth year, and fewer
still who have firmly established it before their fortieth year” (7:294).45
The temporal nature of this revolution is sufficient to show that it is not
identical to the moral revolution described in the Religion. Kant further
distinguishes them when he explicitly describes character in the Anthro-
pology as something of which one is “conscious” (7:294), whereas the
revolution in the Religion is intelligible (6:48) and hence not an object of
possible experience.
But even if this revolution is not identical to the intelligible revolution
in the Religion, it is still a revolution. And one might think that a revolution
at some particular moment would be a good candidate for the temporal
appearance of a moral revolution. Kant rejects this option. But while Kant
makes very clear that the good will must appear as progress rather than
revolution, he is less clear about why this is so. One obvious reason is that
one’s entire life must express one’s revolution, but one might still think
that the early parts of one’s life could represent the evil against which one
revolts, and hence they are part of the revolt itself. To fully understand
why every revolution must express itself as progress rather than as sudden
conversion, one must remember the nature of the evil against which one
If radical evil consisted only of acting on evil maxims, the revolution
against it might consist simply of changing the maxims on which one
Radical Evil and Moral Anthropology 129

acts. But radical evil also involves a propensity to evil according to which
one undermines one’s future choices, establishing a tendency to pursue
further evil. Kant takes this aspect of radical evil to be an empirical fact.
One recognizes not only one’s past evil deeds but also the present and
future effects of past self-manipulation. Kant’s new conception of the
good will must take into account the reality of this propensity to persist in
evil. And this means that one has reason to predict, though never excuse,
that one will do evil in the future. Whatever conversion one undergoes,
one still suffers the effects of one’s past evil choices, including effects on
one’s future behavior. One is fully responsible for these effects, because
they reflect one’s free choice of evil rather than of good. But even if
one commits to moral principles now, this commitment can only begin to
undo the effects of one’s evil actions in the past. The temporal extension
of past decisions does not undermine transcendental freedom, because
the limitations of one’s ability to effect changes in oneself are self-imposed
limitations. But the temporal extension of past decisions does affect the
sort of expression that is possible for a will in revolution. It can at most
appear as a constant struggle against evil.46

How then does an agent in revolution against evil appear in the world?
This question can be applied in the context of either moral evaluation or
deliberation. With respect to evaluation, although Kant does not think
that one can ever be certain that one truly effects a revolution against evil,
he offers some suggestions about the way to think of one’s moral status.
He begins by asking, “What can a human being expect at the end of his life,
or what can he fear, in virtue of his conduct during it?” In response to
this question, Kant allows surprisingly large room for self-knowledge:

A human being must first of all have cognition of his own character, at least to
some extent. Thus, though he may believe that there has been an improvement
in his disposition, he must be equally able to take the old (corrupted) one into
consideration, the one from which he started, and examine what and how much
of that disposition he has cast off, as well as the quality (whether pure or still
impure) and the grade of the supposed new disposition for overcoming the old
one and preventing relapse into it; he will thus have to look at his disposition
throughout his whole life. (6:76)

Kant still insists that one has no direct access to one’s disposition. The
barriers to self-knowledge have not been forgotten. Thus, one “must ex-
tract his disposition from the deed before him” (6:77). So what can one
say about one’s moral status? First, one must consider one’s “whole life”
(6:76), looking for progress throughout that life. Evils of one’s youth are
130 The Solution

important markers of where progress must occur, and there must be a

consistent struggle against evils to which one is particularly prone along
with a steadily building victory in that struggle. Even at the end of life,
one may still in some respects be evil (that is what immortality is for, after
all), but there must be a track record of constant and successful struggle.
A sense of moral goodness comes only from “the assurance of the reality
and constancy of a disposition that always advances in goodness” (6: 67,
emphasis added). This self-evaluation must be based on actions, not on
mere feelings or any supposed sense of oneself. One can never reliably
know that one’s fundamental maxims are good, but unless one avoids
those actions that cannot be done from duty and pursues those most
congruent with it, one can be sure that one is evil. Only good actions and
the constant struggle against one’s propensity to evil provide hope that
one has a good disposition. This is as much as Kant says with respect to
moral evaluation.

In fact, however, Kant is more concerned with deliberation than with

moral evaluation. In his objections to deathbed conversions, Kant focuses
on the effects of various accounts of moral worth on moral deliberation.
From the standpoint of deliberation, the question is not whether or not
one is good but what one must do to be good. And it is from this standpoint
that the practical role of helps and hindrances becomes prominent.
The first and most important thing any agent must do to be morally
good is obey the categorical imperative out of respect for the moral law.
Kant points out that one “cannot allow the previously recognized dispo-
sition to take the place of the deed but . . . must extract his disposition
from the deed before him” (6:77). In the context of evaluation, the “deed
before him” refers to the sum of all the deeds of one’s life. From a deliber-
ative point of view, however, the deed before one is the deed that one is to
perform. Every moment of choice provides a new opportunity to express
a will in revolution against evil. The bare minimum for expressing such
a will in a given choice is that the moral law function as the governing
maxim of one’s free will. Kant points out that “if [virtue] is not rising,
it is unavoidably sinking” (6:409). Any missed opportunity to act on the
basis of the moral law implies that one is sinking, that the progress which
would reflect a good will is lacking. One who does evil “just this once” on
the grounds that he or she can always show progress later errs seriously
in moral deliberation. For although one always can choose rightly in the
future and thereby show progress, what one actually will choose depends
on whether or not one’s will is in revolution against evil. And the most
Radical Evil and Moral Anthropology 131

reliable way to express this revolution is through the action immediately

before one. The only legitimate basis for the hope that one will choose
well in the future is that one chooses well in the present. By choosing
wickedness now, one leaves little hope that one will ever do any better.
And this, of course, means that the primary responsibility of a moral agent
plagued with radical evil but enabled through grace to be good is to
actually be good, right now, with respect to each choice that presents

If that were the only responsibility of moral agents, Kant’s account of

radical evil, although important for theology and interesting from the
standpoint of practical reason, would not have any important implications
for morality itself. For Kant, however, the fact of radical evil has important
implications for what human beings actually need to do to express a good
will. Kant explains, in response to the Stoics, that

Those valiant men [the Stoics] mistook their enemy, who is not to be sought in the
natural inclinations, which merely lack discipline and openly display themselves
unconcealed to everyone’s consciousness, but is rather as it were an invisible
enemy, one who hides behind reason and hence all the more dangerous. They
send forth wisdom against folly, which lets itself be deceived by inclinations merely
because of carelessness, instead of summoning it against the malice (of the human
heart) which secretly undermines the disposition with soul-corrupting principles.

In his footnote to this passage, he says even more explicitly that for the
Stoics “everything was quite correctly apportioned . . . provided that one
attributes to the human being an uncorrupted will” (6:58, emphasis added).
For the Stoics, morality demands only that one prefer the moral law
to sensuous incentives. And this is just what one might expect Kant to
demand, because his initial description of the good will is just this kind
of prioritization of predispositions. But Kant here insists that this is not
sufficient, given the fact of radical evil. As he explains,

To become a morally good human being it is not enough simply to let the germ
of good which lies in our species develop unhindered; there is in us an active and
opposing cause of evil which is also to be combated. (6:57)
We cannot start out in the ethical training of our connatural moral predispo-
sition to the good with an innocence which is natural to us but must rather begin
from the presupposition of a depravity in our power of choice in adopting max-
ims contrary to the original ethical predisposition; and since the propensity to
this [depravity] is inextirpable [in time], with unremitting counteraction against
it. (6:51)
132 The Solution

What ultimately determines the status of one’s will is one’s life as a whole.
And although the best indication of future progress is choosing rightly in
the present, one can never ignore the presence, in oneself, of a propensity
to choose evil. Thus at the same time that one must first and foremost
always choose good, one must also actively combat this evil in oneself.47
This is especially important because a human being who chooses evil
does not merely choose to do evil at a particular time but actively under-
mines the likelihood of choosing good with “soul-corrupting principles.”
To effectively combat this evil, a good person must respond not merely
with good choices but also with soul-correcting principles, with actions that
serve to undermine the influence of evil and increase the strength of the
moral law in one’s future deliberations. In the same way that evil human
beings not only make evil choices but effect through their maxims a
propensity to further evil, a human being who effects a revolution against
evil must not only make good choices but also undermine this propen-
sity to evil and develop the strength of will to continually pursue the
As a result, Kant explains what one can do to effect progress in the pu-
rity and stability of one’s empirical commitment to the moral law. To make
sense of this concern, it is necessary again to point out the relationship
between the free will, the appearance of that will, and empirical influ-
ences on that appearance. From a metaphysical standpoint, the free will is
a noumenal ground of the appearances; and these appearances, as objects
of experience, are also capable of empirical influence. From a practical
standpoint, one considers the free will as grounding the appearances not
as an abstract truth but as the realized condition of deliberation. One’s
commitment to oneself as free ground is manifested as one holds oneself
accountable to the moral law in deliberation. Thus the crucial question
is not the metaphysical possibility of a free self, but what that free self
ought to do, what it ought to realize in the appearances of which it is the
But there is also a perspective of moral anthropology. The need for con-
sidering a moral–anthropological perspective arises from the recognition
that one has a propensity to misuse one’s freedom. This leads, as we have
seen, to a more complicated conception of the appearance in time of the
good will. Whereas one might have thought that one need only do good
deeds now to express a good will, one finds that one must act in such a
way that one will combat the propensity to evil and promote good deeds
in the future. One’s deliberation is always about what to do now, but moral
anthropology provides a means for extending the effects of deliberation
Radical Evil and Moral Anthropology 133

beyond the present. Moral anthropology, then, studies the way in which
natural causes lead to human actions that express most clearly, though
never indubitably, moral goodness. This allows an agent to pursue in the
present those causes which lead to actions that express a good will. And
because the struggle against evil, expressed in consistent moral improve-
ment, is the appearance that one must always seek to realize, one ought
always to pursue these empirical causes.
Moreover, Kant explains that ignorance of one’s moral status intensifies
the importance of helps and hindrances. Given that one can never be
sure whether one is making enough progress and that one can never be
sure that one will continue to do good in the future, one must always seek
to influence a future self that may return to evil. One must remain, as
Kant says, “forever armed for battle” (6:93). In the cool hours in which
one acts from good maxims, one must seek to increase the number and
effectiveness of future cool hours, to combat the deceptions and manip-
ulations of one’s own evil will, and to promote any empirical influences
likely to result in further improvement of one’s actions and observable
intentions. None of this will grant knowledge that one has a good will; noth-
ing can. But the greater the degree of resistance to evil, both present and
future, the more justified one’s hope that one is good. This resistance to
evil, present and future, is in any event what the moral law requires of hu-
man beings, wicked as we are. And the resistance to future evil involves
attention to empirical helps and hindrances.
This ignorance also implies that one cannot simply compare one’s life
with the lives of others. What is required to express a moral revolution
may differ depending on one’s empirical conditions. Of course, one must
always strive to act in accordance with the moral law and to promote aids
to morality. But given the fact of radical evil, one will fall short of always
acting in accordance with the moral law, and one will often fail to promote
aids to morality. And one can never know for sure whether one has gone
far enough. Thus any person committed to moral revolution ought to
express that moral revolution at every opportunity. But when it comes to
evaluating oneself or others, it is reasonable to have different levels of
hope for the same behavior, depending on external conditions.
One who has abundant advantages in terms of natural disposition,
moral education, political stability, and a community that supports ethi-
cal behavior cannot reliably express a moral revolution merely by resisting
some of the temptations to vice that occasionally arise. Given those advan-
tages, such an individual can show his or her struggle against evil only by
consistent obedience to the moral law and by promoting more subtle aids
134 The Solution

to future moral improvement. By contrast, one with a disadvantageous

temperament who lacks moral education or the support of an ethical
community might express moral revolution in the increasing effort to
resist temptation and to promote at least a minimal respect for virtue
in him- or herself and those by whom he or she is influenced. Neither
one with advantages nor one without them is justified in moral compla-
cency or despair. But the ignorance of what is needed to express a moral
revolution does mean that one is not warranted in ascribing to external
conditions a causal influence on one’s free choices.48

Finally, it is important to note that moral anthropology has limits. It is

limited by the constraints of the empirical world, especially laws of human
psychology. One can affect one’s future behavior, but only in ways that
are consistent with the kinds of influence to which human beings are
found to be susceptible. Moral anthropology is also limited by moral
relevance, in the sense that only influences that are morally relevant are
considered part of moral anthropology. One can influence one’s behavior
in ways that are irrelevant to morality; Kant’s pragmatic anthropology
includes a lot of advice that is primarily helpful for increasing future
happiness. But only empirical influences on decisions that are morally
relevant are included within the realm of moral anthropology. Finally,
the range of moral anthropology is limited by transcendental freedom.
One can affect future choices, where a “choice” is something empirical.49
But this “choice” is always only the expression of one’s free will, not the
free will itself. Acting now can affect future choices and can express a
revolution in the free will, but nothing empirical can affect the free will

5. Conclusion
Because one’s actions in the world are susceptible to empirical influence,
and because one is prone to evil, one can and should study those influences
that aid one’s good maxims in getting the upper hand over one’s evil
maxims. This study, because it is undertaken with a practical goal (the
overcoming of evil in oneself) and because it is a study of a person who
is free (though not a study of that person insofar as he or she is free),
is rightly considered a study of “what the human being makes, can, or
should make of himself as a freely acting being” (7:119). At the same time,
because it is a study of appearances and of the causal relations between
appearances, it is a properly empirical science and can yield empirical
Radical Evil and Moral Anthropology 135

knowledge. This discipline, which teaches how one can effect the progress
in appearance from evil to good, is moral anthropology.
Kant’s anthropological observations about empirical helps and hin-
drances to the development of the good will do not undermine the
priority of freedom over nature because the empirical helps and hin-
drances affect only the appearances of the good will in the world. And this,
of course, is exactly what Kant means when he writes, “the resistance or
furtherance is . . . between nature as appearance and the effects of freedom
as appearances in the world of sense” (5:196). But Kant’s anthropology
is also morally relevant because it studies appearances that express one’s
moral status. The empirical influences described by moral anthropol-
ogy are important parts of those expressions because of a propensity to
evil that human beings freely adopt. This propensity affects one’s future
choices because it is a propensity to choose evil in certain circumstances.
The need for empirical aids to combat radical evil does not mark an ex-
ception to Kant’s principle of “ought implies can,” because this need is
due only to choices freely made. The will itself is corrupt and thus needs
help, but it is freely corrupt.
But one will not simply overcome evil through isolated choices. Rather,
one must effect changes in the empirical influences on one’s character.
Effecting these changes expresses one’s moral revolution, and knowledge
of them comes through moral anthropology. Moral anthropology thus
provides human agents empirically informed advice about how best to
express themselves morally, and this advice becomes morally relevant if
such agents freely choose to act on it.

One central problem remains with this account of moral anthropology.

The explanation of anthropological influences as means for expressing
one’s revolution against evil works well for those influences that one
effects in oneself. But the account does not seem to work as well in those
cases where one agent promotes helps or combats hindrances for another.
This limitation seems particularly problematic given Kant’s claims about
moral progress in human history and through moral education. The next
chapter takes up this remaining problem.

Moral Influence on Others

In the previous chapter, I offered a Kantian account of the nature of

the good will and its appearance in the world that makes sense of the
moral significance of empirical helps and hindrances. The account of-
fered there explains why promoting helps and combating hindrances to
moral progress expresses a moral revolution, the only form of good will
that remains possible for one who is radically evil. But many helps and
hindrances do not seem to fit this model. When one struggles against
passions because one recognizes that they are moral hindrances or when
one enters into polite society because one knows that doing so can im-
prove one’s own moral disposition, one expresses a will in revolution by
struggling against the radical evil in one’s nature. But what is going on
when a parent disciplines a child, or a teacher instructs a pupil morally, or
when one is polite to present virtue in the most attractive light to others?
In these cases one does not seem to struggle against one’s own evil, so
one seems not to express a will in revolution. And the child, or pupil,
or those to whom one is polite are not necessarily struggling against evil
themselves, so they may not be expressing a will in revolution. In these
cases, the account in the previous chapter – if it can be relevant at all –
seems to grossly misplace the moral significance of helps and hindrances.
When one promotes aids and resists hindrances for oneself, one’s activ-
ity is morally significant for the one who is helped and hindered. In the
case of interpersonal interaction, the account in the previous chapter
seems to make helps and hindrances to another morally significant for
the helper or hinderer rather than for the one helped or hindered. That
just seems to get the moral importance of helping and hindering all
Moral Influence on Others 137

One might think that the tension between freedom and nature is not
as serious in the interpersonal case as in the individual case because
one need not consider an agent that one seeks to improve from both
the empirical and the practical perspectives at the same time. In one’s
own case, one must consider oneself both from the practical perspective
because one is making a choice and as susceptible to empirical influ-
ence because one is choosing an empirical influence on oneself. But
in a case where one person deliberates about morally helping another,
one might think that only the deliberating person needs to be consid-
ered free while the person helped is considered capable of empirical
influence. Were that true, there would be no conflict in the case of inter-
personal moral influence, because no person would need to be consid-
ered from both the empirical and the practical perspective at the same
But freedom is threatened in the interpersonal case. The notion that
one can consider oneself free while viewing the person to be helped as
an appearance might be acceptable when one seeks merely to improve
another’s physical or even intellectual capacities. But it is inadequate for
cases in which one seeks to affect another for the other’s moral improve-
ment. When one acts on a maxim like “I will do some action in order
to promote the good will of my pupil,” either one holds that the action
will promote the good will of the pupil or that it will not. If one denies
that the action can have any effect on the will of the pupil, then one
acts irrationally in following the maxim, because it violates the norms of
technical reasoning. According to these norms, one should not perform
an action for an end when one believes that the action has no effect
on bringing about the end. But all of one’s actions are, from the stand-
point of the other, empirical influences. Thus if one holds that the action
can affect the will of the other, one grants that empirical influences af-
fect the free will of another. This is the problem of interpersonal moral
In this chapter, I apply the model developed in the previous chapter
to interpersonal moral influence. In section 1, I point out a seeming
conflict between Kant’s account of virtue in the Metaphysics of Morals and
his account of ethical community in the Religion, and I articulate a model
of interpersonal influence in Kant that resolves this conflict. In section 2, I
address the possible objection that my account of interpersonal influence
ends up being self-centered. In section 3, I offer an interpretation of
Kant’s writings on history in the light of my account of interpersonal
moral influence.
138 The Solution

1. Kant on Interpersonal Moral Influence

In the Metaphysics of Morals, Kant presents an explanation of our duties
with respect to others which suggests that interpersonal moral influence
is impossible. In a section entitled “What Are the Ends That Are Also
Duties?” Kant lists only two ends that are duties, “[1] one’s own perfection
and [2] the happiness of others” (6:385). The absence of “the perfection
of others” or “the moral improvement of others” or something of the sort
is noteworthy,1 but more important is the argument that Kant offers for
why the perfection of another cannot be an end:

It is a contradiction for me to make another’s perfection my end and consider myself

under obligation to promote this. For the perfection of another human being, a
person, consists just in this: that he himself is able to set his end in accordance with
his own concepts of duty; and it is self-contradictory to require that I do (make it
my duty to do) something that only the other can do himself. (6:386)

We ought not, because we cannot, do anything to morally improve another,

because moral improvement is precisely something that one must do
for oneself. The view presented in the Metaphysics of Morals seems to be
quite clear, and it is consistent with Kant’s emphasis on freedom as well
as with the account of helps and hindrances presented in the previous
chapter. If one cannot influence others morally, then one can properly
understand every case of empirical influence as an expression of one’s
own struggle against one’s own radical evil, as described in the previous
chapter. Interpersonal moral influence could be dealt with just by denying
that it is possible.
Such an approach suffers from one obvious disadvantage in that it
seems implausible that people cannot have any morally relevant impact
on one another. But more important, it seems to conflict with claims that
Kant makes elsewhere. In the Religion, as well as in his writings and lec-
tures on history and education,2 Kant suggests that interpersonal moral
influence can play an important role in moral life, for either good or ill.
Interpersonal influence can influence one negatively, because the main
“assaults of the evil principle” for control of one’s will come “from the
human beings to whom he stands in relation or association” (6:93). But
others also play an important role in promoting morality, and one seems
to have a responsibility to promote virtue in others because “the domin-
ion of the good principle is not otherwise attainable . . . than through
the setting up and the diffusion of a society in accordance with, and for
the sake of, laws of virtue – a society which reason makes it a task and a
duty . . . to establish” (6:94).
Moral Influence on Others 139

But how can one have a society for the sake of virtue if people are
incapable of affecting one another morally? And how can one have a duty
to establish such a society if the Metaphysics of Morals prohibits seeking to
make the moral perfection of others an end for oneself ?

In fact, by making clear the distinction between influencing the deeds of

others and influencing their moral status itself, one can see several im-
portant roles for interpersonal “moral” influence that fit the strictures
of the Metaphysics of Morals and the account of the good will offered in
Chapter 5. Good deeds are the empirical expression of a good will. As
empirical, these deeds are capable of influence. The good will that they
express, however, is not. In light of this distinction, I present three roles
for interpersonal influence with different degrees of moral significance.
In none of these roles do one’s actions actually determine or even clearly
affect the moral status of another. In that sense, they are all compatible
with Kant’s claims in the Metaphysics of Morals about the limits of interper-
sonal influence. But in all cases one has a duty to promote good deeds
in others. Two reasons for promoting good deeds are taken from the
Metaphysics of Morals itself. The third is based on the discussion of ethical
community in the Religion. In the context of discussing that role for inter-
personal influence, I show how the promotion of good deeds in others
is an important part of one’s own struggle against evil.

A. Promoting Virtue to Promote the Happiness of Self-Esteem

One reason to promote moral behavior in others comes from our imper-
fect duty to promote the happiness of others. In the Metaphysics of Morals,
shortly after pointing out that one ought to have as an end the happiness
of others but not their perfection, Kant adds,

The happiness of others also includes their moral well-being, and we have a duty,
but only a negative one, to promote this. . . . It is my duty to refrain from doing
anything that, considering the nature of a human being, could tempt him to do
something for which his conscience could afterwards pain him. (6:394)

Even if one cannot promote moral perfection in another, one can tempt
others to actions that express an evil will. Kant’s description of the limi-
tations on self-knowledge makes clear that although one can never know
that one has a good will, certain acts and intentions are clear signs of
evil. At least some of these actions, unless one is remarkably adept at self-
deception, will pain one’s conscience. This self-knowledge has beneficial
effects, in that it gives rise to the recognition of radical evil and ushers
140 The Solution

in Kant’s new conception of what it means to have a good will. But the
more direct effect of recognizing one’s evil is displeasure. Doing wicked
things tends to make people unhappy, all else being equal. All people
have a seed of goodness that cannot be extinguished, and that seed in-
cludes a conscience that causes pain at the recognition of moral evil in
oneself. Thus the obligation to promote the happiness of others includes
an obligation to refrain from tempting them to evil deeds for which they
will afterward suffer. The promotion of happiness may include even the
obligation to help others choose to perform more good deeds for which
they will feel the pleasure of self-respect.
It is important to point out that this reason for promoting good deeds
does not imply that one helps others to pursue virtue for the sake of
happiness. A father who helps his children do more good deeds out
of a duty to promote their happiness ought to foster morally upright
behavior at least in part for the sake of his children’s happiness, because
he has a duty to promote their happiness. But his children ought to pursue
morally upright behavior for its own sake.3 Insofar as his children perform
good deeds for the happiness they will derive from such deeds, they fail
to exhibit moral virtue, and thus they lose the proper pleasures of self-
esteem that would have come from doing good for its own sake. Insofar as
a good father cares about the happiness of his children, he will promote
in these children a tendency to act in ways which reflect that they care
more about the moral law than about their own happiness.4
In this context, it is also important to note that the promotion of
good deeds includes good intentions. Intentions, like external actions, are
appearances, and thus they are capable of empirical influence. And they
are appearances that express, but are not identical to, one’s moral status.
Thus one feels moral self-condemnation not only for external actions that
are contrary to the moral law but even for internal states. And one who
seeks to promote the happiness of another will seek to influence not only
the other’s visible actions but also the intentions, thoughts, desires, and
feelings of which only the other is directly conscious. None of this implies
that one affects the free will of the other, because these mental states
are merely objects of inner sense. But this does imply that the “deeds”
with which one is concerned include mental states as well as external

B. Promoting Virtue as Indirect Virtue

When one helps another to act in morally virtuous ways, one promotes not
only the happiness of the one who is morally helped but also the happiness
Moral Influence on Others 141

of third parties. All people have a duty to promote the happiness of

others. By encouraging those on whom one has influence to promote
the happiness of others, one indirectly promotes that happiness oneself.
But if one encourages someone to promote the happiness of others in
ways that conflict with the moral law, then one’s encouragement itself
violates the moral law. Thus one must encourage others to promote the
happiness of others in ways that are morally right. And that is to say, one
must encourage virtue.
To see this more clearly, it is important to see under what circumstances
it would be wrong to promote the happiness of another in a personal case.
Take this maxim:

M1: When I see some person X who can benefit from money or property belong-
ing to another person Y, I will falsely promise to repay in order to borrow the
money or property from Y and give it X, in order to benefit X.

On the one hand, this maxim is consistent with the requirement of virtue
that one make the happiness of others an end. But one who acts on such a
maxim is not virtuous, because this maxim is contrary to the moral law for
the same reasons as the more selfish example of false promising that Kant
discusses in his Groundwork (see 4:422, 4:429–30). Specifically, although
the maxim would probably be to the benefit of X and may not significantly
affect Y in any material sense, the maxim and its universalization cannot
be consistently willed, and one who acts on such a maxim makes use of Y
merely as means of promoting the good of X.
But now imagine a virtuous teacher who considers teaching her stu-
dents to act on maxim M1, or perhaps a refined version of maxim M1
that makes explicit that the transfer from Y to X will increase overall
happiness. She acts on this maxim:

M2: In order to promote the happiness of others, I will teach my students to

falsely promise in order to borrow from some and give to others, in such a way
that overall happiness is increased.

Assuming that the students are smart enough to make such judgments,
acting on this maxim is likely to increase the overall happiness of others.
Thus it would be a suitable maxim for someone who takes the happiness
of others as an end. But like one who acts on M1, a person who acts on M2
is not virtuous, because this maxim is not compatible with the moral law.
Given that human beings are susceptible to instruction, universalizing M2
will cause the same sort of contradiction as universalizing M1. Moreover,
M2 makes use of others as mere means just as much as M1, even if the
142 The Solution

others are not as well specified in each situation. Insofar as a teacher

is truly virtuous, she will teach her students to promote the happiness
of others, because this instruction is an effective and legitimate means
of promoting that happiness. But a virtuous teacher will be careful to
teach her students to promote the happiness of others in morally right
ways, because to teach them otherwise would itself be morally wrong.
But teaching students to promote the happiness of others in morally
appropriate ways just is teaching them virtue. In this context, promoting
true virtue in others is an indirect way of being virtuous oneself.
Both this reason for promoting virtue and the previous one express
one’s own virtue in the sense that they show one’s concern for the happi-
ness of others. Because taking the proper happiness of others as an end
is morally obligatory and promoting good deeds in others is a way of pro-
moting the happiness of others – whether the happiness of those who do
the deeds or the happiness of benefited parties – promoting good deeds
is a means to an obligatory end.

C. Promoting Virtue in Others as an Expression

of a Social Struggle Against Evil
In addition to serving the happiness of others, promoting virtue in others
also serves to promote the other obligatory end of virtue: one’s own
perfection. Unlike the straightforward ways in which the virtue of others
promotes their own happiness and the happiness of third parties, the
effect of the virtue of others on one’s own perfection is complicated. The
virtue of others can nonproblematically promote nonmoral perfections in
oneself. A colleague’s fairness, diligence, and cooperation can help one
cultivate one’s intellectual talents by pointing out fruitful new directions
of thought or beneficial techniques for study and reflection. Because
these perfections are nonmoral, the virtue of others does not in any
sense interfere with one’s freedom but merely enables one to articulate
and pursue one’s ends more effectively.
But in the Religion Kant indicates that the virtue of others can promote
one’s moral perfection. In fact, this is precisely the reason that Kant says
one should strive for a moral community. Kant begins the third part of
the Religion with a recapitulation of the central point of the previous two

[A human being] remains . . . exposed to the assaults of the evil principle; and, to
assert his freedom, which is constantly under attack, he must henceforth remain
forever armed for battle. (6:93)
Moral Influence on Others 143

In Chapter 5, we saw that this struggle against evil is the way in which
human beings express a moral revolution in their free will. But in the
third part of the Religion, Kant goes on to offer a specific source of evil and
corresponding arena for struggle.

If [a human being] searches for the causes and the circumstances that draw him
into this danger [i.e., assault of the evil principle] and keep him there, he can
easily convince himself that they do not come his way from his own raw nature . . .
but rather from the human beings to whom he stands in relation or association.

Kant insists that one “is in this perilous state through his own fault,” and
in the previous chapter we saw that the propensity to evil is itself an out-
growth of one’s own evil choices. But other human beings are among the
most important empirical means by which one fosters one’s propensity to
evil. Thus because one must in general oppose the propensity to evil using
empirical influences, one must oppose the corrupting influence of other
people in one’s society using particular sorts of empirical influences.
One can struggle against the corrupting influence of society only
through the moral reform of that society.

The dominion of the good principle is not otherwise attainable . . . than through
the setting up and the diffusion of a society in accordance with, and for the sake
of, laws of virtue. (6:94)

To engage in a revolution against evil, one must foster a purely ethical

community that has as its purpose not the cultivation of vice but the
promotion of virtue.7 The empirical expression of a will in revolution,
given the social nature of evil, is a will that engages in a social struggle
against evil.
Social influence, of course, does not affect one’s free will. The inter-
dependence of human beings is an empirical claim, and thus the interde-
pendence is at the level of human appearances. Kant even says that this
corrupting influence is possible despite people’s wills’ remaining good:

Even with the good will of each individual, because of the lack of a principle that
unites them, they deviate through their dissensions from the common goal of
goodness, as though they were instruments of evil, and expose one another to
the danger of falling once again under its dominion. (6:97, emphasis added)

Others do not cause one’s radical evil, and they cannot cause one to
undergo a moral revolution, but they can affect the way that moral rev-
olution appears. Insofar as others act wickedly, they encourage one to
act wickedly. And even the mere presence of others around someone
144 The Solution

makes him that person “anxious that other human beings will consider
him poor and will despise him for it” (6:93). People appear in the or-
der of nature, and they act causally on one another in such a way that
desires are fostered on the basis of which people act contrary to the
moral law.
This evil social influence has important implications from the stand-
point of someone wishing to express his or her struggle against evil. Such
a person, recognizing the important interdependence of human beings
at the level of appearance, must promote good deeds among those with
whom he or she comes into contact. Good deeds in others, even if these
deeds are merely appearances and do not express genuine moral good-
ness, promote good deeds in oneself. More important given the social
hindrances to morality, one who is morally good will engage in a strug-
gle against social evils by fostering an ethical community that explicitly
aims at the promotion of morality.8 Only if others act virtuously and ac-
tively promote virtue can one’s moral revolution be fully expressed in
the world. And thus insofar as one truly revolts against evil, one seeks
a community within which that revolution will be expressed. Even if
one can neither make others morally good nor even promote in oth-
ers a genuine interest in promoting morality, one can affect the way
others appear such that they at least seem to be good and to encour-
age goodness. This appearance promotes one’s own good deeds. The
ethical community, even as a mere appearance, is a help to morality.
Thus promoting an ethical community expresses one’s revolution against
This promotion of an ethical community cannot be limited to those
who are likely to affect oneself, nor even merely to the generation in
which one lives. Kant suggests that the promotion of ethical community
involves sharing in a goal of which one will personally experience only
some of the effects. Thus it might seem unreasonable to limit the signifi-
cance of an ethical community to the expression of one’s own revolution
against evil. And of course this community, insofar as it emerges, is not
an expression merely of any individual revolution but the combined ef-
fect of many revolutions, all expressing themselves in part in striving for
ethical community.
Even from the standpoint of the individual, however, there are two
important reasons for promoting ethical community beyond the likely
range of its direct influence on one’s own actions. First, one can never be
sure about what aspects of an ethical community will have a moral effect
on one. The toddler in whom one encourages virtue today might become
Moral Influence on Others 145

an important influence on one later. Second, one’s radical evil promotes

effects that persist beyond one’s death. Insofar as one is radically evil, one
makes use of society to promote and encourage one’s wickedness. This
use of society contributes to general societal corruption. And even if this
corruption cannot actually cause others to be morally evil, it promotes
injustice, unhappiness, and vice. These lingering social evils are at least
partially attributable to one’s past evil choices. In order to fully struggle
against the effects of one’s evil influence on society, one must struggle
to improve society, even if such improvement will benefit only those who
come later.9
In the previous chapter we saw that the presence of radical evil implies
that one’s moral goodness is expressed not only in good actions but
also in a constant struggle against evil. But it turns out that at the level
of the appearance that expresses one’s moral status, one is inextricably
linked with the appearances of others. Thus one’s own moral goodness
expresses itself not merely in good actions combined with an individual
struggle toward goodness but also in a social struggle toward good and
against evil. Because my radical evil makes use of a social network of
promoting evil, the revolution against evil must express itself in social
This emphasis on social transformation and social evil does not imply
that human societies or the human species as a whole has a moral status.10
Even when Kant talks about the human species being evil, he points out
that what it means to say that “character applies to the species” is that
“there is no cause for exempting anyone from it” (6:25–6). Similarly,
moral improvement of a species is the moral improvement of its indi-
viduals. Social transformation is ultimately, at a moral level, widespread
individual transformation.11 In the case of one’s own efforts for social
transformation, one improves oneself morally; one expresses one’s own
moral status. But this expression of moral revolution involves a society that
improves in the performance of good deeds and in the explicit commit-
ment to morality.
The problem of interpersonal influence thus leads to a refinement in
the conception of what the good will looks like in the world. Expressing a
good will involves not only acting rightly and struggling to promote helps
and combat hindrances to one’s own virtue but also participating in a
social struggle by promoting helps and combating hindrances to virtue
in others. This social struggle is not merely a means to an end. Because it
is a means for combating evil in oneself, it becomes constitutive of the
expression of a good will.
146 The Solution

2. Is Moral Influence Morally Self-Centered?

In section 1, I offered three roles for promoting morality in others. All
of them focus on others. The first role leads to happiness for the one
in whom one promotes morality. The second focuses on happiness for
third parties. The third is concerned with increasing the good deeds of
others. But in another sense all three roles are fundamentally concerned
with oneself. None of them actually promotes anything genuinely moral
in anyone else. They all express one’s own moral goodness. For parents,
teachers, moralists, or religious leaders who are deeply concerned with
the moral status of their children, students, audience, or parishioners,
the roles for moral influence offered here can seem grossly insufficient.
In this section I consider how far Kant can go toward allowing for genuine
and ultimate moral influence, and I partially defend the sort of concern
for others that Kant allows.

A. Why One Ought to Be Self-Centered

The first point that Kant can and should make is to remind those who
seek to help others become morally good that this goal reflects a con-
cern that is morally suspect. Kant’s argument in the Metaphysics of Morals
against seeking the moral perfection of others is based on the incapacity
to do for another what only that other can do for him- or herself. In that
argument Kant suggests a couple of dangers of seeking to help others be-
come morally good. First, “the perfection of another consists in this: that
he himself is able to” act in accordance with duty (6:386). True perfection
comes only when it is due to oneself. Second, perfection of another de-
pends on that other’s acting in accordance with “his own concepts of duty”
(6:386, emphasis added). One’s concepts of duty should be universally
valid. One does not arbitrarily choose what is one’s duty, so one’s own
concepts of duty should be the same as any other person’s concepts of
duty. But for Kant it is essential that duty be one’s own in the sense that
one must be autonomous; moral obligations binding on one must have
their source in one’s own practical reason. In this context, promoting
virtue in another is dangerous because even if one avoids the temptation
to promote what is merely one’s own idiosyncratic preference or fixation
as though it were a duty, to promote one’s own concept of duty in the
actions of another is to assign a heteronomous source – oneself – for the
duties of the other. This influence need not be strictly heteronomous.
One can teach another that the moral law must be legislated by him- or
herself. But the relationship of teacher and learner always brings with it
Moral Influence on Others 147

at least the danger that the learner will submit to a heteronomous source
of duty.
A further danger is that the very encouragement by which a good-
hearted parent or moralist seeks to raise another to moral perfection
can in fact lower the other to the level of a mere thing. By seeking to
affect the ultimate moral status of another, one treats the freedom of
the other as something natural, susceptible to empirical influence and
manipulation. What one wants to change in another – his or her moral
status – is just the sort of thing that must be changed only by the other.
To seek to promote ultimate moral change in another really just denies
that other moral autonomy and hence moral dignity.
Thus there is an important sense in which one ought to be morally self-
centered. Being morally self-centered is very different, of course, from
being selfish. One ought to take the happiness and well-being of others
as an end and ought always to behave justly and generously with others.
But one ought not take it upon oneself to ensure that others undergo
fundamental moral change. To desire to affect the ultimate moral status
of another is simply to desire too much. Such a desire is a form of hubris
that sets one’s own capacity to exercise influence in the world above the
freedom of another person.

B. When One Can Be Morally Centered on Another

Of course, the prohibition on making one’s end the moral perfection of
another does not mean that one cannot take a genuine moral concern
in others, but this concern should limit itself to the good deeds of others.
There is no reason why one’s concern for promoting good deeds in others
need always be merely instrumental. I suggested three purposes for which
one might promote good deeds in others, but one might also promote
good deeds for their own sake. One is not obliged to make the good deeds
of others an end. But if one has a fundamental interest in seeing one’s
children or students or friends act in ways that are consistent with the
moral law, then one can promote such action for its own sake. Given that
actions express the ultimate moral status of an agent, the natural desire to
see loved ones be morally good might reasonably take the form of a direct
interest in the appearance of morality in those loved ones. Unlike the
desire to promote moral change at the level of freedom, there is nothing
morally dangerous about this interest. Without doing any harm to others,
one can promote good deeds in them not merely for the sake of happiness
or to express one’s own moral goodness but for the sake of the good deeds
themselves or the good moral appearance of the agents performing them.
148 The Solution

Ultimately, this purpose too is self-interested in a sense, because one

acts to promote one’s own interest in the good deeds of others. In this
sense, a direct interest in the good deeds of others is like the natural
benevolence that Kant describes in the Groundwork (4:398), where one
takes a direct interest in the happiness of others. Such a direct interest is
not morally praiseworthy, but it is not wrong when kept within the limits
of the moral law. And it might seem less morally self-centered than other
reasons for helping others, because one’s interest in others’ good deeds
is direct rather than mediated by any concern with the effects of those
deeds on one’s own moral status.

C. Helping Others and Hope for Others

One might add to this reason for promoting virtue in others a basis for
hope that focuses directly on the ultimate moral status of others. In the
previous chapter, the inscrutability of one’s ultimate moral basis for ac-
tion led to a focus on what is observable – primarily one’s actions –
as the ground for hope in one’s moral goodness. Because one can
hope for grace and future improvement on the basis of present good
deeds and a visible struggle against evil, one effects good deeds and the
struggle against evil at every opportunity. In the case of others, one is
similarly incapable of discerning the ultimate moral ground of the ac-
tions of others. But one is no less capable of discerning this ground
than in one’s own case. One of Kant’s most radical claims in the first
Critique, especially in the B edition, is that one does not have privi-
leged knowledge of oneself. For Kant, one can observe one’s own de-
sires and intentions – in inner sense – in a way that one cannot ob-
serve these in the case of others. It may be easier, at least in some
cases, to recognize one’s own self-deception than deception in others.
But this difference is a difference of degree, and often a slight one. In both
one’s own case and the case of others, one can at most hope for – but never
have knowledge of – moral goodness. And in both cases, one can hope
only on the basis of actions.
This similarity between the case of one’s hope for oneself and one’s
hope for others is important in the context of acting to promote good
deeds in others because it implies that one’s promotion of good deeds in
others can be conjoined with a legitimate hope, as one sees another do
more and more good, that that other is actually morally good. Just as a
person who struggles against evil and does more and more good in life can
legitimately hope for grace and thereby a good will, a parent who raises
children who do more and more good in their lives can legitimately hope
Moral Influence on Others 149

that those children have fundamentally good wills. Acting to promote

good deeds, insofar as this action is successful, can increase one’s hope
in the ultimate moral goodness of those one helps.
But there is a crucial difference when it comes to the implications of
hope between hope in one’s own goodness and hope in the goodness of
others. In the previous chapter we saw that hope in one’s goodness plays
a role in choosing to perform good deeds and to struggle against evil.
Hope in one’s goodness prevents moral despair and provides a reason
for acting in such a way that this hope is legitimate. One acts in such a
way that one can justify hope in one’s worthiness for grace. Because one
is responsible for one’s own moral status, acting such that one justifies
rational hope in one’s own moral goodness is a form of directly expressing
that goodness at the level of one’s freedom. In the case of others, however,
the justification of hope ought not provide a reason for action. When
parents’ efforts to promote morality in their children succeed, the good
deeds of their children can legitimately justify the parents’ hope that
their children are morally good. But this legitimation of hope cannot
be the basis for promoting morality in their children because for it to
provide a basis the parents would have to reason that they have or can
have some responsibility for shaping the ultimate moral status of their
children. In the individual case, it is perfectly appropriate to think that
I have responsibility for my ultimate status. But to think that I have any
responsibility for determining the ultimate moral status of another is to
deny the full freedom and autonomy of the other. To promote good
deeds in another on the ground that I can actually determine – in whole
or in part – the ultimate moral status of that other is to treat that person’s
freedom as an effect in nature. And to treat freedom in that way is to treat
the other as a mere thing.
One ought to promote in others good deeds and even an explicit com-
mitment to the moral law. One can promote these external signs of good-
ness through the empirical aids – such as politeness, moral education,
and the ethical community – that Kant studies in his moral anthropology.
And one can rightly hope that those who express a commitment to the
moral law in word and deed are worthy of grace and thus morally good at
the level of their transcendental freedom. But one cannot make it one’s
intention to directly promote another’s ultimate moral goodness without
disrespecting that other. One’s hope in ultimate moral goodness may
support one’s efforts to promote empirical goodness but can never be the
basis for those efforts. Still, insofar as actions provide the basis for hope in
the ultimate moral goodness of others, and insofar as one promotes these
150 The Solution

actions, this hope reflects an important sense in which one’s efforts to

promote moral goodness in others do reach the ultimate level of having
real and ultimate moral relevance. In addition to promoting the happi-
ness and good deeds of others, it is not impossible that one’s efforts on
behalf of others actually have an effect on their moral status. One cannot
know in a theoretical sense nor even postulate for practical purposes that
one does have such an effect, but it need not be considered impossible.

D. Gestures Toward a Possibility: Real, Ultimate, Moral Influence

A final consideration can give some metaphysical weight to this possibility
of real and ultimate moral relevance. In a discussion of grace in his Reli-
gion, Kant very briefly opens the possibility of God’s affecting human free
choice. In his reconciliation of the antinomy of grace, where he argues
that from a practical perspective grace must be considered consequent
to good works, Kant says,

We can admit an effect of grace [that is, a grace that is the ground rather than
the consequence of good works] as something incomprehensible, but cannot
incorporate it into our maxims for either theoretical or practical use. (6:53)

Later, Kant explains in more detail why we can admit this conception of
grace as something incomprehensible:

The concept of a supernatural intervention into our moral though deficient

faculty . . . – this is a transcendent concept, merely an idea of whose reality no
experience can assure us. – But even to accept it as an idea for a pure practical
intent is very risky and hard to reconcile with reason; for what is to be accredited
to us as morally good conduct must take place not through foreign influence but
only through the use of our own powers. Yet its impossibility (that the two may
not occur side by side) cannot be proven either, because freedom itself, though
not containing anything supernatural in its concept, remains just as incompre-
hensible to us according to possibility as the supernatural [something] we might
want to assume as surrogate for the independent yet deficient determination of
freedom. (6:191)

From the standpoint of experience, it makes no sense to talk about a grace

that pre- or co-determines free choice because neither grace nor freedom
is comprehensible from this standpoint. From the practical standpoint,
one can make sense of both grace and freedom, but giving any kind
of practical priority to grace is “hard to reconcile with reason,” to say
the least, for exactly the reasons already examined with respect to inter-
personal moral influence. One’s moral condition is necessarily due to
oneself. Insofar as someone or something else is a more ultimate ground
Moral Influence on Others 151

of one’s moral status, that other person or thing bears the moral weight.
Only if a moral status is fundamentally the result of one’s own choice can
it be a moral status of one’s own. A goodness that is fundamentally an
effect of grace is not one’s own.12
However, Kant does not rule out the possibility13 of grace. He just
rules out making use of grace from either a theoretical or a practical
perspective. As noted in the previous chapter, Kant does not suggest any
other perspective from which one might make use of grace. But there is
conceptual space for such a new perspective.
One might prefer to leave grace in a position of limbo, without a clear
standpoint from which it makes sense to talk about prevenient grace.
But one can use Kant’s argument against the impossibility of grace to
open a similar space for genuine interpersonal moral influence. This
goes beyond Kant’s own treatment of interpersonal influence, but it is
consistent with his mature moral philosophy overall. Just as the freedom
of a moral agent is “just as incomprehensible to us . . . as the supernatural,”
the freedom of one moral agent is as incomprehensible as the freedom
of the next. In the same way that the incomprehensibility of both grace
and freedom precludes the denial of a sort of relation according to which
grace affects freedom but freedom nonetheless remains the ultimate and
fundamental ground of action, the incomprehensibility of different free
agents precludes the denial of real and ultimate interpersonal moral
influence. The relation between a parent’s incomprehensible free choice
and the free choice of a child may be such that the choice of the parent
somehow affects the choice of the child, but the choice of the child is
nonetheless an uncaused and fundamental ground of the child’s actions.
As in the case of grace, this “possibility” can be put to neither theoretical
nor practical use. Perhaps one could articulate some third perspective –
say, a perspective of hope – from which it would have use. For Kantian
parents, teachers, moralists, and religious leaders who desperately want to
hold open some possibility of genuine and fundamental moral influence,
this possibility may be reassuring.

3. History and Moral Progress

In the previous two sections, I have laid out a Kantian account of interper-
sonal moral influence as fundamentally a matter of promoting good deeds
in others for the sake of increasing happiness in others and expressing
one’s own moral revolution against radical evil. In this section, I show how
Kant’s writings in history are compatible with this general account. Kant’s
152 The Solution

philosophy of history has received extensive discussion in recent years.14

Some commentators insist that Kant’s philosophy of history involves a
sort of moral progress that would be incompatible with the account of
moral influence that I have presented. Louden, for example, claims that
“history, while concerned with visible events that we can see and touch, is
nevertheless ultimately and most fundamentally concerned about moral
progress” (Louden 2000:147).15 In this section, I briefly articulate my
account of Kant’s philosophy of history. I then show how some of the pas-
sages most often taken to support claims that Kant advocates more radical
moral progress can be read as consistent with my overall account.16 For
Kant, culture is progressing, good deeds are becoming more prevalent,
and good states will emerge. All of this takes place in the world of appear-
ances. But there is no reason to think that there are or ever will be more
good free wills than in the past.

A. A Brief Survey of Kant’s Writings on History

Kant writes on history throughout his Critical period, beginning with
an essay in November 1784 for the Berlinische Monatsschrift, “one of the
most aggressive of the secularizing-rationalist journals of the day”
(Zammito 1992:329).17 This essay, “Idea for a Universal History with a
Cosmopolitan Purpose,” offers nine propositions that constitute a prelim-
inary philosophical history of the human species. The essay was followed
the succeeding year with three reviews of Herder’s Ideas on the Philosophy
of the History of Mankind. These reviews were published in installments (in
January, March, and November 1785) in the Allgemeine Literatur Zeitung,
a new journal run by C. G. Schütz. The reviews focus largely on methodo-
logical considerations and do not contribute significantly to Kant’s ac-
count of moral progress. Far more important is another essay published
in the Berlinische Monatsschrift in January 1786. This essay, “Conjectures
on the Beginning of Human History,” is a further response to Herder in
the form of a philosophical interpretation of the Biblical narrative that
describes the origin and fall of humankind.
After this flurry of essays, Kant did not devote further works to philo-
sophical history for almost a decade. During this time he worked on his
Critique of Judgment (published in 1790) and his Religion within the Bound-
aries of Mere Reason (published in 1793). The former book was Kant’s
reason for cutting off his reviews of Herder’s history,18 and it includes
Kant’s most systematic description of the final and ultimate ends of all his-
tory. Thus it provides a conceptual framework within which Kant’s more
explicitly historical writings can be situated. The latter includes Kant’s
Moral Influence on Others 153

discussion of radical evil and the ethical community that one promotes
in one’s struggle against that evil. In the context of this discussion, Kant
includes a brief account of the “historical representation of the gradual
establishment of the dominion of the good principle on earth” (6:124).
About the time that Kant published the Religion (in 1792–3), he also
wrote a pair of essays that show further development in his reflection
on human history. In September 1793, just after the publication of the
Religion, he published in the Berlinische Monatsschrift the essay “On the
Common Saying: ‘This May be True in Theory, but It Does not Apply in
Practice,’” to which I refer simply as “Theory and Practice.” This essay
includes several passages that seem to support those who ascribe to Kant
the view that human history is a history of genuine moral progress. The
following year (1794) Kant published “The End of All Things,” again in
the Berlinische Monatsschrift, an essay that offers a philosophical interpre-
tation of the final end of all things, including human history. In 1795,
Kant published On Perpetual Peace, in which he outlines the role of war in
promoting progress toward peace and political right in human history.
Kant’s last major published work, The Conflict of the Faculties, published
in 1798, includes a major section devoted to “A Renewed Attempt to An-
swer the Question: Is the Human Race Continually Improving?” This essay
marks Kant’s most advanced reflections on the status of moral progress
in history. The same year Kant published The Conflict of the Faculties, he
published his Anthropology, which has its own brief discussion of historical
progress, a discussion that draws from Kant’s lectures on anthropology.

B. The Scope and Limitations of Moral Progress in History

Kant’s Conflict of the Faculties gives his clearest statement of the scope of
history. Kant asks whether “the human race is . . . continually regressing . . .
continually progressing . . . or at a permanent standstill . . . at its present level
of moral attainment” (7:81). He argues for progress, on the grounds of
“an occurrence in our own times which proves this moral tendency of
the human race” (7:84). The occurrence is the French Revolution, and
it shows a moral tendency in the human race because of the enthusiastic
support of the revolution by spectators.
Although Kant describes this progress in terms of “moral attainment,”
there are two important respects in which Kant clearly does not think
that human history reflects genuine moral progress.19 First, while there
is a “moral character” (7:85) that is revealed in people’s reaction to
the French Revolution, this character is a fixed feature of the human
species; one can “credit human beings with . . . a limited will of innate
154 The Solution

and unvarying goodness” (7:84, emphasis added). Second, the progress in

human history is progress in good deeds and in more just political struc-
tures, such as a “national constitution [that is] just and morally good”
(7:85). Kant clearly says,

The profit that will accrue to the human race as it works its way forward will
not be an ever increasing quantity of morality in its attitudes. Instead, the legality
of its attitudes will produce an increasing number of actions governed by duty,
whatever the particular motive behind these actions may be, . . . i.e. from the
external phenomena of man’s moral nature. (7:91)

The empirical history of humans will involve progress, but history gives
no reason to believe that this progress affects the free wills of human
beings. Thus this sort of “moral” progress fits well with the account of
interpersonal moral influence developed in this chapter.

Although not primarily focused on history, the Critique of Judgment fur-

ther clarifies this distinction between genuine moral change and the sorts
of progress that can be studied in an empirical history.20 In particular,
this Critique outlines two “purposes” of nature. One of these, the “final
purpose,” is clearly moral. Kant says, “A final purpose is a purpose that
requires no other purpose as a condition of its possibility” (5:434), and
“only in man, and even in him only as moral subject,” do we find such a
final purpose (5:435). By contrast, the “ultimate purpose” is “the purpose
by reference to which all other natural things constitute a system of pur-
poses” (5:429), that is, the end for the sake of which the rest of nature can
make sense as a means. Humanity is the ultimate purpose of creation as
well, but “this [ultimate] purpose must . . . be man’s aptitude and skill for
various purposes for which he can use nature (outside or within him) . . .,
his culture” (5:429–30). And although human beings constitute both the
ultimate and the final purposes of nature, “we can even prove a priori that
what might perhaps be an ultimate purpose for nature can still, insofar as it
is a natural thing, never be a final purpose” (5:426). For Kant, it is crucial
to distinguish moral goodness from culture in discussions of progress.
In the Critique of Judgment, however, the final purpose is the proper
subject of moral philosophy rather than of empirical history. History is
the study of the ultimate purpose of nature, the culture of humanity. Kant
defines culture as “man’s aptitude in general for setting himself purposes,
and for using nature . . . as a means [for achieving them]” (5:431). This
definition gives rise to a distinction between two types of culture. The
culture of discipline “consists in liberation of the will from the despotism
Moral Influence on Others 155

of desires” so that one can set purposes for oneself (5:432). The culture
of skill cultivates humanity’s abilities to use nature as a means (5:431–
2). Kant’s account of culture as the ultimate end of humanity includes
brief versions of several of Kant’s more specifically historical writings. For
example, Kant claims,

Though war is an unintentional human endeavor (incited by our unbridled pas-

sions), yet it is also a deeply hidden and perhaps intentional endeavor of the
supreme wisdom [Nature], if not to establish, then at least to prepare the way
for lawfulness along with the freedom of states, and thereby for a unified system
of them with a moral basis. . . . [Moreover,] war is one more incentive for us to
develop to the utmost all the talents that serve culture. (5:433)

This claim that war is a means for cultural progress, and in particular
for progress toward a union of states, is a theme throughout Kant’s writ-
ings on history (cf. “Idea” [8:24–6], “Conjectural Beginnings” [8:121],
“Theory and Practice” [8:311], and Perpetual Peace). This connection be-
tween the discussion of culture in the Critique of Judgment and Kant’s more
specifically historical essays suggests that these essays focus on history as
the study of the ultimate and empirical, rather than the final and moral,
purpose of human beings.

But Kant’s historical essays also suggest that culture is not the only
end of human progress.21 We have already seen that the Conflict of the
Faculties focuses more on political progress and increasing good deeds
than on cultural progress. In the Critique of Judgment, political progress
is “the formal condition under which nature can alone achieve this fi-
nal aim [of culture]” (5:432). Perpetual Peace (see 8:365–6) and “The-
ory and Practice” (8:310–13) also emphasize the importance of polit-
ical progress. In Perpetual Peace, political progress is in the service of
peace, rather than the development of culture itself, and in “Theory
and Practice,” political progress is the ultimate end that “the very na-
ture of things” will establish despite us (8:310–11). Thus these later
writings are more interested in progress toward morally good ends –
peace or justice – than is the Critique of Judgment. But this progress is still
far from the final purpose of nature, the cultivation of morally good people.
Some of Kant’s writings on history, however, seem to suggest that gen-
uine moral progress does occur in human history. These seem to go
beyond the limits set by the Conflict of the Faculties. These works seem
to present a history where members of later generations not only do
more good things and live in more just societies but are actually more
156 The Solution

likely to have morally good wills than members of earlier generations. In

the rest of this section, I briefly address four of the most troubling pas-
sages in which Kant seems to suggest that history studies genuine moral

In “Theory and Practice,” Kant writes, “since the human race is con-
tinually progressing in cultural matters (in keeping with its natural
purpose), it is also engaged in progressive improvement in relation
to the moral end of its existence” (8:308–9). Later Kant adds that “various
evidence suggests that in our age, as compared with all previous ages, the
human race has made considerable moral progress” (8:310). Drawing
on this passage, Robert Louden says, “Kant also maintains confidently
in [“Theory and Practice”] that there exists strong empirical evidence
for asserting that humanity is progressing morally” (Louden 2000:147).
Louden is certainly correct that Kant seems to say that not only culture
(the ultimate purpose) but also morality (the final purpose) progresses
throughout human history.23
But Kant’s apparent focus on moral progress is not as straightforward
as it seems. Kant never explicitly identifies the “moral end” toward which
humanity progresses. And while it is natural to read this as a reference
to the genuine moral improvement of individuals, the essay itself sug-
gests that Kant has something else in mind. Shortly after mentioning
this moral end, Kant associates it with Mendelssohn’s goal of promoting
“the enlightenment and the welfare of the nation to which he belonged”
(8:309). And the final paragraph of the essay offers the real key to in-
terpreting the sense in which the human species progresses “morally.”
There Kant explains that political right, not moral virtue, is the “moral
end” of the human species. As he says,

For my own part, I put my trust in the theory of what the relationships between
men and states ought to be according to the principle of right. It recommends to us
earthly gods the maxim that we should proceed in our disputes in such a way that
a universal federal state may be inaugurated, so that we should therefore assume
that it is possible (in praxi). I likewise rely (in subsidium) upon the very nature of
things to force men to do what they do not willingly choose. (8:313)

The principle of right “recommends” – that is, imposes on us as a duty –

that we promote political right in our interactions with one another, and
the historical account in “Theory and Practice” encourages our belief that
this “is possible.” In the Metaphysics of Morals, Kant reiterates that although
one does not have any duty to promote fundamental moral change in
Moral Influence on Others 157

others, one does have a duty to promote political right (see 6:307–13).
And in the Conflict of the Faculties as well, Kant uses language of moral
progress to refer to a political change. Thus “Theory and Practice” fits
well into Kant’s overall conception of historical progress. Politics as well
as culture progresses throughout history. But none of this requires any
deep interpersonal moral influence.

“The End of All Things,” like “Theory and Practice,” makes reference
to “the development of morality” (8:332) and “the moral disposition of
humanity” (8:332) and explicitly refers to the conformity of rational be-
ings to the “final end of their existence” (8:331), which the third Critique
describes as moral rather than merely cultural or political. What is more,
Kant connects this moral disposition to the rewards and punishments of
the last day. If his conception of the highest good and the justice of God
is to make any sense, “The End of All Things” has to be about the moral
disposition of humanity in the deepest sense. The essay cannot be limited
to mere political right, nor even merely to good deeds, but must touch
on the ultimate moral status of human beings. Thus one might think this
essay refers to moral progress in history.24
However, unlike Kant’s other historical essays, “The End of All Things”
is not empirical history. The “end” is not a final moment in time, nor even
an ultimate goal that the species asymptotically approaches. The end of
all things is the end of all temporality, the “transition from time into
eternity,” the thought of a “duratio Noumenon wholly incompatible with
time” (8:327). This end is the thought of human beings “as supersensible,
and consequently as not standing under conditions of time; thus that
duration and its state will be capable of no determination of its nature
other than a moral one” (8:327–8). Immediately after mentioning the
“development of morality,” Kant insists that “here we have to do (or are
playing) merely with ideas” (8:332). He goes further and claims that
there can really be “no alteration” with respect to this moral end (8:333).
Rather, one must

think of the final end as an alteration, proceeding to infinity (in time) in a constant
progression, in which the disposition (which is . . . something supersensible . . .)
remains the same and is persisting. . . . [We must act as if] its disposition (the homo
Noumenon, “whose change takes place in heaven”) were not subject to temporal
change at all. (8:334)

This essay, in other words, discusses the ultimate moral “end” of all things
as the unchanging timeless moral disposition that is the ultimate ground
158 The Solution

of a person’s appearance in the world. This analysis touches deeply on

morality, to be sure, but adds nothing to Kant’s philosophy of empirical
history and does not commit Kant to any kind of historical influence on
one’s moral status.

Like “Theory and Practice” and “The End of All Things,” Kant’s essay
“Conjectures on the Beginning of Human History” seems to suggest that
human beings undergo moral progress.25 Kant says, “I wish merely to
consider the development of human behavior from the ethical point of
view” (8:111) and refers to “man’s development as a moral being” (8:113).
And the final step in Kant’s outline of the development of human beings
is clearly a moral one:

The fourth and last step which reason took, thereby raising man completely above
animal society, was his (albeit obscure) realization that he is the true end of nature,
and that nothing that lives on earth can compete with him in this respect. . . .
This notion implies . . . an awareness of the following distinction: man should not
address other human beings in the same way as animals, but should regard them
as having an equal share in the gifts of nature. This was a distant preparation
for those restrictions which reason would in the future impose on man’s will in
relation to his fellows. (8:114)

This recognition that human beings are ends in themselves deserving of

respect is a deeply moral realization. The “beginning of human history,”
as Kant outlines it in this essay, describes moral progress.
But it is important to note here that this history is a history of the
beginning of human history, and Kant does not commit himself to the
view that the moral progress which accounts for the birth of humans as
moral beings will continue throughout the history of civilization. Kant
explicitly says that “a history of the first development of freedom from its
origins . . . is something quite different from a history of its subsequent
course” (8:109). And although he discusses the development of human
beings after their early beginnings, in those discussions he does not sug-
gest any fundamental moral progress but only progress toward culture,
art, a “civil constitution” (8:119), and perpetual peace (8:121). With re-
spect to properly moral progress, he says that “the human race was . . .
deflected from . . . cultivation of its capacities for goodness,” but we can
remedy these evils by “improving ourselves” (8:120–1).
Even when Kant discusses “moral” progress at the beginning of human
history, this progress does not involve increasing moral goodness. Rather,
Kant outlines how a pre-moral being came to be morally accountable at
all. The progress is from nonmoral to morally responsible rather than from
Moral Influence on Others 159

morally worse to morally better. Thus Kant explains that this progress is a
“release from the womb of nature” (8:114), not from the womb of wicked-
ness, an expulsion “from a garden . . . into . . . unknown evils” (ibid.), not
a release from evil into moral virtue. Prior to this moral progress, “there
were no commandments or prohibitions, so that violations of these were
also impossible” (8:115).
Thus although “Conjectures on the Beginning of Human History”
clearly outlines moral progress of a sort, the essay does not describe a
moral progress that poses any problem for Kant’s account of freedom.
There is no indication that empirical influences can affect one’s moral
status, though they can affect whether one has a moral status. Incapable of
doing good or evil, proto–human beings were happy, but without moral-
ity. After the progress into moral responsibility, human beings became
capable of good and evil. But there is no indication in this essay that the
history of this freedom is a history of progress from evil to good. All we
get here is a conjecture on the beginning of human history.26

One final area in which Kant seems to discuss moral progress is in his An-
thropology. There Kant suggests that historical progress has three parts:
“man is destined . . . to live in a society of other people, and in this
society he has to cultivate, civilize, and moralize himself” (7:324–5, cf.
“Idea” 8:26, and 9:450, 15:641, and 25:847, 1198, 1420). The reference
here to “moralization [moralisieren],” as something distinct from culti-
vation and civilization, suggests that the development of culture is not
what Kant has in mind. Robert Louden, in fact, argues that moralization
involves “the deep structure of an agent’s moral personality” (Louden
Still, Louden rightly points out that with respect to the details of moral-
ization, Kant is “exasperatingly brief” (Louden 2000:43).28 One possible
interpretation of the moralization of the human species is that over time,
more and more people come to be morally good. Unless this is the only
reasonable interpretation, however, it should be rejected. It conflicts with
Kant’s claims in the Conflict of the Faculties about what can be hoped for
from history, and it suggests a sort of interpersonal moral influence that
would be incompatible with Kant’s account of freedom.
Fortunately, there are three other plausible ways to interpret
“moralization” that do not require positing that history marks moral
change in the free wills of human beings. Moralization could be inter-
preted to refer to the just political structures to which Kant refers in the
Conflict of the Faculties and “Theory and Practice.” Kant says that we can
160 The Solution

recognize in the human species “a tendency . . . to bring about sometime

in the future the development of the good out of the evil through its own
efforts” (7:329). When Kant describes this development, which “can be
expected with a moral certainty” (7:329), he says,

As their [human beings’] culture grows, they become more and more aware of the
wrongs which they themselves selfishly inflict upon each other. And since they
see no other remedy for them than subordinating the private interest (of the
individual) to the public interest (encompassing all), they subject themselves,
reluctantly though, to the discipline (of civil restraint), which they only obey,
however, by following laws which they themselves have given. They feel themselves
ennobled by the knowledge that they belong to a species which fits the vocation
of man as reason represents it to him in the ideal. (7:329–30)

Here Kant makes clear both why this progress is “moral,” and the lim-
its of its influence. On the one hand, by becoming subject to just
and representative political structures (“civil restraints”), human be-
ings conform to a structure that reason prescribes as a moral ideal.
On the other hand, because this conformity is possible even when peo-
ple subject themselves “reluctantly [ungern],” this does not imply any
fundamental moral change. Kant expands upon this political progress
later, arguing that people “feel destined by nature to develop, through
mutual compulsion and laws written by them, into a cosmopolitan society”
(7:331), and the Anthropology ends with the hope that humanity can “ad-
vance constantly from evil to good . . . through the progressive organiza-
tion of the citizens of the earth within and toward the species as a system
which is united by cosmopolitical bonds” (7:333).
However, in the Anthropology and “Idea,” Kant seems to think that mor-
alization goes beyond mere political progress. The cosmopolitical bonds
to which Kant summons his readers at the end of the Anthropology seem to
be the means for a more significant moral progress. Likewise in the “Idea”
Kant says that “the human race will remain in this condition [civilized
and cultured but not moralized] until it has worked itself out of the
chaotic state of political relations” (8:26). Of course, even the Conflict of
the Faculties goes beyond merely political progress. There Kant claims that
“an increasing number of actions governed by duty” will be an effect of
political progress (7:91). Moralization could refer to this increase of good
The discussion of moralization in the Anthropology might even be read
to suggest that moralization goes beyond mere progress in good deeds
and includes the kind of resistance to evil discussed in the Religion.29 Kant
introduces the notion of moralization after a discussion of “the moral
Moral Influence on Others 161

predisposition,” in which he very briefly recapitulates the first two books

of the Religion, explaining that while every human being “sees himself . . .
subject to a moral law . . ., experience also shows that in man there is a
propensity to . . . evil” (7:324). Moralization is thus set as a task: “man
[must] moralize himself” (7:324). Kant goes on to explain moralization
as the process whereby man will “make himself worthy of humanity by
actively struggling with the hindrances that cling to him from the crudity
of his nature” (7:325).30 This reference to moralization as a struggle
against obstacles echoes the language of struggle in the Religion, where
this struggle is manifested in the promotion of an ethical community.
And in the Religion too, this ethical community is distinguished from
mere cultural or political progress (6:94–7), and even from progress
in terms of good deeds. Moralization could refer, then, to that explicit
commitment to and encouragement of virtue which characterizes the
ethical community. But as we have seen, this ethical community need not
refer to any fundamental moral change.
Thus moralization can plausibly be read as progress toward more just
political structures, more good deeds, or an ethical community. None of
these implies fundamental moral change. Likewise Kant’s other historical
writings all refer to one of these sorts of progress (in the case of “Idea,”
“Theory and Practice,” Perpetual Peace, and the Conflict of the Faculties), or
to the historical development of moral responsibility itself (in the case of
“Conjectures”), or to morality itself, rather than history strictly speaking
(in the case of “The End of all Things”). Some have read other passages
in Kant’s historical writings as implying the kind of deep moral progress
that would be incompatible with this view. Ultimately, these can all be
dealt with in ways that are similar to my treatments of difficult passages in
this section.31 Kant’s writings on history, like his more general accounts of
interpersonal moral influence in the Metaphysics of Morals and the Religion,
reflect a commitment to empirical helps and hindrances that are morally
significant within the bounds of his theory of freedom.

4. Conclusion
Kant’s account of interpersonal influence allows one to salvage many
common-sense views of the importance of such influence. One ought to
act for the moral improvement of others because such efforts can have
positive results. One can increase the extent to which others do good
things, which will often provide both them and others with better lives.
Moreover, we are all part of communities such that each can profoundly
162 The Solution

affect the way others behave, and this provides a further reason to cultivate
virtue not only in oneself but also in others.
Still, on Kant’s account there is no practical or theoretical basis for
believing that one person can determine, change, or even affect the fun-
damental moral status of another. Some might wish for such a possibility,
but it is not clear that such a demand is made by the common-sense
conviction that we should try to improve others morally, and that we can
improve them. Kant allows that there is a sense, and a rather important
sense, in which we can and should work toward the moral betterment of
others. All people have a responsibility to promote morality in others in
the sense that all people should promote the performance of good deeds
and an explicit commitment to virtue.
Kant can even go part of the way toward allowing for even more ro-
bust moral influence. One can rationally hope that those who perform
good deeds and express commitments to morality are actually morally
good, and this can give rise to a direct interest in the explicit goodness of
others. And Kant may even have room for an inscrutable and practically
important possibility of mutual influence at the deepest level. Thus there
is room for serious social cooperation and corporate moral responsibil-
ity, and for the cultivation of moral community and mutual encourage-
ment in virtue. And all of this is possible within the bounds of Kant’s ac-
count of transcendental freedom as a condition of the possibility of moral

Incorporating Moral Anthropology

and Defending Kantian Moral Philosophy

Near the end of Kant’s Impure Ethics, Robert Louden asks whether Kant
is “saved by impurity” (Louden 2000:167). That is, he asks whether the
anthropological and other “impure” elements of Kant’s moral philoso-
phy save it from the objections to which it has been subject over the past
two centuries. In this book, my task has been more limited. I am con-
vinced that Kant’s moral anthropology strengthens his ethical theory.
Kant rightly recognized the possibility of effecting changes in human
agents, especially oneself. And he saw that these changes can be morally
beneficial when they are directed to the cultivation of a good and strong
character. One’s moral status is not merely a matter of the choices that
one makes about the present but involves the sort of person that one is.
And becoming a certain sort of person involves cultivating and promot-
ing certain dispositions, emotional responses, interactions with others,
habits of thought and action, and social structures. These aspects of hu-
man life are clearly important parts of morality, and Kant’s moral theory
is stronger for having recognized them.
But Kant’s moral theory is stronger for having recognized the impor-
tance of anthropology only if the theory remains consistent. The impor-
tance of freedom – both practical autonomy and the transcendental free-
dom which underlies that autonomy – is still the core insight of Kantian
moral theory. Many moral philosophers have recognized the importance
of anthropological considerations. What makes Kant both distinctive and
attractive remains his emphasis on freedom from empirical influence in
determining the content of the moral law and acting on the basis of
it. Without these Kantian emphases, one may as well turn to Aristotle,
Aquinas, Smith, Nietzsche, or any of a dozen other moral theorists. Unless
164 Epilogue

Kant’s account of moral anthropology is consistent with his account of

freedom, the benefits for ethics of those anthropological insights cannot
rightly be counted as advantages for Kant’s ethics. Any defense of Kant’s
moral theory that depends on the value of Kant’s moral anthropology
must show how that moral anthropology is consistent with Kant’s account
of freedom. Showing this consistency has been the task of this book.
This book thus serves as a prolegomenon to anthropological defenses
of Kant’s moral theory. Schleiermacher objected to any such defense on
the grounds that Kant’s anthropology and his moral philosophy cannot
be made consistent. Although I showed that there are good prima facie
reasons for this objection, I ultimately argued that Kant can respond to
it. By showing that moral anthropology is consistent with Kant’s account
of freedom, I thus legitimate attempts to defend Kant by drawing on his
moral anthropology. The concrete examples of moral anthropology in
Chapter 3 can serve as initial responses to those who argue that Kant fails
to take into account the importance of politeness, the complexity of hu-
man emotions, or the role of character in moral life. There is much more
work to be done fleshing out a robust Kantian anthropology, however.
This book provides only a framework within which such work is possible.
At the same time, I have shown that there are limits to how far Kant can
go in allowing for morally significant empirical influences. Moreover, the
framework within which these influences must fit involves a conception
of the role of such influences that is not shared by all those who make
use of anthropology in moral reflection. In contrast to Robert Louden’s
account of Kant, for example, I insist that no empirical influence can
“cut through the surface causal network” (Louden 2000:59) to effect a
change of one’s moral status. This limitation does not imply that acts in
the world are irrelevant to one’s moral status. As we have seen, one can
effect changes in future choices through empirical influences, and effect-
ing these sorts of changes is an important part of expressing a good will.
But the empirical influences do not cause one to have a good will. At a
practical level, this fact does not give one any reason to refrain from pro-
moting good future choices through empirical influences. But it changes
the philosophical understanding of the relationship between these influ-
ences and one’s moral status. One concrete implication of this different
understanding, explained in detail in Chapter 6, is that one cannot seek
to change the moral status of another through one’s actions in the world.

This book leaves some unresolved issues. I have said very little about the
accuracy of Kant’s particular anthropological observations. As I noted
Defending Kantian Moral Philosophy 165

in Chapter 2, Kant’s anthropology is fallible. He recognized mistakes

over the years, and he made many mistakes that he did not recognize.
As investigations in the human sciences have progressed over the past
two hundred years, many of Kant’s claims have been called into question.
Even in Kant’s own time, many of his anthropological claims were contro-
versial. The relationship between politeness and virtue was a particularly
hot topic in the eighteenth century, and claims about race and gender,
the emotions, and the role of churches and political structures were and
are still widely disputed. Some of these issues can now be thought through
more carefully by philosophers. Many are being discussed among those
involved in fields such as psychology, sociology, anthropology, and cog-
nitive science. Today we have access to more data than Kant had about
human behavior, motivation, and mutual influence. This new data will
allow us to revise the details of Kant’s anthropology. But for Kant, no
amount of data can determine the relationship between empirical data
and the freedom that grounds moral responsibility. No data can show the
ultimate significance of interpersonal influence. And no data can spec-
ify the proper structure of a moral theory. These philosophical insights
about the limits of empirical inquiry remain a Kantian legacy for today.
But the details of Kant’s anthropology may be in need of serious revision.

Moreover, there are remaining philosophical problems with the ac-

count offered here. The limitation on interpersonal moral influence
might be unsatisfying for many, even after my attempts in Chapter 6
to make this limitation more palatable. The role of grace in Kant’s reso-
lution to the problem of radical evil still might seem to raise significant
problems, and these may infect my integration of anthropology and free-
dom. Some might still have metaphysical worries about Kant’s apparent
dualism. However one understands the relationship between nature and
freedom, the dichotomy can seem troubling, and I have said little in this
book to ease purely metaphysical troubles.
Even if these worries are solved, Kant’s argument for transcendental
freedom can still be challenged on grounds that do not relate to an-
thropology per se. Even if Kant’s account of transcendental freedom is
consistent with his moral anthropology, that in itself does not provide
any reason for thinking that one should accept transcendental freedom.
Kant’s argument for such freedom, at least as I recounted it in Chapter 1,
depends on a “fact of reason” about moral responsibility. Many might
reasonably question whether there is such a “fact,” or whether such a fact
involves any commitment to something as strong as Kant’s transcendental
166 Epilogue

freedom.1 What I have shown in this book is that two aspects of Kant’s
moral theory – his account of freedom and his moral anthropology – are
consistent. But even if Kant’s moral theory is consistent, it might still be

Aristotelians, Humeans, Anti-theorists, Utilitarians, and other moral the-

orists need not hang up their hats and go back to Kant. But if the reason
they reject Kant is his insufficient attention to moral anthropology, they
have reason to give him another look. I have shown that one common
objection to Kant is not decisive. Kant can reconcile his strong conception
of freedom with a robust moral anthropology. This shifts the debate over
Kant’s moral theory back to issues that Kant considered more central than
anthropology, issues regarding the nature of moral obligation and the im-
plications of moral obligation for our conception of human freedom.

1. See the bibliography to Thomas Sturm’s “Kant und die Wissenschaft des
Menschen,” unpublished manuscript.
2. Schleiermacher was not the only one to question Kant’s success in combining
systematicity and popularity. One of Kant’s most important early interpreters,
Karl Reinhold, sought to improve on Kant by providing a version of transcen-
dental idealism that could be both popular and systematic. For more on the
ways in which Reinhold misinterpreted Kant and thereby inaugurated the
ambitious programs of later German Idealists, see Ameriks 2000 and Franks
3. Ultimately, Schleiermacher does not think that any particularly strong con-
ception of freedom is necessary to make sense of normativity. In fact, his
critique of Kant’s Anthropology is part of a larger project of replacing Kant’s
notion of transcendental freedom with a soft determinism about human free-
dom. Still, Schleiermacher recognizes the need to give an account of the nor-
mative if one seeks to be a determinist about human choices. Much of On
Freedom (Schleiermacher 1992) is designed to give just such an account, for
both epistemological and moral normativity.
4. Even if Kant does throw out one or more aspects of his moral anthro-
pology, the dilemma arises in other areas. The dilemma permeates Kant’s
practical philosophy whenever he seeks to articulate means for promot-
ing advancement toward greater autonomy. For example, in “What Is
Enlightenment?” Kant insists that “Enlightenment is man’s emergence from his
self-incurred immaturity” (8:35). But the essay also raises the question of how
man is to emerge from self-incurred immaturity. “Laziness and cowardice” lead
people to “gladly remain immature for life” (8:35). Kant goes on to shift
responsibility for promoting enlightenment from the individuals whose im-
maturity is self -incurred to those “guardians who have kindly taken upon
them the work of supervision [and who] see to it that by far the largest part of
mankind . . . should consider the step forward to maturity not only as difficult

168 Notes to Pages 2–7

but as highly dangerous” (8:35). An essay that begins as a call to individuals

to take responsibility for immaturity that is self-incurred turns out to
be a call to leaders to promote maturity in others. It is apparently up to the
leaders to help their subjects develop the courage to think for themselves,
and subjects can apparently take responsibility for themselves only with the
right kinds of people responsible for them (8:41). This case is not as prob-
lematic as moral anthropology in general. There is no strict inconsistency
between the ideal of people thinking for themselves and the fact that they
will realize this ideal only under certain leaders. The means toward enlight-
enment are not necessary parts of enlightenment. While members of an
enlightened public can think freely regardless of their leader, for a pub-
lic to become enlightened, a certain sort of leader may be necessary. Still,
“What Is Enlightenment?” exposes two aspects of Kant’s notion of enlight-
enment that, though consistent, do not sit comfortably with each other. And
in the case of moral anthropology, the conflict appears to be even more
5. See for instance Rehberg’s early criticism (in Schulz 1975) and Hegel’s
famous criticisms in the Phenomenology.
6. The best article for describing both objections to Kant and Kantian re-
sponses is Wood 1984. See, too, Allison 1983, 1990, and 1996; Beck 1987;
and Korsgaard 1996a.
7. Schleiermacher does pose his dilemma in general terms. He suggests that
there is a problem with helps or hindrances to any “mental faculties
[Gemüthsvermögen]” (Schleiermacher 1984: v.1, p. 366). In fact, however,
Kant is committed to transcendental freedom only for human choices, and
then only for those choices for which one can be held morally responsi-
ble. Thus the real dilemma for Kant is raised not by his anthropology in
general but by his specifically moral anthropology. Schleiermacher’s descrip-
tion of the dilemma as a conflict between claims about nature and choice
shows his primary focus on moral helps and hindrances. Schleiermacher’s
On Freedom (Schleiermacher 1992), written shortly before the review of Kant’s
Anthropology, makes even clearer that moral issues are foremost in Schleier-
macher’s mind at this time.
8. This is discussed in detail in Chapter 3.
9. Strictly speaking, one might still promote polite society for the sake of the
happiness of others, even if doing so does not affect anyone’s moral status,
but this is not the particular reason that Kant gives for polite society. Instead,
he argues that politeness is a duty because the “illusion resembling virtue”
can give rise to genuine virtue.
10. See for example Baron 1995, Chapter 6.
11. For neo-Aristotelian approaches, see MacIntyre 1984 and 1988, Hursthouse
1999, Foot 2001, and Anscombe 1958, just to mention a few. Humeans in-
clude Annette Baier (see Baier 1987, 1991, 1995) and, more recently, Simon
Blackburn (Blackburn 1998). The most prominent Anti-theorist is Bernard
Williams (see especially Williams 1981, 1985, 1995).
12. One must be careful in reading too much of virtue ethics into Kant, however.
Kant often describes very different aspects of moral life using terms that are
Notes to Pages 7–14 169

similar to those in virtue theories. When Kant discusses character, disposition,

virtue, or emotion, he often does not mean by these terms what an
Aristotelian or Humean would mean. For some specific examples, see the
discussions of passions, affects, and character in Chapter 3. See too Munzel

Chapter 1
1. If one prefers more explicitly two-perspective language, one should read this
to say that the categories and forms of intuition cannot be used in order to
have knowledge of how things are when considered from a practical perspec-
2. Actually, Kant’s own language here is significantly different. The antinomy
is between a thesis that “causality in accordance with laws of nature is not
the only causality from which the appearances of the world can one and all
be described. To explain these appearances it is necessary to assume that
there is another causality, that of freedom” and an antithesis that claims that
“there is no freedom; everything in the world takes place solely in accordance
with laws of nature” (A444–5/B472–3). The tension between these claims is
different from the tension between the claims that there is and that there is
not freedom, but it is not necessary to discuss these differences for the purpose
of the present discussion. This antinomy has been discussed in detail in the
secondary literature. See for example Strawson 1966, Allison 1983, Pippin
1982, and Kemp Smith 1992.
3. There is some ambiguity in the third antinomy about the sort of priority the
cause must have. Understanding demands a cause that is prior in some sense,
and the only sense in which the understanding can describe that priority
is temporal. Kant’s resolution of the antinomy ultimately depends on the
possibility (as far as we can tell) of a nontemporal cause. Such a cause meets
the demand for a cause, but not the limitations imposed by the understanding
on the sort of cause that is available. In that sense, reason goes beyond the
4. I intend this formulation to be neutral between two-object and two-aspect the-
ories of the relationship between things and their appearances. A two-object
theorist should read this to refer to two distinct “things” – the thing in itself
and the appearance – that are in some kind of causal interaction. The two-
perspective theorist should just read this to refer to a single thing insofar as it is
considered either “in itself” – that is, from some non-empirical perspective –
or as an object of experience. For further discussion of the distinction be-
tween two-object theories and two-aspect theories, see Ameriks 1982, Pippin
1982 (especially pp. 194–215), Prauss 1983, Allison 1983, and Nelkin 2000.
The best defense of a two-object view has been offered by Aquila (especially
in Aquila 1979). The best defense of the two-aspect theory is presented by
Allison (in Allison 1983 and Allison 1996). The two-perspective view has been
applied to issues of freedom in moral philosophy in Allison 1999 and 1996
and in Korsgaard (especially 1996a). For an excellent overview of these ac-
counts, see Nelkin 2000.
170 Notes to Pages 14–19

5. See section 4 of this chapter for more on the inscrutability of freedom and
the analogical use of the categories.
6. I take the argument of the second Critique to reflect Kant’s most considered
views about the justification of freedom. For an excellent discussion of the
differences between the argument in the Groundwork and the argument in the
second Critique, see Ameriks 1981. The clearest articulation of the argument
I paraphrase here is given in the second Critique at 5:28–30.
7. It is easy to mistake this for the claim, which Kant is not making, that context
is not relevant to what is morally required. Context is often a relevant feature
of maxims that are tested for moral permissibility. But the context itself does
not determine the way in which the maxims themselves will be evaluated.
Rather, it is insofar as the content or context of a maxim affects its form that
it becomes morally relevant.
8. Note that this does not mean that I am totally free of any influence. I am
free of any empirical influence, which means that my desires and beliefs do
not determine the way I act. But to act independent of these influences is to
act on the basis of a moral law that provides a reason for action. I can be free
of empirical causes, for Kant, but I always act for some reason.
9. This does not mean that one cannot give causal explanations of one’s actions.
But these explanations are always only conditioned. The ultimate, uncondi-
tioned explanation of any act is the free agent, the self-in-itself.
10. Possibility must be understood loosely here. In the first Critique, Kant claims,
“the fact that a concept does not contradict itself by no means proves the
possibility of its object” (A596/B624). He adds in a footnote:

A concept is always possible if it is not contradictory. . . . But it may none the less
be an empty concept, unless the objective reality of the synthesis through which the
concept is generated has been specifically proved; and such proof, as we have shown
above, rests on principles of possible experience. . . . This is a warning against arguing
directly from the logical possibility of concepts to the real possibility of things. (A596/

“Real possibility” is possibility as an object of possible experience, and freedom is

not possible in that sense. But freedom is not merely logically consistent. It is
“possible” in the sense that it is consistent with the world of experience, but
only as something not included in that world. That is, it “does not conflict
with” nature (A558/B586).
11. In fact, it is not clear that Korsgaard actually thinks there is any asymmetry.
In the Sources of Normativity (Korsgaard 1996c:95–7), she seems to endorse
a thoroughly symmetrical account.
12. It is important to recognize, however, that this is not a methodological prior-
ity. Theoretical reason is methodologically prior to practical reason. The con-
viction that people are free, a conviction established on the basis of practical
reason, would be illegitimate without the possibility opened for freedom by
theoretical reason. Thus theoretical reason holds a methodological priority
over practical reason, such that the first Critique is a necessary preparation
for the second. For Kant’s insistence on this sort of priority of the theoretical,
see the first Critique, Bxxix.
Notes to Pages 21–23 171

13. Strictly speaking, theoretical reason holds open the logical consistency of
and need for freedom as ground of experience. See footnote 10.
14. One passage in the first Critique may seem to raise a problem for this account.
In his resolution of the Third Antinomy, Kant seems to suggest that the
practical perspective – which in this context is the perspective from which one
discusses the “intelligible ground” of appearances – need not be considered
at all in empirical investigation. He says,
This intelligible ground does not have to be considered in empirical enquiries;. . . . We
have to take their strictly empirical character as the supreme ground of explanation,
leaving entirely out of account their intelligible character (that is, the transcendental
cause of their empirical character) as being completely unknown, save in so far as the
empirical serves for its sensible sign. (A545–6/B573–4)

This passage might be read to imply that from the standpoint of theoretical
understanding or empirical inquiry the status of the intelligible does not
even arise. When Kant says that it “does not have to be considered” and
should be left “entirely out of account,” he lends himself to this interpreta-
tion. But in the final part of the passage, Kant makes clear that what he in fact
opposes is not the consideration of the relationship between the practical
and theoretical perspectives but the use of insights from the practical per-
spective in empirical explanation. One can affirm, from a purely theoretical
standpoint, that if there is a practical perspective, then it provides a more
ultimate explanation than the empirical. But one cannot know that there
is such a perspective, and one certainly cannot know anything about such a
perspective, on the basis of theoretical reason. The practical perspective is
useless for empirical explanation, even if theoretical reason can and even
needs to establish that if there is such a perspective, it is prior to – in the sense
of being the ultimate ground of – empirical explanation.
15. As mentioned earlier (see footnote 12), there is still a sense in which the the-
oretical perspective has methodological priority. The priority of the practical
perspective is a priority of the moral ground over the empirically described
ground as a proper explanation of the ultimate basis of an action.
16. This “realm of experience” is not limited to the realm of outer experience.
One may evaluate desires, intentions, and plans that are experienced only
by oneself, and only through introspection. But even this introspection gets
merely at one’s appearance in experience, not at the noumenal ground
of those experiences. See Kant’s discussions of self-knowledge in the first
Critique (Bxl–xli, B68, B275f., A362–3).
17. There are some variations on these two ways of interpreting Kant that I have
not considered. Most of these resolve themselves into either a two-object or
a two-perspective interpretation. For example, Karl Ameriks has pointed out
that even if Kant has a two-object view of the relationship between things in
themselves and appearances in general, he may have a two-perspective view
with respect to the self. Thus Ameriks suggests that
at least some of the passages appealed to in defense of the two-perspective view in-
volve the special context of freedom, where it is important (and proper . . .) for Kant to
maintain that there can be an identity between the being who appears in phenomenal
172 Notes to Pages 23–26

and empirically determined action and the one who in a transcendent way is an abso-
lute source of this action. Such a view does not require that in general all phenomena
and associated noumena are to be said to involve the very same objects. (Ameriks

Such a view poses no special problem for the arguments presented here, be-
cause my arguments make a point only about the special case of freedom. For
someone who holds that in general there is a distinction between phenome-
nal objects and their noumenal counterparts, but in the case of freedom the
difference is merely one of perspective, my arguments about the asymmetry
between perspectives should be sufficient.
Another option, suggested by Ameriks in conversation and by Kant in
his resolution to the Fourth Antinomy (see especially A561/B589), is that
while the self is a single phenomenal object, its causality is free and therefore
noumenal, but not in any sense an object. This view fits into a metaphysical
middle ground between standard two-object and two-perspective accounts
of the self, but its development is beyond the scope of my discussion here,
and it is likely that the sorts of arguments developed to show asymmetry in
the case of two-object and two-aspect accounts will apply in this case as well.
18. One passage in the Critical Kantian corpus suggests some form of soft deter-
minism. In the Canon in the Critique of Pure Reason, at the end of a discussion
of human freedom through reason, Kant says,

Whether reason is not, in the actions through which it prescribes laws, itself again
determined by other influences, and whether that which, in relation to sensuous in-
fluences, is entitled freedom may not, in relation to higher and more remote operating
causes, be nature again, is a question which in the practical field does not concern
us, since we are demanding of reason nothing but the rule of conduct; it is a merely
speculative question, which we can leave aside so long as we are considering what
ought or ought not to be done. (A803/B831)

This passage has inspired a flurry of creative interpretations. (See for exam-
ple Beck 1960:190, Carnois 1987:29, Kemp Smith 1992:569–70, and Allison
1990:55–7 for some prominent examples.) The tricky phrase is “in relation to
higher and more remote operating causes,” which makes it sound as though
freedom may not be the ultimate explanation of an action. What is important
for the purposes of this section is that whatever view Kant held at the time
he wrote this passage – and it is not even clear when exactly that was – he
came to strongly reject soft determinism by the time of the second Critique.
19. In the Religion, Kant seems to take a more restrained position with respect
to the possibility of one’s freedom being determined by God at an ultimate
level. See 6:191. I discuss this in Chapters 5 and 6. See, too, Quinn 1990 and
Mariña 1997.
20. I will take up some of these issues in my discussion of the timeless moral
self in Chapter 5. Wood 1984 offers a detailed account of many of the most
important problems that arise from Kant’s position on the free self.
21. I do not mean to imply here that he goes against these limits. Kant’s argu-
ments for the reality of freedom in the first Critique do not contradict his
limitations on theoretical knowledge precisely because his arguments are
Notes to Pages 26–34 173

practical. But as practical arguments, they go beyond speculative reason and

fit awkwardly in the first Critique.
22. In “What Is Orientation in Thinking,” Kant takes the importance of imagi-
nation even further than I suggest here. There he explains,

However exalted we may wish our concepts to be, and however abstract we may make
them in relation to the realm of the sense, they will continue to be associated with
figurative notions. The proper function of these is to make such concepts, which are
not in other respects derived from experience, suitable for use in the experiential world.
For how else could we endow our concepts with sense and significance if we did not
attach them to some intuition (which must ultimately always be an example derived
from some possible experience). (8:133)

23. As we will see in Chapter 5, this is particularly important with respect to the
notion of the “moral revolution” that supposedly takes place for the noumenal
free agent.

Chapter 2
1. Brandt 1997, for example, argues that the anthropology of the lectures and
Anthropology is not moral anthropology, while Stark 1997, Louden 2000, and
others argue that at least part of Kant’s “pragmatic” anthropology is properly
moral. I take up this debate in the next chapter.
2. Robert Louden interprets differently the significance of both the empha-
sis on practical application and the problems Kant raises for anthropology.
Whereas I point out the ways in which these confirm the empirical nature of
anthropology, Louden suggests that they point to something non-empirical
about anthropology. Thus he refers to “Kant’s conviction that accurate ex-
periments and observations of human beings are not possible” (Louden
2000:68) and claims that “the goal of an observation-based science of behav-
ior is not one he [Kant] embraces” (67). If Louden just means that Kant’s
anthropology is not based on observation of human activity in the exter-
nal world but that it also includes observations in inner sense, then his ac-
count and my own are consonant. But insofar as Louden makes the stronger
claim – that something beyond any observation is useful or necessary in
anthropology – Louden’s account varies from my understanding of Kant’s
own assertions about his project.
3. There are, of course, important moral dimensions to Kant’s discussions of
the character of the sexes and the nations. Kant points out moral dangers
and advantages of various national characters, discusses various virtues and
vices for each sex (25:719), and points out the role that women play in
morally refining men. Still, the discussions do not have the systematic place
that the accounts of the character of the individual and of the species do.
Kant says that the character of the individual “shows what man is prepared
to make of himself,” reiterating the distinguishing feature of his moral an-
thropology (7:285, cf. 7:119). By contrast, the character of the sexes is fo-
cused on “Nature’s foresight” (7:303) in the distinction between men and
women and ends with an apology that “I have spent longer on the subject
174 Notes to Pages 34–36

of characterization [of sexes] than may seem proportionate to the other

divisions of anthropology” (7:310).
Kant has two remarks that suggest a similar caution with respect to his
treatment of the character of nations. He says near the beginning of that
discussion that “hereditary maxims . . . are only . . . attempts to classify the
varieties in the natural tendencies of entire peoples more empirically for the
geographer than rationally for the philosopher” (7:312). And near the end
of his discussion he says that “we are speaking here about innate natural char-
acter, which, so to speak, lies in the composition of a person’s blood” (7:319).
Kant’s account of character is a “classification of varieties” and reads much
like his geographical accounts. This suggests, though it does not decisively
show, that Kant considered his account of the character of nations peripheral.
There are several nonsystematic reasons for Kant to include these accounts
of the character of the sexes and of the nations, however. For one, they are
important for those who will travel, or for those involved in politics or the
military, to teach them “what each [nation] has to know about the other, and
how each could use the other to its own advantage” (7:312). Moreover, they
fit Kant’s goal, expressed in his letter to Herz, of including in his anthro-
pology lectures “so many observations . . . that listeners . . . find the lectures
entertaining and never dry” (10:146). As dated as these passages are for us
today, they would no doubt have been among the most entertaining parts of
Kant’s lectures in his day.
4. Erwin Hufnagel (Hufnagel 1988:47–53) does discuss the notion of “transcen-
dental anthropology,” but his focus is on the role that the notion of Anlage
plays in Kant’s philosophy of education. Hufnagel explicitly draws much of his
account from the Religion, where Kant’s most systematic discussion of Anlagen
is found. Although there are some differences between Allison and Hufnagel,
both argue for the a priori status of universal claims on the basis of central
concepts from the Religion rather than more general concerns from Kant’s
5. Others have followed Allison’s example. See Herman 1996:42–3; Anderson-
Gold 1984; Hufnagel 1988; and Munzel 1999:135f., 170, 278.
6. In Chapter 5, I show in more detail why it is crucial that Kant’s claim of radical
evil be both universal and empirical. Here my goal is simply to show that such
judgments are possible.
7. This is discussed in more detail in Chapter 5.
8. Similarly, Kant writes in his Reflexionen that “in order to know whether some-
one has a character, many observations are needed; but to know that a cer-
tain character is not present, only a single observation is required. For, since
character consists in Denkungsart, namely acting from principles, a single ex-
ception is proof that the basis of the action was not a universal maxim”
(RA 1230, see Munzel 1999:86). Munzel has rightly pointed out in conversa-
tion that this brings an ineliminable transcendental component into Kant’s
argument. In particular, one cannot directly perceive any propensity to evil in
human nature. The inference from evil deeds to evil maxims, and even more
so from evil maxims to an evil propensity, is a transcendental argument. There
is a sense in which “radical evil” and the “propensity to evil” are answers to the
Notes to Pages 36–40 175

transcendental question about the conditions of the possibility of evil in

human nature. But to even motivate that question, Kant needs to have an
empirical basis for the claim that there is any evil in human nature that needs
to be explained. For more, see Munzel 1999, Hufnagel 1988, and Anderson-
Gold 2001.
9. “The claim . . . is to be taken as a priori; indeed, as a postulate, it must
be synthetic a priori. Consequently, it requires some sort of deduction or
justification; and since Kant fails to provide one, we must attempt to do so
for him” (Allison 1990:155).
10. Given his excellent analysis of the way in which one must incorporate incen-
tives into one’s maxims, it is surprising that Allison in this discussion describes
the influences of self-love and the moral law in such passive terms. The moral
law is a stronger incentive precisely because it is taken to be stronger – that is,
because it is chosen.
11. Another example of a universal empirical claim is Kant’s theory of matter
in the Metaphysical Foundations of Natural Science. See Walker 1971, in which
Walker argues against those who propose that this theory must be grounded
a priori. For a very different account in the realm of natural science, see
Friedman’s views about the status of gravity (Friedman 1992). Friedman
argues that Kant’s theory of gravity is in a sense empirical, but unlike my
account of radical evil, Friedman’s account of gravity involves the claim that
Kant’s theory of gravity is meant to be held with strict – rather than merely
comparative – universality. See Friedman 1992:157–210.
12. To see this in the case of the Religion, see Wood 1970:224–6. Wood does not
lay out the arguments as progressing along three different lines, but he does
present arguments of all three sorts. I discuss Kant’s argument for radical
evil in more detail in Chapter 5 of this book.
13. This sort of universality applies only to the most general anthropological
claims. Kant often makes narrower claims about particular sorts of people –
people of different sexes, races, or temperaments. These claims have a more
restricted range of application than properly “universal” anthropological
claims. See too footnote 3.
14. These two difficulties are the following:

2. If [someone] wants to observe himself, he will reach a critical point with regard to
his emotional state where, generally, no further concealment is possible; that is to say
he is not consciously watching himself when impelling forces are in action, and that
he is observing himself when the impelling forces are at rest.
3. Conditions of time and place, when lasting, result in habits which, it is said, consti-
tute second nature, which makes man’s judgment of himself more difficult. (7:121)

Kant reiterates these later, saying that “knowledge of man through inner
experience” is difficult because “one who investigates his inner conscious-
ness easily carries many things into his own self-consciousness instead of
simply observing what is there” (see 7:143). In addition to these general
concerns about the difficulty of knowing oneself, Kant also suggests that
certain kinds of focus on inner sense can lead to dangerous fanaticism
(cf. 7:132–4, 161–2). None of these concerns, however, takes from the fact
176 Notes to Pages 40–46

that for Kant introspection is “of great importance” (7:143) as the primary
source of anthropological knowledge. The cautions simply point out how
difficult anthropology is and warn his readers away from improper sorts of
15. Thomas Sturm has recently argued that “from at least the 1780s on, [Kant]
advances a methodological claim against introspection as the primary method
of knowing the human mind” (Sturm 2001:174; cf. Sturm’s “Kant und die
Wissenschaft der Menschen,” unpublished manuscript). While I agree that
Kant is cautious in his use of inner sense as a source of anthropological
knowledge and that Kant insists upon various corrections to inner sense,
Sturm goes too far in his condemnation of inner sense as a source of Kantian
anthropology. As we have seen, inner sense remains primary in the published
Anthropology, even if it has important dangers and must be supplemented with
other sources.
16. Literature and experience are not the only ways in which one can correct
and refine one’s judgment. Cross-cultural studies – the descendants of the
travel writings of which Kant was so fond – are also important. Moreover,
historical-genealogical projects – such as those of Nietzsche and Foucault –
can be rightly appropriated within the context of a Kantian anthropology.
Kant could not go as far as Foucault in treating these sorts of inquiries as un-
earthing a “historical a priori” (Foucault 1970:xxii), but he can and should
endorse the important empirical project of clarifying what anthropological in-
sights are historically local. Kantians should appreciate the ways in which
Nietzsche, Foucault, and others help refine anthropological judgment about
what features of human beings are universal, even if Kant would insist that
there are at least some legitimate universal claims about human beings.
17. See for example Chapter 3, footnote 17, where I point out a shift in Kant’s
views about the ability of humans to sustain passions for long periods of time.
18. At least two candidates come immediately to mind. Given the structure of
the arguments in Perpetual Peace, one might think that he is using the term
anthropological loosely here and really means that they need to adopt a moral
vantage point. If he does mean anthropology as he normally does, this higher
vantage point is probably the perspective of inner experience, within which
an honest politician will admit his or her own capacity for doing good.
19. Granted, Kant’s anthropology has its own errors. As has already been noted,
he saw some of these and changed some of his views over the course of his
life. Other errors (about women, other races, and the deaf, for example) he
never corrected.
20. One might think here of the failure to attend to women in traditional psy-
chology, or the consideration of women as merely an afterthought. See for
example Gilligan’s critiques of Kohlberg and others in Gilligan 1982. For a
different approach, with more attention to the philosophical implications
of a more inclusive focus in psychology, see Noddings 1984. The shrinking
global village also presents more possibilities for harm from anthropologi-
cal myopia, such as interventions of the U.N., the World Bank, or various
NGOs in cultures that have been excluded from consideration in the devel-
opment of the psychological theories put to use to help them. Still, while
Notes to Pages 46–50 177

globalization increases the dangers of acting on myopic views, it also makes

those views more difficult to sustain. Neither extreme optimism nor extreme
pessimism is warranted here. People have made some very serious mistakes
about what is truly universal, sometimes with devastating consequences. But
many people continue to come to deeper and deeper understandings of
others, which often lead to fruitful and mutually beneficial interaction.
21. With respect to the postulates, Kant sometimes – especially in the second
Critique – seems to consider them merely as rational requirements for mak-
ing sense of moral obligation. In this context, they need not function as aids
to morality at all, much less empirical aids. At other times – for instance, in
the Critique of Judgment (see 5:451–3) – Kant seems to consider belief in the
postulates to function as an aid to morality, going so far as to suggest that
without such belief, one will be incapable of sustaining a commitment to
the moral law. One might think, though, that even in this case, the postu-
lates are not an empirical aid but a purely rational one, the psychological
importance of which can be recognized by empirical means. My own view
is that in these cases Kant focuses on an empirical aid, the belief itself as a
possible object of inner sense, rather than a purely rational influence; but
defending that claim is unnecessary for the purposes of this book.
22. Kant does think that people can make justified claims beyond the limits of
experience, but these claims cannot be based on any experiential evidence. In
particular, the fact of reason gives access to one’s non-empirical self precisely
because it is not a fact of experience.

Chapter 3
1. Some recent neokantian appropriations of Kant’s moral anthropology have
failed to make this distinction explicit. Thus Nancy Sherman, for example,
says of emotions that they are “primarily a matter of moral anthropology for
Kant – a way of applying the Categorical Imperative and its a priori motive
congenially to the human case” (Sherman 1990:149). I agree with Sherman
that emotions are an important part of moral anthropology but add that
this is not merely a matter of applying the Categorical Imperative to the
human case. (Sherman herself often has a more sophisticated account of
the way in which emotions fit into anthropology. I discuss her account in
detail in Chapter 4.) Robert Louden is more explicit about distinguishing
the anthropology needed to apply the moral law from the specific task of
moral anthropology. See Louden 2000:13–14.
2. Although Kant speaks specifically of a “power of judgment sharpened by
experience,” the context of this passage is an explanation of the respective
roles of pure moral philosophy and moral anthropology. Thus it is clear that
Kant conceived of “experience” here not primarily as general life experience
but as experience that is elucidated in anthropology.
3. For an example of his use of practical to describe parts of his anthropology,
see 25:1367.
4. In his introduction to the lectures on anthropology, Brandt mentions some
of the points to which I turn for evidence of Kant’s moral anthropology
178 Notes to Pages 50–52

but says nonetheless that “pragmatic anthropology is in none of its phases

of development identical with the anthropology that Kant repeatedly looks
forward to as the completion of his moral doctrine after 1770” (25:xlvi).
He goes further, claiming that “pragmatic anthropology is not systematically
integrated into Kant’s moral philosophy” (25:xlvii). Both of these claims are
true in a sense. Pragmatic anthropology includes more than mere moral
anthropology, so it is not identical to it. And Kant never explicitly points out the
systematic place of pragmatic anthropology in his moral theory. But moral
anthropology is included in pragmatic anthropology. Moreover it does fit into
his overall moral theory, and Kant makes specific mention in his lectures on
both anthropology and ethics of how it fits, even if these mentions do not
rise to the level of systematic integration. Thus insofar as Brandt suggests that
moral anthropology is not to be found in the Anthropology, he goes too far.
5. Others have recently begun to do more work on Kant’s pragmatic anthro-
pology, and some of these studies (in particular Wood 1999 and Louden
2000) include accounts of different senses of pragmatic in Kant. Although
there is some overlap between their accounts and my own, my ideas were
developed independently and have some important differences that should
become clear in the course of the chapter. Wood, in particular, misses one
of the most important senses of pragmatic (my third) in his enumeration of
four senses of pragmatic. One very important sense of pragmatic that I will not
discuss in this chapter refers primarily to the civilizing influence of culture
on the species as a whole. This sense is not discussed specifically in Wood or
Louden, but it can be found in Kant’s works. See the Groundwork (4:417fn),
Anthropology (7:323 and throughout the last section), and Kant’s writings on
6. Even here, Kant attaches to his description of pragmatic a footnote that makes
clear his reservations about considering this a strict definition of the term.
7. For an interesting application of a similar anthropological principle in con-
temporary political philosophy, see Fleischacker 1999, especially Chapter 5
on “Proper Pleasures.”
8. This is not to say that Kant actually thinks that we are determined in matters of
prudence. But his argument for our freedom implies only that we must be free
to act in accordance with the moral law and thus does not directly show any
capacity to choose for or against dictates of prudence. Still, even when one
is simply choosing between two morally acceptable alternatives, one ought
(and therefore can) always make this choice only once one is assured that
the alternatives are in fact morally acceptable. In that sense every choice is
free. Moreover, once one accepts that one is free to choose the moral law over
some particular preference, there is no reason to doubt that one can freely
choose between sensuous interests, even when both are morally acceptable.
But this latter sort of choice is not required by Kant’s argument. For more on
Kant’s theory of prudence, see Kain (in press) and Kain 1998.
9. As noted earlier (see footnote 4), Brandt does not take this to allow for a
significant moral dimension in pragmatic anthropology. Stark, by contrast,
insists that the anthropology is systematically linked with Kant’s moral philos-
ophy. See Stark 1997.
Notes to Pages 52–58 179

10. For more on the relationship between Kant and Platner, see Sturm, “Kant
on the Human Sciences” (unpublished manuscript), and Sturm, “Kant und
die Wissenschaften vom Menschen,” unpublished manuscript. Sturm also
discusses the relationship between Kant’s anthropology and the disciplines
of rational and empirical psychology. For this discussion, see too Sturm
2001. It is important to recognize, in this context, that Kant’s distinction
between pragmatic and merely theoretical anthropology does not merely
distinguish his anthropology from Platner’s particular brand of psychology,
but more generally from any version of rational psychology, and many ver-
sions of empirical psychology.
11. The other purposes are:

1. The better we know people, the better we know how to arrange our actions so as to
be fitting with theirs.
2. It teaches how one should win [against] people.
3. It teaches self-satisfaction, when one has for oneself the good that one finds in
others. (25:734)

All of these uses can relate to morality. The first clearly aids in the fulfill-
ment of imperfect duties to promote the happiness of others and conduct
oneself without interfering with others. The second relates to one’s ability
to influence others and will be discussed next as a distinct sense of the term
pragmatic. The third might seem to imply that by observing others one can
come to see what promotes their happiness and thereby acquire that one-
self. But given Kant’s association of self-satisfaction with moral goodness (see
5:38, 117–18; 6:391), one might reasonably read it instead as suggesting that
anthropology teaches one how to acquire for oneself the moral good that
one finds in others.
12. Although the context here, where pragmatic seems to be distinguished from
moral and technical, might lead one to read this prudence in the first sense (as
related to happiness), the fuller context of the passage makes clear that it has
more to do with influence on others, which could be used to any practical
(including moral) ends.
13. Kant also mentions another use of the term prudence. When used in the sense
of “private wisdom,” prudence refers to “the sagacity to combine all these
purposes for one’s own lasting advantage” (4:416fn.). This sense of prudence
simply identifies it with skill in promoting one’s happiness. A pragmatic
anthropology that developed this skill would be pragmatic in the first sense
covered above. While I do not deny that this sense of pragmatic is at play in
Kant’s anthropology, it is not the primary sense of prudence there, as the rest
of this section will show.
14. For more on politeness, see Brender 1997 and 1998.
15. Kant gives numerous examples of how this illusion is supposed to work in the
Anthropology and lectures, but his focus in those books is on distinguishing this
“illusion” from deceit, in order to provide the anthropological resources for
claiming that politeness can be morally acceptable. Basically, the distinction
is that the effect of deceit depends on getting another to actually believe the
content of the deception, whereas illusions depend for their effect on an
180 Notes to Pages 58–65

appearance, such that the effect is the same whether or not one believes the
illusion to be true. For more on this, see Brender 1997 and 1998.
16. This example has been discussed in much greater detail in Frierson 2000
and Sorenson (In process).
17. Kant’s accounts of affects and passions change over the course of his lectures
in anthropology. Specifically, in his early lectures (e.g., the 1776 Friedlander
notes, 25:589–91, 612), neither passions nor affects are lasting conditions.
In later lectures (e.g., the 1789 Busolt notes, 25:1526) and the published
Anthropology, one of the most important differences between passions and
affects is that passions last and affects pass quickly. In my account of passions
and affects, I focus on the account in the Anthropology and supplement it with
complementary details from the lectures. For more on the differences, see
Frierson 2000.
18. There are, of course, important moral dimensions to Kant’s discussion of the
character of the sexes and the nations. The character of the races is simply
too short for Kant to make much comment in the way of its moral relevance.
But Kant points out the moral dangers and advantages of various national
characters, and with respect to the character of the sexes, Kant discusses
different virtues and vices for each sex (e.g., at 25:719), prescribes certain
sorts of relations between the sexes, and points out the moral role of women
in refining and morally improving men.
19. This is a variant reading (see note 4 in the Academy edition, 25:631) that
makes more sense in the context. The Friedlander lectures actually read
“hard wax” here.
20. Another important social help to the development of character is culture.
Kant’s accounts of the importance of culture are linked with his discussions
of politeness (mentioned earlier) and with his more general accounts of the
development of the human species. Munzel 1999 examines this in detail.
21. See Chapter 6 for more on Kant’s view of historical progress and its relation
to his anthropology.
22. He never lectured on ethics without also lecturing on anthropology during
the same year. For more on the circumstances of Kant’s lectures, see Brandt
1997 and Stark 1997.
23. In fact, in his review of Kant’s Anthropology, Schleiermacher suggests that
there is little in it that is not merely peripheral. He calls the Anthropology
a “collection of trivia” and a “clear portrayal of peculiar confusion.” Kant’s
anthropology is particularly prone to this criticism in part because, as he
proudly announces when he first writes about his course in anthropology,
he includes “so many observations of ordinary life that listeners [of his lec-
tures] have constant occasion to compare their ordinary experience with
[his] remarks and thus, from beginning to end, find the lectures enter-
taining and never dry” (Kant’s letter to Herz, 10:146). This desire to make
his lectures popular infects even Kant’s published Anthropology, and so one
frequently finds tidbits of anthropological information that hold one’s atten-
tion but that do not really deserve the attention they receive. The extensive
explanations of the character of nations and of the differences between
men and women probably fall into this category. In addition to making the
Notes to Pages 65–70 181

Anthropology somewhat jumbled in general, these accounts can often obscure

the particular systematic place of moral anthropology.
24. There is one clear sense in which all of pragmatic anthropology relates di-
rectly to morality. Pragmatic anthropology teaches one how to develop all
of one’s faculties (of cognition, feeling, and desire). The cultivation of all
of these is an imperfect duty, the duty of self-perfection. So all of pragmatic
anthropology is morally relevant in that sense. And of course, one of the
most important examples I gave – politeness – came from the middle of the
section on cognition, so there is no need to rule all of Kant’s remarks there
out of the scope of moral anthropology strictly speaking. But this can by no
means be said of the whole of the Anthropology.
25. Kant’s own explanation of why he combines his discussion of affects and
passions is in a note in the section on feeling: “Note: in this section we
also ought to consider the affective excitement as a feeling of pleasure and
displeasure, which transgresses the limits of man’s inner freedom. But since
it is frequently confused with the passions, which will be discussed in another
section, that is, the one on the faculty of desire, and since it is really closely
related to desires, I shall address myself to the affective excitement at the
proper time in the third section” (7:235).

Chapter 4
1. To be fair, neither Herman nor Sherman is specifically concerned with rec-
onciling Kant’s anthropology with his theory of freedom. Nancy Sherman, in
particular, explicitly says, in a comment at the beginning of Making a Necessity
of Virtue that applies to all her work on Kant, that she will “not talk about
the thorny problems of Kant on freedom” (Sherman 1997a:21). Moreover,
although Sherman clearly sees her work on Kant as part of a revival of inter-
est in “what Kant would call moral anthropology” (Sherman 1995:369), her
conception of moral anthropology is considerably broader than Kant’s own.
Sherman seems to realize the difference between what she means by moral
anthropology and what Kant means by it. When she discusses Kant’s own
explicit description in the Metaphysics of Morals of the distinction between
moral anthropology and a metaphysics of morals, Sherman suggests that
“this [distinction] misses the force of Kant’s own fuller theory” (1997a:134).
Sherman finds fault with Kant here because Sherman’s own general concep-
tion of moral anthropology is so broad that it includes any anthropological
considerations that are needed to apply the categorical imperative to human
situations. I argued in Chapter 3 that it is more appropriate to read moral
anthropology in a narrower sense, dealing only with helps and hindrances
to having a good will. But Sherman generally uses the term more broadly.
Sometimes, however, Sherman does seem to include the emotions within
moral anthropology in the narrower sense (see footnote 3).
2. Louden even claims (see fn 57, p. 198) that Wood influenced his views. For
further parallels, see below, footnote 22.
3. But see above, footnote 1. In my treatment of Sherman through the next
three sections of this chapter, I look at her recommendations about the
182 Notes to Pages 70–73

roles that emotions can play in moral life as possibilities for the role of helps
and hindrances more generally. As I argue in Chapter 3, Kant conceives of
passions and affects, which Sherman sometimes includes among the emo-
tions, as important hindrances to moral life. Sherman focuses, however, on
emotions as empirical aids to moral choice. Moreover, although her concep-
tion of moral anthropology is often broader than mine, at times Sherman
clearly does recognize the role of emotions in moral anthropology narrowly
construed as providing subjective conditions that help or hinder having a
good will. She says, for example,
The emphasis in The Doctrine of Virtue is on who will act in morally worthy ways from a
pure attitude of virtue. And the thought is that what we can do to increase our chances
to be part of that pool is to cultivate emotions that do not battle with our duty motive
and that positively promote it. Such emotions are not themselves expressive of the
purest attitude of virtue in that they are not the ultimate source of adequate reasons for
doing what is required or for determining what is morally permissible. But they are a
layer of character that can, nonetheless, best support moral motivation. (1997a:144;
cf. Sherman 1997b:272 and Herman 1993:13)
In this passage, emotions that support moral motivation are connected ex-
plicitly with increasing the “chances” that one will be morally good. Emotions
help and hinder having a good will. Insofar as they have this role, it is rea-
sonable to insist that Sherman explain how they can serve that role. In that
context, it is proper to look to her recommendations regarding the place of
emotions in moral life in order to find explanations of how one class of helps
and hindrances promotes the good will.
4. See Sherman 1997a:144 and 1997b:273. I focus on the formulations in
Sherman’s most recent work. Here she outlines three roles for the emotions
but commits Kant to only two of them. She is more ambivalent about the
third, the “motivational” role, as a matter of Kant interpretation, though she
clearly thinks that it should play a role in a good moral theory. In an earlier
essay, “The Place of Emotions in Kantian Morality” (Sherman 1990), Sherman
outlines five different roles for the emotions. Three of these are the same as
those in the later essays, though the motivational role is described in terms of
emotions under a “regulative constraint” (Sherman 1990:161). The two roles
discussed in that essay but dropped in her more recent work are the role of
emotions as a “provisional morality” (Sherman 1990:158) and as an “aesthetic
ornament” (160). Sherman does not reject either of these functions in her
later work and even (in Sherman 1995) reiterates her commitment to the role
of emotions in provisional morality, but they seem to play a less significant
5. Of course, this does not mean that all such failures are free of moral fault on
the part of the agent. In the United States today, for example, it is conceivable
that many people could innocently fail to recognize that referring to women
as “girls” is demeaning, but it is hard to imagine a white adult who innocently
thinks that there is nothing wrong with referring to African-Americans as
6. Sometimes Herman seems to think that greater epistemic awareness is all that
is necessary for people to be morally good. Thus her account of character
Notes to Pages 73–78 183

development is primarily an account of how one comes to judge rightly, and

to have the right desires, and she explicitly denies that the cultivation of a good
will involves developing any sort of moral muscle (see Herman 1996:43; for
a response to Herman, see Brender 1997:39n17).
7. Herman considers the possibility that emotions can play these roles (see
Herman 1993:82), but her overall treatment is more general and focuses
more centrally on moral education – another important empirical influence.
8. See Chapter 3.
9. In fact, as I argue in Chapter 5, Kant does think this!
10. Again, I argue in Chapter 5 that Kant himself struggles with the latter option.
Seeing that he is evil, Kant must find some way to combat the moral despair
that seems to follow from this recognition.
11. Sherman connects this role of the emotions with politeness in particular.
(See 1997a:147f; 1997b:273f.) In this, she seems to confuse the importance
of emotions for carrying out particular duties, say of benevolence, and the
particular role of politeness as a part of moral anthropology. It is certainly
true that service with a smile is more conducive to the happiness of others
than begrudging fulfillment of one’s duty to help others. But when Kant dis-
cusses the particular importance of the graces in making virtue attractive, he
has in mind (as indicated in the previous chapter) the way in which the graces
promote the virtue, and not the happiness, of oneself and others. Sherman
notes this difference in her attention to the “educative role” (1997a:148;
1997b:274) of emotions, but she fails to offer an account of how such an
educative role is possible.
12. Louden, by contrast, reads Kant’s account of temperament to imply sharp
limits on one’s control over emotions. See Louden 2000:81.
13. The importance of distinguishing between striving for certain emotions and
actually having those emotions is emphasized in Baron 1995. She concludes
with a wonderful summary of what an agent can be held accountable for,
even if not for emotions themselves:

In short, it may be that merely having the feeling is not something that Kant sees to
be morally objectionable – or at least something for which the agent may legitimately
be held responsible. But he does recognize responsibility for harboring, cultivating,
or failing to cultivate certain desires or feelings and for taking desires to be reasons to
act accordingly. (Baron 1995:199)

14. As noted in footnote 11, Sherman’s treatment of politeness in the context of

the instrumental role of emotions is particularly confusing in this context.
The “educative role” of the graces is extremely problematic, but Sherman
seems to think that it can be explained in ways similar to the way in which
the expression of sympathy and compassion serves to help comfort a crying
15. The cultivation of sensitivity to the pains of others could serve even evil ends
in one who seeks to torture others for pragmatic purposes or who just enjoys
the pain of others.
16. Of course, all kinds of caveats are necessary here. If one needs to work
fourteen hours a day to feed one’s children without violating the rights of
184 Notes to Pages 78–81

those around one, and if one is naturally insensitive, such that the cultivation
of sympathy would be very difficult and time consuming, it may be morally
irresponsible to cultivate one’s emotional capacities. One who lacks basic
necessities need not seek out the sickroom or the debtor’s prison in an
attempt to share in the pains of others.
17. In the previous three sections, I have had to glean general accounts of the
role of empirical influences from contemporary Kantians who are concerned
with more particular issues. By contrast, Louden is explicitly concerned with
problems of integrating moral anthropology into Kant’s account of freedom.
Kant’s Impure Ethics (Louden 2000) offers an explanation and integration of
Kant’s impure ethics into his moral philosophy. This integration is not the
primary purpose of the book (as Louden has made clear to me in corre-
spondence). Louden’s emphasis is on pointing out the moral importance of
Kant’s work in anthropology, as well as related work in aesthetics, religion,
history, and pedagogy. But even if the primary purpose of the book is to
draw attention to aspects of Kant’s moral philosophy that have not been suf-
ficiently discussed in other accounts of Kant’s ethics. Louden also explicitly
addresses the philosophical problems with which this book is concerned. He
explains that impure ethics has “the aim of locating factors that will help or
hinder the development and spread of morality within human life” (Louden
2000:13), and he even explicitly raises the problem of reconciling moral an-
thropology with freedom: “A deeper doubt is whether the idea of a moral
anthropology ‘to which empirical principles belong’ is itself consistent with
Kant’s own dualistic views concerning transcendental freedom and nature”
(Louden 2000:7). He asks, “What real impact can the various empirical pre-
cepts of practical anthropology and moral education have? Won’t they affect
only our empirical character, and not our intelligible character?” (19). In
response, Louden suggests that “Kant did not address these issues, and in or-
der for me to do so it will be necessary to offer conjectures that occasionally
go beyond his texts” (Louden 2000:19). Louden’s account, then, is explic-
itly designed to solve the problem that I address in this book and that Kant
supposedly left unsolved.
18. Except for the term indispensable on page 116, all italics are my own.
19. One aspect of Louden’s view that I will not discuss in detail is a particular sug-
gestion about how empirical aids act as necessary preparations for morality.
Louden occasionally explains that certain empirical influences help by mak-
ing the moral law more accessible to the senses. Thus he says that “aesthetic
experience offers human beings tangible access to concepts of pure ethics,
and we need this access if we are going to make ethics comprehensible”
(Louden 2000:116), that “a large part of the project of impure ethics con-
sists in finding concrete ways to make the message of pure ethics graspable
to human beings” (127) and that “Kant’s account of historical progress is
of a piece with the larger tapestry of impure ethics; it is yet another way of
making morality graspable to human beings” (152). This notion of making
morality “graspable” is similar to suggestions made by Paul Guyer (see Guyer
1993) about the way in which aesthetic feeling can facilitate moral feeling.
With respect to this particular way in which empirical influences can play a
Notes to Pages 81–82 185

role as preparation for morality, there are two important problems. First, it
is not clear why a clearer grasp of morality will make it more likely that one
will choose good rather than evil. At most, it might be necessary to make it
possible for one to recognize duties, which is necessary to be a moral agent
at all (as noted in section 1, above). These influences could play an epistemo-
logical role that enables moral responsibility, but there is no explanation of
how they facilitate virtue. In his discussions of aesthetics in particular, Kant
seems more interested in describing how beauty functions to give people
a better awareness of morality (something cognitive), rather than a greater
commitment to it (something genuinely moral). Kant may – and almost cer-
tainly does – think there is a connection between these, but Louden never
explains what this connection is nor how it is consistent with transcendental
20. One could of course hold that earlier generations are morally responsible
and that these conditions really are necessary, if one denied that “ought
implies can.” Louden might want to do this, but if so, he is giving up the core
of Kant’s moral theory.
21. He mentions two other possible objections, but they relate less directly to
the issue of the relationship between freedom and empirical influence: the
problem of “perfecting others,” and the suggestion that there is a kind of
hubris – hard to justify after the twentieth century – in claiming that we are
morally better than previous generations.
22. Louden footnotes Allen Wood for “discussion on this topic” (fn 57, p. 198).
This is no surprise, because Wood’s recent book (Wood 1999) falls into many
of the same problems that Louden’s does. Throughout most of his discussion,
Wood is more careful (or more ambiguous). He can be read as suggesting
not that social and historical forces are necessary conditions of individuals’
coming to have good wills, but only of moral responsibility in general or the
moral progress of the species as a whole. But at times he seems to suggest
something stronger, as when he says that “Kant’s ethical thought is funda-
mentally about the human race’s collective, historical struggle to develop
its rational faculties and then through them to combat the radical propen-
sity to evil that alone makes their development possible” (Wood 1999:296).
Following this bold summary of Kant’s ethical thought, Wood suggests that
history can be divided into an “epoch of nature” that “continues today”
(Wood 1999:296) and an “epoch of freedom,” to which “all ethical duties
belong” (325). Thus Wood finally says, “Kant does not think that I can ever
achieve this inner revolution [from evil to good] entirely on my own. The
origin of evil is social, and so must be the struggle against it. . . . The pursuit
of my own morality can be distinguished from the moral progress of the
human species, but . . . the two ends are necessarily linked in their pursuit”
(Wood 1999:314). In Chapter 6, we will see that there is a sense in which
the moral revolution by which one has a good will is necessarily linked with
one’s actions to promote progress in the species. But whereas Wood, like
Louden, sees the origin of this link in the necessary social preconditions of
“my own morality,” I argue that social influences must be considered expres-
sions of a good will. Wood, like Louden, needs some account of how these
186 Notes to Pages 82–96

influences can cut through the causal network to affect one’s freedom, and
this account will get him into the same problems that Louden encounters
23. Kant does recognize that there are some historical prerequisites for moral
responsibility. He discusses some of these in his essay “Conjectures on the
Beginning of Human History,” discussed in Chapter 6. He also presumably
can allow that some human beings, such as infants and the insane, are not
morally responsible. But most of the aids to morality he discusses are not
aids of this sort.
24. Munzel makes use of a similar argument. She says, for example, “the pedagog-
ical process complete with propaedeutic functions that prepare and develop
these aptitudes, even awakening them to a state of moral responsiveness, is
deemed by Kant to be essential ” (Munzel 1999:340–1, emphasis added). Rec-
ognizing the potential threat to freedom, she then explains that “however
many external agencies are involved (nature, parents, teacher, philosopher,
constitution) on the behalf of human moral purpose, individual freedom is
not jeopardized. The others’ efforts of discipline and cultivation stop short
of the indispensable act of formation,” which can be effected only by free
choice (Munzel 1999:341–2). Munzel even uses the language of a “neces-
sary (albeit not sufficient) condition” (Munzel 1999:340). Insofar as Munzel
takes these passages to save freedom in the light of empirical influences, my
criticisms of Louden apply to her as well. In the next section, I argue that
she need not fall into this danger.
25. Ironically, the main metaphor I do not discuss (because it is not as fruitful)
is Munzel’s detailed horticultural metaphor. Munzel runs into several of the
problems that I raised with Louden’s account in the course of her analysis
of the horticultural metaphor. For one thing, empirical aids are represented
by the horticulturist and his activity, which activity is absolutely essential for
the success of the grafting process, which represents moral formation. This
leads Munzel into thinking of empirical aids as necessary but not sufficient
conditions of moral development, a mistake to which her other illustrations
are less susceptible.

Chapter 5
1. For those who prefer two-aspect accounts of the will, this could be explained
as the distinction between the empirical aspect of the will and the free aspect
of the will. For those who prefer two-perspective language, the distinction is
between the will considered as empirically determined and the will consid-
ered as free.
2. In part, this footnote is a response to criticisms raised in the first reviews
of the second Critique. See for example reviews by Pistorius, Rehberg, and
others. Several of these are collected in Bittner 1975. Rehberg’s criticism is
found in Schulz 1975.
3. I follow Pluhar’s translation here, though with some reservations. Pluhar
interprets jener in the German to refer to nature, but it could also refer to
freedom. Ultimately, the reference of this term does not affect my overall
Notes to Pages 96–102 187

point, because Kant clearly says that the appearance of the self is subject to
both the laws of nature and the laws of freedom.
4. This partially explains why most contemporary “Kantian” solutions to the
problem are quite different from Kant’s solution. Following the example of
Rawls, many contemporary Kantians are deeply resistant to making use of
this distinction (or any other aspects of Kant’s metaphysics) in their moral
theories. Again, although Kant’s language here suggests a two-object account
of the self, one could translate it into two-aspect or even two-perspective
terms. But one cannot get around the sorts of metaphysical considerations I
discuss in Chapter 1. In particular, one must take seriously the nature of the
asymmetry between nature and freedom.
5. There are metaphysical objections to this account of freedom and nature,
and these have been discussed by Wood, Allison, and others. As noted in the
Introduction, I am primarily concerned with the practical problems raised by
moral anthropology.
6. Christine Korsgaard has described this very clearly, though without sufficient
attention to the role of situation, in Korsgaard 1996a:75.
7. This notion of “expression” has some similarity to Charles Taylor’s notion of
expression in his discussion of “The Expressivist Turn” in The Sources of the
Self. Taylor explains,
To express something is to make it manifest in a given medium. I express my feelings
in my face; I express my thoughts in the words I speak or write. I express my vision of
things in some work of art, perhaps a novel or a play. . . . But to talk of ‘making manifest’
doesn’t imply that what is so revealed was already fully formulated beforehand. (Taylor

There are two crucial differences between this Kantian notion of expression
and Taylor’s Romantic notion. First, while expression in a Kantian sense does
not imply that what is revealed was preexistent, this is not because the ex-
pression “shapes” what it expresses (Taylor 1989:376) but because what is
expressed is necessarily something expressed, such that it “exists” only as
something expressed. Second, Romantic “expression” is the expression of
a highly individual self, such that one can express any of a vast number of
possible “selves” and thereby define oneself in any of a vast number of dif-
ferent ways. By contrast, for Kant one is fundamentally either good or evil,
and this moral status is what is expressed in one’s life. While the expressions
may be different depending on the situations in which one finds oneself, the
underlying moral disposition is, for each of us, either good or evil.
8. The restriction to voluntary actions is meant to exclude reflex actions or
actions that are the results of affects. The restriction to morally relevant
actions excludes considering eating one flavor of ice cream over another
to be expressive of one’s free will. Such nonmoral choices might be free,
but Kant is not committed to considering them expressions of one’s free
9. As Kant says, “the human being knows how to distort even inner declarations
before his own conscience” (“On the Miscarriage of all Philosophical Trials in
Theodicy,” 8:270). In this attention to self-deception, Kant fits into a tradition
188 Notes to Pages 102–107

that includes Bishop Butler, Adam Smith, and Rousseau. For more on the
influence of Smith on Kant, see Fleischacker 1999 and 1991. For more on
the influence of Rousseau, see Velkley 1978 and 1989.
10. It is particularly important to point out Kant’s awareness of this problem
of self-deception in the context of those who criticize Kant for misunder-
standing the complexities of maxim formation. In fact, Kant was well aware
not only of the possibility that the same actions could be represented under
different maxims but also of the human tendency to use this fact as a means
of justifying evil deeds.
11. Of course, one’s imperfect duty of self-perfection does require promoting
helps and preventing hindrances to morality. But this requirement is due
to the moral effects that these helps and hindrances have independent of
their place as objects of an imperfect duty. Unless helps and hindrances
affect morality in a way other than being objects of an imperfect duty, the
imperfect duty to promote them cannot get off the ground. (See my criticism
of Sherman in Chapter 4, section 3.)
12. For other reasons as well, the Religion is a good place to look for an account of
the relation between the good will and its appearances that can make sense
of the importance of moral anthropology. The first half of the work involves
an extensive discussion of the nature of the good will and its appearance,
and Kant moves in the second half to his most extensive discussion of any
single empirical aid to morality – ethical community. The first and second
Critiques and the Metaphysics of Morals include significant accounts of the rela-
tionship between freedom and its appearance in the world. The Anthropology
and essays on history include extended treatments of anthropological con-
siderations. But the Religion most clearly combines an account of nature and
freedom with an account of helps and hindrances to morality.
13. This limitation of scope applies primarily to those discussions of inscrutability
that occur in a practical context. In theoretical discussions, where Kant argues
that one cannot have theoretical knowledge of the self, he does not restrict
his scope to knowledge of one’s goodness. One cannot know theoretically
whether one is good or evil.
14. For more on Kant’s empirical argument for radical evil, see Chapter 2,
section 2, where I respond to the objection that Kant’s argument cannot
be empirical because radical evil is supposed to be universal. Here I argue
against a different objection, that Kant’s account of radical evil runs afoul of
limitations on theoretical knowledge of the self.
15. One could read this passage in a slightly different way. In particular, when
Kant says that he omits “all actions recognized as contrary to duty,” he could
just mean that an action that is recognized by a particular agent as contrary to
duty cannot be performed by that agent from duty. This would not require his
positing any inherently wicked actions. Even in this case, however, whether
or not an action is recognized by an agent as contrary to duty is an answerable
empirical question, at least by that agent.
16. That said, Kant is rightly concerned about the tendency to dwell on one’s
weaknesses in a way that undermines hope in the possibility of moral reform.
See Brender 1997 and 1998.
Notes to Pages 107–110 189

17. Kant’s claim that humans are radically evil applies to both individuals and to
the species as a whole, but in this section I focus entirely on its application to
individuals. Although Kant is interested in the role of evil at the level of the
species, for his moral philosophy I take the individual case to be fundamental.
What matters is that I, as a human being, am evil, and thus I need to struggle
against that evil. Even his argument for the importance of social solutions to
the problem of evil is rooted in the fact that each human being recognizes
that evil is furthered through bad social influences. This is discussed further
in Chapter 6. For accounts that focus on the social dimensions of evil, see
Wood 1999 and Anderson-Gold 2001.
18. There is one exception. Insofar as one acts from affect, one’s will is not
involved. Hence it is conceivable that one could perform “actions” contrary
to right without having an evil will. Unless one has lived an exceptional life,
however, it would be incredible to ascribe to this cause all one’s acts that
are contrary to right. For most of us, to do so would verge on denying one’s
own freedom. See Chapter 3 for more on affects. See too Frierson 2000 and
Sorenson (in process).
19. Throughout, it is important to remember that for Kant self-love can include
actions that are normally considered altruistic. The point is simply that one
does what is in one’s sensuous interests, either in that it immediately satisfies
some desire that one happens to have or that it is conducive to the satisfaction
of desires at some later date. Thus someone who acts from a good-hearted
desire to see others flourish is acting from “self-love” in this Kantian sense.
20. Several recent commentators (see, e.g., Allison 1990:146–61 and Wood
1999:283–5, but contrast Munzel 2000) misunderstand Kant’s insistence
that all humans have a propensity to evil as the end of the story for Kant. In
fact, the propensity for evil, because it is our own fault, is itself an indication
of actual evil in human nature. Only in this context can one make sense of
the problems that Kant raises in his discussions of grace, and in particular
in his account of substitionary atonement.
21. Even in the Anthropology, when Kant points briefly to the fact that humanity
is trapped in a “labyrinth of evil,” he quickly points out that it wanders into
this labyrinth “through its own fault” (7:326).
22. Kant does not make clear the extent to which the notion of prior experi-
ence is relevant to the propensity to moral evil. That is, he does not say that
one must have some prior experience of violating the moral law for such
violation to be incorporated into one’s maxims. Moreover, it is hard to imag-
ine that such an experience could be a necessary precondition. Whereas it
is possible to experience an object for which one has no particular prior
desire, one could not similarly experience moral evil, which is a matter of
choice, without actually choosing it. Thus the conditions of actualization of
the propensity to evil must be different from the conditions of actualization
of other propensities (for objects of potential inclination). What is relevant,
for Kant, is that human beings have a potential for deviation from the moral
law, a potential that is actualized in them given appropriate conditions.
23. For more on this, recall the discussion of Allison in Chapter 2. The “necessity”
that one with a propensity to evil will perform evil is not a logical necessity, nor
190 Notes to Pages 110–114

a morally relevant necessity, but akin to a natural necessity. Unlike a natural

necessity, the propensity to evil reflects a commitment of choice rather than
something out of one’s control, and in that sense one brings it upon oneself
that one will do evil, in addition to merely bringing it upon oneself that one
does evil.
24. This is moral antecedence, not temporal. The choice of the person is re-
sponsible for the propensity.
25. There may be degrees of this frailty, but for Kant the degrees are less sig-
nificant than the fact of frailty itself. Thus some may be able to act on their
adopted maxims much of the time but occasionally falter. Others may aban-
don principle for the slightest reasons. In both cases, one is incapable of
being morally good, because goodness requires an unbending commitment
to moral principles. However, frailty to a lesser degree may be less dangerous
in that it is easier to overcome. The more consistent one is, the more one
can make use of that consistency to further combat one’s frailty. If I cannot
stick to any principles at all, it will be very difficult to improve. If I stick to
principles for the most part, I can also stick for the most part to principles
that will strengthen my character itself. See Ameriks 1989 for more.
26. Self-deception is neither a necessary nor a sufficient condition of impurity.
There are forms of self-deception that do not involve moral motives at all,
and there are forms of impurity that are not rooted in self-deception. But
self-deception is typical of the way in which one seeks to cultivate impurity
in oneself.
27. Kant explains in his arguments for rigorism in the Religion (6:22–5) that
moral status depends upon one’s life as a whole. I discuss this throughout
the rest of this chapter.
28. There are obvious connections between Kant’s account of evil and the
Christian doctrine of original sin. Although I do not read the Religion as
merely a rational reconstruction of Christian doctrine, the similarities are
undeniable. For more on this theme, see Quinn 1984, 1986, and 1990; Adams
1998; and Hare 1996.
29. The one exception to this general approach is in the Groundwork, where Kant
(in the first two sections) consistently brackets the question of whether one
is bound by morality at all and (in the last section) “proves” that humans
are morally obligated by showing that they are free. Between the Groundwork
and his second Critique, however, Kant abandons this approach in favor of
the “fact of reason,” according to which moral philosophy simply takes for
granted the fact of moral obligation and explores its conditions of possibility.
See Ameriks 1981 for an explanation of this transition.
30. There are important differences between Kant’s defense of freedom and his
defense of the postulates of God and immortality. See Rawls 2000:317–25
for a discussion of these differences.
31. Although this account of grace is not central to my account of moral an-
thropology, it is important to be clear about the significance of grace for
Kant. Without recognizing the importance of grace, it is easy to think that
moral anthropology and a revised conception of the human good will pro-
vide Kant’s entire solution to the problem of radical evil. Natalie Brender,
Notes to Pages 114–116 191

for example, claims that “unsatisfactory as it is in the context of rational re-

ligion, the hypothesis of belief in divine aid is manifestly unavailable in the
context of secular discussions” (Brender 1997:54). She seems to think that
Kant is interested in carrying out a “secular discussion” in his solution of
the problem of radical evil. This conception of Kant’s project leads Brender
to make stronger claims on his behalf than Kant would make, such as the
claim that “we must also remain convinced that we can overcome” our will’s
corruption (Brender 1997:92, emphasis added). Thus she goes so far as to
say that “Kant’s solution to this crisis of belief and motivation [caused by
radical evil] is to direct an agent’s attention away from the impossibility of
inner validation toward the sort of validation to be found in outer experi-
ence” (Brender 1997:118). Although this kind of attention to polite society
provides an important part of moral anthropology, it is crucial to recognize
that for Kant these sorts of anthropological considerations are only part of
the story. Because he has a role for divine grace, Kant does not need to lean as
heavily on his altered conception of the human good will to escape the prob-
lems raised by radical evil. Because his account is not “secular” in Brender’s
sense, Kant has some problems that Brender does not have. But moral an-
thropology also carries less of a burden for Kant than for Brender. While it
may be worthwhile to attempt a purely secular solution to the problem of
radical evil – and my account in the next section provides resources for such
a solution – it is important to realize that Kant did not make such an attempt.
32. Several commentators have discussed Kant’s account of grace in more de-
tail. See Mariña 1997; Adams 1998; Quinn 1984, 1986, and 1990; and Hare
1996. Ultimately, I take seriously Kant’s claim that “only common moral-
ity is needed to understand the essentials of this text” (6:14), and so I see
Kant’s account of grace as first and foremost “anthropological, not theolog-
ical,” to use Allen Wood’s phrase in a slightly different context (see Wood
1999:291). Still, many of the issues that arise in the context of specifically
Christian accounts of grace arise for Kant’s as well, as Mariña, Adams, Quinn,
and Hare point out. And this makes sense, given Kant’s claim that his reli-
gion is a “pure religion of reason” that is supposedly contained within the
“wider sphere of [Christian] faith” (6:12). Because I am primarily concerned
with how grace makes possible the account of the good human will that I de-
velop in the next section, however, I do not discuss these theological issues in
33. See too 6:45, 6:61, and 7:328, where Kant makes clearer the connection
between our duty to morally improve and our need for divine grace.
34. Mariña distinguishes these two different sorts of grace (as well as one other)
in Mariña 1997. Kant distinguishes between these notions of grace through-
out the Religion (see, e.g., 6:143) and in the Conflict of the Faculties (see
35. John Hare nicely summarizes the challenge:
If we want to keep morality as demanding as Kant says it is, and if we want to concede
what Kant says about our natural propensity not to live by it, and if we want at the same
time to reject these traditional Christian doctrines [of grace], then we will have to find
some substitute for them. (Hare 1996:37)
192 Notes to Pages 116–123

Hare does not include (1) in this description, although denying (1) is a
way to avoid Christian doctrines or their substitutes. One can simply admit
the rigor of morality and our propensity not to live by it but claim that we
have freely fallen short and now are morally evil without any further hope
of reform. This would be an option, but it is not one that Kant is willing
to accept. For some recent versions of this option, see Scheffler 1992 and
Kagan 1998.
36. The lectures on religion both before and after the Religion emphasize the
inscrutability of atoning grace and give no further details. In the Religion,
although Kant does point out that grace is ultimately inscrutable, he offers
some details about the way in which atoning grace works. See Quinn 1986
for a discussion of these details. Even in the Religion, sanctifying grace is
thoroughly mysterious.
37. The Religion was published in parts during 1792–3. The Conflict of the Faculties
was written, at least in part, by December 1793 (see letter to Kiesewetter, Ak.
11:456), but it was not published until 1798.
38. This interpretation contrasts with Storr, who drew closer parallels be-
tween the postulates and grace, though see Rawls 2000, Chapter 10, for
a defense of the claim that even the postulates of God and immoral-
ity depend on practical concerns that are not essential for pure practical
39. See Chapter 1, especially footnote 12, for more. Hare uses an argument of
this sort to motivate Kant’s account of grace. As Hare explains,
On Kant’s principles it is perfectly fair for him to refuse to give an account of how
supernatural assistance might work to enable a person who is not naturally good to
make herself a good person. It would indeed be inconsistent for him to presume to
give such an account, which would go beyond the proper limits of our understanding.
But it is not legitimate, on Kant’s principles, simply to say that a bad person’s becoming
good is simply possible and leave it at that. This is not legitimate because it leaves us
with an antinomy, with the apparent conclusion that the revolution of will both is
possible (because obligatory) and impossible (because the ground of our maxims is
corrupt). (Hare 1996)
A concept of grace is necessary to show that this apparent contradiction is
only apparent. Of course, this concept of grace raises problems of its own and
must be shown to be coherent before it can legitimately be used to resolve
Kant’s antinomy.
40. Later philosophers and theologians do attempt to offer such a Critique. In
different ways and with different debts to Kant, Fichte (especially in his
Critique of All Revelation), Schleiermacher, Novalis, and Kierkegaard all artic-
ulate perspectives within which grace can be described. More recently, John
Hare (Hare 1996) has offered a philosophical analysis of the problems and
some attempted solutions, raised by the “moral gap” between the demands
of morality and the problem of radical evil.
41. For more details, see Hare 1996, Quinn 1984, Mariña 1997, and Adams
42. Kant does not use the phrase “will in revolution” in the Religion. However, the
revolution that he describes is a revolution that takes place in the noumenal
Notes to Pages 123–134 193

will, whereby a will “becomes” a good will. Because this revolution is nontem-
poral, however, it is appropriate to think of it as the status of the will rather
than as a change that happens to it.
43. For some reason, Wood and DiGiovanni translate Denkungsart as both “mode
of thought” and “attitude of mind,” even in the same sentence. This serves
to obscure important connections in Kant’s line of thinking. I translate
Denkungsart as “mode of thought” throughout. For an extensive discussion
of Kant’s notion of Denkungsart, see Munzel 1999.
44. G. Felicitas Munzel seems to take these passages to describe the same revolu-
tion that Kant discusses in the Religion. See Munzel 1999:27n13, 160–4, and
330. As will become clear, I disagree with this reading.
45. This conception of establishing character is not something that Kant comes
up with only after writing the Religion. Throughout his anthropology lectures
Kant reiterates the theme. In the Friedländer lectures of 1775–6, sixteen years
before the Religion was published, Kant is reported to have said, “character is
fixed very late, around [ohngefehr] the fortieth year” (25:654). Mrongovius,
in 1784–5, reports Kant’s saying that “character comes with ripe old age”
(25:1385). And by the time of the published Anthropology in 1798, Kant’s
views on the late establishment of character appear not to have changed
(see 7:294).
46. There is another, less important reason that a will in revolution cannot be
expressed as a life divided neatly into an evil phase and then a good one that
follows a sudden conversion. Kant’s explanation of atoning grace involves
the claim that one can suffer “punishment . . . executed in the situation of
conversion itself” (6:73, cf. 6:73ff.). Kant depends on a temporal coexistence
of the self that is radically evil and the self that is morally good, because
only on this basis can the morally good self properly function as an atoning
sacrifice for the wickedness of radical evil. For one way of explaining this
account, see Quinn 1986:450f. Quinn does not work out in sufficient detail
the relationship between what he calls a “lack of moral identity between the
pre-revolutionary and the post-revolutionary persons” and the unity of the
physical person. For both practical and metaphysical reasons, it is important
to understand that the lack of moral identity between the radically evil and
the morally good selves must be merely figurative, but this analysis is beyond
the scope of this book.
47. That combating one’s propensity to evil is possible is a condition of the pos-
sibility of our obligation to become better, an obligation that one recognizes
by the fact of reason. How it is possible cannot be comprehended, though
Kant makes use of the category of “grace” to describe the possibility of it, as
noted in section 3.
48. This also helps explain how Kant can respond to claims about moral un-
fairness or moral luck. Louden 2000 (pp. 55–6, 141) explains this problem
in several ways. The most troubling form of the problem is the claim that
whether or not one is morally good is a matter of luck or social conditions or
historical accident. For Kant, luck, social conditions, and historical accident
can affect the way in which one’s moral revolution is expressed, but they
cannot affect the status of the will itself.
194 Notes to Pages 134–143

49. Kant allows for empirical choices in the Metaphysics of Morals, where his gen-
eral notion of Willkühr (choice) is a faculty of desire (something empirical,
in this context) combined with “consciousness of the ability to bring about
its object by one’s action” (6:213), where such consciousness is something
that can be observed in inner sense. The empirical nature of this choice is
clear later when Kant contrasts “human choice” with “animal choice” (6:213).
There is no basis for positing anything about animals as things in themselves,
so the notion of “animal choice” can apply only to something empirical, an
empirical choice. Similarly, although there is a basis for positing a noumenal
grounding choice in human beings, this choice grounds an appearance that
includes numerous empirical “choices.”

Chapter 6
1. The absence of a duty to promote one’s own happiness is less important.
Kant’s reason for omitting such a duty is that “what everyone wants unavoid-
ably of his own accord does not come under the concept of duty” (6:386).
2. In the rest of this section, I focus on the Religion, because it provides the most
general account of the importance of interpersonal moral influence. I deal
with the writings on history in section 3.
3. The motivational structure here is more complicated than merely pursu-
ing virtue for its own sake, or for the sake of duty. See Herman 1993 and
Korsgaard 1996a (especially Chapter 2) and 1996b for discussions of the
distinctions between acting from a particular motivation and for a particular
4. Moreover, it is important for parents to promote their children’s moral self-
esteem in a way that does not do injustice or harm to others. This requirement
is imposed by the categorical imperative, which prohibits the promotion even
of the happiness of others if that end is achieved by immoral means. In the
next section this argument is expanded with respect to the happiness of third
parties. With specific reference to promoting the happiness gained by proper
self-esteem, this requirement has the particular implication of ruling out any
strategy according to which one dulls the conscience of another, rather than
promoting their virtue, to prevent them from the pain of self-condemnation.
5. Kant uses the term deed in this broad sense when he refers to the sensible
“deed contrary to law . . . that resists the law materially and is then called
vice,” a deed that is “sensible, empirical, [and] given in time” (6:31). Just as
sensible deeds contrary to the law are called “vice,” one can refer to deeds
that reflect goodness as “virtue.” In general, Kant reserves the term virtue for
good deeds that actually reflect a good will. In this section, however, I use
the term virtue primarily to refer to good deeds themselves, whether or not
they reflect an underlying good will. Thus one can promote “virtue,” in this
sense, without promoting a good will itself.
6. In Chapter 3, we looked at passions and affects as hindrances to morality. In
this section of the Religion, Kant specifically highlights passions as a hindrance
that has a social origin. He writes, “The passions . . . assail his nature, which
on its own is undemanding, as soon as he is among human beings” (6:94).
Notes to Pages 143–145 195

7. Of course, like all duties imposed by moral anthropology, the obligation

to promote moral community is an imperfect duty. As such, the effort one
should put into the promotion of such a community will depend on other
duties – both perfect and imperfect – that one has.
8. Of course, one does not promote moral revolutions in others, just as they
cannot determine that one undergoes a moral revolution. Kant makes this
clear when he describes the “historical representation of the gradual es-
tablishment of the dominion of the good principle on earth” (6:124). In
describing progress toward the ethical community, which he here associates
with a universal Church, Kant insists that “pure moral faith . . . is not a pub-
lic condition” and thus cannot be the object of progress. Instead, one can
give only “a universal historical account of ecclesiastical faith” (6:124), the
progress by which the Church gradually comes to endorse and promote “an
autonomous principle [of morality] which is one and the same for all human
beings for all times” (6:124). The details of Kant’s account of the relation-
ship between the Church and the ethical community are unimportant here.
The main point is that what one seeks to promote is a community focused
on teaching and affirming the value of the moral law (see too 6:132). One
does not seek and cannot expect to promote a moral revolution, or “pure
moral faith,” in another.
9. Unlike Yovel (see Yovel 1980), I do not take Kant to advocate the pursuit of
the moral community as a “highest good” that involves the moral goodness
of all people united with their happiness. As a specifically moral end, the
“highest good” as described in the Critique of Practical Reason (5:110) involves
only one’s own virtue in a world in which happiness is distributed in pro-
portion to virtue. As an end of theology, the founding of a “moral people of
God is . . . a work whose execution cannot be hoped for from human beings”
(6:100). But as an end of moral anthropology, the highest good is “an ethical
community . . . greatly scaled down . . ., namely to an institution which [is]
at best capable of representing with purity only the form of such a [truly
moral] community” (6:100). And while Yovel may be right that this ethical
community involves “an intersubjective system of attitudes that embraces,
ideally, the whole human race” (Yovel 1980:139), these attitudes are, finally,
only appearances and not good wills themselves, and this end is an end for
me primarily because it expresses my own struggle against radical evil.
10. At times, Anderson-Gold (e.g., Anderson-Gold 2001:26, 45) seems to go too
far in this direction.
11. Kant does occasionally use language that is misleading in this respect, as
when he discusses the “duty sui generis, not of human beings toward human
beings but of the human race toward itself ” (6:97). There is a sense, of
course, in which an entire community, or even the entire human race, can
fall short morally. A state can be unjust, for example, if its institutions are
fundamentally unjust, if property rights are not respected, if criminals are
not punished, and so on. Kant even goes so far as to say that a society that
disbands would be morally required to kill any imprisoned murderers. These
demands of justice are responsibilities that a political community (and only
a political community) can and must fulfill. Falling short results in the state’s
196 Notes to Pages 145–152

being unjust. Similarly, moral communities (churches, etc.) have the moral
obligation to promote virtue. And the human race itself can be said to have
the obligation to form political and moral communities. In all of these cases,
however, the moral status of a community is not independent of the moral
status of its members, and the ultimate responsibility always lies with the
individual members of the state, the church, or the human race.
12. There is one sense in which a goodness that is the result of grace can be
one’s own, at least in part. Although Kant is adamant that from a practical
perspective, one’s choice must precede grace, the form of this choice is a
deliberate receptivity to grace. Grace can be a necessary condition of goodness,
as long as the presence of this grace is dependent on some practically prior
choice (temporal priority is not an issue here). In this case, one’s goodness
is not entirely one’s own, because grace was required. This is one way in
which a will in revolution differs from a morally untainted will. But because
receptivity is both up to oneself and the ground of receiving grace, one’s
goodness is in an important sense “ultimately” one’s own. See Mariña 1997
for more on the importance of receptivity in Kant’s account of grace.
13. Again, because of Kant’s peculiar conception of possibility, it would be better
to say that he does not commit himself to the impossibility of grace. For Kant,
to claim that something is “possible” is to make a very strong claim. All he is
willing to say with regard to prevenient grace is that we cannot know that it
is impossible.
14. See especially Yovel 1980, Despland 1973, Fackenheim 1957, Galston 1975,
Michaelson 1979, Kleingeld 1995 and 1999, and Anderson-Gold 2001. Wood
1999, Louden 2000, Munzel 1999, and Michaelson 1990 also include exten-
sive treatments of history in the context of more general accounts of Kant’s
moral philosophy. Louden 2000, in particular, has a very good overview of
the way in which different accounts understand moral progress in Kant’s
writings on history (see Louden 2000:144–9).
15. For other accounts which to varying degrees argue that historical progress
involves moral progress that cannot be reduced to appearances, see Guyer
2000: Chapter 11; van der Linden 1988; Wood 1999; and Yovel 1980.
16. My criticism of Louden in Chapter 4 presents several philosophical reasons
to reject readings of Kant that suggest any sort of historical influence on
the moral status of free human beings. In this section, I focus on exegetical
issues, dealing with what Kant actually said, which I show, in most cases, is
just what he should have said.
17. This survey covers those writings that I discuss in the following sections. “What
Is Orientation in Thinking,” “On the Miscarriage of all Philosophical Trials
in Theodicy,” and parts of Kant’s first and second Critique s have relevance
to Kant’s general account of history, but none of these focuses primarily on
history, and none contributes to the issue of whether or to what extent Kant
is concerned with moral progress. For these reasons I omit them from my
account here.
18. Herder continued to publish his Ideas through 1791, but Kant said that he
could no longer review the work because he was working on his Critique of
Judgment: “The third part of Herder’s Ideen will now probably be reviewed by
Notes to Pages 152–158 197

another . . . for I lack the time for it, since I must soon get to the Grounding
of the Critique of Taste” (Letter to Christian Schütz, 25 June 1787, 10:467).
19. Robert Louden (Louden 2000:147–8), following Friedrich Kaulbach
(Kaulbach 1975:83), claims that Kant’s reference in the Conflict of the Facul-
ties to a “moral cause” influencing human events is a sign that Kant thinks
there is “strong empirical evidence . . . that humanity is progressing morally”
(Louden 2000:147). However, Kant immediately explains that this “moral
cause is twofold: first, that of the right [to a civil constitution] . . . and second,
that of the end . . . that that same constitution alone be just” (7:85). In both
cases, the “moral cause” turns out to be a purely political one.
20. See van der Linden 1988:135f; Wood 1999:310; and Louden 2000:141–2
for other examples of using the third Critique to clarify Kant’s conception of
21. They also give considerably more detail about how culture (and other ends) is
promoted in history. I do not provide the details of these discussions, because
they are not directly relevant to Kant’s views about the end of historical
22. There are other passages that I do not discuss, though these can be dealt
with in the way that I deal with one of the four passages below. For examples
of such passages and their use to support a strong notion of moral progress,
see Louden 2000:56, 147–8; Guyer 2000: Chapters 11 and 12; and Wood
1999: Chapter 9. There are other, more overarching interpretations of Kant’s
treatment of history that raise challenges for my view, but a full discussion
of these approaches is beyond the scope of this book. In this context, see
especially Yovel 1980 and Kleingeld 1995.
23. Kant’s justification for this assumption is even more provocative:

I base my argument upon my inborn duty of influencing posterity in such a way that
it will make constant progress (and I must thus assume that progress is possible), and
that this duty may rightfully be handed down from one member of the series to the
next. (8:309)

Kant seems to be saying that one can know – or at least postulate – that the
human race will make moral progress because one is obligated to promote this
progress. Were this the real import of this passage, it would provide a clear
example of the importance of intersubjective moral assistance for Kant.
24. See Louden 2000:147 for an example of using this passage to support a
reading of Kant according to which there is genuine moral progress.
25. It is not altogether clear how far this essay can be taken to be a serious repre-
sentation of Kant’s own views. Herder’s Ideen has strong religious overtones,
and Kant may have used a philosophical reconstruction of the Biblical narra-
tive as merely a means for rebutting Herder’s own supposedly pious history.
Thus even this essay, which deals directly with the development of the hu-
man species, may assume for the sake of argument that human beings have
progressed and argue about how that progress occurred. Still, this section will
show that even if one takes “Conjectures” to represent Kant’s own view about
human progress, the essay does not show that the human species progresses
morally, in the sense of its becoming more and more morally good.
198 Notes to Pages 159–161

26. This also helps makes sense of at least one way in which “nature . . . prepare[s]
man for what he himself must do in order to be a final purpose” (5:431).
For a different interpretation of this passage, see van der Linden 1988:142.
27. Louden draws heavily from Kant’s lectures on pedagogy (especially 9:450)
for his interpretation of moralization. While these lectures are generally
accepted as loosely based on Kantian ideas (see Louden 2000:33–6 and Beck
1978, but contrast Weisskopf 1975), they are unreliable as primary sources
of Kant’s view of moralization. Still, even these lectures can be interpreted
in a way that is compatible with the account of moralization offered here.
28. Louden says this specifically with regard to the “good ends” that one who is
moralized will pursue, but the point holds for all of the details of moraliza-
29. Likewise in the “Idea,” Kant specifically mentions the promotion of a “morally
good character [moralisch-gute Gesinnung],” in contrast to mere “semblances
of morality” (8:26).
30. Kant even makes oblique reference to the problem of evil and the role
of grace, saying that “this human . . . is expected to bring about what he
himself is still in need of” (7: 325) and thus “only expects [moralization]
from Providence” (7:328). Yet just as in the Religion, Kant ends up claiming
that practically one must take the stance that “the human species should and
can create its own good fortune” and thus we should “further (each to his
best ability) with all good sense and moral inspiration the approach to this
goal” (7:329).
31. See footnote 29 for a troubling text from the “Idea.” Further suggestions
of troubling texts are found in Louden 2000, Guyer 2000, van der Linden
1988, Wood 1999, and Yovel 1980.
Beyond his strictly historical writings, Kant’s discussions of education in
the Anthropology, anthropology lectures, the lecture on pedagogy, the Critique
of Practical Reason, and the Metaphysics of Morals have been taken to imply
interpersonal moral influence that goes beyond what my account allows. A
full exploration of Kant’s philosophy of education is beyond the scope of my
discussion here, but some indication of how I would deal with it is provided
by my treatment of Kant’s historical writings. L. W. Beck points out,

The ages of an individual life correspond to the stages in the history of the world. By
reading this fuller treatment of the philosophy of history we discover a key to the less
well organized treatment of the philosophy of education. (Beck 1978:197–8)

In particular, the reference to the “civilization, cultivation, and moralization”

of the human species in the Anthropology is connected there with the claim
that “man must be educated to the good” (7:325), and a similar tripartite
structure of progress appears in Kant’s lectures on education (see Louden
2000:38–44). The same sorts of moral progress that are possible in history
can be furthered through education. Just as Kant gives in the “Conjectures”
a historical account of humankind’s emergence into moral responsibility,
certain forms of education are necessary for children to become morally
responsible at all. And just as Kant describes the cultural progress of human
beings in the third Critique and “Idea,” some forms of education simply
Notes to Pages 161–166 199

further “cultural” development. Finally, just as Kant allows for the promotion
of good deeds, and even the fostering of an ethical community explicitly
committed to virtue, he points out in his writings on education how such
commitment to virtue can be fostered in children.

1. Bernard Williams, for example, has argued that although Kant was right
about the connection between morality as he understood it and freedom,
he was wrong about morality, and thus about freedom as well (see Williams
1995). R. Jay Wallace (in Wallace 1994), Simon Blackburn (in Blackburn
1998), and others have worked out sentiment-based moral theories that do
not seem to depend on the kind of transcendental freedom that Kant insists
is necessary for moral responsibility. And even Schleiermacher himself (see
especially On Freedom, Scheiermacher 1992) tried to define a notion of moral
accountability according to which transcendental freedom was unnecessary.

Kant’s Works
References to Kant’s works are cited by volume: page number of Kant’s Gesammelte
Schriften. 1900–. Ed. The Royal Prussian (later German) Academy of Sciences.
Berlin: Georg Reimer (later Walter de Gruyter).
I have made use of the following translations, making changes where necessary:

Anthropology from a Pragmatic Point of View, trans. Victor Lyle Dowdell. Carbondale:
Southern Illinois University Press, 1978.
Critique of Judgment, trans. Werner Pluhar. Indianapolis: Hackett Publishing Com-
pany, 1987.
Critique of Practical Reason, trans. Mary Gregor. New York: Cambridge University
Press, 1997.
Critique of Pure Reason, trans. Norman Kemp Smith. London: Macmillan,
Critique of Pure Reason, trans. Paul Guyer and Allen Wood. New York: Cambridge
University Press, 1998.
Groundwork of the Metaphysics of Morals, trans. H. J. Paton. New York: Harper Torch-
books, 1964.
Kant’s Philosophical Correspondence 1759–99. Ed and trans. Arnulf Zweig. Chicago:
University of Chicago Press, 1970.
Lectures on Ethics/Immanuel Kant, Ed. J. B. Schneewind and trans. Peter Heath.
Cambridge: Cambridge University Press, 1992.
Lectures on Metaphysics/Immanuel Kant, Ed. and trans. Karl Ameriks and Steven
Naragon. Cambridge: Cambridge University Press, 1997.
Political Writings, Ed. Hans Reiss. Cambridge: Cambridge University Press,
Religion within the Boundaries of Mere Reason, ed. Allen Wood and George di
Giovanni. Cambridge: Cambridge University Press, 1998.

References 201

Secondary Sources
Adams, Robert Merrihew. 1998. “Introduction” to Religion within the Boundaries of
Mere Reason. Cambridge University Press.
Allison, Henry. 1996. Idealism and Freedom. Cambridge University Press.
1990. Kant’s Theory of Freedom. Cambridge University Press.
1983. Kant’s Transcendental Idealism: An Interpretation and Defense. Yale University
Ameriks, Karl. 2000. Kant and the Fate of Autonomy: Problems in the Appropriation of
the Critical Philosophy. Cambridge University Press.
1981. “Kant’s Deduction of Freedom and Morality.” Journal of the History of
Philosophy 19:53–79.
1989. “Kant on the Good Will.” Der Begriff der Politik: Bedingungen und Gründe
politischen Handelns. Ed. Otfried Höffe. Stuttgart: J. B. Metzler.
1982. “Recent Work on Kant’s Theoretical Philosophy.” American Philosophical
Quarterly 19:1–24.
Anderson-Gold, Sharon. 1986. “Kant’s Ethical Commonwealth: The Highest
Good as a Social Goal.” International Philosophical Quarterly 26:23–35.
1984. “Kant’s Rejection of Devilishness: The Limits of Human Volition.” Ideal-
istic Studies 14:35–48.
2001. Unnecessary Evil. SUNY Press.
Anscombe, G. E. M. 1958. “Modern Moral Philosophy.” Philosophy 33:124.
Aquila, Richard E. 1979. “Things in Themselves and Appearances: Intentionality
and Reality in Kant.” Archiv für Geschichte der Philosophie 61:293–307.
Baier, Annette. 1995. Moral Prejudices. Harvard University Press.
1991. A Progress of Sentiments: Reflections on Hume’s Treatise. Harvard University
1987. “Hume, The Women’s Moral Theorist.” Women and Moral Theory. Ed. Eva
Feder Kittay. Rowman-Littlefield.
Baron, Marcia. 1995. Kantian Ethics (Almost) without Apology. Cornell University
Beck, Lewis White. 1960. A Commentary on Kant’s Critique of Practical Reason.
University of Chicago Press.
1978. “On Education.” Essays on Kant and Hume. Yale University Press.
1987. “Five Concepts of Freedom in Kant.” Philosophical Analysis and Reconstruc-
tion. Martinus Nijhoff.
Bittner, Rüdiger, and Konrad Cramer, eds. 1975. Materialien zu Kants >Kritik der
praktischen Vernunft<. Suhrkamp.
Blackburn, Simon. 1998. Ruling Passions. Oxford University Press.
Brandt, Reinhard, and Werner Stark. 1997. “Einleitung” in Kants gesammelte
Schriften 25:vii–cli. Walter de Gruyter & Co.
Brender, Natalie. 1997. “Precarious Positions: Aspects of Kantian Moral Agency.”
Doctoral Dissertation. Johns Hopkins University.
1998. “Keeping Up Appearances: Kantian Moral Agency and the Fragility of
Respect.” Presented at the NAKS Session of the Pacific APA.
Carnois, Bernard. 1987. The Coherence of Kant’s Doctrine of Freedom. Trans. David
Booth. University of Chicago Press.
202 References

Despland, Michel. 1973. Kant on History and Religion. McGill–Queen’s University

Fackenheim, Emil. 1957. “Kant’s Concept of History.” Kant-Studien 48:381–98.
Fleischacker, Samuel. 1991. “Philosophy in Moral Practice: Kant and Adam
Smith.” Kant-Studien 82:249–69.
1999. A Third Concept of Liberty: Judgment and Freedom in Kant and Adam Smith.
Princeton University Press.
Foot, Philippa. 2001. Natural Goodness. Clarendon.
Foucault, Michel. 1970. The Order of Things. Random House.
Franks, Paul. 2000. “All or Nothing: Systematicity and Nihilism in Jacobi, Rein-
hold, and Maimon.” Cambridge Companion to German Idealism. Ed. Karl
Ameriks. Cambridge University Press.
Friedman, Michael. 1992. Kant and the Exact Sciences. Harvard University Press.
Frierson, Patrick. 2000. “Kant on Affects and Passions.” Presented at the Eastern
Division of the APA.
Galston, William. 1975. Kant and the Problem of History. University of Chicago Press.
Gilligan, Carol. 1982. In a Different Voice: Psychological Theory and Women’s Develop-
ment. Harvard University Press.
Guyer, Paul. 1993. Kant and the Experience of Freedom: Essays in Aesthetics and Morality.
Cambridge University Press.
2000. Kant on Freedom, Law, and Happiness. Cambridge University Press.
Hare, John. 1996. The Moral Gap: Kantian Ethics, Human Limits, and God’s Assistance.
Clarendon Press.
Herman, Barbara. 1996. “Making Room for Character.” Aristotle, Kant, and the
Stoics: Rethinking Happiness and Duty. Ed. Stephen Engstrom and Jennifer
Whiting. Cambridge University Press.
1993. The Practice of Moral Judgment. Harvard University Press.
Hufnagel, Erwin. 1988. “Kant’s pädagogische Theorie.” Kant-Studien 79:43–56.
Hursthouse, Rosalind. 1999. On Virtue Ethics. Oxford University Press.
Kagan, Shelly. 1998. Normative Ethics. Westview Press.
Kain, Patrick. 1998. “Self-legislation and Prudence in Kant’s Moral Philosophy:
A Critical Examination of Some Constructivist Interpretations.” Doctoral
Dissertation. University of Notre Dame.
In press. “Prudential Reason in Kant’s Anthropology.” Essays on Kant’s Anthro-
pology. Ed. Brian Jacobs and Patrick Kain. Cambridge University Press.
Kaulbach, Friedrich. 1975. “Welchen Nutzen gibt Kant der Geschichtsphiloso-
phie?” Kant-Studien 66:65–84.
Kemp Smith, Norman. 1992. Commentary to Kant’s “Critique of Pure Reason.” Hu-
manities Press. (First printed 1918.)
Kleingeld, Pauline. 1999. “Kant, History, and the Idea of Moral Development.”
History of Philosophy Quarterly 16:59–80.
1995. Fortschritt und Vernuft: Zur Geschichtsphilosophie Kants. Würtzburg.
Korsgaard, Christine. 1996a. Creating the Kingdom of Ends. Cambridge University
1996b. “From Duty and for the Sake of the Noble: Kant and Aristotle on Morally
Good Action.” Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty. Ed.
Stephen Engstrom and Jennifer Whiting. Cambridge University Press.
References 203

1996c. The Sources of Normativity. Cambridge University Press.

Louden, Robert. 2000. Kant’s Impure Ethics: From Rational Beings to Human Beings.
Oxford University Press.
1986. “Kant’s Virtue Ethics.” Philosophy 61:473–89.
MacIntyre, Alasdair. 1984. After Virtue. University of Notre Dame Press.
1988. Whose Justice? Which Rationality? University of Notre Dame Press.
Mariña, Jacqueline. 1997. “Kant on Grace: A Reply to His Critics.” Religious Studies
Michaelson, G. E., Jr. 1990. Fallen Freedom: Kant on Radical Evil and Moral Regener-
ation. Cambridge University Press.
1979. The Historical Dimensions of a Rational Faith: The Role of History in Kant’s
Religious Thought. University Press of America.
Munzel, G. Felicitas. 1999. Kant’s Conception of Moral Character: The “Critical”
Link of Morality, Anthropology, and Reflective Judgment. University of Chicago
Nelkin, Dana K. 2000. “Two Standpoints and the Belief in Freedom.” The Journal
of Philosophy 157:564–76.
Noddings, Nel. 1984. Caring: A Feminine Approach to Ethics and Moral Education.
University of California Press.
O’Neill, Onora. 1989. Constructions of Reason: Explorations of Kant’s Practical Philos-
ophy. Cambridge University Press.
Pippin, Robert. 1982. Kant’s Theory of Form. Yale University Press.
Prauss, Gerold. 1983. Kant über Freiheit als Autonomie. Vittorio Klosterman.
Quinn, Philip L. 1986. “Christian Atonement and Kantian Justification.” Faith and
Philosophy 3:440–62.
1984. “Original Sin, Radical Evil, and Moral Identity.” Faith and Philosophy
1990. “Saving Faith from Kant’s Remarkable Antinomy.” Faith and Philosophy
Rawls, John. 2000. Ed. Barbara Herman. Lectures on the History of Modern Philosophy.
Harvard University Press.
Scheffler, Samuel. 1992. Human Morality. Oxford University Press.
Schleiermacher, Friedrich. 1984–. Kritische Gesamtausgabe. Walter de Gruyter.
1992. On Freedom. Trans. Albert L. Blackwell. E. Mellen Press.
1998. “Review of Kant’s Anthropology.” Trans. Jacqueline Mariña. In Schleierma-
cher on the workings of the knowing mind: new translations, resources, and under-
standings. Ed. Ruth Drucilla Richardson. E. Mellen Press.
Schulz, Eberhard Günter. 1975. Rehbergs Opposition gegen Kants Ethik. Bohlau.
Sherman, Nancy. 1998. “Concrete Kantian Respect.” Social Philosophy and Policy
1997a. Making a Necessity of Virtue. Cambridge University Press.
1997b. “Kantian Virtue: Priggish or Passional.” Reclaiming the History of Ethics:
Essays for John Rawls. Ed. Andrews Reath, Barbara Herman, and Christine
Korsgaard. Cambridge University Press.
1990. “The Place of Emotions in Kantian Morality.” Identity, Character, and Moral-
ity: Essays in Moral Psychology. Ed. Owen Flanagan and Amelie Oksenberg
Rorty. The MIT Press.
204 References

1995. “Reasons and Feelings in Kantian Morality.” Philosophy and Phenomenolog-

ical Research 55:369–77.
Sorenson, Kelly. In process. “Kant’s Taxonomy of the Emotions.” Kantian Review.
Stark, Werner. 1997. “Kant’s Anthropology Lectures.” Presented at the NAKS
Session of the Eastern APA.
Strawson, P. F. 1966. The Bounds of Sense: An Essay on Kant’s Critique of Pure Reason.
Sturm, Thomas. 2001. “Kant on Empirical Psychology: How Not to Investigate the
Human Mind” in Watkins 2001.
“Kant und die Wissenschaften vom Menschen.” Unpublished manuscript.
“Kant and the Human Sciences.” Unpublished manuscript.
van der Linden, Harry. 1988. Kantian Ethics and Socialism. Hackett.
Velkley, Richard Lee. 1978. “Kant as Philosopher of Theodicy.” Doctoral Disser-
tation. Pennsylvania State University, 1978.
1989. Freedom and the End of Reason. University of Chicago Press.
Walker, Ralph C.S. 1971. “The Status of Kant’s Theory of Matter.” Synthese 23:
Wallace, R. Jay. 1994. Responsibility and the Moral Sentiments. Harvard University
Weisskopf, Traugott. 1970. Immanuel Kant und die Pädagogik: Beiträge zu einer Mono-
graphie. Editio Academica.
Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Harvard University
1981. Moral Luck. Cambridge University Press.
1995. Making Sense of Humanity. Cambridge University Press.
Wood, Allen. 1999. Kant’s Ethical Thought. Cambridge University Press.
1970. Kant’s Moral Religion. Cornell University Press.
1984. “Kant’s Compatibilism.” Self and Nature in Kant’s Philosophy. Ed. Allen
Wood. Cornell University Press.
1991. “Unsociable Sociability: The Anthropological Basis of Kantian Ethics.”
Philosophical Topics 19:325–51.
Yovel, Yirmiahu. 1980. Kant and the Philosophy of History. Princeton University
Zammito, John H. 1992. The Genesis of Kant’s Critique of Judgment. University of
Chicago Press.
Index of Kant’s Works

For the purpose of this index, I have listed Kant’s works according to the
order in which they appear in the Academy Edition. The Critique of Pure
Reason, for which I do not use Academy Edition pagination, is listed first.
Page numbers in italic indicate pages on which the work is quoted but
not mentioned by name.
Ai/Bi–A855/B883: Critique of Pure Reason, 191n33, 191n34, 192n36, 192n37,
4–5, 14–15, 17, 19–21, 25–8, 34–5, 36, 192n42, 193n44, 193n45, 193n46,
38, 41, 46, 100, 118, 123, 148, 169n2, 194n2, 194n5, 194n6, 195n8, 195n9,
170n10, 170n12, 171n14, 171n16, 195n10, 198n30
172n17, 172n18, 172–3n21, 188n12, Ak. 6: 203–493: Metaphysics of Morals, 1, 5,
196n17 6, 31, 33, 49, 51, 55, 57–9, 61, 106,
Ak. 1: 215–368 Allgemeine Naturgeschichte, 137–9, 146, 156–7, 161, 179n11, 181n1,
33 188n12, 194n49, 194n1, 198n31
Ak. 4: 384–463: Groundwork of the Ak. 6: 202–372: Doctrine of Right (Part I of
Metaphysics of Morals, 5, 7, 15, 20, 31, 32, the Metaphysics of Morals), 106
33, 48–9, 51, 53, 97–8, 100–2, 104, 106, Ak. 6: 373–493: Doctrine of Virtue (Part II of
141, 148, 170n6, 178n5, 179n13, 190n29 the Metaphysics of Morals), 77, 182n3
Ak. 4: 464–559: Metaphysical Foundations of Ak. 7: 1–116: Conflict of the Faculties, 117,
Natural Science, 175n11 153, 154, 155, 157, 159–61, 191n34,
Ak. 5: 1–163: Critique of Practical Reason, 192n37, 197n19
4–5, 15, 17, 18, 20–4, 26–7, 31, 114, 118, Ak. 7: 117–33: Anthropology from a Pragmatic
120–1, 125,170n6, 170n12, 172n18, Point of View, 1–2, 5–6, 31, 32, 33, 34, 35,
177n21, 179n11, 186n2, 188n12, 39–41, 42, 43, 44, 49–62, 63, 64, 65,
190n29, 195n9, 196n17, 198n31 66–7, 68, 127–8, 134, 153, 154, 159–60,
Ak. 5: 165–486: Critique of Judgment, 1, 29, 161, 167n3, 168n7, 173n1, 173–4n3,
31, 32, 39, 61, 96, 135, 152, 154–5, 157, 174n4, 175–6n14, 176n15, 178n4,
177n21, 196n18, 197n20, 198n31 178n5, 179n15, 180n17, 180–1n23,
Ak. 6: 1–202: Religion within the Boundaries 181n24, 181n25, 188n12, 189n21,
of Mere Reason, 1, 8, 27, 31, 32, 35, 36, 39, 193n45, 198n31
55, 67, 68, 77, 101–2, 104, 105–12, 113, Ak. 8: 15–31: Idea for a Universal History
114, 115, 117–18, 119, 120, 121–7, 128, with a Cosmopolitan Purpose, 152, 155,
129–31, 133, 137–9, 142–3, 144–5, 150, 159, 160–1
152–3, 160–1, 172n19, 174n4, 175n12, Ak. 8: 33–42: What is Enlightenment?,
188n12, 190n27, 190n28, 191n32, 167–8n4

206 Index of Kant’s Works

Ak. 8: 107–23: Conjectures on the Beginning of Ak. 9: 439–99: Lectures on Education, 31, 55,
Human History, 152, 155, 158–9, 161, 159, 198n31
186n23, 197n25 Ak. 10–12: Correspondence, 52, 180n23,
Ak. 8: 131–47: What is Orientation in 196–7n18
Thinking, 20, 173n22, 196n17 Ak. 15–16: Reflexionen, 34, 52, 159,
Ak. 8: 252–71: On the Miscarriage of all 174n8
Philosophical Trials in Theodicy, 187n9, Ak. 25: Lectures on Anthropology, 32, 40, 42,
196n17 43, 52, 53–5, 60–1, 62, 63–4, 65, 66, 68,
Ak. 8: 272–313: On the Common Saying: ‘This 159,173n3, 177n3, 177n4, 179n11,
May be True in Theory, but it Does not Apply 180n17, 180n18, 180n19, 180n22,
in Practice’, 153, 155–9, 161 193n45, 198n31
Ak. 8: 326–39: The End of All Things, 153, Ak. 27, 29: Lectures on Ethics, 18, 27, 31, 49,
157–8, 161 51, 55, 59, 115–16, 117
Ak. 8: 340–86: On Perpetual Peace, 45, 153, Ak. 28, 29: Lectures on Metaphysics,
155, 161, 176n18 27, 55
Name Index

Adams, Robert Merrihew, 190n28, 191n32, Eliot, George, 44

Allison, Henry, 15, 35–8, 168n6, 169n2, Fackenheim, Emil, 196n14
169n4, 172n18, 174n4, 174n5, 175n9, Fichte, J. G., 3, 192n40
175n10, 187n5, 189n20, 189n23 Fleischacker, Samuel, 178n7, 188n9
Ameriks, Karl, 15, 167n2, 169n4, Foot, Philippa, 168n11
170n6, 171–2n17, 190n25, Foucault, Michel, 176n16
190n29 Franks, Paul, 167n2
Anderson-Gold, Sharon, 32, 55, 68, Friedman, Michael, 175n11
174n5, 175n8, 189n17, 195n10, Frierson, Patrick, 180n16, 180n17,
196n14 189n18
Anscombe, G. E. M., 168n11
Aquila, Richard E., 16, 169n4 Galston, William, 196n14
Aquinas, Thomas, 163 Gilligan, Carol, 176n20
Aristotle, 163 Guyer, Paul, 68, 78, 184n19, 196n15,
Austen, Jane, 44 197n22, 198n31

Baier, Annette, 168n11 Hare, John, 190n28, 191n32, 191–2n35,

Baron, Marcia, 168n10, 192n39, 192n40, 192n41
183n13 Hegel, G.W. F., 3, 168n5
Beck, Lewis White, 15, 168n6, 172n18, Herder, J. G., 152, 196n18, 197n25
198n27, 198n31 Herman, Barbara, 6, 69, 71–4, 79, 91,
Bittner, Rüdiger, 186n2 174n5, 181–2n1, 182n6, 183n7,
Blackburn, Simon, 168n11, 199n1 194n3
Brandt, Reinhard, 50, 52, 173n1, 177–8n4, Herz, Markus, 52, 174n3, 180n23
178n9, 180n22 Hufnagel, Erwin, 174n4, 174n5, 175n8
Brender, Natalie, 179n14, 180n15, 183n6, Hursthouse, Rosalind, 168n11
188n16, 190–1n31
Brontë, Charlotte, 44 Kagan, Shelly, 192n35
Brontë, Emily, 44 Kain, Patrick, 178n8
Butler, Bishop, 188n9 Kaulbach, Friedrich, 197n19
Kemp Smith, Norman, 169n2, 172n18
Carnois, Bernard, 172n18 Kierkegaard, Søren, 192n40
Kleingeld, Pauline, 196n14, 197n22
Despland, Michel, 196n14 Kohlberg, Lawrence, 176n20
DiGiovanni, George, Korsgaard, Christine, 18–9, 168n6, 169n4,
193n43 170n11, 187n6, 194n3

208 Name Index

Louden, Robert, 6, 8, 32, 55, 68–70, 79–87, Richardson, John, 43–4

91, 152, 156, 159, 163–4, 173n1, 173n2, Rousseau, Jean Jacques, 188n9
177n1, 178n5, 181n2, 183n12, 184–5n17,
184n19, 185n20, 185–6n22, 186n24, Scheffler, Samuel, 192n35
186n25, 193n48, 196n14, 196n16, Schleiermacher, Friedrich, 1–4, 6, 8–9, 32,
197n19, 197n20, 197n22, 197n24, 48, 56, 67, 95–6, 164, 167n2, 167n3,
198n27, 198n28, 198n31 168n7, 180n23, 192n40, 199n1
Schulz, E. G., 168n5, 186n2
MacIntyre, Alasdair, 168n11 Schütz, C. G., 152
Mariña, Jacqueline, 172n19, 191n32, Shakespeare, William, 43–4
191n34, 192n41, 196n12 Sherman, Nancy, 6–8, 68–71, 73–8, 90, 91,
Mendelssohn, Moses, 156 177n1, 181n1, 181–2n3, 182n4, 183n11,
Michaelson Jr., G. E., 196n14 183n14, 188n11
Molière, 43 Smith, Adam, 163, 188n9
Montaigne, Michel de, 43–4 Sorenson, Kelly, 180n16, 189n18
Mrongovius, 193n45 Stark, Werner, 52, 173n1, 178n9,
Munzel, G. Felicitas, 6, 8, 32, 53, 68–70, 180n22
87–91, 169n12, 174n5, 174n8, 180n20, Storr, Gottlob Christian, 192n38
186n24, 186n25, 189n20, 193n43, Strawson, P. F., 169n2
193n44, 196n14 Sturm, Thomas, 53, 167n1, 176n15,
Nelkin, Dana K., 169n4
Nietzsche, Freidrich, 163, 175n16 Taylor, Charles, 187n7
Noddings, Nel, 176n20
Novalis, 192n40 van der Linden, Harry, 196n15, 197n20,
198n26, 198n31
O’Neill, Onora, 6 Velkley, Richard Lee, 188n9

Pillau, 52 Walker, Ralph C. S., 175n11

Pippin, Robert, 169n2, 169n4 Wallace, R. Jay, 199n1
Pistorious, W. F., 186n2 Weisskopf, Traugott, 198n27
Platner, Ernst, 52–3, 56, 58, 179n10 Williams, Bernard, 168n11, 199n1
Pluhar, Werner, 186n3 Wood, Allen, 6, 16, 23, 37, 55, 68, 70,
Prauss, Gerold, 169n4 168n6, 172n20, 175n12, 178n5, 181n2,
185–6n22, 187n5, 189n17, 189n20,
Quinn, Philip L., 118–19, 172n19, 190n28, 191n32, 193n43, 196n14, 196n15,
191n32, 192n36, 192n41, 193n46 197n20, 197n22, 198n31

Rawls, John, 15, 19, 187n4, 190n30, Yovel, Yirmiahu, 195n9, 196n14, 196n15,
192n38 197n22, 198n31
Rehburg, August Wilhelm, 3, 168n5, 186n2
Reinhold, Karl, 167n2 Zammito, John H., 152
Subject Index

Aesthetic taste, 31, 78, 79, 81, 184–5n19 Desire, faculty of, 59–61, 65, 181n24,
Affects, 59–61, 65–6, 169n12, 180n17, 181n25, 194n49
181n25, 182n3, 189n18
Analogy, 29 Education, 6, 7, 31, 54–5, 63–4, 67, 70, 71,
Antinomy, fourth, 172n17; of grace, 120, 74, 81–4, 87, 134, 135, 138, 149, 174n4,
150–1, 192n39; third, 4, 14, 17, 19, 28, 183n7, 183n11, 183n14, 184n17,
169n2–3, 171n14 198–9n31. See also Pedagogy
Athaeneaum, 1 Emotions, 7, 43, 59, 68, 70–8, 91, 163–4,
Autonomy, 24, 147, 163, 167n4 165, 169n12, 175n14, 177n1, 181n1,
182n3–4, 183n7, 183n11–13, 184n16.
Beauty. See Aesthetic taste See also Affects; Passions
Beneficence, 76 Ethical community, 134, 137, 139, 143–4,
149, 153, 161, 188n12, 195n8–9, 199n31.
Categories, 14, 17, 27–30, 170n5 See also Moral community
Categorical imperative, 7, 48, 55, 63, 97, Evil. See Radical Evil; Will, Evil
122, 130, 177n1, 181n1, 194n4 Expression, 8–9, 87, 95, 97, 99–100, 103,
Character, 1, 6–7, 9, 54, 60, 61–7, 68–9, 104, 110, 115, 124–5, 129, 134–5, 138,
87–91, 101, 111, 127–9, 135, 153, 163–4, 142–5, 185n22, 187n7–8
169n12, 174n8, 180n18, 180n20,
180n23, 182n3, 182n6, 190n25, 193n45; Fact of reason, 15, 114, 165, 177n22,
Intelligible vs. empirical, 25–6, 35, 83, 190n29, 193n47
86, 89, 122, 171n14, 173–4n3, 184n17 Feelings, 31, 41, 43, 44, 57, 59–60, 75, 101,
Children, 64, 85, 136, 140, 146, 147, 149, 130, 140, 181n25, 183n13, 187n7.
151, 218n4, 198–9n31 See also Affects; Emotions
Christian, 114, 118, 190n28, 191n32, Feeling, faculty of, 65, 181n24
191–2n35 Frailty, 111, 113, 190n25
Civilization, 67, 79, 88, 158, 159, 160, Freedom, Asymmetry of, 4, 13–20, 83, 88,
178n5, 198n31 103, 170n11, 172n17, 187n4; Kant’s
Compatibilism, 3, 16, 23–5, 167n3, 172n18. argument for, 4–5, 14–15, 24;
See also Soft determinism Transcendental, 4, 49, 67, 70, 74, 80–7,
Culture, 55, 61, 64, 79, 80, 81, 84, 86, 152, 91, 97, 129, 134, 149, 162, 163, 165–6,
154–60, 176n20, 178n5, 180n20, 167n3, 168n7, 184n17, 185n19, 199n1
197n21 Free Will. See Will, free

Desire(s), 40–1, 58, 98, 108–9, 140, 144, God, 20, 114, 115, 117, 118–19, 120, 122,
148, 155, 170n8, 171n16, 183n6, 124, 150, 172n19, 190n30, 192n38,
183n13, 189n19, 189n22 195n9

210 Subject Index

Grace, 31, 95, 114–22, 123, 126, 131, Nature, final vs. ultimate purpose of,
148–51, 165, 189n20, 190–2n31–6, 154–5, 157–9, 198n26; laws of, 169n2,
192n38–40, 193n46–7, 196n12–13, 187n3
Observation, 5, 33–4, 39, 42, 173n2,
Happiness, 3, 37, 46, 49, 50–2, 54, 59, 65, 174n8. See also Introspection
73, 75–8, 83, 89, 112, 125–6, 134, Original sin, 190n28
138–42, 145, 146, 147, 148, 150, 151,
168n9, 179n11–13, 183n11, 194n1, Passions, 6, 59–61, 64, 65–7, 136, 155,
194n4, 195n9 169n12, 176n17, 180n17, 181n25, 181n1,
Highest good, 157, 195n9 195n6. See also Emotions
History, 1, 6, 32, 55, 64, 68, 79–83, 135, Peace, 64, 67, 82, 85, 153, 155, 158. See also
137, 138, 151–61, 178n5, 184n17, War
185n22, 186n23, 196n14, 196n17, Pedagogy, 54, 63, 87. See also Education
197n20–2, 197n25, 198n31 Pleasure, 51, 59–60, 65, 72, 98, 125–6, 140,
178n7, 181n25
Illusion, 57–8, 60, 180n15. See also Politeness, 3, 6, 7, 57–8, 61, 64, 67, 73, 149,
Politeness 164, 165, 168n9, 179n14–15, 180n20,
Incentive, 38, 102, 106, 108, 110–12, 131, 181n24, 183n11, 183n14. See also Illusion
175n10 Possibility, Kant’s notion of, 15, 27, 170n10
Inclinations, 40, 54, 55, 59–61, 64, 89, 98, Postulates, 20, 31, 46, 117, 120, 177n21,
108–10, 111, 112, 113, 122, 126, 131 190n30, 192n38
Inner sense, 39–44, 46, 140, 148, 173n2, Pragmatic, 5, 48, 50–6, 65–7, 78, 134,
175–6n14–15, 177n21, 194n49. See also 173n1, 178n4–6, 178n9, 179n10–13,
Introspection 181n24, 183n15. See also Prudence
Innocence, 131 Progress, 31, 64, 79–83, 124, 126, 128–33,
Inscrutability, 170n5 135, 136, 151–61, 184n19, 185n22,
Introspection, 39–44, 101–2, 107, 171n16, 195n8, 196n14–15, 197n19, 197n21–5,
176n14–15. See also Inner sense 198n31
Propensity, 35–8, 61, 63, 105, 107–15, 122,
Judgment, 43–5; logical forms of, 17; 124, 129–32, 135, 143, 161, 174n8,
moral, 3, 48, 71, 177n2 185n22, 189n20, 189–90n23–4, 192n35,
Justice, 76, 115–16, 119, 145, 155, 195n11 193n47. See also Radical Evil
Providence, 198n30
Literature, 43–4, 176n16 Prudence, 51–6, 85, 178n8, 179n12–13.
Love, 31, 58, 61. See also Self-love See also Pragmatic
Psychology, 6–7, 134, 165, 176n20, 179n10
Mahabharata, 44
Maxims, 35–6, 38, 63, 70–4, 75, 78–9, 98, Races, racism, 34, 62, 175n13, 176n19,
101–11, 113–14, 116, 122, 124, 127, 128, 180n18
130, 131–3, 134, 137, 141–2, 170n7, Radical Evil, 8–9, 35–9, 95, 101, 103–23,
174n8, 175n10, 188n10, 189n22, 192n39 125, 126, 128–9, 131, 133, 135, 136,
Moral Community, 55, 142, 162, 195n7–8, 138–9, 143, 145, 151, 153, 165, 174n6,
196n11. See also Ethical Community 174n8, 175n11–12, 188n14, 189n17,
Moral feeling, 184n19 190–1n31, 192n40, 193n46, 195n9
Moral law, 15, 20, 24, 26, 27, 37–8, 46, Respect, 31, 55, 73, 98–9, 102–4, 107, 130,
48–9, 55–6, 69, 71–4, 80, 81, 85, 96, 134, 158
97–9, 102, 103–4, 106–12, 113, 115–17,
119, 121, 122–3, 125–7, 130–4, 140, 141, Spectator, English, 43
144, 147, 149, 161, 163, 170n8, 175n10, Self-deception, 57, 102, 104, 107, 111–13,
177n21, 177n1, 178n8, 184n19, 189n22, 131, 133, 140, 148, 187–8n9–10,190n26
195n8 Self-knowledge, 41, 73, 101, 103, 129, 139,
Moral revolution, 9, 122–35, 136, 143–5, 171n17. See also Inner Sense;
151, 173n23, 185n22, 193n48, 195n8. Introspection
See also Will in revolution Self-love, 37–8, 104, 107, 108, 111, 175n10,
Moralization, 55, 67, 79–82, 84, 159–61, 189n19
198n27–8, 198n30–1 Self-mastery, 58, 59, 66
Subject Index 211

Self-respect, 138–9 Virtue, 7, 13, 57–64, 75–7, 102, 123,

Sexes, character of, 62, 173n3, 175n13, 137–46, 161–2, 184–5n19, 194n5
180n18. See also Woman, Women Virtue ethics, 7, 68, 168–9n12
Stoics, Stoicism, 131
Sympathy, 6, 31, 75, 183n14, 184n16 War, 79, 107, 153, 155. See also Peace
Will, evil, 8, 63, 80, 106–7, 110–12, 126,
Tale of Genji, 44 133, 139, 189n18; empirical, 8–9, 36, 95,
Teleology, 82. See also Nature 194n49; free, 8–9, 28, 30, 86, 88, 95, 97,
Temperament, 34, 63, 99, 134, 175n13, 99–104, 122, 124, 130, 132, 134, 137,
183n12 140, 143, 152, 154, 159, 187n8; in
Transcendental idealism, 14–15, 118, revolution, 9, 123–5, 127, 129–30, 136,
167n2; Two-world/two-perspective 143, 192–3n42, 193n46, 196n12
theories of, 16–23, 25, 30, 31, Woman, women, 34, 42, 70, 73, 173n3,
169n4 176n19–20, 180n18, 180n23, 182n5.
Travel, 39–40, 42–3, 176n16 See also Sexes, character of