Habermas - Justification and Application

Justification and Application
Remarks on Discourse Ethics
J iirgen Habermas
translated by Ciaran Cronin
The MIT Press, Cambridge, Massachusetts, and London, England
Third printing, 2001

First MIT Press paperback edition, 1994
This edition 1 993 Massachusetts Institute of Technology. This English version

includes three essays that were published in Erliiuterungen zur Diskursethik, 1 9 9 1
Suhrkamp Verlag, Frankfurt a m Main, Germany, and two additional pieces. The
essay "To Seek to Salvage an Unconditional Meaning without God Is a Futile
Undertaking: Reflections on a Remark of Max Horkheimer" was prepared for a
festschrift in honor of Alfred Schmidt. The interview with Torben Hviid Nielsen
was published in Die nachholende Revolution, volume 7 of Habermas's Kleine Polilische
Schriften, 1 990 Suhrkamp Verlag, Frankfurt am Main, Germany.
All rights reserved. No part of this book may be reproduced in any form or by any
electronic or mechanical means (including photocopying, recording, or information
storage and retrieval) without permission in writing from the publisher.
This book was set in Baskerville by DEKR Corporation and was printed and bound
in the United States of America.
Library of Congress Cataloging-in-Publication Data
Habermas, Jiirgen.
J ustification and application : remarks on discourse ethics I
Jiirgen Habermas; translated by Ciaran Cronin.
p. em. - (Studies in contemporary German social thought)
"Includes three essays that were published in Erlauterungen zur
Diskursethik . . . and two additional pieces"-T.p. verso.
I ncludes bibliographical references and index.
Contents: On the pragmatic, the ethical, and the moral employments
of practical reason - Remarks on discourse ethics - Lawrence
Kohlberg and neo-Aristotelianism - To seek to salvage an
unconditional meaning without God is a futile undertaking Morality, society, and ethics : an interview with Torben Hviid
Nielsen.
ISBN 0-262-08217-9 (HB), 0-262-58136-1 (PB)
I. Ethics. 2. Habermas, Jiirgen-Interviews. 3. Habermas, Jiirgen
Ethics. I. Habermas, Jiirgen. Erlauterungen zur Diskursethik.

II. Title. I I I . Title: Discourse ethics. IV. Series.
BJ 1 1 1 4. H 264 1 993
92-36599
1 70--dc 20
C IP
Contents
Preface
VII
Translator's Note
IX
Translator's I ntroduction
x1
I On the Pragmatic, the Ethical, and the Moral

Employments of Practical Reason
2 Remarks on Discourse Ethics

3 Lawrence Kohlberg and Neo-Aristotelianism
1
19
113
4 To Seek to Salvage an Unconditional Meaning

Without God Is a Futile Undertaking: Reflections on a
Remark of Max Horkheimer
5 Morality, Society, and Ethics: An Interview with
133
Torben Hviid Nielsen
147
Notes
177
Index
189
Preface
With this book I continue the investigations set forth in Moral Con
sciousness and Communicative Action ( 1990). The background to the
discussion is formed primarily by objections against universalistic
concepts of morality that can be traced back to Aristotle, Hegel, and
contemporary [ethical] contextualism. Going beyond the sterile op
position between abstract universalism and a self-contradictory rela
tivism, I endeavor to defend the primacy of the just (in the
deontological sense) over the good. That does not mean, however,
that ethical questions in the narrow sense have to be excluded from
rational treatment.
It is my hope that these essays reflect a learning process. This holds
at any rate for the explicit distinction between moral and ethical
discourses. It is worked out for the first time in the Howison Lecture
[which appears here under the title "On the Pragmatic, the Ethical,
and the Moral Employments of Practical Reason"] delivered at Berke
ley in 1988 and dedicated to my daughter Judith. Since then it would
be more accurate to speak of a "discourse theory of morality," but
I retain the term "discourse ethics," which has become established
usage.
The "Remarks on Discourse Ethics" consutute the main text and
derive from notes made during the years 1987 to 1990. They rep
resent a confrontation with competing theoretical programs and are
offered as a global critical evaluation of the relevant literature.
viii
Preface
The discussions of the working group on legal theory that took

place under the auspices of the Leibniz-Programm of the Deutsche
Forschungsgemeinschaft contributed to clarifying my thoughts; I am
indebted to the participants in the Thursday afternoon seminars.
Translator's Note
This book is a partial translation of Jiirgen Habermas's book Erliiu

terungen zur Diskursethik (Frankfurt, 199 1). Chapters I, 2, and 3 cor
respond, respectively, to chapters 5, 6, and 4 of the German text.*
Chapter 4 is a translation of "Einen unbedingten inn zu retten ohne
Gott, ist eitel. Reflexionen iiber einen Satz von Max Horkheimer,"
which appeared in Matthias Lutz-Bachmann and Gunzelin Schmidt
Noerr (eds.), Kritischer Materialismus. Zur Diskussion eines Materialismus
derPraxis (Munich, 199 1), pp. 125- 142. Chapter 5 is a translation of
"Interview mit T. Hviid Nielsen" from Habermas's Die Nachholende
Revolution. Kleine Politische Schriften VII (Frankfurt, 1990), pp. 1 14145. It consists of Habermas's written replies to questions posed by
Nielsen. An anonymous translation previously appeared under the
title ''Jiirgen Habermas: Morality, Society and Ethics : An Interview
with Torben Hviid Nielsen," in Acta Sociologica 33 ( 1990), 2:92- 1 14.
Although it deviates significantly from the German version, I have
benefited from it at a number of points and have adopted its title
and critical apparatus.
*Of the remaining three chapters of the German text, chapter I has appeared in
translation as "Morality and Ethical Life: Does Hegel's Critique of Kant Apply to
Discourse Ethics?" in Jurgen Habermas, MOTal Consciou.smss and Comunicative Action,
trans. C. Lenhardt and S.W. Nicholsen (Cambridge, Mass. , 1990), pp. 195-215, and
chapter 3 as "Justice and Solidarity: On the Discussion Concerning 'Stage 6"' in
Thomas Wren (ed.), The MOTa[ Domain: Essays in the Ongoing Discussion between Philosophy
and the Socio.l Scimces, (Cambridge, Mass., 1990), pp. 224-251. To date chapter 2, "Was
macht eine Lebensform rational?" has not appeared in English.
Translator's Introduction
Habermas's discourse theory of morality represents one of the most

original and far-reaching attempts to defend a cognitivist, deontolog
ical ethical theory in contemporary moral philosophy. 1 His declared
goal is to find a middle ground between the abstract universalism
with which Kantian ethics is justly reproached and the relativistic
implications of communitarian and contextualist positions in the tra
c:lition of Aristotle and Hegel. In pursuing this theoretical project
Habermas is rowing against the prevailing tide of skepticism con
cerning the possibility of universally valid claims in ethics.2 In the
present work he undertakes a comprehensive defense of discourse
ethics against its critics, especially those in the neo-Aristotelian camp,
and in the process develops incisive criticisms of some of the major
competing positions. Since the precise nature and strength of Ha
bermas's ethical claims have so often been misunderstood, this intro
duction begins with a sketch of the argument on which discourse
ethics rests. The second part addresses the main points of contention
with several competing positions, with a view to situating Habermas's
project in relation to important currents in contemporary Anglo
American moral thought. My goal is to show that he has philosoph
ically robust responses to the (often serious) theoretical concerns
underlying the criticisms commonly brought against discourse ethics.
I
While self-consciously Kantian in its cognitivism and its commitment

to a universalistic interpretation of impartiality and autonomy, dis-
xii
course ethics represents a sustained critique of the central role Kan

tian ethics has traditionally accorded individual reflection. Kant
argued that reflection on what is implicit in everyday moral experi
ence and judgment shows tha the autonomous exercise of the will
unconditioned by extraneous empirical motives-and hence the
spontaneous activity of a noumenal self unencumbered by such mo
tives-is a necessary precondition of genuinely moral action. For
human agents who are affected by sensuous desires and inclinations,
to act morally is to act for the sake of duty alone, which translates
into the requirement that I reflect on whether I can consistently will
that every other agent should act on my maxim of action as though
it were a universal law. Understood as an elucidation of the grounds
of validity of moral principles and judgments, the categorical imper
ative assumes that the meaning of moral validity can be adequately
grasped from the perspective of an individual reflecting on his or
her motives of action. Discourse ethics, however, is based on the
conviction that, in the wake of the irreversible shift in philosophical
concern from individual consciousness to language, monological re
flection can no longer fulfill the foundational role accorded it by
Kant. Once consciousness and thought are seen to be structured by
language, and hence essentially social accomplishments, the deliber
ating subject must be relocated in the social space of communication
where meanings-and hence individual identity which is structured
by social meanings-are matters for communal determination
through public processes of interpretation. 3
For Habermas, however, this paradigm shift does not license a
devaluation of the role of rational autonomy in ethical thought as
urged by Aristotelians and Hegelians who subordinate the individual
will to an encompassing communal ethical life, or Sittlichkeit, borne
by the supraindividual forces of custom and tradition. For Habermas
autonomy remains a central concept in ethical theory; it is defining
for the social and political project of modernity to which his thought
as a whole remains committed. With the historical transition from
traditional to modern society-mirrored at the level of individual
psychological development in the transition from conventional to
postconventional moral consciousness4-religious and metaphysical
worldviews lose their capacity to provide consensual justification of
norms of social interaction and the autonomous individual becomes
xiii
the center of the moral universe. 5 In light of this ineluctable historical

transformation, the principal alternatives to rational autonomy as a
source of moral validity seem to be (a) an arbitrary affirmation of
one's own-or adopted-traditions and ways of life and the values
underlying them as unconditionally valid, (b) a moral order based on
a contractual agreement among self-interested utility calculators
whose mutual solidarity would lack sufficient normative foundation
to sustain communal goals, or (c) an unrestricted relativism of values
and ways of life whose logical consequence would be complete prac
tical disorientation. Given his commitment to a social theory that
affords a normative standpoint for criticizing unjust social arrange
ments and their ideological justifications, none of these alternatives
is viable for Habermas. Hence in his discourse ethics he undertakes
to reconceptualize the notions of autonomy and practical reason with
the goal of vindicating the cognitivist and universalist claims of Kant's
moral theory within a dialogical framework.
This reappropriation of Kantian themes can be reconstructed in
terms of three fundamental theoretical orientations : (i) a commu
nicative theory of meaning, rationality, and validity that analyzes
language in pragmatic terms; (ii) a "transcendental-pragmatic" elu
cidation of the validity-basis of moral judgment; and (iii) a procedural
approach to moral justification.
(i) In contrast to a view that has wide currency in contemporary
analytic philosophy of language, Habermas holds that meaning can
not be adequately understood in terms of semantic rules specifying
truth conditions of proposition but must be viewed pragmatically in
terms of acceptability conditions of utterances in which speakers raise
different kinds of claims to validity. 6 The basic unit of meaning on
this account is not the sentence, statement, or proposition but the
speech act, whose primary function is to mediate ongoing commu
nicative interaction. Speech acts structure social interactions through
their illocutionary binding force. 7 This approach derives its power in
part from the connections it establishes between meaning, rationality,
and validity within a theoretical framework that ties them inextricably
to human action. Habermas concurs with Wittgenstein and the prag
matists in viewing meaning as inseparable from the role of language
in structuring practices and social interactions. His superordinate con
cept of validity allows for a more differentiated account of the inter-
xiv
relation between meaning and standards of validity than is possible

on the dominant semantic views. Because they elucidate meaning in
terms of truth conditions, semantic accounts accord preeminence to
the assertoric use of language. But on Habermas's account truth is
just one of a number of rationally criticizable validity claims raised
in speech, and this permits a distinction crucial to his defense of
ethical cognitivism.
Traditionally the issue of the objectivity of moral discourse has
been understood to be whether moral judgments express claims that
admit of truth and falsity. In Habermas's view this reflects a crucial
misunderstanding of moral discourse that has led to fruitless inves
tigations into the possibility of moral knowledge.8 The claim raised
in moral judgments, he argues, is not one to factual truth at all; the
question of the cognitive status of moral discourse turns, rather, on
identifying a distinctive validity claim raised in moral judgments,
which, however, also admits of rational criticism on the basis of pub
licly intelligible reasons. This he characterizes as the claim to normative
rightness, and the specific goal of his ethical theory is to show how it
can be rationally redeemed, that is, adjudicated on publicly intelligible
grounds in argumentative discourse. Neither the truth of factual
statements nor the rightness of norms can be decided in a deductive
fashion or by direct appeal to evidence or intuition. The only forum
where such issues, once raised, can be decided without coercion and
on a mutually acceptable basis is public discourse in which arguments
and counterarguments are competitively marshaled and critically
evaluated. Logically speaking, we can make sense of the notion of
objectivity only in terms of the kinds of reasons that can be offered
in argumentation for or against a validity claim, and in this respect
claims to rightness are on a par with truth claims. 9
Within the framework of his general theory, Habermas distin
guishes between communicative action and discourse proper. For the
most part communicatively mediated interaction proceeds on a con
sensual basis of accepted facts and shared norms. Indeed, commu
nication is conceivable only against the background of broad
agreement concerning the basic features of the natural and social
worlds within which human life unfolds, since it is impossible to
problematize all factual or normative claims simultaneously. 10 But
where disagreements arise concerning the truth of assertions or the
XV
rightness of norms, consensual interaction is disrupted and can be

resumed only when agreement on the contentious issues has been
restored. 1 1 In such cases, restoring a disrupted consensus calls for a
transition to a higher level of discourse where factual and normative
claims are subjected to critical scrutiny in a process of argumentation
freed from the imperatives of action. 12 Hence, on Habermas's ac
count, truth and normative rightness are essentially discursive matters.
Elucidating truth and rightness in terms of the conditions of ra
tional acceptability in critical discourse demands that rigorous ideal
izing conditions be set on such discourse. Truth and normative
rightness cannot be identified without further ado with the rational
consensus reached in any factual process of argumentation, since
factual agreements are fallible in principle. Regardless of our assur
ance that a particular consensus is rational, it can always transpire
that it involved ignoring or suppressing some relevant opinion or
point of view, that it was influenced by asymmetries of power, that
the language in which the issues were formulated was inappropriate,
or simply that some evidence was unavailable to the participants. 15
These considerations lead Habermas-taking his orientation from
Peirce's notion of truth as the opinion fated to survive critical ex
amination in an unlimited community of researchers-to elucidate
validity in terms of the conditions of an "ideal speech situation," that
is, the conditions that would ideally have to be satisfied by a form of
communication free of the kinds of distortions that impede the ar
gumentative search for truth or rightness. Clearly these ideal condi
tions of discourse-such as the absence of all forms of coercion and
ideology and the unrestricted right of all competent subjects to par
ticipate-can never be realized fully in any real argumentation. Yet
the notion of consensus under ideal conditions of discourse is not an
empty ideal without relation to real discursive practices. Habermas
maintains that the ideal has concrete practical implications because,
insofar as participants in real discourses understand themselves to be
engaging in a cooperative search for truth or rightness solely on the
basis of good reasons, they must, as a condition of the intelligibility
of the activity they are engaged in, assume that the conditions of the
ideal speech situation are satisfied to a sufficient degree. And it is
this normative presupposition that Habermas exploits in developing
his "quasi-transcendental" grounding of a basic moral principle.
xvi
(ii) For Habermas, as for Kant, the goal of moral theory is to

establish a basic principle of moral deliberation and judgment in
terms of which the validity of moral norms can be decided. But the
dialogical orientation of discourse ethics imposes distinctive require
ments on such a basic principle: unlike the categorical imperative, it
cannot take the form of a principle of private moral deliberation.
Rather, it functions as a bridging principle in practical argumentation
permitting participants to reach consensus on the validity of nor
mative arrangements, with a view to their implications for the satis
faction of the needs and interests of all those potentially affected by
them. Specifically, the moral principle takes the form of a procedural
principle of universalization, 'U', which states that valid moral norms
must satisfy the condition that "All affected can accept the conse
quences and the side effects its general observance can be anticipated
to have for the satisfaction of everyone's interests (and these conse
quences are preferred to those of known alternative possibilities for
regulation)."14 A central question for discourse ethics is how the
universal validity of such a principle could be established without
recourse to the metaphysical assumptions Kant relied on in elucidat
ing the categorical imperative. Habermas's justification strategy takes
the form of a number of interlocking "transcendental-pragmatic"
arguments.
Broadly speaking, transcendental arguments take some features of
experience or practice accepted as indubitable or indisputable and
argue to what must be the case if the features in question are to be
possible. 15 Habermas, taking his lead from Karl-Otto Apel, employs
such an argument to defend normative conclusions, specifically, the
claim that argumentation necessarily involves pragmatic presuppo
sitions from whose normative content a basic moral principle can be
derived. 1 6 He presents his argument in the rhetorical form of a
refutation of a moral skeptic who attempts to argue for the relativity
of moral values. Already by engaging in argumentation, Habermas
argues, the skeptic unavoidably makes certain presuppositions as a
matter of the logic of the activity he or she is engaged in, presup
positions whose normative content contradicts the position he or she
is explicitly defending, and thereby falls into a performative or prag
matic contradiction. 17 The success of this argumentative strategy de
pends on identifying appropriate features of a realm of experience
xvii
or practice demonstrably unavoidable for us, in the sense that we

cannot conceive of ourselves apart from it. Kant thought that the
objective character of our experience and knowledge provided just
such a ground from which to argue for conclusions concerning the
necessary structure of human understanding; analogously, Habermas
argues that practical argumentation constitutes a sphere of practice
that is unavoidable for human agents. Communicative action, by its
very structure, is oriented to discourse as the mechanism for repair
ing disruptions in the consensual basis of communicative interac
tion. 18 Hence, as social beings who are dependent on practical
interactions for the preservation and reproduction of our identities,
we are already implicitly committed to the normative presuppositions
of argumentative discourse. 19
(iii) It is not possible here to go into the details of the justification
of the principle of universalization, but it is important to clarify some
points concerning its logical status. 20 'U' is intended as a procedural
principle of practical argumentation that shows how a determinate
range of practical issues can be decided in a way mutually acceptable
to all participants. Its procedural character may be seen as a reinter
pretation of the formal character of the categorical imperative: while
it does not directly entail any particular normative principles, it spec
ifies the condition such principles must meet in order to be justified.
In doing so, it preserves the central role of autonomy by rejecting
sources of moral authority external to the wills of rational agents,
though autonomy is now construed in intersubjective terms as each
participant's impartial concern with ends that can be willed in common.
The structure imposed on practical argumentation by U compels
each participant to adopt the perspectives of all others in examining
the validity of proposed norms, for it is their consequences for the
needs and interests of those affected that ccnstitute the relevant
reasons in terms of which the issue of normative validity must be
decided.2 1 Now, clearly, not all practical questions admit of resolution
in this manner since they do not necessarily involve potentially com
mon interests. But practical discourse regulated by U is not envis
aged as a decision procedure for dealing with all kinds of practical
questions and hence it is not coextensive with practical reason as
such. Habermas differentiates between three distinct kinds of prac
tical questions-pragmatic, ethical, and moral-which are correlated
'
'
'
'
xviii
with different employments of practical reason. 22 Pragmatic questions

address the technical issue of appropriate strategies and techniques
for satisfying our contingent desires, ethical questions the prudential
issue of developing plans of life in light of culturally conditioned self
interpretations and ideals of the good ; neither can be answered in
universally valid terms, and the scope of the correlative notions of
practical rationality-respectively, the strategic and the prudential
is correspondingly limited . Only questions of the just regulation of
social interaction-in other words, issues of the right-admit of uni
versally valid consensual regulation, whereas ethical questions con
cern who I am (or we are) and who I (or we) want to be, and this
cannot be abstracted from culturally specific notions of identity and
the good life. Habermas treats the sphere of the moral as coextensive
with questions of justice and hence excludes from its purview much
of what has traditionally been included under the rubric of the
ethical.
One final point is important for understanding Habermas's model
of practical argumentation : while it involves strong counterfactual
idealizations, it should not be understood in the manner of social
contract constructions as a hypothetical model from which conclu
sions concerning valid principles of justice can be drawn in private
reflection. Rawls's contractualist theory of justice provides a suitable
contrast. In his more recent writings he has characterized the theo
retical status of the original position variously as a "model-concep
tion" and a "device of representation"23 in terms of which we, as
members of a modern liberal democracy, can clarify our intuitions
concerning the right and justify basic principles of justice. On Ha
bermas's approach to the theory of justice, by contrast, we cannot
anticipate the outcome of real discourses concerning proposed prin
ciples of justice among those potentially affected by their observance.
Participants alone are ultimately competent to adjudicate claims con
cerning their needs and interests, and only a consensus achieved in
argumentation that sufficiently approximates to the conditions of the
ideal speech situation can legitimately claim to be based on rational
considerations, and hence to be valid. Thus the discourse theory of
ethics demands that we go beyond theoretical speculations concern
ing justice and enter into real processes of argumentation under
sufficiently propitious conditions. 24
XIX
II
In order to situate Habermas's approach within the context of con

temporary English-language debates in moral philosophy, I note
some fundamental points of conflict between discourse ethics and
neo-Aristotelian ethics and indicate briefly the burden of proof borne
by either side.25 Perhaps the strongest thread uniting thinkers of a
neo-Aristotelian bent is a deep suspicion of what might be called the
project of modernity in ethical theory. Their suspicions are nourished
by the conviction that the modern ethics of autonomy cleave to an
individualistic understanding of the self at odds with a substantive
notion of community. In contrast to Aristotle, who saw the commu
nity, in the shape of the polis, as the bearer of the values and practices
that alone enable an agent to orient his deliberation and action to
practical goals and ideals of character, in the modern period the
individual comes to be viewed as an independent source of value
bound only by the dictates of his or her rational will. With this
individualistic turn, practical reason undergoes a profound transfor
mation : it can no longer rely completely on a sustaining background
of values embodied in communal traditions and ways of life; indeed,
the practical interest in autonomy precludes any final appeal to such
substantive values as something extraneous to the rational will and
hence, in Kantian terms, heteronomous. Practical reason thereby
finds itself burdened with the task of generating decontextualized,
and hence unconditional, moral demands in a purely immanent fash
ion from formal requirements on practical deliberation, such as those
Kant expressed in the various formulations of the categorical
imperative.
Viewed through the lens of Aristotelian ethical concerns, these
theoretical orientations seem fundamentally misguided and lead in
evitably to empty formalism at the level of moral principles, sterile
rigorism or impotence at the level of individual deliberation and
action, and incoherence and practical disorientation at the communal
level. The latter point encapsulates a communitarian critique of mod
ernity that sees the tendencies toward fragmentation, alienation, an
omie, and nihilism in modern societies as symptoms of the loss of a
coherent sense of community. Thus Alasdair Macintyre paints a bleak
picture of the incoherent state of our moral culture: the currency of
XX
contemporary moral debate, he suggests, is nothing but the debased

remnants of conceptual schemes that have long since been severed
from the totalities of theory and practice from which they originally
derived their point; under such conditions moral disputes are vitiated
by conceptual incommensurability and are fated to continue inter
minably, the participants lacking shared criteria in terms of which
they could mediate their emphatic claims and counterclaims.26 But
while the pathologies of contemporary life may lend a certain plau
sibility to Macintyre's critical posture, a critique of our moral lan
guage that depicts us as systematically deluded concerning the import
of our own moral judgments would have to show that the modern
ideal of autonomy is empty and that the philosophical project of
grounding morality in requirements of practical reason is intrinsically
untenable. Macintyre's historical narrative of decline, which draws
parallels between alleged inconsistencies in that project and inco
herences in modern ethical culture, apart from exaggerating the
importance of moral philosophy, is scarcely adequate to the task. It
is open to a defender of modernity like Habermas to counter this
story of the decay of a grand tradition in ethics extending from
Aristotle through the Middle Ages with one in which Kant's moral
theory marks the uncovering of an autonomous dimension of prac
tical reason that remained implicit in the thought of his predecessors.
Indeed, Habermas is here on relatively strong ground: against neo
Aristotelian critiques of the normative incoherence of modern life he
can bring to bear the full weight of a sophisticated analysis of pro
cesses of social and cultural rationalization (grounded in his theory
of communicative action) to argue that the modern period marks the
culmination of an irreversible historical process of increased differ
entiation of spheres of validity and discourse.27 As we have seen,
Habermas maintains that communicative action-action oriented to
reaching understanding on the basis of criticizable validity claims
is essential to social order and that claims to normative rightness
constitute one of the dimensions of validity that structure commu
nication. This enables him to paint a compelling picture of modernity
as involving the emergence of forms of social organization explicitly
structured by such claims. Moreover, he can counter that under
conditions of irreducible pluralism, consensus concerning basic values
and notions of the good life has permanently receded beyond the
XXI
horizon of possibility, and hence that neo-Aristotelian appeals to

tradition and community as a basis for coordinating social action
simply fly in the face of historical reality. Under such circumstances
we are left with no alternative except to locate the normative basis
for social interaction in the rational structure of communication itself.
But ultimately the construction of competing interpretations of
history cannot be decisive since they necessarily presuppose a guiding
normative standpoint, as both Habermas and Macintyre acknowl
edge. Viewed in this light, the issue between discourse ethics and
neo-Aristotelianism comes down to the philosophical question of the
internal coherence of their respective accounts of practical reason.
Here neo-Aristotelians draw on a long tradition of powerful critiques
of the apparent abstractness of the modern autonomous subject and
the inevitable emptiness of formal principles grounded solely in the
constraints of an unsituated reason. These and related criticisms can
be traced back to one form or another of Aristotle's distinction be
tween the realm of theoria, the unchanging realities of which we can
have universal knowledge, and that of praxis, the changing social
situations in which our actions unfold. The fact that the agent must
always take account of the shifting features of practical situations in
deliberating on how to act--or, in Aristotle's terms, the fact that action
is necessarily rooted in the particular-means that theoretical cogni
tion of universal truths, or episteme, has strictly limited relevance for
practical reflection. Since theoretical knowledge can at best take ac
count of the universal features of practical situations, action calls for
a different form of cognition-prudential deliberation or phronesis
that cannot attain a high level of certainty or generality because it
must remain sensitive to particulars. 28 Moreover, since maxims of
prudence cannot be applied solely on the basis of intellectual insight,
phronesis must be inculcated through training and practical experi
ence and sustained through a stable personality structure comprising
fixed traits of character. Thus practical reason for Aristotle essentially
presupposes a background of communal traditions embodying ideals
of individual virtue, and it is only through induction into the asso
ciated practices and forms of communal life that the individual ac
quires the capacity for ethical agency.
Because this account of practical reason gives expression to endur
ing insights concerning human agency, it provides ammunition for
xxii
potentially damaging attacks on moral theories in the Kantian tra

dition. Viewed in Aristotelian terms, the primacy Kant accords the
justification of universal principles of action, for example, must lead
either to formalism and practical impotence, since unconditionally
universal principles cannot presume to capture all of the practically
relevant features of action situations, or to a sterile rigorism where
principles are applied in a rigid fashion without regard to relevant
contextual features. Correlative problems arise regarding the subject
of deliberation and action and the sources of moral motivation. On
Kant's account our grounds for acting morally must be immanent to
practical reason as such, understood as independent of socially or
naturally conditioned desires or prudential considerations of the in
dividual good. Human nature or social context cannot provide points
of application for the moral will, which must be viewed as generating
moral value from within itself. This seems to presuppose a radically
unsituated moral subject who can formulate coherent practical inten
tions in isolation from natural desires and a socially conditioned
identity. But even aside from the intractable problem of how the
yawning gap between such a faculty of reason and concrete intentions
and actions could possibly be bridged (the problem of application),
this position seems to render the sources of moral motivation inscrut
able by divorcing questions of morally right action from considera
tions of the individual good. And once Kant's own seemingly
boundless faith in reason is shaken, it is a short step to the voluntarist
idea that moral values are grounded in free decisions of individual
wills.
Whatever the merits of these criticisms of Kantian ethics in general,
they cannot be applied to discourse ethics without significant quali
fications that tend to neutralize their destructive potential. This be
comes evident once we consider its treatment of the practical subject.
One of the cornerstones of discourse ethics is its emphatic rejection
of the unsituated notion of the subject criticized by neo-Aristotelians:
it regards the capacity for agency as the result of socialization into
forms of life structured by communicative action; hence autonomy
and freedom are for Habermas essentially social matters. Even more
significantly, discourse ethics goes beyond both Kant and the Aris
totelian tradition in understanding practical reason from the per
spective of the interaction of a plurality of subjects rather than that
xxiii
of the individual deliberating subject. On this account, thinkers such

as Bernard Williams simply fail to comprehend the point of the
modern notion of morality by accepting the ancient understanding
of the issue of how one should live as essentially an individual prob
lem.29 In modern societies, where agents can no longer coordinate
their actions solely by appeal to a background of shared values, the
question of how one should live inevitably raises the question of how
we should re[!;Ulate our interactions. But the meaning of this question is
such that it cannot in principle be elucidated from the perspective of
the Aristotelian deliberating subject. It demands that individuals look
beyond their own needs and interests and take account of the needs
and interests of others-that is, that they go beyond the egocentric
perspective of prudence. In addition, it requires that each adopt a
perspective whose basic feature is captured in the universalization
test of the categorical imperative, that is, the impartial perspective of
principles of action that all could will. Impartiality in matters of the
regulation of social interaction, Habermas claims, can only be
achieved through a process of practical deliberation and reasoned
agreement among all those potentially affected by a proposed norm
of justice. In thus reinterpreting moral-practical reason as essentially
communicative, and hence intersubjective, discourse ethics can legit
imately claim to put the Kantian project on a new footing.50
It might nevertheless be objected that discourse ethics remains
vulnerable to modified, though no less damaging, forms of the crit
icisms of emptiness and formalism. On Habermas's model, practical
argumentation is a procedure for deliberating upon the validity
claims of proposed principles of justice at a remove from the exigen
cies and constraints of action. Must not the same yawning gap be
tween valid principles and real contexts of action that threatens to
engulf Kant's construction again open up here? Moreover, Haber
mas's analysis of the normative presuppositions of practical argu
mentation only yields a procedural principle governing discourse but
no substantive principles of justice as such. What practical guidance
could agents hope to derive from such an abstract principle, and how
can it claim validity beyond the sphere of discourse it regulates?
Habermas responds to the first concern by insisting on a dear
distinction between discourses of justification and discourses of ap
plication.31 An irreducible duality attaches to the notion of a valid
XXIV
norm : on the one hand, it should be capable of commanding the

rational assent of all potentially affected by its observance and, on
the other, its observance should be appropriate in all situations in
which it is applicable. But these two requirements cannot be satisfied
simultaneously because participants in a practical argumentation de
signed to test the validity of a proposed norm cannot take account
of the relevant features of all possible situations in which the norm
in question might be applicable. Thus if it is to be possible for finite
subjects to reach any justified normative conclusions-the alternative
being complete practical paralysis-the principle of universalization
can demand at most that they take account of the consequences that
the general observation of a norm can be anticipated to have on the
basis of their present knowledge. 52 But this means that all conclusions
concerning the validity of norms are open to reinterpretation in the
light of unforeseen situations of application and that questions of
their appropriateness to particular situations must be answered sep
arately from the question of justification. In other words, application
calls for a new discursive procedure, governed by a principle of
appropriateness, which addresses the question of whether a norm
should be observed in a particular situation in light of all of the
latter's relevant features. Only the principles of universalization and
appropriateness together do complete justice to the notion of impar
tiality underlying discourse ethics.
To the modified objection of formalism-that the proposed pro
cedural moral principle does not generate any substantive principles
of justice and can give no concrete guidance to action-Habermas
responds that the very meaning of the notion of autonomy, as rein
terpreted in intersubjective, discursive terms, dictates that philosophi
cal reflection on the moral cannot itself generate substantive moral
principles. Such reflection itself stipulates that questions of validity
can be answered only through real processes of argumentation among
those involved. Because the meaning of impartiality is elucidated in
terms of adopting the perspective of everyone affected, and because
this notion is given an operational interpretation in terms of a dis
cursive procedure in which each participant has the opportunity to
express his or her needs and interests, it is only by actually engaging
in discourse with others that one can attain a rational conviction
concerning the validity of a normative proposal. 55 As to the question
XXV
of how discourse ethics can provide concrete guidance for action,

this has already been addressed in part by the analysis of the problem
of application. Moreover, Habermas never suggests that practical
discourse could generate concrete practical principles from out of
itself in an a priori fashion. Practical discourses respond to disrup
tions in normative consensus and hence, notwithstanding their ideal
izing presuppositions, are always situated within the life-world
horizon of some particular group of people from which they derive
the contents to be tested. 34
Insofar as it bears on the issue of motivation, however, this criticism
raises another problem for Habermas. Thus Herbert Schnadelbach
has objected that in its one-sided cognitivist orientation and its anxiety
to exorcize the ghost of decisionism, discourse ethics underestimates
the significance of volition and decision in moral life.35 Habermas's
response is that the issue of motivation cannot be addressed at the
level of moral theory. Nor can adherence to valid norms itself be
assured by the outcomes of practical discourse. Argumentation can
generate rational conviction concerning the validity of norms of in
teraction, but it cannot ensure that they will in fact be acted upon.
Moral motivation has its sources in the affective psychological devel
opment of individuals, which is contingent on socialization into forms
of communal life that foster and reinforce sensitivity and openness
to the claims of others. In Habermas's words, "any universalistic
morality is dependent on a form of life that meets it halfway. There
has to be a modicum of congruence between morality and the prac
tices of socialization and education. The latter must promote the
requisite internalization of superego controls and the abstractness of
ego identities. "36
A more global criticism, which speaks to a sense of unease inspired
in some by Habermas's unabashed advocacy of a universalist notion
of practical reason, is that Kantian moral theory involves an ideal of
public reason that strives for unlimited transparency in human life
by demanding that all evaluative commitments be understood as
voluntary commitments that are publicly justifiable.37 The role dis
course ethics assigns public argumentation would seem to make it
particularly vulnerable to such criticism. But Habermas's concern
with openness and publicity is motivated neither by an aspiration to
unlimited explicitness nor by the mistaken assumption that all valid
XXVI
evaluative commitments must be entered into voluntarily (in the sense

that they should ideally be accepted only on the basis of rational
convictions resulting from discursive examination). Rather, he limits
the demand for consensual legitimation to one clearly circumscribed
sphere of practical questions-those concerning just norms of social
interaction-where such an ideal is not merely appropriate but his
torically unavoidable. When confronted with the question of which
norms should govern our interactions (itself inescapable given the
character of life in modern industrial societies) , we have no choice
but to look to public norms to which all mature agents could freely
assent, since we can no longer count on a shared ethos to sustain our
interactions. But it would be a dangerous illusion to think that we
could completely transform the normative parameters of our exis
tence in this manner and that the moral community might thereby
become coextensive with human life as such. Though as social actors
we are under a moral obligation to adopt an impartial perspective
on the needs and interests of all affected, such a demand is clearly
inappropriate when it comes to deciding the ethical questions of who
I am and who I want to be-what career I wish to pursue, who I
wish to associate with in the sphere of intimate relations, and so forth.
The network of identity-sustaining loyalties and evaluative commit
ments into which we are born and socialized is something that re
mains substantially untouched by the outcomes of practical
discourses, except in the negative sense that we must renounce or
modify commitments and loyalties that conflict with our moral obli
gations toward others. 3 8
Neo-Aristotelian contrasts between abstract rights and principles
and substantive ethical life, and between rational autonomy and the
situated practical subject-to the detriment of the former term in
each case-must be reconsidered in light of the intersubjective turn
imparted the Kantian project by discourse ethics. At times Habermas
stresses the discontinuity between the moral point of view operation
alized in practical argumentation and the internal perspective of
concrete ethical life from which issues of the individual and collective
good are thematized : under the impartial moral gaze factual norms
and values take on a merely problematic status and are examined as
to their abstract validity.39 By opposing the moral to the evaluative
xxvii
in such a stark fashion, he seems to lend substance to the view that

morality as construed by discourse ethics is ultimately alien to the
identities and interests of particular individuals. But a closer exami
nation of his position reveals this impression to be at very least one
sided. We have already noted several points of mediation between
universal principles and concrete contexts of action in discourse eth
ics : the issues addressed in practical discourse have their origin in
contexts of interaction structured by existing norms and values; dis
courses of justification have to be supplemented by discourses of
application sensitive to relevant, though unforeseeable, features of
situations of action ; and moral principles are dependent for trans
lation into action on complementary sources of motivation rooted in
structures of identity that are the result of socialization into appro
priate forms of social life. Thus moral discourse is tied back into the
lifeworld of socialized subjects both at the outset and in its issue.
Moreover, Habermas goes some way toward accommodating the
neo-Aristotelian concern with community in terms of a moral com
mitment to solidarity. Since personal identity can be achieved only
through socialization, the moral concern with autonomy and equal
respect is inextricably bound up with an interest in the preservation
and promotion of intersubjective relationships of mutual recognition,
and hence of forms of communal life in which they can be realized.40
Thus morality must be supplemented by a political ethics whose goal
is to mediate between abstract principles of justice and collective
identities via positive law and public policy. Nor is morality merely
an arbitrary imposition of alien normative standards onto a recalci
trant substratum of communal forms of life : the lifeworld we mod
erns inhabit is already pervaded through and through by the
universal principles of justice and corresponding abstract personality
structures outlined by discourse ethics: "Because the idea of coming
to a rationally motivated mutual understanding is to be found in the
very structure of language, it is no mere demand of practical reason
but is built into the reproduction of social life . . . . To the extent that
normative validity claims become dependent on confirmation
through communicatively achieved consensus, principles of demo
cratic will-formation and universalistic principles of law are estab
lished in the modern state."41
xxviii
Notes
1. The most important systematic exposition of his approach is "Discourse Ethics:
Notes on a Program of Philosophical Justification" (henceforth "DE") in Mural Con

sciowness and Communicative Action, trans. C. Lenhardt and S. W. Nicholsen (Cambridge,
Mass., 1990), pp. 43-115, to which the present work may be seen as a companion
volume.
2. Indeed he confronts the issue head on by casting his exposition in the form of a
demonstration of the self-defeating character of ethical skepticism. Cf. DE, pp. 7677.
3. Habermas's repeated criticisms of the "monological" character of the reflective
procedure enjoined by the categorical imperative-in contrast with the dialogical

procedure of practical argumentation central to discourse ethics-reflect his conviction
that the paradigm shift from the "philosophy of consciousness" or "philosophy of the
subject" to the philosophy of language and action marks an undeniable advance in
our understanding of the central problems of modern philosophy. Cf. The Philosophical
Discourse of Modernity, trans. F. Lawrence (Cambridge, Mass., 1987), especially pp. 296
ff.
4. Habermas regards Kohlberg's stage theory of moral-psychological development as
providing essential empirical confirmation of his discourse theory of ethics. For a

comprehensive treatment see "Moral Consciousness and Communicative Action," in
Habermas, Mural Consciowness, pp. 1 16-194, and the third essay in the present volume.
5. In J. B. Schneewind's words, the transition to the modern period in moral and
political thought is marked by "a movement from the view that morality must be
imposed on human beings towards the belief that morality could be understood as
human self-governance or autonomy." "Modern Moral Philosophy," in P. Singer, ed.,
A Companion to Ethics (Oxford, 1991), p. 147.
6. Cf. below pp. 55-56, 145-146, 162-163.
7. Cf. Habermas "What is Universal Pragmatics?" in Communication and the Evolution of
Society, trans. T. McCarthy (Boston, 1979), pp. 65-68. For an illuminating discussion
of Habermas's formal pragmatics in relation to analytic theories of meaning, see

Kenneth Baynes, The Nurmative Grounds of Social Criticism: Kant, Rawls, and Habermas
(Albany, NY, 1992), pp. 88-108.
8. Cf. his criticisms of moral intuitionism and value ethics, DE, pp. 50-57 and Mural
Consciowness, p. 196. Bernard Williams's discussion of the objectivity of ethical judg
ments in terms of the question of the possibility of ethical knowledge is also open to
this criticism-d. Ethics and the Limits of Philosophy (Cambridge, Mass., 1985), pp. 132
ff.
9. Hence Habermas's insistence that the validity clajm raised in moral judgments, while
not a claim to truth, is analogous to a truth claim-see DE, pp. 57-62. Discourse ethics
rejects the opposition governing the recent metaethical debate in analytic ethics con
cerning realist and anti-realist interpretations of moral discourse by implying that the
question of whether or not there exist moral 'facts' described in moral judgments
presupposes a mistaken interpretation of the logic of moral discourse on the model
of factual discourse. See, for example, Michael Smith, "Realism," in Singer, ed., Com
panion, pp. 399-410, and G. Sayre- McCord, ed., Essays on Moral Realism (Ithaca, 1988).
xxix
10. Habennas elaborates this fundamental insight in his theory of the 1ifewor1d-cf.
The Theory of Communicative Action Vol. 2, trans. T. McCarthy (Boston, 1987), pp. l l 3
ff. and Philosophical Discourse, pp. 298-299, 342ff.
11. There are, of course, other possibilities. Interaction may be broken off altogether
an option of limited scope given the practical imperatives of communal coexistence
or it may continue on a curtailed consensual basis, where disputed factual issues are
bracketed or a compromise is negotiated concerning disputed normative issues. Alter
natively, belief and compliance can be assured through various forms of deception or
coercion (e.g. , propaganda, psychological manipulation, or straightforward threats),
but such pseudo-consensus, apart from being morally and politically objectionable, is
inevitably an unstable basis for ongoing interaction.
12. While communicative action and discourse are very closely interrelated-Habermas
describes discourse as a reflective form of communicative action-only in discourse is
the issue of validity thematized in a universalistic manner that transcends the limits of
a particular community. Cf. Habermas, Moral Consciousness, pp. 201-202 and ''justice
and Solidarity: On the Discussion Concerning 'Stage 6'," in Michael Kelly, ed . , Her
meneutics and Critical Theory in Ethics and Politics (Cambridge, Mass., 1990), p. 48.
13. An important feature of Habermas's account of validity claims often overlooked
by critics is how it combines a nonrelativistic defense of the objectivity of truth and
normative rightness with a thoroughgoing fallibilism concerning particular factual and
normative claims, however well supported by real argumentation. This applies to his
own theoretical claims as well: he explicidy ties the fate of discourse ethics to recon
structions of implicit knowledge and competences that he acknowledges are fallible,
and hence contestable, in principle. Cf. Habermas, Moral Consciousness, p. 119 and
"Justice and Solidarity," n. 16, p. 52.
14. DE, p. 65.
1 5. Cf. Charles Taylor, "The Validity of Transcendental Arguments" in Proceedings of
the Aristotelian Society 79 ( 1 978-1979), pp. 151-165.
16. On the logical status of this "transcendental-pragmatic" argument, see DE, pp. 8386.
17. Cf. DE, pp. 77ff.
18. On the concept of communicative action, see Habermas, "Remarks on the Concept
of Communicative Action," in Gottfried Seebass and Raimo Tuomela, eds. , Social Action
(Dordrecht, 1985), pp. 1 5 1-178 and The Theory of Communicative Action Vol. 1, trans.
T. McCarthy (Boston, 1984), pp. 94- 10 1 .
19. Cf. Habermas, Moral Consciousness, p. 130, and this volume, pp. 3 1, 83-84.
20. On the justification of 'U', cf. DE, pp. 86ff. , and William Rehg, "Discourse and the
Moral Point of View : Deriving a Dialogical Principle of Universalization," in Inquiry
34 (1991), pp. 27-48.
21. This construction represents a synthesis of Mead's notion of ideal role-taking and
Peirce's discursive notion of truth. On the notion of ideal role-taking, see Habermas,
"Justice and Solidarity," pp. 38-40.
XXX
22. See the first essay of the present volume, "On the Pragmatic, the Ethical, and the
Moral Employments of Practical Reason . "
23. See, respectively, "Kantian Constructivism i n Moral Theory," journal of Phirosophy
77 (1980), pp. 520-522, and ':Justice as Fairness: Political not Metaphysical," Phirosophy
and Public Affairs 14 (1985), pp. 236-237.
24. For Habermas's views on Rawls, see chapter 2, pp. 25ff., 92ff. This emphasis on
public discourse is a development of a theme already present in his early historical
sociological account of the bourgeois public sphere (now belatedly available in English),
The StnJ.t:tural Transf01'71UJtion of the Public Sphere, trans. T. Burger and F. Lawrence
(Cambridge, Mass., 1989). In it he analyzes the legitimating function of public discus
sion concerning matters of general interest in the bourgeois public sphere which
developed in seventeenth-century England and eighteenth-century France and traces
its internal contradictions and vicissitudes up to its occlusion with the emergence of
the social-welfare state in the nineteenth and twentieth centuries.
25. For present purposes the term neo-Aristotelian is used to designate ethical positions
structured by recognizable successors to the fundamental orientations of Aristotle's
ethics: the central role accorded communally shaped ideals of character and the human
good, the distinction between theory and practice, and the distinctions between praxis
and poiesis and between phronesis and techne. On this use of the term, see Herbert
Schnii.delbach, "What is Neo-Aristotelianism?" Praxis Intemational 7 (1987188), pp. 225237.
26. Cf. After Virtue (Notre Dame, I nd., 1984), especially chapter 2.
27. For a concise statement, see Habermas, Philosophical Discourse, pp. 342-349.
28. Phronesis involves a kind of situational appreciation which Aristotle assimilates to
perception and which does not admit of codification in terms of general rules or
criteria of judgment. On this dimension of Aristotle's account of practical reason, see
David Wiggins, "Deliberation and Practical Reason," in Needs, Values, Truth (Oxford,
1 991), pp. 215-237.
29. Cf. Williams, Ethics, pp. 52-53, 67-69. Such criticisms as Sandel's against Rawls
that contemporary liberalism repeats the error of its classical predecessors in presup
posing "unencumbered" selves do not apply to discourse ethics, which views indivi
duation from the outset as a product of socialization; but far from prejudging the
issue against an ethics and politics of autonomy, Habermas argues, the social embed
dedness of the subject demands that individual autonomy be reconceptualized in
intersubjective terms.
30. Habermas rejects Macintyre's and Williams's criticisms of attempts to derive a
moral principle from the structure of human action as such on the grounds that they
are based on a version of the argument (i.e. Alan Gewirth's) that remains tied to an
individualistic notion of agency and a correspondingly restricted conception of prac
tical reason. Cf. Macintyre, After Virtue, pp. 66ff. and Williams, Ethics, pp. 55ff.
3 1 . See chapter 2, pp. 35ff. In clarifying this distinction he draws on Klaus Gunther's
study Der Sinn fur Angemessenheit (Frankfurt, 1988). For a summary of the argument
of that work, see Gunther, "Impanial Application of Moral and Legal Norms: A
Contribution to Discourse Ethics," in David Rasmussen, ed., Universalism vs. Commu.
nitarianism: Contemporary Debates in Ethics (Cambridge, Mass., 1990), pp. 199-206.
xxxi
32. Cf. Gunther, "Impartial Application," p. 200.

33. Cf. DE, p. 67. Habermas also avoids the narcissistic connotations of the Kantian
concern with purity of motive in moral judgment y incorporating into his basic
principle the discursive examination of the consequences of proposed moral norms,
thereby accommodating valid consequentialist intuitions within a deontological ethical
theory.
34. Cf. DE, p. 1 03 and "Morality and Ethical Life: Does Hegel's Critique of Kant
Apply to Discourse Ethics?" in Habermas, Mural Consciousness, p. 204.
35. On the criticism of ethical intellectualism, see Schnadelbach, "Was ist Neoaristo
telismus ?" in Wolfgang Kuhlmann, ed. , Muralitii.t und Sittlichlleti (Frankfurt, 1986),
pp. 57-59. (This discussion does not appear in the translation cited above, n. 25.)
36. Habermas, Moral Consciousness, p. 207.
37. Cf. Williams, Ethics, pp. 18, 37-38, 1 00-1 0 1 , 174ff. Fredric jameson expresses a
related unease of postmodernists when he (somewhat tendentiously) attributes to
Habermas a "vision of a 'noisefree,' transparent, fully communicational society,'' in his
introduction to jean-Franois Lyotard, The Postmodem Condition (Minneapolis, 1 984),
p. vii.
38. Cf. DE, p. 1 04 and chapter I of this volume.
39. See especially DE, pp. 1 07- 1 09.
40. Cf. Habermas, "Justice and Solidarity," pp. 47ff. As he says in another place, "the
free actualization of the personality of one individual depends on the actualization of

freedom for all," Moral Consciousness, p. 207.
4 1 . Habermas, Theory of Communicative Action Vol. 2, p. 96; cf. also Philosophical Dis
course, pp. 344-345.
1
On the Pragmatic, the Ethical, and the
Moral Employments of Practical Reason
For judith
Contemporary discussions in practical philosophy draw, now as be

fore, on three main sources : Aristotelian ethics, utilitarianism, and
Kantian moral theory. Two of the parties to these interesting debates
also appeal to Hegel who tried to achieve a synthesis of the classical
communal and modern individualistic conceptions of freedom with
his theory of objective spirit and his "sublation" (Aufhebung) of mo
rality into ethical life. Whereas the communitarians appropriate the
Hegelian legacy in the form of an Aristotelian ethics of the good and
abandon the universalism of rational natural law, discourse ethics
takes its orientation for an intersubjective interpretation of the cate
gorical imperative from Hegel's theory of recognition but without
incurring the cost of a historical dissolution of morality in ethical life.
Like Hegel it insists, though in a Kantian spirit, on the internal
relation between justice and solidarity. It attempts to show that the
meaning of the basic principle of morality can be explicated in terms
of the content of the unavoidable presuppositions of an argumenta
tive practice that can be pursued only in common with others. The
moral point of view from which we can judge practical questions
impartially is indeed open to different interpretations. But because
it is grounded in the communicative structure of rational discourse
as such, we cannot simply dispose of it at will. It forces itself intuitively
on anyone who is at all open to this reflective form of communicative
action. With this fundamental assumption, discourse ethics situates
itself squarely in the Kantian tradition yet without leaving itself vul
nerable to the objections with which the abstract ethics of conviction
2
On the Employments of Practical Reason
has met from its inception. Admittedly, it adopts a narrowly circum

scribed conception of morality that focuses on questions of justice.
But it neither has to neglect the calculation of the consequences of
actions rightly emphasized by utilitarianism nor exclude from the
sphere of discursive problematization the questions of the good life
accorded prominence by classical ethics, abandoning them to irra
tional emotional dispositions or decisions. The term discourse ethics
may have occasioned a misunderstanding in this connection. The
theory of discourse relates in different ways to moral, ethical, and
pragmatic questions. It is this differentiation that I propose to clarify
here.
Classical ethics, like modern theories, proceeds from the question
that inevitably forces itself upon an individual in need of orientation
faced with a perplexing practical task in a particular situation: how
should I proceed, what should I do? 1 The meaning of this "should"
remains indeterminate as long as the relevant problem and the aspect
under which it is to be addressed have not been more clearly speci
fied. I will begin by taking the distinction between pragmatic, ethical,
and moral questions as a guide to differentiating the various uses of
practical reason. Different tasks are required of practical reason un
der the aspects of the purposive, the good, and the just. Correspond
ingly, the constellation of reason and volition changes as we move
between pragmatic, ethical, and moral discourses. Finally, once moral
theory breaks out of the investigative horizon of the first-person
singular, it encounters the reality of an alien will, which generates
problems of a different order.
I
Practical problems beset us in a variety of situations. They "have to

be" mastered ; otherwise we suffer consequences that are at very least
annoying. We must decide what to do when the bicycle we use every
day is broken, when we are afflicted with illness, or when we lack the
money necessary to realize certain desires. In such cases we look for
reasons for a rational choice between different available courses of
action in the light of a task that we must accomplish if we want to
achieve a certain goal. The goals themselves can also become prob
lematic, as, for example, when holiday plans fall through or when
we must make a career decision. Whether one travels to Scandinavia

or to Elba or stays at home or whether one goes directly to college
or first does an apprenticeship, becomes a physician or a salesper
son-such things depend in the first instance on our preferences and
on the options open to us in such situations. Once again we seek
reasons for a rational choice but in this case for a choice between the
goals themselves.
In both cases the rational thing to do is determined in part by what
one wants: it is a matter of making a rational choice of means in the
light of fixed purposes or of the rational assessment of goals in the
light of existing preferences. Our will is already fixed as a matter of
fact by our wishes and values; it is open to further determination
only in respect of alternative possible choices of means or specifica
tions of ends. Here we are exclusively concerned with appropriate
techniques-whether for repairing bicycles or treating disease-with
strategies for acquiring money or with programs for planning vaca
tions and choosing occupations. In complex cases decision-making
strategies themselves must be developed ; then reason seeks reassur
ance concerning its own procedure by becoming reflective-for ex
ample, in the form of a theory of rational choice. As long as the
question "What should I do?" has such pragmatic tasks in view,
observations, investigations, comparisons, and assessments under
taken on the basis of empirical data with a view to efficiency or with
the aid of other decision rules are appropriate. Practical reflection
here proceeds within the horizon of purposive rationality, its goal
being to discover appropriate techniques, strategies, or programs. 2 It
leads to recommendations that, in the most straightforward cases,
are expressed in the semantic form of conditional imperatives. Kant
speaks in this connection of rules of skill and of counsels of prudence
and, correspondingly, of technical and pragmatic imperatives. These
relate causes to effects in accordance with value preferences and prior
goal determinations. The imperative meaning they express can be
glossed as that of a relative ought, the corresponding directions for
action specifying what one "ought" or "must" do when faced with a
particular problem if one wants to realize certain values or goals. Of
course, once the values themselves become problematic, the ques
tion "What should I do?" points beyond the horizon of purposive
rationality.
4
In the case of complex decisions-for example, choosing a career

it may transpire that the question is not a pragmatic one at all.
Someone who wants to become a manager of a publishing house
might deliberate as to whether it is more expedient to do an appren
ticeship first or go straight to college; but someone who is not clear
about what he wants to do is in a completely different situation. In
the latter case, the choice of a career or a direction of study is bound
up with one's "inclinations" or interests, what occupation one would
find fulfilling, and so forth. The more radically this question is posed,
the more it becomes a matter of what life one would like to lead, and
that means what kind of person one is and would like to be. When
faced with crucial existential choices, someone who does not know
what he wants to be will ultimately be led to pose the question, "Who
am I, and who would I like to be?" Decisions based on weak or trivial
preferences do not require justification ; no one need give an account
of his preferences in automobiles or sweaters, whether to himself or
anyone else. In the contrasting case, I shall follow Charles Taylor in
using the term strong preferences to designate preferences that concern
not merely contingent dispositions and inclinations but the self-un
derstanding of a person, his character and way of life; they are
inextricably interwoven with each individual's identity.3 This circum
stance not only lends existential decisions their peculiar weight but
also furnishes them with a context in which they both admit and
stand in need of justification. Since Aristotle, important value decisions
have been regarded as clinical questions of the good life. A decision
based on illusions-attaching oneself to the wrong partner or choos
ing the wrong career--<:an lead to a failed life. The exercise of
practical reason directed in this sense to the good and not merely to
the possible and expedient belongs, following classical usage, to the
sphere of ethics.
Strong evaluations are embedded in the context of a particular
self-understanding. How one understands oneself depends not only
on how one describes oneself but also on the ideals toward which one
strives. One's identity is determined simultaneously by how one sees
oneself and how one would like to see oneself, by what one finds
oneself to be and the ideals with reference to which one fashions
oneself and one's life. This existential self-understanding is evaluative
in its core and, like all evaluations, is Janus faced. Two components
5
are interwoven in it: the descriptive component of the ontogenesis

.of the ego and the normative component of the ego-ideal. Hence,
the clarification of one's self-understanding or the clinical reassurance
of one's identity calls for an appropriative form of understanding
the appropriation of one's own life history and the traditions and
circumstances of life that have shaped one's process of development. 4
If illusions are playing a role, this hermeneutic self-understanding
can be raised to the level of a form of reflection that dissolves self
deceptions. Bringing one's life history and its normative context to
awareness in a critical manner does not lead to a value-neutral self
understanding; rather, the hermeneutically generated self-descrip
tion is logically contingent upon a critical relation to self. A more
profound self-understanding alters the attitudes that sustain, or at
least imply, a life project with normative substance. In this way, strong
evaluations can be justified through hermeneutic self-clarification.
One will be able to choose between pursuing a career in manage
ment and training to become a theologian on better grounds after
one has become clear about who one is and who one would like to
be. Ethical questions are generally answered by unconditional imper
atives such as the following: "You must embark on a career that
affords you the assurance that you are helping other people." The
meaning of this imperative can be understood as an "ought" that is
not dependent on subjective purposes and preferences and yet is not
absolute. What you "should" or "must" do has here the sense that it
is "good" for you to act in this way in the long run, all things consid
ered. Aristotle speaks in this connection of paths to the good and
happy life. Strong evaluations take their orientation from a goal
posited absolutely for me, that is, from the highest good of a self
sufficient form of life that has its value in itself.
The meaning of the question "What should I do?" undergoes a
further transformation as soon as my actions affect the interests of
others and lead to conflicts that should be regulated in an impartial
manner, that is, from the moral point of view. A contrasting com
parison will be instructive concerning the new discursive modality
that thereby comes into play. Pragmatic tasks are informed by the
perspective of an agent who takes his preferences and goals as his
point of departure. Moral problems cannot even be conceived from
this point of view because other persons are accorded merely the
6
status of means or limiting conditions for the realization of one's own

individual plan of action. In strategic action, the participants assume
that each decides egocentrically in accordance with his own interests.
Given these premises, there exists from the beginning at least a latent
conflict between adversaries. This can be played out or curbed and
brought under control; it can also be resolved in the mutual interest
of all concerned. But without a radical shift in perspective and atti
tude, an interpersonal conflict cannot be perceived by those involved
as a moral problem. If I can secure a loan only by concealing pertinent
information, then from a pragmatic point of view all that counts is
the probability of my deception's succeeding. Someone who raises
the issue of its permissibility is posing a different kind of question
the moral question of whether we all could will that anyone in my
situation should act in accordance with the same maxim.
Ethical questions by no means call for a complete break with the
egocentric perspective ; in each instance they take their orientation
from the telos of one's own life. From this point of view, other
persons, other life histories, and structures of interests acquire im
portance only to the extent that they are interrelated or interwoven
with my identity, my life history, and my interests within the frame
work of an intersubjectively shared form of life. My development
unfolds against a background of traditions that I share with other
persons; moreover, my identity is shaped by collective identities, and
my life history is embedded in encompassing historical forms of life.
To that extent the life that is good for me also concerns the forms of
life that are common to us. 5 Thus, Aristot:).e viewed the ethos of the
individual as embedded in the polis comprising the citizen body. But
ethical questions point in a different direction from moral questions :
the regulation of interpersonal conflicts of action resulting from op
posed interests is not yet an issue. Whether I would like to be someone
who in a case of acute need would be willing to defraud an anony
mous insurance company just this one time is not a moral question,
for it concerns my self-respect and possibly the respect that others
show me, but not equal respect for all, and hence not the symmetrical
respect that everyone should accord the integrity of all other persons.
We approach the moral outlook once we begin to examine our
maxims as to their compatibility with the maxims of others. By max
ims Kant meant the more or less trivial, situational rules of action by
which an individual customarily regulates his actions. They relieve

the agent of the burden of everyday decision making and fit together
to constitute a more or less consistent life practice in which the agent's
character and way of life are mirrored. What Kant had in mind were
primarily the maxims of an occupationally stratified, early capitalist
society. Maxims constitute in general the smallest units in a network
of operative customs in which the identity and life projects of an
individual (or group) are concretized ; they regulate the course of
daily life, modes of interaction, the ways in which problems are
addressed and conflicts resolved, and so forth. Maxims are the plane
in which ethics and morality intersect because they can be judged
alternately from ethical and moral points of view. The maxim to
allow myself just one trivial deception may not be good for me-for
example, if it does not cohere with the picture of the person who I
would like to be and would like others to acknowledge me to be. The
same maxim may also be unjust if its general observance is not equally
good for all. A mode of examining maxims or a heuristic for gen
erating maxims guided by the question of how I want to live involves
a different exercise of practical reason from reflection on whether
from my perspective a generally observed maxim is suitable to reg
ulate our communal existence. In the first case, what is being asked
is whether a maxim is good for me and is appropriate in the given
situation, and in the second, whether I can will that a maxim should
be followed by everyone as a general law.
The former is a matter for ethical deliberation, the latter for moral
deliberation, though still in a restricted sense, for the outcome of this
deliberation remains bound to the personal perspective of a partic
ular individual. My perspective is structured by my self-understand
ing, and a casual attitude toward deception may be compatible with
my preferred way of life if others behave similarly in comparable
situations and occasionally make me the victim of their manipulations.
Even Hobbes recognizes a golden rule with reference to which such
a maxim could be justified under appropriate circumstances. For him
it is a "natural law" that each should accord everyone else the rights
he demands for himself.6 But an egocentrically conceived universal
izability test does not yet imply that a maxim would be accepted by
all as the moral yardstick of their actions. This would follow only if
my perspective necessarily cohered with that of everyone else. Only
8
if my identity and my life project reflected a universally valid form

of life would what from my perspective is equally good for all in fact
be equally in the interest of alP
A categorical imperative that specifies that a maxim is just only if
all could will that it should be adhered to by everyone in comparable
situations first signals a break with the egocentric character of the
golden rule ("Do not do unto others what you would not have them
do unto you") . Everyone must be able to will that the maxims of our
action should become a universal law. 8 Only a maxim that can be
generalized from the perspective of all affected counts as a norm that
can command general assent and to that extent is worthy of recog
nition or, in other words, is morally binding. The question "What
should I do?" is answered morally with reference to what one ought
to do. Moral commands are categorical or unconditional imperatives
that express valid norms or make implicit reference to them. The
imperative meaning of these commands alone can be understood as
an "ought" that is dependent on neither subjective goals and pref
erences nor on what is for me the absolute goal of a good, successful,
or not-failed life. Rather, what one "should" or "must" do has here
the sense that to act thus is just and therefore a duty.
II
Thus, the question "What should I do?" takes on a pragmatic, an

ethical, or a moral meaning depending on how the problem is con
ceived. In each case it is a matter of justifying choices among alter
native available courses of action, but pragmatic tasks call for a
different kind of action, and the corresponding question, a different
kind of answer, from ethical or moral ones. Value-oriented assessments
of ends and purposive assessments of available means facilitate ra
tional decisions concerning how we must intervene in the objective
world in order to bring about a desired state of affairs. This is
essentially a matter of settling empirical questions and questions of
rational choice, and the terminus ad quem of a corresponding prag
matic discourse is a recommendation concerning a suitable technol
ogy or a realizable program of action. The rational consideration of
an important value decision that affects the whole course of one's life
is quite a different matter. This latter involves hermeneutical clarifi-
cation of an individual's self-understanding and clinical questions of

a happy or not-failed life. The terminus ad quem of a corresponding
ethical-existential discourse is advice concerning the correct conduct
of life and the realization of a personal life project. Moral judgment
of actions and maxims is again something different. It serves to clarify
legitimate behavioral expectations in response to interpersonal con
flicts resulting from the disruption of our orderly coexistence by
conflicts of interests. Here we are concerned with the justification
and application of norms that stipulate reciprocal rights and duties,
and the terminus ad quem of a corresponding moral-practical discourse
is an agreement concerning the just resolution of a conflict in the
realm of norm-regulated action, .
Thus, the pragmatic, ethical, and moral employments of practical
reason have as their respective goals technical and strategic directions
for action, clinical advice, and moral judgments. Practical reason is
the ability to justify corresponding imperatives, where not just the
illocutionary meaning of "must" or "ought" changes with the practical
relation and the kind of decision impending but also the concept of
the will that is supposed to be open to determination by rationally
grounded imperatives in each instance. The "ought" of pragmatic
recommendations relativized to subjective ends and values is tailored
to the arbitrary choice (Willkur) of a subject who makes intelligent
decisions on the basis of contingent attitudes and preferences that
form his point of departure; the faculty of rational choice does not
extend to the interests and value orientations themselves but presup
poses them as given. The "ought" of clinical advice relativized to the
telos of the good life is addressed to the striving for self-realization
and thus to the resoluteness (Entschluflkraft) of an individual who has
committed himself to an authentic life; the capacity for existential
decisions or radical choice of self always operates within the horizon
of a life history, in whose traces the individual can discern who he is
and who he would like to become. The categorical "ought" of moral
injunctions, finally, is directed to the free will (freien Willen), emphat
ically construed, of a person who acts in accordance with self-given
laws ; this will alone is autonomous in the sense that it is completely
open to determination by moral insights. In the sphere of validity of
the moral law, neither contingent disp ositions nor life histories and
personal identities set limits to the determination of the will by prac-
10
O n the Employments o f Practical Reason
tical reason. Only a will that is guided by moral insight, and hence is
completely rational, can be called autonomous. All heteronomous
elements of mere choice or of commitment to an idiosyncratic way
of life, however authentic it may be, have been expunged from such
a will. Kant confused the autonomous will with an omnipotent will
and had to transpose it into the intelligible realm in order to conceive
of it as absolutely determinative. But in the world as we experience
it, the autonomous will is efficacious only to the extent that it can
ensure that the motivational force of good reasons outweighs the
power of other motives. Thus, in the plain language of everyday life,
we call a correctly informed but weak will a "good will."
To summarize, practical reason, according to whether it takes its
orientation from the purposive, the good, or the just, directs itself in
turn to the choice of the purposively acting subject, to the resolute
ness of the authentic, self-realizing subject, or to the free will of the
subject capable of moral judgment. In each instance, the constellation
of reason and volition and the concept of practical reason itself
undergo alteration. Not only the addressee, the will of the agent who
seeks an answer, changes its status with the meaning of the question
"What should I do?" but also the addresser, the capacity of practical
deliberation itself. According to the aspect chosen, there result three
different though complementary interpretations of practical reason.
But in each of the three major philosophical traditions, just one of
these interpretations has been thematized. For Kant practical reason
is coextensive with morality ; only in autonomy do reason (Vernunft)
and the will attain unity. Empiricism assimilates practical reason to
its pragmatic use; in Kantian terminology, it is reduced to the pur
posive exercise of the understanding (Verstand). And in the Aristo
telian tradition, practical reason assumes the role of a faculty of
judgment ( Urteilskraft) that illuminates the life historical horizon of a
customary ethos. In each case a different exercise is attributed to prac
tical reason, as will become apparent when we consider the respective
discourses in which they operate.
III
Pragmatic discourses in which we justify technical and strategic rec

ommendations have a certain affinity with empirical discourses. They
11
serve to relate empirical knowledge to hypothetical goal determina

tions and preferences and to assess the consequences of (imperfectly
informed) choices in the light of underlying maxims. Technical or
strategic recommendations ultimately derive their validity from the
empirical knowledge on which they rest. Their validity does not
depend on whether an addressee decides to adopt their directives.
Pragmatic discourses take their orientation from possible contexts of
application. They are related to the actual volitions of agents only
though subjective goal determinations and preferences. There is no
internal relation between reason and the will. In ethical-existential
discourses, this constellation is altered in such a way that justifications
become rational motives for changes of attitude.
The roles of agent and participant in discourse overlap in such
processes of self-clarification. Someone who wishes to attain clarity
about his life as a whole-to justify important value decisions and to
gain assurance concerning his identity-cannot allow himself to be
represented by someone else in ethical-existential discourse, whether
in his capacity as the one involved or as the one who must weigh
competing claims. Nevertheless, there is room here for discourse
because here too the steps in argumentation should not be idiosyn
cratic but must be comprehensible in intersubjective terms. The in
dividual attains reflective distance from his own life history only
within the horizon of forms of life that he shares with others and
that themselves constitute the context for different individual life
projects. Those who belong to a shared lifeworld are potential par
ticipants who can assume the catalyzing role of impartial critics in
processes of self-clarification. This role can be refined into the ther
apeutic role of an analyst once generalizable clinical knowledge comes
into play. Clinical knowledge of this sort is first generated in such
discourses.9
Self-clarification draws on the context of a specific life history and
leads to evaluative statements about what is good for a particular
person. Such evaluations, which rest on the reconstruction of a con
sciously appropriated life history, have a peculiar semantic status, for
"reconstruction" here signifies not just the descriptive delineation of
a developmental process through which one has become the individ
ual one finds oneself to be; it signifies at the same time a critical
sifting and rearrangement of the elements integrated in such a way
12
On the Employments o f Practical Reason
that one's own past can be accepted in the light of existing possibilities
of action as the developmental history of the person one would like
to be and continue to be in the future. The existential figure of the
"thrown projection" (geworfener Entwurf ) illuminates the Janus-faced
character of the strong evaluations justified by way of a critical ap
propriation of one's own life history. Here genesis and validity can
no longer be separated as they can in the case of technical and
strategic recommendations. Insofar as I recognize what is good for
me, I also already in a certain sense make the advice my own; that is
what it means to make a conscious decision. To the extent that I have
become convinced of the soundness of clinical advice, I have also
already made up my mind to transform my life in the manner sug
gested. On the other hand, my identity is only responsive to-even
at the mercy of-the reflexive pressure of an altered self-understand
ing when it observes the same standards of authenticity as ethical
existential discourse itself. Such a discourse already presupposes, on
the part of the addressee, a striving to live an authentic life or the
suffering of a patient who has become conscious of the "sickness unto
death." In this respect, ethical-existential discourse remains contin
gent on the prior telos of a consciously pursued way of life.
IV
In ethical-existential discourses, reason and the will condition one

another reciprocally, though the latter remains embedded in the life
historical context thematized. Participants in processes of self-clari
fication cannot distance themselves from the life histories and forms
of life in which they actually find themselves. Moral-practical dis
courses, by contrast, require a break with all of the unquestioned
truths of an established, concrete ethical life, in addition to distancing
oneself from the contexts of life with which one's identity is inextric
ably interwoven. The higher-level intersubjectivity characterized by
an intermeshing of the perspective of each with the perspectives of
all is constituted only under the communicative presuppositions of a
universal discourse in which all those possibly affected could take
part and could adopt a hypothetical, argumentative stance toward
the validity claims of norms and modes of action that have become
problematic. This impartial standpoint overcomes the subjectivity of
13
the individual participant's perspective without becoming discon

nected from the performative attitude of the participants. The ob
jectivity of the so-called ideal observer would impede access to the
intuitive knowledge of the lifeworld. Moral-practical discourse rep
resents the ideal extension of each individual communication com
munity from within. 10 In this forum, only those norms proposed that
express a common interest of all affected can win justified assent. To
this extent, discursively justified norms bring to expression simulta
neously both insight into what is equally in the interest of all and a
general will that has absorbed into itself, without repression, the will of
all. Understood in this way, the will determined by moral grounds
does not remain external to argumentative reason; the autonomous
will is completely internal to reason.
Hence, Kant believed that practical reason first completely comes
into its own and becomes coextensive with morality in its role as a
norm-testing court of appeal. Yet the discourse-ethical interpretation
of the categorical imperative we have offered reveals the one-sided
ness of a theory that concentrates exclusively on questions of justifi
cation. Once moral justifications rest on a principle of universalization
constraining participants in discourse to examine whether disputed
norms could command the well-considered assent of all concerned,
detached from practical situations and without regard to current
motives or existing institutions, the problem of how norms, thus
grounded, could ever be applied becomes more acute. 1 1 Valid norms
owe their abstract universality to the fact that they withstand the
universalization test only in a decontextualized form . But in this
abstract formulation, they can be applied without qualification only
to standard situations whose salient features have been integrated
from the outset into the conditional components of the rule as con
ditions of application. Moreover, . every justification of a norm is
necessarily subject to the normal limitations of a finite, historically
situated outlook that is provincial in regard to the future. Hence a
forteriori it cannot already explicitly allow for all of the salient features
that at some time in the future will characterize the constellations of
unforeseen individual cases. For this reason, the application of norms
calls for argumentative clarification in its own right. In this case, the
impartiality of judgment cannot again be secured through a principle
of universalization ; rather, in addressing questions of context-sensi-
14
O n the Employments o f Practical Reason
tive application, practical reason must be informed by a principle of

appropriateness (Angemessenheit). What must be determined here is
which of the norms already accepted as valid is appropriate in a given
case in the light of all the relevant features of the situation conceived
as exhaustively as possible.
Of course, discourses of application, like justificatory discourses,
are a purely cognitive undertaking and as such cannot compensate
for the uncoupling of moral judgment from the concrete motives
that inform actions. Moral commands are valid regardless of whether
the addressee can also summon the resolve to do what is judged to
be right. The autonomy of his will is a function of whether he is
capable of acting from moral insight, but moral insights do not of
themselves lead to autonomous actions. The validity claim we asso
ciate with normative propositions certainly has obligatory force, and
duty, to borrow Kant's terminology, is the affection of the will by the
validity claim of moral commands. That the reasons underlying such
validity claims are not completely ineffectual is shown by the pangs
of conscience that plague us when we act against our better judgment.
Guilt feelings are a palpable indicator of transgressions of duty, but
then they express only the recognition that we lack good reasons to
act otherwise. Thus, feelings of guilt reflect a split within the will itself.
v
The empirical will that has split off from the autonomous will plays
an important role in the dynamics of our moral learning processes. 1 2
The division o f the will is a symptom o f weakness o f will only when
the moral demands against which it transgresses are in fact legitimate
and it is reasonable (zumutbar) to expect adherence to them under the
given circumstances. In the revolt of a dissident will, there all too
often also come to expression, as we know, the voice of the other
who is excluded by rigid moral principles, the violated integrity of
human dignity, recognition refused, interests neglected, and differ
ences denied.
Because the principles of a will that has attained autonomy embody
a claim analogous to that associated with knowledge, validity and
genesis once again diverge here as they do in pragmatic discourse.
Thus, behind the facade of categorical validity may lurk a hidden,
15
entrenched interest that is susceptible only of being pushed through.

This facade can be erected all the more easily because the rightness
of moral commands, unlike the truth of technical or strategic rec
ommendations, does not stand in a contingent relation to the will of
the addressee but is intended to bind the will rationally from within.
Liberating ourselves from the merely presumptive generality of se
lectively employed universalistic principles applied in a context
insensitive manner has always required, and today still requires, social
movements and political struggles; we have to learn from the painful
experiences and the irreparable suffering of those who have been
humiliated, insulted, injured, and brutalized that nobody may be
excluded in the name of moral universalism-neither underprivi
leged classes nor exploited nations, neither domesticated women nor
marginalized minorities. Someone who in the name of universalism
excludes another who has the right to remain alien or other betrays
his own guiding idea. The universalism of equal respect for all and
of solidarity with everything that bears the mark of humanity is first
put to the test by radical freedom in the choice of individual life
histories and particular forms of life.
This reflection already oversteps the boundaries of individual will
formation. Thus far we have examined the pragmatic, ethical, and
moral employments of practical reason, taking as a guide the tradi
tional question, "What should I do?" But with the shift in horizon of
our questions from the first-person singular to the first-person plural,
more changes than just the forum of reflection. Individual will for
mation by its very nature is already guided by public argumentation,
which it simply reproduces in foro interno. Thus, where moral life
runs up against the boundaries of morality, it is not a matter of a
shift in perspective from internal monological thought to public dis
course but of a transformation in the problem at issue ; what changes
is the role in which other subjects are encountered .
Moral-practical discourse detaches itself from the orientation to
personal success and one's own life to which both pragmatic and
ethical reflection remain tied. But norm-testing reason still encoun
ters the other as an opponent in an imaginary-because counterfac
tually extended and virtually enacted-process of argumentation.
Once the other appears as a real individual with his own unsubstitut-
16
able will, new problems arise. This reality of the alien will belongs to
the primary conditions of collective will formation.
The fact of the plurality of agents and the twofold contingency
under which the reality of one will confronts that of another generate
the additional problem of the communal pursuit of collective goals,
and the problem of the regulation of communal existence under the
pressure of social complexity also takes on a new form. Pragmatic
discourses point to the necessity of compromise as soon as one's own
interests have to be brought into harmony with those of others.
Ethical-political discourses have as their goal the clarification of a
collective identity that must leave room for the pursuit of diverse
individual life projects. The problem of the conditions under which
moral commands are reasonable motivates the transition from mo
rality to law. And, finally, the implementation of goals and programs
gives rise to questions of the transfer and neutral exercise of power.
Modern rational natural law responded to this constellation of
problems, but it failed to do justice to the intersubjective nature of
collective will formation, which cannot be correctly construed as in
dividual will formation writ large. Hence, we must renounce the
premises of the philosophy of the subject on which rational natural
law is based. From the perspective of a theory of discourse, the
problem of agreement among parties whose wills and interests clash
is shifted to the plane of institutionalized procedures and commu
nicative presuppositions of processes of argumentation and negotia
tion that must be actually carried out. 13
It is only at the level of a discourse theory of law and politics that
we can also expect an answer to the question invited by our analyses:
Can we still speak of practical reason in the singular after it has
dissolved into three different forms of argumentation under the
aspects of the purposive, the good, and the right? All of these forms
of argument are indeed related to the wills of possible agents, but as
we have seen, concepts of the will change with the type of question
and answer entertained. The unity of practical reason can no longer
be grounded in the unity of moral argumentation in accordance with
the Kantian model of the unity of transcendental consciousness, for
there is no metadiscourse on which we could fall back to justify the
choice between different forms of argumentation. 14 Is the issue of
whether we wish to address a given problem under the standpoint
17
of the purposive, the good, or the just not then left to the arbitrary
choice, or at best the prediscursive judgment, of the individual?
Recourse to a faculty of judgment that "grasps" whether a problem
is aesthetic rather than economic, theoretical rather than practical,
ethical rather than moral, political rather than legal, must remain
suspect for anyone who agrees that Kant had good grounds for
abandoning the Aristotelian concept of judgment. In any case, it is
not the faculty of reflective judgment, which subsumes particular
cases under general rules, that is relevant here but an aptitude for
discriminating problems into different kinds.
As Peirce and the pragmatists correctly emphasize, real problems
are always rooted in something objective. The problems we confront
thrust themselves upon us; they have a situation-defining power and
engage our minds with their own logics. Nevertheless, if each prob
lem followed a unique logic of its own that had nothing to do with
the logic of the next problem, our minds would be led in a new
direction by every new kind of problem. A practical reason that saw
its unity only in the blind spot of such a reactive faculty of judgment
would remain an opaque construction comprehensible only in phe
nomenological terms.
Moral theory must bequeath this question unanswered to the phi
losophy of law ; the unity of practical reason can be realized in an
unequivocal manner only within a network of public forms of com
munication and practices in which the conditions of rational collective
will formation have taken on concrete institutional form.
2
Discourse ethics has met with objections directed, on the one hand,
against deontological theories generally and, on the other, against
the particular project of offering an explication of the moral point
of view in terms of universal communicative presuppositions of ar
gumentation . Here I here take up some of these objections and
discuss them in a metacritical fashion by way of explicating once
again, though in an unsystematic fashion, the theoretical program I
share, in its essentials, with Karl-Otto Apel.
In the following sections I shall refer to theses of Bernard Williams,
John Rawls, Albrecht Wellmer, Klaus Gunther, Ernst Tugendhat,
Stephen Lukes, Charles Fried, Charles Taylor, Apel, Thomas Mc
Carthy, Alasdair Macintyre, and Gunther Patzig, discussing in succes
sion the following topics :
1. The relation between theoretical and practical reason.

2. Similarities and differences between the mode of validity of truth
claims and that of norms.
3 . Some interrelations between rationality and morality.
4. The relation between the justification and the application of

norms.
5. The relation between the validity of norms, sanctions, and self
respect.
6. The discourse-ethical interpretation of the moral point of view.
20
7. The role of idealizations in this explication of the moral point of
vtew.
8. The distinction between negative and positive rights and duties.
9. The attempt to develop a postmetaphysical ethics of the good .
1 0 . The meaning of "ultimate justifications" in moral theory.

1 1 . The primacy of the right over the good.
1 2. The relation between tradition and modernity constitutive for
the concept of "postconventional moral consciousness."
1 3. The challenge posed by an ecological ethics for an anthropocen
tric conception.
1 . The cognitivism of Kantian ethics has repeatedly met with the
incomprehension of those who judge practical reason by the stan
dards of what Kant called the understanding. Thus, empiricism dis
putes whether moral questions can even be decided in a rational
manner. Normal language use-so runs a prima facie plausible objec
tion-should already make the cognitivist wary: when we act immor
ally, we are not necessarily behaving irrationally. This is indeed
indisputable if we understand "rational" in terms of intelligent, prag
matically astute, and, hence, purposively rational action. But then, of
course, our way of using language can no longer serve as an unbiased
witness, since it is already informed by an outlook that limits the
rational to the sphere of purposive action. Certainly we cannot simply
assimilate moral insight to epistemic knowledge without further ado,
for the former tells us what we ought to do, whereas we only know
something, strictly speaking, when we know how things stand in the
world. Practical questions do not seem to admit of theoretical treat
ment. In fact, our everyday moral intuitions neither depend on an
ethical theory nor can they in the normal course of events derive
much benefit from one. But it does not follow that intuitively mas
tered everyday knowledge is not knowledge at all. On the contrary,
our practices of criticizing immoral actions and of disputing moral
questions by appealing to reasons suggest rather that we associate a
cognitive claim with moral judgments. Kant too shows no small re
gard for the "moral knowledge of common human reason" and is
21
Remarks o n Discourse Ethics
cognizant of the fact "that neither science nor philosophy is needed

in order to know what one has to do." 1
It must be asked, therefore, whether, in the light of the "moral
knowledge of common human reason," moral theory is not itself
subject to narrow constraints. Moral judgments give authors like
Bernard Williams occasion to reflect on the "limits of philosophy."2
They acknowledge that moral reflection is indeed cognitive in char
acter but only in the weak sense of a reflective confirmation of the
familiar conditions under which we live or would like to live. This
corresponds to an Aristotelian approach that views practical reason
as limited essentially to ethical self-understanding and consequently
to the sphere of the good. Aristotle advocated the thesis that expres
sions such as "moral judgment" and "moral justification" have a
specific, nonempirical meaning and held that ethics is not a matter
of knowledge in the strict sense but of practical deliberation.
Aristotle defined this faculty of phronesis (prudentia, "prudence") in
a negative fashion in contrast to the strong claims of episteme-the
faculty of knowledge concerned with the universal, necessary, and
supratemporal dimension of existence and, ultimately, of the cos
mos-but without completely denying its cognitive status. However,
modern Aristotelians can no longer uncritically appeal to such a faculty
of metaphysical knowledge as a point of contrast. The fallible con
ception of knowledge that informs the sciences involves the renun
ciation of all metaphysical aspirations, and it is not clear that
significant modifications could still be made to this weak, postmeta
physical conception of knowledge without jeopardizing its funda
mental cognitive status. On the other hand, the theoretical knowledge
secured by the modern empirical sciences can no longer be employed
in genuinely practical contexts; at best, it permits calculations of
means and ends (technical and strategic recommendations) that are
indifferent to moral concerns. On these premises it becomes ques
tionable whether our everyday ethical knowledge can be viewed as
genuine knowledge.
Modern Aristotelians can circumvent this difficulty by appealing
to the distinction between naive, contextual, everyday knowledge, on
the one hand, and generalized, theoretical, reflective knowledge, on
the other. Williams expounds the thesis that we can speak of ethical
as well as scientific knowledge, because the former enables us to orient
22
ourselves in the social world, just as the latter enables us to orient

ourselves in the objective world of things and events. Ethical knowl
edge retains its capacity to provide orientation, however, only within
the horizon of the established everyday practice of individuals so
cialized into a specific culture, whereas empirical knowledge becomes
prey to illusions precisely in everyday contexts and can be shown to
be universally valid factual knowledge only from the detached per
spective of scientific reflection . The empirical sciences adopt a critical
attitude toward the kind of everyday intuitions on which we imme
diately rely in our moral judgments. On the other hand, we would
destroy our ethical knowledge by submitting it to scientific exami
nation, because theoretical objectification would dislodge it from its
proper place in our life.
Williams recognizes that Aristotelian reflections such as these lead
us up a blind alley. Modern life is characterized by a plurality of
forms of life and rival value convictions. For this reason-and not on
account of the empty misgivings of moral theorists-the traditional,
established knowledge of concrete ethical life is drawn into a dynamic
of problematization that no one today can elude. This awareness of
contingency also pervades ethical knowledge and compels it to reflect
upon itself: "the urge to reflective understanding of society and our
activities goes deeper and is more widely spread in modern society
than it has ever been before . . . . There is no route back from reflec
tiveness."3 In view of this situation, the attempt to shield traditional
powers and institutions from the pressure of reflection, in the manner
of an Arnold Gehlen, is hopelessly reactionary. Equally implausible,
on the other hand, is the decisionistic attempt to evade the growing
contingency besetting value convictions by making certainty a func
tion of pure decision. Like other noncognitivist proposals, decision
ism is counterintuitive, for a moment of passivity always attaches to
convictions, which take shape gradually and are not produced by us
like decisions. And, finally, if we do not cynically reject the phenom
ena as they force themselves upon us from the participant perspective
in favor of a relativism informed by the observer perspective, and if
we refuse to follow Nietzsche and the historicists in simply repudiat
ing the clear language of our moral feelings, then we are faced with
an acute dilemma: How can we appropriate naive, everyday ethical
knowledge in a critical fashion without at the same time destroying
23
it through theoretical objectification? How can ethical knowledge

become reflective from the perspective of the participants
themselves?
The answer Williams offers points in the direction of ethical self
reflection. just as an individual can reflect on himself and his life as
a whole with the goal of clarifying who he is and who he would like
to be, so too the members of a collectivity can engage in public
deliberation in a spirit of mutual trust, with the goal of coming to an
understanding concerning their shared form of life and their identity
solely through the unforced force of the better argument. In such
ethical-political discourses, as I propose to call them, participants can
clarify who they are and who they want to be, whether as members
of a family, as inhabitants of a region, or as citizens of a state. The
strong evaluations that shape the self-understanding of the person
or of the community as a whole are here up for discussion. An
individual life history or an intersubjectively shared form of life is
the horizon within which participants can critically appropriate their
past with a view to existing possibilities of action. Such processes of
self-understanding lead to conscious decisions that are judged ac
cording to the standard of an authentic way of life. Insofar as any
kind of theoretical knowledge can be of any help in these processes,
it is generalized therapeutic knowledge rather than philosophical
knowledge: " How truthfulness to an existing self or society is to be
combined with reflection, self-understanding, and criticism is a ques
tion that philosophy itself cannot answer. It is the kind of question
that has to be answered through reflective living. The answer has to
be discovered, or established, as a result of a process, personal or
social, which essentially cannot formulate the answer in advance,
except in an unspecific way. Philosophy can play a part in the process,
as it plays a part in identifying the question, but it cannot be a
substitute for it."4 Philosophy can at best clarify the most general
features of ethical self-reflection and the form of communication
appropriate to it.5
But if this is the task that Williams assigns to philosophy, then
philosophy must also be in a position to differentiate specifically
moral questions from ethical ones and to give them their proper due.
Williams does accord moral questions in a narrower sense-those
dealing with rights and duties-a special status and even a certain
24
urgency, but his differentiations are not sufficiently incisive. He does

not make clear that morality is not oriented to the telos of a successful
life with a view to answering the question, "Who am I , (or who we
are) and who would I (or we) like to be?" Rather, it is concerned with
the categorially different question of the norms according to which
we want to live together and of how practical conflicts can be settled
in the common interest of all. The peculiarly moral problematic
detaches itself from the egocentric (or ethnocentric) perspective of
each individual's (or our) way of life and demands that interpersonal
conflicts be j udged from the standpoint of what all could will in
common. A moral theory can accomplish no less with this question
than in Williams's view it is supposed to accomplish in the ethical
case: clarification of the conditions under which the participants
could find a rational answer for themselves. In the Kantian tradition,
this is called the explication of the moral point of view, that is, a point
of view that permits the impartial treatment of questions of justice.
In moral argumentation, as in the case of ethical discourse, it must
be left to the participants themselves to find concrete answers in
particular cases ; it cannot be known in advance. Moral questions, like
ethical questions, must be addressed from the perspective of the
participants if the questions and answers are not to be robbed of
their normative substance and their binding force. For both dis
courses, the proposition holds equally: "If the agreement were to be
uncoerced, it would have to grow from inside human life."6
The moral point of view, however, requires that maxims and con
tested interests be generalized, which compels the participants to
transcend the social and historical context of their particular form of
life and particular community and adopt the perspective of all those
possibly affected. This exercise of abstraction explodes the culture
specific lifeworld horizon within which processes of ethical self-un
derstanding take place. Furthermore, it places the neo-Aristotelian
demarcation of theoretical from everyday practical knowledge in
question for a second time. It was already recognized that ehical
knowledge had cast off the naivete of everyday knowledge and at
tained reflective status. But moral knowledge that raises a claim to
universal validity must in addition detach itself from the contexts in
which ethical knowledge remains embedded (though with the quali
fications to which all discursive knowledge is subject).
25
This step is incompatible with Williams's paradoxical attempt to

accord practical knowledge a status that divorces it from strict knowl
edge on an analogy with the way in which phronesis was once divorced
from episteme. Today all discursive knowledge is taken to be fallible
and more or less context dependent, more or less general, more or
less rigorous ; correlatively, it is not just the nomological knowledge
of the objectifying empirical sciences that raises a claim to universal
validity. Logic, mathematics, and grammar are also sciences that re
construct the intuitive knowledge of competent judging and speaking
subjects. In an analogous fashion, moral theory engages in a task of
rational reconstruction when it elicits from everyday moral intuitions
the standpoint of the impartial judgment of interpersonal practical
conflicts. In this reflection, of course, it cannot abandon the perfor
mative attitude of participants in interaction ; only in this way can it
maintain contact with the intuitive knowledge acquired through so
cialization that makes moral judgments possible. To this extent, the
connection to the pretheoretical knowledge of everyday life remains
intact. Williams does not allow for this possibility because he remains
committed to an empirically truncated concept of theoretical knowl
edge: "I do not believe, then, that we can understand reflection as a
process that substitutes knowledge for beliefs attained in unreflective
practice. We must reject the objectivist view of ethical life as in that
way a pursuit of ethical truth. "7 Williams fails to recognize that theory
does not necessarily take the form of objectifying knowledge that
explains everyday knowledge in terms of prior dispositions instead
of reconstructing it in terms of the underlying generative knowledge
of the participants.
2. With his method of "reflective equilibrium," John Rawls has de
veloped just such a reconstructive theory of morality and justice that
takes its orientation from everyday situations.8 He also addresses the
question of the relation between theoretical and practical reason. He
wants to justify principles of justice, though he understands his jus
tification in constructivist rather than strictly empirical terms. He
develops a contract theory of the validity of moral commands because
in this way he can bracket the question of "moral truth" and avoid
committing himself to either realism or subjectivism concerning val
ues. In his view, these are the only alternatives because he regards
26
truth exclusively as a property of assertoric propositions. Proposi

tional truth concerns the existence of states of affairs; assertoric
propositions say what is the case. But if the meaning of the assertoric
mode were the only model in terms of which we could interpret the
meaning of normative propositions, and thus the validity of "moral
truths," a cognitivist interpretation of morality would present us with
a choice between two equally counterintuitive interpretations. Either
we would have to accept something like moral facts and understand
"moral truth" in the sense of a correspondence theory of truth, as the
conformity of propositions with an antecedent realm of value objects
that is ultimately independent of the self-understanding and the
needs of agents, or we would have to deny that normative proposi
tions can be true or false at all and hold that behind the apparent
validity claim to moral truth there is concealed something purely
subjective-feelings, attitudes, or decisions that we attribute to our
selves. The former contradicts the grammatical intuition that we can
express neither the existence of things nor their actual configurations
by means of normative propositions. The latter alternative contradicts
another grammatical intuition : that we do not merely express what
we feel, wish, intend, or prefer by means of normative propositions.
Rawls rightly regards this alternative as unacceptable because, while
moral commands, unlike constative utterances, do not relate to any
thing in the objective world, yet like them they have something ob
jective in view. What ought to be is neither an entity nor a mere
experience.
In an attempt to escape this alternative Rawls brings, in addition
to the objective and subjective worlds, the concept of a social world
into play-a world produced by the actors themselves but in accor
dance with standards that are not at their disposition and that, in a
similar though less rigid manner to the existence of states of affairs,
are independent of them : "What justifies a conception of justice is
not its being true to an order antecedent to and given to us, but its
congruence with our deeper understanding of ourselves and our
aspirations, and our realization that, given our history and the tra
ditions embedded in our public life, it is the most reasonable doctrine
for us. We can find no better basic charter for our social world.
Kantian constructivism holds that moral objectivity is to be under
stood in terms of a suitably constructed social point of view that all
27
can accept. Apart from the procedure of constructing the principles

of justice, there are no moral facts."9 Both moments-the moment
of passivity inscribed in reason and the moment of activity attribut
able to the will-must be such that they can be related to one another
in the concept of a procedural morality. We do not determine the
procedure through which norms can be judged and accepted as
valid-it imposes itself upon us; at the same time, the procedural
practice performs the function of generation or construction no less
than that of discovery, that is, of moral cognition of the principles of
a correctly regulated communal life. This procedure admits of dif
ferent characterizations and takes on a different meaning as we high
light one or the other moment of the procedural practice. If the
procedure is interpreted on the model of an agreement between
private contracting subjects, the moment of voluntary construction
comes to the fore, whereas the model of argumentation oriented to
justification suggests an overhasty assimilation of moral cognition to
forms of knowledge.
Rawls opts for the model of the social contract and develops a
constructivist account of the rational production of principles of jus
tice: "It recasts ideas from the tradition of the social contract to
achieve a practicable conception of objectivity and justification
founded on public agreement in judgment on due reflection. The
aim is free agreement, reconciliation through public reason." 10 Like
many earlier formulations familiar from A Theory ofJustice, this state
ment is in an interesting way ambiguous. Rawls shares this ambiguity
with the tradition of rational natural law in which the justification of
principles of natural law took on a different meaning according to
the manner in which the autonomy of the contracting parties was
understood. Parties such as those envisaged by Hobbes who are
equipped only with freedom of choice can justify their contractual
agreements exclusively on purposive grounds, with the result that
their reasons remain tied to the contingent interests and preferences
of participants. The agreement that they reach is, in accordance with
the model of civil law, essentially an act of will of subjects who possess
power. Parties such as those of Kant, by contrast, who are equipped
with freedom of will must justify their contractual agreements from
the moral point of view-and thus by recourse to the moral law
with the result that their reasons become independent from the
28
egocentric perspectives of participants and are bound up with the

discovery of norms that admit of general assent and the shared
interests that underlie them. The agreement in this case rests on the
insight of morally judging subjects into what they could all will in
common.
In A Theory of justice Rawls still mistakenly took the two readings
to be compatible but later explicitly adopted the Kantian reading. 1 1
I n fact, he had already integrated the determinations of practical
reason into the procedure of will formation under the guise of the
specific limitations to which the parties in the original position are
subject.
Nevertheless, Rawls fails to distance himself from the voluntaristic
implications of a pure contractualist model for the justification of
principles of justice. Since the latter are constructed rather than
discovered, the corresponding procedure cannot be understood ep
istemologically as a procedure for discovering truth. Rawls does not
merely differentiate the procedure of rational will formation from
theoretical cognition but goes so far as to dissociate it from processes
of belief formation oriented to truth in a way similar to that in which
neo-Aristotelians dissociate prudence or practical deliberation from
knowledge as such. What still sets him apart from the neo-Aristote
lians is a stronger Kantian concept of practical reason ; but on his
present conception, this should no longer be introduced as a proce
dure of rational will formation. The proposed procedure no longer
owes its rationality directly to the idealized conditions of a commu
nicative practice that makes agreement in the sense of rationally
motivated assent possible, as was still the case in the A Theory ofjustice.
Rather, this procedure is now supposed to derive its rationality from
the rational capacities of the participants. As a consequence, the
concept of a person now bears the full explanatory weight in dem
onstrating the normative content of practical reason. Everyday moral
intuitions presuppose the existence of persons who are so constituted
that they possess a sense of justice, form conceptions of the good,
regard themselves as sources of legitimate claims, and accept the
conditions of fair cooperation. In short, the theoretical problem of
justification is shifted from characteristics of procedures to qualities
of persons. But since a substantive normative concept of the person
cannot be justified straightforwardly in anthropological terms, Rawls
29
in his more recent publications is of two minds as to whether he

should give up the claim to moral-theoretical justification in favor of
a political ethics. He is now widely interpreted as attempting to found
his postmetaphysical, political concept of justice on the self-under
standing of a particular political tradition: the two-hundred-year-old
American tradition of the constitutional state. 12
Regardless of where one stands on this question, a neo-Aristotelian
retreat from the strong claims to justification of a Kantian theory of
justice would be consistent with Rawls's fear of an epistemological
assimilation of practical to theoretical reason. But his apprehension
becomes groundless once we dissociate the idea of a rationally
grounded consensus from a mistaken concept of truth. 13 Here I
cannot go into the difficulties of the correspondence theory of truth
that have been repeatedly raised since Peirce. But if we understand
propositional truth as a claim raised in constative speech acts that can
be redeemed discursively only under the exacting communicative
presuppositions of argumentation, the claim to rightness raised in
regulative speech acts, which is analogous to the claim to truth, can
be freed from assumptions concerning correspondence. The concept
of a validity claim is of a higher level of generality and leaves open
the possibility of specifying a number of different validity claims. A
validity claim says that the conditions of validity of an utterance-be
it an assertion or a moral command-are satisfied, something that
cannot be shown by direct appeal to decisive evidence but only
through discursive redemption of the claim to propositional truth or
normative rightness. The conditions of validity that are not directly
accessible are interpreted in terms of reasons that can be advanced
in discourse, and the kinds of reasons relevant to discursive redemp
tion of a validity claim cast light on the specific meaning of the validity
claim raised in a given instance. Just as the assertoric mode of utter
ance can be explicated in terms of the existence of the states of affairs
asserted, so too the deontological mode can be explicated in terms of
the actions enjoined being equally in the interest of all possibly
affected.
Moreover, this interpretation of the notion of validity in terms of
the logic of argumentation finds support in epistemological consid
erations. The epistemological view with which the proposed theory
of validity claims accords best is undoubtedly a constructivist one, but
30
Remarks
on
Discourse Ethics
this constructivism applies equally to practical and to theoretical rea

son. The objectifying knowledge of the empirical sciences is also
contingent on the constitutive and meaning-disclosing accomplish
ments of the expert community of researchers; such accomplishments
are by no means the prerogative of the public communication com
munity of citizens. Pragmatism, genetic structuralism, and episte
mological anthropology have highlighted in their respective ways the
phenomenon described in an ontological fashion by Heidegger as
"being-ahead-of-oneself" in a "thrown projection." The anticipatory
character of understanding is universal ; the moments of projection
and discovery complement each other in aU cognitive activities. In
this connection, Peirce, Piaget, and Merleau-Ponty can appeal to
Kant, Marx, and Nietzsche. The constellations of elements do indeed
vary; at one time, the passive moment of experience through which
the world acts upon us predominates, and at another the active
moment of an anticipation of possible effects upon us ; but both
moments, those of discovery and construction, intermesh, and the
relative proportions vary already within the sphere of theoretical
reason. From physics to morality, from mathematics to art criticism,
our cognitive accomplishments form a continuum within the com
mon, though shifting, terrain of argumentation in which validity
claims are thematized.
3. Empiricist objections to cognitivist approaches in moral theory can
be explained in part as a reflection of restricted concepts of knowl
edge, rationality, and truth that are oriented to the modern empirical
sciences and eliminate practical reason in the Kantian sense. From
this perspective, moral judgments are assimilated to either feelings,
attitudes, or decisions or to strong evaluations resulting from pro
cesses of self-clarification. 14 A different kind of objection is directed
against the specific justification strategy of discourse ethics: grounding
the moral principle in the normative content of our practice of ar
gumentation. Albrecht Wellmer maintains that moral obligations can
not be derived from such implicitly presupposed conditions of
rationality: "Obligations to rationality are concerned with arguments
regardless of who voices them, whereas moral obligations are con
cerned with people regardless of their arguments." 1 5 Although this
formulation is striking, it can count as an objection only if one at-
31
tributes a mistaken premise and a false conclusion to discourse

ethics. 16
Anyone who seriously engages in argumentation must presuppose
that the context of discussion guarantees in principle freedom of
access, equal rights to participate, truthfulness on the part of partic
ipants, absence of coercion in adopting positions, and so on. If the
participants genuinely want to convince one another, they must make
the pragmatic assumption that they allow their "yes" and "no" re
sponses to be influenced solely by the force of the better argument.
This must be distinguished from the institutional arrangements that
obligate specific groups of people to engage in argumentation, and
consequently to accept the rationality assumptions alluded to, in ad
dressing certain topics and on certain occasions-for example, in
university seminars, in court, or in parliamentary hearings. One could
concur with Wellmer in holding that such institutions impose "obli
gations to rationality," since norms alone-here the norms through
which discourses are institutionalized-<an ground obligations to be
have in a more or less rational fashion. But Wellmer blurs an impor
tant distinction. The general pragmatic presuppositions that must
always be made by participants when they enter into argumentation,
whether institutionalized or not, do not have the character of practical
obligations at all but that of transcendental constraints. Even prior
to institutionalization, argumentation leaves participants without a
choice; just in virtue of undertaking to engage in such a practice as
such, they must accept certain idealizations in the form of presup
positions of communication.
The latter have "normative" content in a broad sense that cannot be
equated with the obligatory force of norms of interaction. Presup
positions of communication do not have regulative force even when
they point beyond actually existing conditions in an idealizing fash
ion . Rather, as anticipatory suppositions they are constitutive of a
practice that without them could not function and would degenerate
at the very least into a surreptitious form of strategic action. Pre
suppositions of rationality do not impose obligations to act rationally;
they make possible the practice that participants understand as
argumentation.
The program of justification pursued by discourse ethic s sets itself
the task of deriving from suppositions of rationality of this kind a
32
rule of argumentation for discourses in which moral norms can be

justified. It attempts to show that moral questions can be decided
rationally as a general rule. Among the premises of such a "deriva
tion," moreover, belong not only the suppositions of rational argu
mentation as such (expressed in the form of rules) but also a more
detailed specification of what we intuitively appeal to when we wish
to justify a moral action or an underlying norm. Knowing what us
tification" signifies in this context is not of itself to prejudge the
further question of whether moral justifications and justificatory dis
courses are indeed possible. This further issue can be resolved only
by specifying a rule of argumentation that can perform a role in
practical discourse similar to, for example, that played by the prin
ciple of induction in empirical-theoretical discourses.
The controversies concerning assertions have made clear what jus
tifications consist in and are generally supposed to accomplish. They
resolve disputes about facts-disputes, that is, concerning the
truth of corresponding assertoric propositions-through arguments
and thereby lead to argumentatively achieved consensus. Further
more, everyday life teaches us what disputes concerning the rightness
of normative sentences involve. We have an intuitive mastery of the
language game of norm-guided action in which agents adhere to or
deviate from rules while possessing rights and duties that can clash
with one other and lead to practical conflicts understood in normative
terms. Thus, we are also aware that moral justifications resolve dis
putes concerning rights and duties, that is, concerning the rightness
of the corresponding normative statements. If this is the (weak) sense
of normative justification17 and if anyone who engages in a corre
sponding argumentative praxis must make idealizing presuppositions
of the sort indicated, then it follows from the normative content of
these suppositions of rationality (openness, equal rights, truthfulness
and absence of coercion) that, insofar as one's sole aim is to justify
norms, one must accept procedural conditions that implicitly amount
to the recognition of a rule of argumentation, (U) : "Every valid norm
must satisfy the condition that the consequences and side effects its
general observance can be anticipated to have for the satisfaction of
the interests of each could be freely accepted by all affected (and be
preferred to those of known alternative possibilities for
regulation)." 1 8
33
This moral principle yields a precise specification of the validity

claim that attaches to obligatory norms of interaction. The obligatory
character of justified norms involves the notion that they regulate
problems of communal life in the common interest and thus are
"equally good" for all affected . For this reason, moral obligations
relate, on the one hand, to "persons regardless of their arguments,"
if by this one understands "without taking into account egocentric
convictions that may be bound up with generally valid arguments
from the perspective of individual persons." On the other hand, the
moral principle owes its rigorously universalistic character precisely
to the assumption that arguments deserve equal consideration re
gardless of their origin and, hence, also "regardless of who voices
them."
Furthermore, the opposition between rationality and morality Well
mer presents gains superficial plausibility from the erroneous as
sumption that cognitivist ethical positions assert or are committed to
asserting that moral insight is already a sufficient motive for moral
action. But it is part of the cognitivist understanding of morality that
justified moral commands and corresponding moral insights only
have the weak motivating force of good reasons. No direct action
regulating force outside the context of argumentation may (or need)
be ascribed to the "normative" content of presuppositions of argu
mentation that cannot be denied without falling into a performative
contradiction or to the moral principle based upon them. The moral
principle performs the role of a rule of argumentation only for
justifying moral judgments and as such can neither obligate one to
engage in moral argumentation nor motivate one to act on moral
insights. A valid moral judgment does indeed signify in addition an
obligation to act accordingly, and to this extent every normative
validity claim has rationally motivating force grounded in reasons.
Hence, for Kant too, only a will determined by moral insight counts
as autonomous. But insight is compatible with weakness of will. With
out the support of complementary processes of socialization and
structures of identity, without a background of complementary insti
tutions and normative contexts, a moral judgment that is accepted as
valid can establish only one thing: that the insightful addressee then
knows he has no good reason to act otherwise. The weak motivating
force of moral insights is manifested empirically in the fact that
34
someone who acts against his better judgment must not only face the
moral rebukes of others but is also prey to self-criticism, and thus to
"bad conscience." Hence Wellmer is simply asserting a consequence
of the cognitivist understanding of morality, and not making an
objection against discourse ethics, when he asserts "that the effective
ness of moral arguments remains dependent on preconditions which
are not only cognitive, but also affective in nature. . . . A rational
equivalent to a moral agreement supported by sacred or religious
authority is only possible in so far as a successful adaptation to con
ditions of mutual recognition between persons-in both cognitive and
affective terms-has taken place. " 19
The uncoupling of moral judgment from moral action may initially
appear counterintuitive because judgments of obligations, like asser
toric judgments, are associated with an unconditional validity claim.
We say that moral commands are "right" or "wrong" and understand
this in a sense analogous to truth. It is no coincidence that we speak
of "moral truths" to express the categorical character of normative
validity, but with this validity claim, reason affects a will whose con
tingency consists in its ability to choose to act differently. A will that
lets itself be bound by moral insight, though it could choose otherwise,
is autonomous. Kant mistakenly identified this quality with the act of
liberating the will from all empirical motives. This residuum of Pla
tonism disappears once we abandon the idealistic conception of the
catharsis of a will purging itself of all earthly impurities. Then the
autonomous will is not eo ipso a repressive will that suppresses incli
nations in favor of duties.
Since Schiller, the rigidity of the Kantian ethics of duty has been
repeatedly and rightly criticized. But autonomy can be reasonably
expected (zumutbar) only in social contexts that are already themselves
rational in the sense that they ensure that action motivated by good
reasons will not of necessity conflict with one's own interests. The
validity of moral commands is subject to the condition that they are
universally adhered to as the basis for a general practice. Only when
this condition is satisfied do they express what all could will. Only
then are moral commands in the common interest and-precisely
because they are equally good for all-do not impose supererogatory
demands. To this extent rational morality puts its seal on the abolition
of the victim. At the same time, someone who obeys the Christian
35
commandment to love one's neighbor, for example, and makes sac

rifices that could not reasonably be morally required of him, is de
serving of our moral admiration. Supererogatory acts can be
understood as attempts to counteract the effects of unjust suffering
in cases of tragic complication or under barbarous living conditions
that inspire our moral indignation.
4. Wellmer touches on another aspect of the relation between ration
ality and morality in his objection regarding the alleged inapplicability

of the principle of universalization in the form proposed by discourse
ethics. Universalism seems to overtax the limited capacities of our
rational faculty and to necessitate the operations of a divine intellect.
If we understand the fundamental moral question "What ought I (or
we) to do?" immediately as a concrete question that arises for me (or
for us) in a context-dependent manner in a determinate situation, it
is indeed unclear how the application of the rule of argumentation
(U) could lead to an unambiguous solution. Wellmer assumes that
we wish to determine "what is the right way of acting under the given
circumstances" by directly addressing the particular case and that we
propose to answer this question by appeal to a corresponding singular
command by means of a discursively generated operation of gener
alization. Then it must be acknowledged that "this increases enor
mously the difficulty of the task of determining the consequences
and side-effects of a universal observance of norms for each individual
and, beyond that, of finding out whether all would be able to accept
without coercion these consequences and side effects, as they would
arise for each individual."20
This characterization, however, misrepresents the role of the prin
ciple of universalization in the logic of argumentation, which is solely
that of j ustifying generalized behavioral expectations or modes of
action, that is, ofjustifying the norms that underlie a general practice.
(U) belongs properly to justificatory discourses in which we test the
validity of universal precepts (or their simple or double negations
prohibitions and permissions). Since Kant neglects the problem of
application, his formulations may suggest another view, or at least a
misunderstanding of his view. Discourse ethics has learned from this
and makes a careful distinction between the validity-{)r justice-of
36
norms and the correctness of singular judgments that prescribe some

particular action on the basis of a valid norm. Analytically, "the right
thing to do in the given circumstances" cannot be decided by a single
act of justificationr within the boundaries of a single kind of
argumentation-but calls for a two-stage process of argument con
sisting of justification followed by application of norms.
Klaus Gunthe r has drawn on this conclusion to rebut Wellmer's
objection convincingly.2 1 Moral rules claim validity for an abstract
state of affairs, for a way of regulating some practical matter. But
the meaning of the validity claim in question can be differentiated in
two ways: in terms of the rationally motivated assent of all potentially
affected that a valid norm earns and in terms of the totality of possible
situations to which the norm capable of commanding assent in this
manner can be applied : "Does not recognizing a norm as valid for
each participant in discourse mean that he regards its observance in
all situations in which the norm is applicable as appropriate?"22 Hence
the idea of impartiality, which is expressed in the moral point of view
and gives determinate meaning to the validity claim of moral judg
ments, demands that we take into account a norm's rational accep
tance among all those possibly affected with reference to all situations
of application appropriate to it. Gunther formulates this duality in the
following manner: "A norm is valid and appropriate whenever the
consequences and side effects of its general observance for the inter
ests of each individual in every particular situation can be accepted
by all."23 Of course, participants in argumentation could apply this
formula properly only if they had unlimited time at their disposal or
were privy to complete knowledge that enabled them to predict re
liably all situations that could possibly arise. But the principle of
universalization, as a rule of argumentation, must retain a rational,
and thus operational, meaning for finite subjects who make judg
ments in particular contexts. Hence it can demand at most that in
justifying norms, those consequences and side effects be taken into
account that general adherence to a norm can be anticipated to have
for the interests of each on the basis of the information and reasons
available to them at a particular time.
Clearly, only situations actually used by participants, on the basis
of their state of knowledge, for purposes of paradigmatically expli
cating a matter in need of regulation can be taken into account in
37
the conditional components of a valid norm. The principle of univer

salization must be formulated in such a way that it does not impose
impossible demands; it must relieve participants in argumentation of
the burden of taking into account the multitude of completely un
foreseeable future situations in justifying norms. Hence, Gunther
proposes the formula: "A norm is valid if the consequences and side
effects of its general observance for the interests of each individual
under unaltered circumstances can be accepted by all."24 The rebus sic
stantibus clause here expresses the qualification that the validity claim
of a norm that has withstood the universalization test bears a "time
and knowledge index." This reservation ensures that justificatory
discourses cannot completely exhaust the notion of impartiality but
can only specify its meaning in relation to universal and reciprocal
worthiness of recognition. Prima facie valid norms remain open to
further interpretation in the light of particular constellations of un
foreseeable situations of application. The question of whether norms
determined to be valid with reference to anticipated typical situations
cited as exemplars are also appropriate for similar situations actually
occurring in the future in the light of the relevant features of these
situations is left unanswered by justificatory discourses. This question
can be answered only in a further discursive step, specifically, from
the changed perspective of a discourse of application.
In discourses of application, the principle of appropriateness takes
on the role played by the principle of universalization in justificatory
discourses. Only the two principles taken together exhaust the idea
of impartiality : "In justification only the norm itself, independently
of its application in a particular situation, is relevant. The issue is
whether it is in the interest of all that everyone should follow the
rule . . . . In application, by contrast, the particular situation is rele
vant, regardless of whether general observance is also in the interest
of all (as determined by the prior discursive examination). The issue
here is whether and how the rule should be followed in a given
situation in light of all of the particular circumstances . . . . What must
be decided is not the validity of the norm for each individual and his
interests but its appropriateness in relation to all of the features of a
particular situation ."25 Discourses of application bring to bear the
hermeneutic insight that the appropriate norm gains concrete signif
icance in the light of the salient features of the situation, and the
38
situation is described in turn in the light of the conditions specified

in the norm.
Here I do not need to go into the principle of appropriateness and
the logic of discourses of application, since these matters have been
investigated in detail by Gunther. 26 The problem to which both re
spond becomes apparent in the case of conflict between norms, for
in such cases it must be determined which of the prima facie valid
norms that are candidates for application proves to be the one most
appropriate to a situation, described as exhaustively as possible in all
of its relevant features. The norms that are eclipsed by the norm
actually applied in a given case do not thereby lose their validity but
form a coherent normative order together with all other valid rules.
From the standpoint of coherence, the relations within this order
shift with each new case that leads to the selection of the "single
appropriate norm." Thus, it is the system of rules as a whole that
ideally permits just one correct solution for every situation of appli
cation. Conversely, it is the particular situation whose appropriate
interpretation first confers the determinate shape of a coherent order
on the unordered mass of valid norms.
Moreover, this result enables us to account for an unsettling asym
metry between the treatment of moral-practical questions, on the one
hand, and of empirical-theoretical questions, on the other. In justi
fying factual claims, we find no analogue of the peculiar division of
the impartial judgment of moral conflicts of action into the steps of
justification and application. Although the discursive redemption of
assertoric validity claims is subject to the fallibilistic qualification that
we cannot know definitively whether the assertion taken to be true
will withstand all future objections, in this case a justification does
not stand in need of supplementation in the same way as in the case
of the prima facie validity of a norm; valid empirical knowledge is
not logically contingent on the resolution of questions of application.
Practical knowledge, by contrast, is of its very nature related to
action. This fact provides an explanation of the asymmetry only if it
is understood in a particular way. Given its relation to action, moral
knowledge of how things should go in the social world is influenced
differently by history from empirical knowledge of how things do go
in the objective world. The fallibilism that characterizes all knowl
edge, and hence also the fruits of moral discourses of justification
39
and application, amounts to the acknowledgment of the critical po

tential of superior future knowledge, that is, of history in the shape
of our own unforeseeable learning processes. The specific reservation
that expresses itself in the fact that we take well-grounded norms of
action to be prima facie valid only in a provisional sense can indeed
be explained in terms of the limitedness of our knowledge but not
in terms of its fallibility. The more far-reaching reservation concern
ing incompleteness cannot be explained in terms of cognitive provin
ciality in view of potentially better future knowledge; it is rather a
function of existential provinciality resulting from historical trans
formations in the objects themselves, and thus in the contexts in
which future actions will be determined by rules accepted at present.
The social world toward which we are oriented in the normative
attitude is historical in a different sense from the laws and regularities
that constitute the realm of describable events and states of affairs in
the objective world. The incompleteness of what can be accomplished
by discourses of moral justification can be ultimately explained by
the fact that the social world, as the totality of legitimately ordered
interpersonal relations, has a different ontological constitution from
the objective world. Whereas in the objectifying attitude we pre
suppose the objective world as the totality of existing states of
affairs, the social world as such has a historical character. Gunther's
normative concept of coherence seeks to do justice to this "intrinsic"
historicality : "If every valid norm is dependent on coherent supple
mentation by all others in situations in which norms are applicable,
then their meaning changes in every situation. In this way we are
dependent on history, since it first produces the unforeseeable situ
ations that compel us in each instance to produce a new interpretation
of all valid norms."27 Deontological ethical conceptions assume in the
final analysis only that the moral point of view remains identical; but
neither our understanding of this fundamental intuition, nor the
interpretations we give morally valid rules in applying them to un
foreseeable cases, remain invariant.
5. Only cognitivist basic assumptions can do justice to the phenomena
and the experiences of a posttraditional morality that has detached
itself from the religious and metaphysical contexts from which it
arose. Kantian ethics derives its plausibility not from the justification
40
of the categorical imperative in particular, or from the construction

of a Kingdom of Ends, and definitely not from the architectonic of
the two-worlds theory as a whole but from moral intuitions to which
a cognitivist interpretation can appeal. Moral experiences are a re
sponse to the violation of what Kant called duty, respect, and free
will; they crystallize around the harm inflicted on a person by im
moral action-the humiliation and abasement of a person whose
integrity has been violated, and the reproaches that the injurious
action as well as the wrongdoer bring upon themselves. With this,
the peculiar mode of validity of moral injunctions becomes the focus
of interest as a phenomenon in need of explanation. Hence, the
interpretation of the mode of validity of normative statements on an
analogy with the truth of assertoric statements is opposed (a) to the
empiricist notion that the illocutionary force of moral commands
rests on mere feelings of obligation originating in the internalization
of threatened sanctions and (h) to the noncognitivist notion that the
reciprocal interest in the observance of norms can ultimately be
traced back to an interest in self-respect.
(a) "Ought" sentences expressing obligations are the primary lin
guistic form in which morality finds expression. Duties prescribe
actions or omissions. Prohibitions are the negations of permissions,
permissions the negations of prohibitions. Obligations have their ex
periential basis not in perceptions but, as Strawson has shown, in
moral feelings. The latter point as a general rule to violations of
duties, transgressions against norms from which duties and rights
(i.e., legitimate expectations concerning actions in accordance with
duties) can be derived. Feelings of offense and resentment are sec
ond-person reactions to violations of our rights by others; feelings of
shame and guilt are reactions to our own transgressions; and outrage
and contempt are reactions of one present but not directly involved
to the violation of a recognized norm by a third person. Thus, these
affective states correspond to the perspectives and roles of the par
ticipants in interaction--ego and alter-and of a neutral party who
is not presently involved but whose perspective should not be con
fused with that of a mere observer, his view being that of a represen
tative of universality. They all belong to a community in which
interpersonal relations and actions are regulated by norms of inter-
41
action and can be judged in the light of these norms to be justified

or unjustified.
These affective responses to violations that find expression in turn in
reproaches, confessions, condemnations, and so forth and can lead
to accusations, justifications, or excuses constitute the experiential basis
of obligations, though they do not exhaust their semantic meaning.
The normative sentences in which these obligations are expressed
point to a background of normatively generalized behavioral expec
tations. Norms regulate contexts of interaction by imposing practical
obligations in a justifiable manner on actors who belong to a shared
community. Conventions are norms of interaction that define recip
rocal behavioral expectations in such a way that their content does
not need to be justified. "Mere" conventions bind, so to speak, in a
groundless fashion by custom alone; we do not associate a moral
claim with them. Duties, by contrast, derive their binding force from
the validity of norms of interaction that claim to rest on good rea
sons. We feel obligated only by norms of which we believe that, if
called upon to do so, we could explain why they both deserve and
admit of recognition on the part of their addressees (and of those
affected) .
The internal connection between norms and justifying grounds
constitutes the rational foundation of normative validity. This can be
confirmed at the phenomenological level by the corresponding sense
of obligation. Duties bind (binden) the will but do not bend (beugen) it.
They point the will in a certain direction and give it orientation but
do not compel it as impulses do; they motivate through reasons and
lack the impulsive force of purely empirical motives. Hence the em
piricist notion that norms obligate only to the extent that they are
backed up by well-founded expectations of sanctions neglects the
fundamental intuition that the noncoercive binding force is trans
ferred from the validity of a valid norm to the duty and the act of
feeling obligated. Only the affective reactions to the violation and the
perpetrator-resentment, outrage, and contempt-are expressed in
the sanctions that result from transgressions of norms.
But the violation of legitimate expectations, to which these feelings
are reactions, already presupposes the validity of the underlying
norms. Sanctions (however much they are internalized) are not con
stitutive of normative validity; they are symptoms of an already felt,
42
and thus antecedent, violation of a normatively regulated context of

life. Hence Kant correctly presupposes the primacy of the "ought"
over sanctions-as indeed does Durkheim-in explaining the original
phenomenon of insight into, and the moral feeling of, being obligated
to do something in terms of the interrelation between autonomy of
the will and practical reason. 2 8
We do not adhere to recognized norms from a sense of duty
because they are imposed upon us by the threat of sanctions but
because we give them to ourselves. Of course, this preliminary reflec
tion does not provide an adequate basis for developing a notion of
self-legislation. Norms we give to ourselves may express our own
orders, and thus mere choices (Willkiir), in which case they lack the
very quality that would make them binding norms. It is not because
recognized norms are certified by custom and tradition that we observe
them from a sense of duty but because we take them to be justified.
But even these reflections viewed in themselves are not a sufficient
basis for developing the concept of a norm-testing reason. For ex
ample, we might want to justify norms as we do facts, thereby over
looking precisely what makes reason practical. Only by combining
both reflections do the concepts of "autonomous will" and "practical
reason" emerge as coeval with one another.
Only a will that is open to determination by what all could will in
common, and thus by moral insight, is autonomous; and that reason
is practical which conceives of everything that is justified in accor
dance with its impartial judgment as the product of a legislating will.
Voluntas and ratio are interwoven in a remarkable fashion in these
two concepts without the one being reduced to the other. These
moments no longer confront each other abstractly as, respectively,
an active faculty that intervenes in the world and a passive faculty
that mirrors facts. An autonomous will gives itself only rationally
grounded laws, and practical reason discovers only laws that it simul
taneously formulates and prescribes. A cognitive moment inheres as
much in self-determination as a constructive moment does in norm
testing reason. Even Kant could not ultimately give a satisfactory
explanation of this perplexing interrelation; it becomes intelligible
only when we cease to regard freedom and reason as merely subjec
tive faculties.
43
Remarks
on
Discourse Ethics
From the standpoint of the theory of intersubjectivity, autonomy

does not signify the discretionary power of a subject who disposes of
himself as his own property but the independence of a person made
possible by relations of reciprocal recognition that can exist only in
conjunction with the correlative independence of the other. The
intersubjective character of freedom and practical reason becomes
manifest when we analyze the role an expression such as "respect"
assumes in the language game of morality.
(b) In the various stages of its development, Ernst Tugendhat's
moral theory exhibits a tendency to combine the empiricist conception
of normative validity just discussed with an intersubjective conception
of morality through the mediation of the concept of self-respect.29
Tugendhat understands morality as a "system of norms that exist in
a society as a product of social pressure."30 This description is in
tended to hold not only from the observer's perspective but also from
that of the participant. Hence, he believes that the phenomenon of
normative validity can be elucidated by a suitable description of the
social sanctions consequent on the transgression of moral norms. This
approach yields a noncognitivist conception of morality, since im
moral actions cannot be viewed as irrational : "What happens when
one violates a moral order is rather that one experiences a social
sanction."3 1 For this reason Tugendhat thinks that we must give up the
attempt to justify morality "in the strict sense." Rationality in the
practical sphere he takes to be synonymous with purposive rationality.
At the same time Tugendhat takes issue with theories that appeal to
these empiricist assumptions as a basis for reducing morality to pur
posive considerations, with the goal of demonstrating that everyone
has good reasons grounded in premoral or natural interests to accept
a certain system of externally imposed constraints of reciprocal con
cern. Mackie, for example, with his appeal to enlightened self-interest
and prudent accommodation to external promises of rewards and
threats of punishment, completely ignores the obligating sense of
norms that is clearly expressed in moral feelings. Tugendhat outlines
the intuition he wishes to oppose to this reductionism as follows: "In
the one case we have a community constituted by norms that subserve
reciprocal utility, in the other a community constituted by (substan
tially the same) norms, but their meaning now consists in the fact
that in them reciprocal respect comes to expression."32
44
The phenomenon of reciprocal respect must here be understood in

terms of the empiricist premises that Tugendhat shares. The key to
this is provided by inner sanctions, that is, the feelings of guilt and
shame that result from the internalization of outer sanctions. Tug
endhat understands moral shame and guilt as reactions to the loss of
one's sense of self-worth; thus, it is ultimately my self-respect that is
endangered when I act immorally. One who violates moral norms
not only exposes himself to the contempt of others but also feels
contempt for himself because he has internalized this sanction. A
norm counts as 'justified," therefore, only insofar as it is in the
interest of each, viewed from his own perspective, that everyone
should engage in a practice regulated by the exchange of signs of
respect.
If we follow Tugendhat in assuming that everyone has an interest
in self-respect and being respected as a person by others, we can
explain why moral norms are good for me only when they are good
for all, for self-respect requires reciprocal respect, since I can affirm
myself only if I am valued by those who behave in such a way that
they, in turn, are worthy of respect and can also be valued by me.33
The outcome of Tugendhat's reflections is a morality of mutual re
spect that seems to boil down to the familiar universalistic principle
of equal respect for all. But the premises of this position come into
conflict with precisely this intuition. If my esteem for other persons
and their respect for me are ultimately rooted in the fact that each
individual can respect himself only if he is respected by others whom
he does not hold in contempt, then there is something purely instru
mental about the mutuality of recognition : respect for others is me
diated by the concern with self-respect. But my respect for others
cannot be made conditional on the satisfaction of my interest in self
respect if relations of mutual respect are to generate the perfectly
symmetrical structures of recognition commensurate with our intuitive
understanding of noninstrumental relations between autonomously
acting persons. The egocentric character of the underlying need to
be respected by others postulated as primary is transmitted to the
structures of recognition based upon it and undermines the complete
reciprocity of relations of recognition.34
Reciprocal respect represents a necessary pragmatic precondition
of participants in interaction ascribing themselves rights and duties
45
only when understood in the sense of complete reciprocity. Tugendhat

has since adopted a new approach in an effort to free his project
elucidating the phenomenon of reciprocal recognition from the per
spective of each individual's sense of self-respect-from the lingering
suspicion of egocentricity. In his most recent work, he links the
central capacity to experience moral shame and guilt from the outset
with the status of membership in a community. My understanding of
myself as a person is interwoven with my social identity in such a way
that I can value myself only if the community of which I view myself
as a part and whose authority is binding for me confirms me in my
status as a member. What I have internalized as a sanction is the fear
of expulsion from a community with which I have identified. Tug
endhat introduces this foundation of self-respect in antecedent social
relations of recognition through an account of what it means to
respect someone as a person.
Respect in the sense of esteem is not always a moral matter. We
hold someone in esteem as an athlete or as a scientist because of his
outstanding achievements. We value someone as a colleague or a
friend for his competence or reliability-in short, on account of some
outstanding personal qualities he possesses. The example of a friend
need be altered only slightly for us to recognize that we can also
value someone for his moral qualities-as someone, for example,
who refrains from acting improperly, or even opportunistically, in
difficult situations and who is willing to make a sacrifice or, in extreme
cases, even to sacrifice himself. In all such cases respect can be trans
formed into admiration, since respect here is a function of greater
or lesser estimation of actions and qualities of character. By contrast,
respect for a person as a person admits of no gradations; we respect
a person as such not on account of some outstanding characteristic
or other. We respect a person as such on account of his capacity to
act autonomously, that is, to orient his actions to normative validity
claims ; we respect him solely on account of the accomplishment or
quality that makes him a person. One cannot possess this constitutive
capacity to a greater or lesser degree; it is definitive of what it means
to be a person as such. We do not respect someone as a person
because he impresses us or because he is worthy of esteem in some
way or other--or even because he is a good person or lives a good
life-but because he is, and by his conduct shows himself to be,
46
fundamentally capable of being a "member of a community," that is,

capable of observing norms of communal life as such.
With this the concept of social membership assumes the privileged
status previously enjoyed by the concept of self-respect. Self-respect
cannot be an original phenomenon for the simple reason that it is
unclear what the respectworthiness of the isolated subject is supposed
to consist in prior to all socialization. Tugendhat initially thought that
the intrinsic worth that qualified the subject by his very nature, so to
speak, to claim the respect of others could be understood as the value
that this subject ascribes to his life as a whole.85 But clearly the dignity
of a person cannot be reduced to the value he confers on his life,
since on occasion we may risk our life in order to preserve our self
respect. In fact, the self of self-respect is tied to an extremely vul
nerable personality structure; but the latter first emerges in the con
text of relations of reciprocal recognition. The unconditional
relations of mutual respect in which individuals confront one another
as responsible acting persons are coeval with the phenomenon of self
respect, and thus with the consciousness of being worthy of the
respect of others. Hence, Tugendhat adopts a conception of the social
constitution of the self from which it follows that nobody adequately
understands his own identity who does not derive his sense of self
respect from his status as a member of a community, a status that is
recognized by all other members.36
On this account, the phenomenon of reciprocal recognition is no
longer explained in terms of an original interest in self-respect or of
an initial fear of the inner sanction of loss of one's sense of self-worth
The self is no longer the primary phenomenon but is viewed as the
product of a process of socialization that itself already presupposes
the structure of relations of reciprocal recognition. What is funda
mental is the idea of a community "in which each member derives
his sense of self-worth from the observance of the norms that make
a community possible and demands the same of others. Thus reci
procity does not here consist in exchange relations but in reciprocally
understanding one another in a certain way and in reciprocally de
manding such understanding of one another."37 But in that case it
may be asked whether the meaning of the moral "ought" grounded
in the reciprocal demand for mutual recognition is still to be sought
"in the inner sanction," as Tugendhat would have it, or whether the
.
47
central feelings of shame and guilt, which are coeval with outrage
and contempt, are not secondary phenomena to the extent that they
are reactions to the violation of legitimate expectations grounded ul
timately in the reciprocity of the structures of recognition underlying
communities in general. In short, Tugendhat confuses genesis and
validity. He is misled by the observation that in the process of social
ization, conscience is formed through the internalization of external
sanctions to suppose that, even from the participant perspective of the
conscientious individual who has been socialized in this manner, be
hind the moral "ought" there is concealed a sanction, the inner
sanction of loss of self-respect, instead of the unforced force of the
good reasons in terms of which moral insights impress themselves on
consciousness as convictions.
The understanding of postconventional morality that Tugendhat
proposes itself reveals the cognitive meaning of the mode of validity
of moral norms, which cannot be analyzed in terms of inner sanctions.
In traditional societies, moral norms are indeed so closely bound up
with religious worldviews and shared forms of life that individuals
learn what it means to enjoy the status of membership in a community
thus founded through identification with the contents of this estab
lished concrete ethical life. But in modern societies, moral norms
must detach themselves from the concrete contents of the plurality
of attitudes toward life that now manifest themselves ; they are
grounded solely in an abstract social identity that is henceforth cir
cumscribed only by the status of membership in some society, not in
this or that particular society. This explains the two salient features
of a secularized morality that has transcended the context of an over
arching social ethos. A morality that rests only on the normative
content of universal conditions of coexistence in a society (founded
on mutual respect for persons) in general must be universalistic and
egalitarian in respect of the validity and sphere of application of its
norms ; at the same time, it is formal and empty in the content of its
norms. But from its formal and empty character there follows a
consequence that is incompatible with a noncognitivist understanding
of morality.
The generalized structure of the reciprocal recognition of subjects
who confront each other simultaneously as nonreplaceable individ
uals and as members of a community henceforth amounts only to
48
the viewpoint of the impartial judgment of practical questions that

remain to be decided by argument among those affected. When
Tugendhat also speaks ultimately of the "enlightened perspective"
from which "only that which can be agreed upon is decisive" and
comes to the conclusion that "the viable core of all morality reduces
to the basic stock of natural or rational norms to which contractualism
also appeals and without which there could not be anything like a
community,"38 then he abandons his empiricist premises and falls
back on an intersubjectivist interpretation of Kantn a construction
such as that of Rawls, for example, who analyzes moral obligation,
and thus the validity of moral principles and institutions, in terms of
the idea that they can be rationally justified under conditions of
impartial judgment. The general structure of the relations of rec
ognition that make the understanding of self as a person and as a
member of a community simultaneously possible is presupposed in
communicative action and is preserved in the communicative presup
positions of moral argumentation. Just this structure compels each
participant in argumentation to adopt the perspective of all others.
6. The fact of a form of life structured by communicative action can

no more be circumvented in moral judgment than in moral action.
This becomes manifest once we analyze the rule specifying how one
can regard something "from the moral point of view." This is the
name we give the standpoint from which moral questions can be
judged impartially. Interestingly enough, all attempts to characterize
this standpoint as that of a neutral observer have miscarried, whether
the latter is represented as surveying the moral world as a whole
from a transcendental vantage point (as in Kant) or as an empirical
observer who occupies a position in the world but is equipped with
ideal knowledge (as in utilitarian theories). The observer standpoint
seems to guarantee an exceptional degree of objectivity of judgment
because of its third-person perspective, but in fact it is unsuitable for
judging the question of whether actions or norms are in the common
interest or contribute to the common good. The ideal observer op
erates as an isolated subject, collecting and assessing his information
in the light of his own individual understanding of the world and of
himself. Impartiality of judgment, by contrast, is essentially depen
dent on whether the conflicting needs and interests of all participants
49
Remarks on Discou rse Ethics
are given their due and can be taken into consideration from the
viewpoint of the participants themselves. The privileged position,
which initially appears to be advantageous because it promises to
liberate the observer from the perspectival interpretations of the
disputing parties, has the disadvantage that it isolates him in a mon
ological fashion from the interpretive horizons of the participants
and denies him hermeneutic access to an intersubjectively shared
moral world that reveals itself only from within.
One who wishes to examine something from the moral point of
view must not allow himself to be excluded from the intersubjective
context of participants in communication who engage in interper
sonal relations and who can understand themselves as addressees of
binding norms only in this performative attitude. Controversial ques
tions of normative validity can be thematized only from the first
person plural perspective, that is, in each instance "by us" ; for nor
mative validity claims are contingent on "our" recognition. We cannot
attain an impartial standpoint by turning our back on the context of
linguistically mediated interaction and by abandoning the participant
perspective completely, but only by extending the individual partici
pant perspective in a universal fashion. 39 Each of us must be able to
place himself in the situation of all those who would be affected by
the performance of a problematic action or the adoption of a ques
tionable norm. What G. H. Mead recommends with his notion of
ideal role taking cannot be performed privately by each individual
but must be practiced by us collectively as participants in a public
discourse.
Once it becomes clear that the goal of such an inclusive process of
communication (uncoerced agreement) can be achieved only through
the mediation of good reasons, the reflective character of what Mead
calls "universal discourse" takes on sharper contours. It should not
be conceived simply as a kind of net of communicative action that
encompasses all potentially affected. On the contrary, as the reflective
form of communicative action, it is a form of argumentation in the
strict sense.
Viewed in this way, Mead's construction loses the quality of a mere
projection, for in every real process of argumentation the participants
unavoidably undertake such a "projection." They must make a prag
matic presupposition to the effect that all affected can in principle
50
freely participate as equals in a cooperative search for the truth in

which the force of the better argument alone can influence the out
come. On this fact of universal pragmatics is founded the fundamen
tal principle of discourse ethics : only moral rules that could win the
assent of all affected as participants in a practical discourse can claim
validity. In virtue of the idealizing assumptions that everyone who
seriously engages in argumentation must make as a matter of fact,
rational discourse can play the role of a procedure that explicates the
moral point of view. Practical discourse may be understood as a
communicative process that induces all participants simultaneously to
engage in ideal role taking in virtue of its form, that is, solely on the
basis of unavoidable universal presuppositions of argumentation.
This explanation of the moral point of view privileges practical
discourse as the form of communication that secures the impartiality
of moral j udgment together with universal interchangeability of par
ticipant perspectives. It also explains why there is a reasonable pros
pect of deriving the fundamental principle of universalistic morality
from the necessary practical presuppositions of argumentation in
general without committing a naturalistic fallacy. Discourses overlay
communication oriented to reaching understanding as their reflective
form; in the symmetry conditions and reciprocity expectations of
everyday speech oriented to reaching understanding, there already
exist in nuce the basic notions of equal treatment and general welfare
on which all morality turns, even in premodern societies. The ideas
of justice and solidarity are already implicit in the idealizing presup
positions of communicative action, above all in the reciprocal recog
nition of persons capable of orienting their actions to validity claims.
Of course, the normative obligations that children assume in virtue
of the mere form of socializing interaction do not of themselves point
beyond the limits of a concrete lifeworld (of the family, the clan, the
city, or the nation). These barriers must first be breached in rational
discourse. Arguments by their very nature point beyond particular
individual lifeworlds; in their pragmatic presuppositions, the
normative content of presuppositions of communicative action is gen
eralized, abstracted and enlarged, and extended to an ideal com
munication community encompassing all subjects capable of speech
and action (as Apel claims following Peirce).
51
The standpoint of morality differs from that of concrete ethical

life in its idealizing extension and reversal of interpretive perspectives
that are tied to particular, established cultural lifeforms and are the
result of individual processes of development. This transition to the
idealizing presuppositions of a socially and spatiotemporally unlim
ited communication community in every actually carried out dis
course is an anticipation of, and approximation to, a regulative idea.
At the same time, actually engaging in discourse belongs to that very
idea, since otherwise the solipsistic character of an examination of
norms undertaken "in the privacy of individual reflection" would
reassert itself. Even an advocatory discourse, or one pursued in mente,
is still different from the essentially monological procedure of exam
ining maxims in accordance with the categorical imperative in that it
clarifies the meaning and burden of proof of the principle of univ
ersalization. Admittedly, Kant too makes a distinction between an
ego-oriented mode of generalization in the weak sense of the Golden
Rule and the rigorously universalistic operation of generalization
commensurate with the pragmatic implications of forming a general
will.40 But as long as the isolated subject, in his role as custodian of
the transcendental, arrogates to himself the authority to examine
norms on behalf of all others, the difference between his supposition
concerning a general will and an intersubjective agreement concerning
a common will never comes to light. The perspective from which I test
the generalizability of all concerned interests remains in each instance
an unquestioned background assumption, for as long as each auton
omous will can regard itself as one with all other intelligible beings
in the Kingdom of Ends, the maxims that I (in the light of my own
self-understanding and view of the world) can will to be rules gov
erning a general practice are valid. But once we abandon the meta
physical doctrine of two separate spheres of reality, subjects
encounter each other as individuals who can no longer rely on this
antecedent transcendental agreement. They, and hence we, must
confront the problem of reaching an adequate intersubjective un
derstanding of the interests of each individual.
In such a situation the principle "Do unto others as you would
have them do unto you" will be found wanting because it remains
bound in an egocentric fashion to an unthematized individual un
derstanding of self and world. Only an intersubjective process of
52
argumentation in which all potentially affected could participate

makes radical generalization at once possible and necessary-in other
words, only a test of the generalizability of modes of action that
excludes all surreptitious privileging of individual viewpoints and
demands the coordination of all of the interpretive perspectives that
tend toward individualism and pluralistic fragmentation, at least in
modern societies. Given the communicative presuppositions of an
inclusive and noncoercive discourse among free and equal partners,
the principle of universalization requires each participant to project
himself into the perspectives of all others; at the same time, it remains
possible for each participant to test whether he can will a disputed
norm as a general law from his own point of view on the basis of
reciprocal criticism of the appropriateness of interpretive perspec
tives and need interpretations.
The notion that ideal role taking-that is, checking and reciprocally
reversing interpretive perspectives under the general communicative
presuppositions of the practice of argumentation-becomes both pos
sible and necessary loses its strangeness when we reflect that the
principle of universalization merely makes explicit what it means for
a norm to be able to claim validity. Already in Kant the moral prin
ciple is designed to explicate the meaning of the validity of norms;
it expresses, with specific reference to normative propositions, the
general intuition that true or correct statements are not valid just for
you or me alone. Valid statements must admit of justification by
appeal to reasons that could convince anyone irrespective of time or
place. In raising claims to validity, speakers and hearers transcend
the provincial standards of a merely particular community of inter
preters and their spatiotemporally localized communicative practice.
The moral principle articulates a notion of transcendence or self
overcoming already contained in assertoric validity claims. Thus,
Peirce, in explicating the meaning of truth, already engages in exten
sions or idealizations of the social and temporal dimensions of the
practice of communication similar to those required by an intersub
jective principle of universalization.41
The world as the totality of possible facts is always constituted by
some community of interpreters whose members come to an under
standing with one another, within an intersubjectively shared life
world, about something in the world. What is "real" can be expressed
53
in true statements, where "true" can be explained in terms of the

claim that an individual addresses to others when he makes a state
ment. With the assertoric meaning of his utterance, the speaker raises
a criticizable claim to the validity of the asserted proposition; and
since we have no direct access to uninterpreted conditions of validity,
"validity" (GiUtigkeit) must be understood epistemically as "validity
(Geltung) that is established for us." Every justified truth claim ad
vocated by someone must be capable of being defended with reasons
against the objections of possible opponents and must ultimately be
able to command the rationally motivated agreement of the com
munity of interpreters as a whole. Here an appeal to some particular
community of interpreters will not suffice. Hence, Peirce explicates
truth by appealing to the counterfactual redemption of criticizable
validity claims under the communicative conditions of a community
of interpreters extended ideally in social space and historical time.
The projection of an unlimited communication community serves to
substitute for the infinite (or transtemporal) character of "uncondi
tionality" the idea of an open but goal-directed process of interpre
tation that transcends the boundaries of social space and historical
time from within, from the perspective of a finite existence situated
within the world. In time, according to Peirce, the learning processes
of the unlimited communication community should erect a bridge
that spans all temporal distance; in the world, the conditions that
must be assumed to be sufficiently satisfied as a precondition of the
unconditional claim of transcending validity claims sho1,1ld be realiz
able. That degree of fulfillment counts as "sufficient" which qualifies
our argumentative practice in a given instance as a spatiotemporally
localized component of the universal discourse of an ideally extended
community of interpreters that we necessarily presuppose. This pro
jection has the effect of shifting the tension between facticity and
validity into presuppositions of communication, which, despite the
fact that their ideal content can only ever be approxi mately realized,
must a.s a matter of fact be made by all participants every time they
assert or dispute the truth of a statement and undertake to justify
this validity claim in argumentation.
Peirce's discursive conceptio n of truth shows how the concept of
normative validity, understood on the model of truth, can be expli
cated in terms of a moral principle that links the requisite notion of
54
ideal role taking to the communicative form of practical discourse .

Correct normative propositions do not refer to existing states of
affairs in the objective world as do true statements but to legitimately
ordered interpersonal relations in the social world. But in addition
to being constructed, just social orders are also discovered. The role
of the idealizations to which an intersubjective explanation of the
moral principle repeatedly appeals is, of course, in need of further
explanation.
7. I f normative rightness, like validity (Giiltigkeit) in general, is con

strued as the three-place relation "validity (Geltung)-of-something-for
someone," then the counterfactual meaning of rational acceptability
cannot be reduced to that of acceptance within a community of inter
preters. On the other hand, the idea of an ideally extended com
munication community is paradoxical in that every known
community is limited and distinguishes members from nonmembers
through rules of inclusion. Precisely this difference would have to be
effaced through reflective extension of social space and historical
time. Even the image of an ideal audience (Chaim Perelman) is still
too concrete. The model of a public sphere accessible to all partici
pants, issues, and contributions comes closest to the notion envisaged,
for these pictures merely serve to express in tangible form limiting
concepts for specifying an ideal structure of communication. The
idealizations relate to the possible structure of a communicative prac
tice that operates in a self-reflexive and self-correcting manner and
results in a progressive derelativizing of the conditions under which
validity claims that are raised in a particular context, but are context
transcending in their meaning can be redeemed.
We speak of "idealizations" in a variety of different contexts. The
ideal generality of concepts and meanings generates at first sight few
difficulties because it seems to establish itself in the rarefied medium
of thought. But in fact the universality of meaning already depends
on the linguistic idealizations that enable us to recognize the same
sign type in individual sign tokens and the same grammatically struc
tured sentence in different utterances. Geometrical figures that we,
like Euclid, can realize only approximately in the sand or on the
blackboard exhibit a different kind of ideal generality. With them we
associate the idea of perfection or perfectibility-we can draw a circle
55
only with a greater or lesser degree of exactitude. Physical measuring

operations also rest on the counterfactual assumption that we can
approximate ideal quantities to an arbitrary degree of exactitude; as
a practical matter, we cannot produce a perfectly level plane. In a
similar way experiments may be based on the assumption of an ideal
gas, a perfect vacuum, or such like. Simulations have a different,
though related, significance. We can remove in thought the operation
of gravity and simulate on earth conditions of weightlessness actually
encountered in outer space. Finally, with the concept of a regulative
idea, the mathematical model of infinite approximation is transposed
from the sphere of instrumental action to interactions. These few
comparative illustrations must here suffice to lend plausibility to the
notion that to idealize forms of communication, that is, to think of pro
cesses of communication as if they took place under idealized con
ditions, is not per se absurd.
The idealizations undertaken in the sphere of communicative
praxis have an air of arbitrariness about them only when viewed from
the perspective of a semantically abridged conception of language.
Once we reflect on the use of linguistic expressions in the perfor
mative attitude of speakers and hearers, we experience idealizations
as those simultaneously unavoidable and trivial accomplishments that
sustain communicative action and argumentation. We attribute iden
tical meanings to expressions, attach context-transcending signifi
cance to validity claims, and ascribe rationality or accountability to
speakers. These and similar attributions remain for the most part
implicit and have their proper locus in pragmatic presuppositions.
In performing an act of communication, we can assume in a self
reflexive fashion that a pragmatic precondition of this communica
tion is fulfilled even though this assumption may not be objectively
correct; when our communication does not actually fulfill the as
sumed condition, we have acted under a counterfactual presupposition.
An idea whose comprehension is a major source of difficulties is
the transcendental-pragmatic assumption that language use oriented
to reaching understanding, whether in communicative action or in
argumentation, always demands that participants make certain formal
pragmatic presuppositions if the practice is to exist at all-and this
independently of whether they turn out post hoc to be counterfactual.
We know at least intuitively that certain of these presuppositions
56
cannot be fulfilled under normal empirical restrictions, yet we must

nevertheless assume that these idealizing presuppositions are suffi
ciently fulfilled. It is in this sense that in rational discourse, where the
speaker seeks to convince his audience through the force of the better
argument, we presuppose a dialogical situation that satisfies ideal
conditions in a number of respects, including, as we have seen, free
dom of access, equal rights to participate, truthfulness on the part of
participants, absence of coercion in taking positions, and so forth. It
must be shown for each of these conditions of a so-called ideal speech
situation (through the demonstration of performative self-contradic
tions) that they belong to the unavoidable presuppositions of argu
mentation. Every speaker knows intuitively that an alleged argument
is not a serious one if the appropriate conditions are violated-for
example, if certain individuals are not allowed to participate, issues
or contributions are suppressed, agreement or disagreement is ma
nipulated by insinuations or by threat of sanctions, and the like.
The general presuppositions of argumentation cannot be easily
fulfilled because of their rigorous idealizing content. Rational dis
courses have an improbable character, existing like islands in the sea
of everyday practice. At the same time, they constitute an irreducible
order and cannot be replaced by functional equivalents. In the con
temporary world, discourses have specialized social functions-such
as the production of knowledge, therapy, art criticism, the adminis
tration of justice, political will formation, and so forth-and have to
be institutionalized. The rules in accordance with which discourses
are institutionalized in modern societies are for the most part legal
norms-in the first instance, organizational and procedural norms
that specify qualifications for participation, areas of jurisdiction, sub
jects for discussion, deliberative procedures, decision processes, ob
ligation to justify decisions, and so on. Rules regulating the order of
business in deliberative bodies provide an example of how delibera
tions are regulated with the goal of ensuring the probability of ade
quate fulfillment of demanding communicative presuppositions
under temporal, social, and practical limitations. The institutionali
zation of specific regions of discourse such as these relieves partici
pants of the considerable motivational and cognitive burdens that
every presupposition-laden cooperative quest for the truth imposes
when it is pursued in an ad hoc fashion in the form of unregulated,
57
spontaneous argumentation. From the perspective of the legislator

who institutionalizes rules of discourse, the degenerate forms in
which such quasi-natural, informal deliberative practices appear con
stitute the resistant substrate in which the "geometrical forms" of an
ideal deliberative process must be realized in an approximate fashion.
The most serious misgivings do not concern the possibility of ef
fectively institutionalizing rational discourse, which is a matter for
a discourse theory of law and politics; skeptical objections are di
rected against the idealizations involved in the discourse model as
such.42
(a) Thus Steven Lukes asks in what form actors could satisfy the
preconditions for participating in a practical discourse that is sup
posed to offer the prospect of a rationally motivated agreement
among all affected on the basis of ideal role taking. A consensus on
generalizable interests can be expected, he claims, only if the theory
postulates either homogeneous societies or abstract discourse partic
ipants such as Rawls's parties in the original position and thereby
assumes that flesh-and-blood actors are surreptitiously transformed
into intelligible beings under the communicative presuppositions of
rational discourse. But participants in argumentation who must alter
their identity to such an extent that they no longer bear any discern
able relation to the agents involved in conflicts in need of regulation
would become disengaged from their own interests and problems. A
form of argumentation uncoupled from real life in this way would
necessarily lose sight of its true object: "Even if it were so, the greater
the change from the actual agents to the ideally rational agents,
capable of reaching the requisite consensus, the less relevant would
be the deliberations of the latter to the purpose at hand-which is to
establish how the actual participants would think and feel, were al
leged structures of domination to be overthrown ."43 In a similar
context, Williams objects that the dream of an ideal communication
community-by which he means Kant's Kingdom of Ends-is too far
removed from social and historical reality to preserve any relevance
for concrete issues.44 But these objections rest on a mistaken concep
tion of idealization.
Anyone who seriously engages in argumentation must indeed pre
suppose that the conditions of an "ideal speech situation" alluded to
are sufficiently realized. But this idealization does not bear on the
58
objects treated in argumentation; it leaves the identity of the partic

ipants and sources of conflict originating in the lifeworld untouched.
The moral point of view calls for the extension and reversibility of
interpretive perspectives so that alternative viewpoints and interest
structures and differences in individual self-understandings and
worldviews are not effaced but are given full play in discourse. Klaus
Gunther's operational interpretation of the idea of impartiality
underlines what is essential to the meaning of the validity of singular
moral judgments : just one of the norms that are equally in the interest
of all must be appropriate to a given situation when all of its relevant
features are taken into consideration. In justificatory discourses it is
necessary to abstract from the contingent contextual embeddedness
of a proposed norm only to ensure that the norm, assuming it with
stands the generalization test, is sufficiently open to context-sensitive
application. It must not be so selective that it admits only of standard
interpretations and excludes the differentiations of a maximally ex
haustive situation description that must be taken into account in
discourses of application.
(b) A further objection against procedural interpretations of the
moral principle is that practical discourse cannot stipulate a sufficiently
determinate procedure. The procedure of argumentation cannot en
sure the choice of correct answers solely on the basis of presupposi
tions of communication. It would have to have access to independent
criteria for assessing reasons. The procedurally correct practice of an
intersubjective exchange of arguments could not be the decisive fac
tor in this regard, for in the final analysis only the substantive grounds
that each participant can just as well ponder for himself are signifi
cant.45 Objections of this kind rest on either a pragmatically truncated
concept of argumentation or a semantically truncated concept of
justification, or both.
I understand argumentation as a procedure for the exchange and
assessment of information, reasons, and terminologies (or of new
vocabularies that make a revision of descriptions possible). Of course,
the procedure cannot itself generate these elements; its task is to
ensure that the argumentative exchange can proceed on the basis of
all relevant information and reasons available at a particular point in
time and within the most fruitful and appropriate descriptive frame
work in each instance. I follow Peirce in holding that arguments are
59
essential components of reflective learning processes that for their

part certainly cannot be explicated solely in terms of argumentation.
But procedures and reasons are so closely interwoven with each other
that there cannot be any evidence or criteria of assessment that are
completely prior to argumentation that do not have to be justified in
turn in argumentation and validalied by rationally motivated agree
ment reached in discourse under the presuppositions of argumen
tation. Deductive arguments alone are not informative, and
experiences or moral feelings do not constitute a foundation inde
pendent of interpretation. Because there cannot be "ultimate" evi
dence or "decisive" arguments in dealing with substantive questions,
we must appeal to the pragmatics of the procedure of argumentation
in order to explain why we can even think ourselves capable of raising
and redeeming context-transcending validity claims.
(c) Finally, it is asked whether we can expect exclusively correct an
swers in argumentation and in particular in practical argumenta
tion.46 Because of the ideal content of the central presuppositions of
argumentation in general, we can never be certain that the statements
we take to be true or correct at this stage of discussion belong among
the statements that will withstand all future criticism. But this falli
bilism built into the theory of discourse is merely the converse side
of the postulate that every sufficiently precise question admits of just
one valid answer. Even if we find this convincing for assertoric prop
ositions, things appear to be otherwise in the case of normative ques
tions. Is there only one correct answer to the abortion question, for
example? At this stage of the debate, both sides in this dispute appear
to have good, perhaps even equally good, arguments. For the time
being, therefore, the issue remains undecided. But insofar as what is
at issue is in fact a moral matter in the strict sense, we must proceed
from the assumption that in the long run it could be decided one
way or the other on the basis of good reasons. However, a forteriori
the possibility cannot be excluded that abortion is a problem that
cannot be resolved from the moral point of view at all. From this
point of view, what we seek is a way of regulating our communal life
that is equally good for all. But it might transpire that descriptions
of the problem of abortion are always inextricably interwoven with
individual self-descriptions of persons and groups, and thus with
their identities and life projects. Where an internal connection of this
60
sort exists, the question must be formulated differently, specifically,

in ethical terms. Then it would be answered differently depending
on context, tradition, and ideals of life. It follows, therefore, that the
moral question, properly speaking, would first arise at the more
general level of the legitimate ordering of coexisting forms of life.
Then the question would be how the integrity and the coexistence of
ways of life and worldviews that generate different ethical concep
tions of abortion can be secured under conditions of equal rights.
In other cases it is possible to deduce from the inconclusive out
come of practical discourses that the problems under consideration
and the issues in need of regulation do not involve generalizable
interests at all; then one should not look for moral solutions but
instead for fair compromises.
8. The communication-theoretical interpretation of morality and the
discourse-ethical version of the moral principle have the advantage
of avoiding a confusion that creeps in under the premises of the
philosophy of the subject. A narrow moral concept tailored to the
private sphere of the individual and initially removed from the public
sphere of socially organized communal life follows from the deon
tological distinction between normative questions of right action and
evaluative questions of the good life only if one takes as one's un
derlying assumption a narrow, individualistic concept of the person.
By contrast, once we take a concept of the individual as essentially
socialized as our point of departure and regard the moral point of
view as implicit in the structure of reciprocal recognition among
communicatively acting subjects, there is no longer a distinction in
principle between private morality and public justice but only in
respect of the degree of organization and institutional mediation of
interaction.47 It then becomes clear that persons, understood as mu
tually respecting individuals, are morally obligated in precisely the
same way as persons understood as members of a community engaged
in the activity of realizing collective goals.
The narrow, individualistic conception of morality rests on a neg
ative reading of the categorical imperative understood as being ap
plied in a monological fashion. This is meant to serve the justification
of moral injunctions by way of prohibiting nongeneralizable modes
of action. Negative duties such as "You should not lie" or "You should
not inflict harm on anyone" provide models for this form of justifi-
61
cation. If I cannot will that the maxim that permits me to deceive

and inflict harm on others in order to procure advantages for myself
should become a general norm, then we have a duty not to act in
this manner. If, on the other hand, I come to the conclusion that a
maxim is fit to hold as a general law, it does not follow that we have
a duty, but only that we are permitted, to act in accordance with this
maxim.48 This reading derives its plausibility from the fact that it
gives precedence to negative over positive duties and the fact that
negative duties seem to express the deontological meaning of moral
injunctions with special force. Categorical prohibitions such as "You
should not kill" or "You should not lie" are addressed in the second
person to all subjects capable of action and demand that we refrain
from specific actions regardless of the consequences. These norms
owe their force (a) to the unconditional character of their validity,
(b) the determinateness of their content, and (c) the lack of ambiguity
concerning their addressees. Only when these three conditions are
fulfilled can each individual know unequivocally what satisfies and what
violates the norm.
From the perspective of a potential victim, to every duty there
corresponds a right-for example, to the duty to refrain from killing
corresponds the right to the inviolability of one's life and body. A
moral principle that makes possible only the justification of general
prohibitions gives precedence to the class of negative rights and
duties as the essential core of morality. This coheres with liberal
convictions, for the rights that correspond to the negative duties
circumscribe the spheres of freedom in which, according to the for
mulation of the Kantian principle of right, the choice (Willkilr) of
each can coexist with the freedom of all in accordance with a general
law. If this followed conclusively from the deontological approach
itself, the discourse-ethical principle, whose application extends be
yond the formation of a negative general will, would be too unspecific
and in reality a basic principle not of moral but of political will
formation. For this reason Albrecht Wellmer, for example, distin
guishes a negatively understood and privately applicable moral prin
ciple from a public democratic principle of legitimacy that leads to
the formation of a common positive will: "In either case the way of
distinguishing between 'right' and 'wrong' involves an appeal to the
idea of a common will formed in the absence of coercion . . . . But
62
the nature of this appeal to a common will should be understood

differently in eacn case. In the case of moral judgment we are con
cerned with finding a way of acting in concrete situations which we
should be able, in B. Gert's terminology, to 'publicly advocate' as a
generalizable one. . . . The question that arises in each specific in
stance is whether we-as rational beings-are able to will that a
particular way of acting should become universal. And only the neg
ative answer to this question constitutes a moral 'ought. " '49 Against
this view I will show that the precedence accorded negative rights
does not follow from the deontological approach as such but only
from it in conjunction with a narrow, individualistic concept of the
person.
To the deontological distinction between the right and the good
corresponds the distinction between normative judgments about what
we ought to do and evaluative judgments about something in the
world that is more or less good or bad for us. This distinction draws
initial support from the fact that the grammar of commands is related
internally to action, which is not the case with the grammar of eval
uations. Normative judgments always relate to the choice between
alternative possibilities of action; evaluations relate to objects or states
of affairs independently of whether they are described as ends or
goods from the perspective of a purposively acting subject. More
important is the nonteleological significance of commands. Norm
guided actions do indeed have good or bad consequences for those
directly involved and those affected. But a moral judgment about an
action can claim general validity only because it is made with refer
ence to an underlying valid norm of action and does not depend on
an ad hoc evaluation of the consequences of action judged more or
less desirable by the standards of someone directly involved or some
collection of individuals affected. Evaluations express what in a given
case is more or less good and useful or bad and harmful "for me" or
"for us." Binary encoded commands prescribe which modes of action
are right or wrong, without reference to specific addressees and
without considering their contingent preferences or evaluative ori
entations. Murder and deceit are not wrong merely because they are
not good for those whom they victimize. As norms of action they are
wrong in general because they do not express a generaliwble interest.
Hence the unconditional or categorical character of normative valid-
63
ity would be compromised if the obligatoriness of impartially

grounded actions and norms of action were not divorced from values
and preferences that arise only from the evaluative perspective of
particular persons or reference groups. In this respect, normative
judgments are oriented to rules rather than ends.
Rules of action coordinate the plans of a variety of agents; ends
are the intended results of the interventions of an individual or
collective agent in the world. One who wishes to act morally decides
between alternative available courses of action in the light of norms
he believes to be right, and the correctness of these norms is deter
mined by whether they are equally in the interest of all potentially
affected . One who wishes to act purposively decides between alter
native available courses of actions in the light of preferred goals and
selects his goals in the light of values. From a deontological viewpoint,
therefore, moral deliberations must be kept completely free from
goal-directed reflections. Because general prohibitory norms un
doubtedly satisfy this condition, they appear to be capable of ground
ing moral duties in an exemplary fashion. Against this I will show
under the three above headings (a) through (c) that positive norms
of action can also ground duties in the strict sense.
In regard to (a), prohibitions of the form "You should not kill"
create the impression that such conduct is forbidden "uncondition
ally" in the sense of strict generality-that is, for everyone under all
circumstances and for all time, in short, categorically. This, at least, is
how Kant understood the validity of the maxims justified as general
laws by the categorical imperative; the meaning of moral obligation,
the meaning of the moral "ought," is expressed with special force in
the notion of unconditional or absolute validity. On closer inspection,
though, we find that negative rights and duties can no more claim
"absolute" validity than can positive duties.
An untruthful statement that saves the life of another is no less
morally commanded than killing in cases of self-defense or refraining
from offering assistance to avoid a greater evil are morally permis
sible. Valid norms are valid only in a "prima facie" sense. Regardless
of whether they rest on double negations, all rights and duties play
the same role in discourses of application, namely, that of reasons.
In cases of conflict between norms, it can be shown only on the basis
of a maximally complete description of all relevant features of the
64
given situation which of the competing norms is appropriate to a

particular case.so Norms that are overruled by this "single appropri
ate" norm do not lose their validity because they are not "pertinent."
They remain valid even if they do not apply to the given case. This
already entails that the deontological force of normative validity can
not be interpreted simply as an unconditional or absolute ought, as it
is by an ethics of conviction.
On the contrary, normative validity has the intersubjective sense
that a behavioral expectation is equally good "for all," not that it has
desirable consequences for a particular addressee: a connection with
good and bad is not categorically excluded, but the norm is qualified
through generalization as a regulation that is equally good for all.
Because of these intersubjective relations inscribed in moral norms,
no norm, regardless of whether it involves negative or positive rights
and duties, can be justified or applied privately in the solitary mon
ologue of the soul with itself. It is not a foregone conclusion that
maxims generalizable from my point of view must also be acknowl
edged to be moral obligations from the perspective of others, let
alone of all others. Kant could disregard this fact because, as I noted,
he assumed that all subjects in the Kingdom of Ends share the same
conception of themselves and of the world. With this abstract preun
derstanding at the level of "consciousness as such" is correlated, at
the level of the world of appearances, the assumption of an abstract
equality of interest among individual persons who are conceived, on
the model of possessive individualism, as standing in a relation of
ownership to themselves.5 1 Once we abandon these premises it be
comes imperative to submit all norms--and not just general prohib
itions-to a public, discursive, generalization test that necessitates
reciprocal perspective taking.
In regard to (b), prohibitions of the form "You should not kill"
have the merit of determinateness. With positive duties, goal orienta
tions that require agents to intervene in the world with a view to
consequences are imported into the moral sphere. Whereas it is
always within the power of an agent to refrain from performing an
action, the realization of an intended state of affairs is never solely
dependent on the good will of the agent alone but also on the con
tingencies of available means, the causal connections with other
events, and the limitations of a situation of action. Positive norms of
65
action render prognostication indispensable, as well as a concrete

expenditure of effort, energy, and attention; and with this there arise
the problem of imputing responsibility for the consequences and side
effects of action, the problem of the reasonableness of imposing
moral demands (Zumutbarkeit) and of the division of moral labor
between different agents, and in general the problem of assessing
results in the light of goals that admit of greater and lesser degrees
of fulfillment. Whether an act of providing assistance is appropriate,
whether it can be reasonably required of someone, whether it should
be regarded as sufficient in a given instance: these issues depend on
the circumstances. The difficulties that must be dealt with in dis
courses of application increase as (more or less specific) ends are
integrated into the content of norms.
Their relative freedom from the burden of problems of application
establishes at most that negative duties enjoy a heuristic advantage,
not that positive duties should be completely excluded from the
sphere of what is morally justifiable, for problems of application can
never be entirely ignored in the case of negative duties either. A
categorical privileging of negative rights would call for a different
justification. Let us consider, for example, the list of negative duties
to which Bernard Gert accords privileged status:52
1. You should not kill anyone.
2. You should not cause pain to anyone.

3. You should not prevent anyone from developing his abilities.
4. You should not deprive anyone of freedom or opportunity.
5 . You should not restrict anyone's possibilities of satisfying his

needs.
6. You should not deceive anyone.

With such intuitively evident examples in mind, Charles Fried out
lines corresponding categorical rights as the core of a universalistic
morality. In his view negative basic norms are defined by precisely
those duties that have as their content the protection of the integrity
of a person as a freely acting being: "What we may not do to each
other, the things which are wrong, are precisely those forms of per
sonal interaction which deny to our victim the status of a freely
66
choosing, rationally valuing, specially efficacious person, the special

status of moral personality."53
Here Fried expresses an intuition deeply rooted in the liberal
tradition: the integrity of the person seems to be adequately protected
only by the general, unconditional, and negative duties that assure
everybody the same zones of inviolability of subjective freedom by
imposing obligations to refrain from certain actions. However, this
conclusion follows only given Fried's underlying assumption of an
individualistic concept of the person that privileges action guided by
interests and informed by the rationality of choice, and hence a type
of action that has in fact become dominant under modern social and
economic conditions. Freedom of choice, rational determination of
ends in the light of analytically clarified preferences, informed choice,
and efficient employment of appropriate means-these notions cir
cumscribe competences demanded by the success-oriented actions of
an egocentric, self-regarding agent. Interestingly enough, these con
ditions are far from sufficient for defining the status of the person
deemed worthy of protection from the moral point of view. For we
encounter a person as essentially worthy of moral respect only when
we adopt the performative attitude toward a second person. Only
when at least two people encounter each other in the context of an
intersubjectively shared lifeworld with the goal of coming to a shared
understanding about something can-and must-they mutually rec
ognize each other as persons capable of taking responsibility for their actions
(zurechnungsfiihige Personen) . They then impute to each other the
capacity to orient themselves to validity claims in their actions.
From this viewpoint the privileged status accorded basic duties such
as "You should not lie" can be accounted for in a different way. To
these core duties belong only those that can be understood as aspects
of the general demand : "Act with an orientation to mutual under
standing and allow everyone the communicative freedom to take
positions on validity claims." They are fundamental because they are
oriented to respect for the integrity of communicatively acting sub
jects. But these norms do not just have the force of purely negative
duties. In behaving truthfully I do not merely refrain from deception
but at the same time perform an act without which the interpersonal
relation between performatively engaged participants in interaction
dependent on mutual recognition would collapse. The norms that
67
prescribe the fulfillment of the necessary pragmatic presuppositions

of communicative action as a duty are strangely indifferent regarding
the distinction between negative and positive duties: by showing re
spect for another person, I at the same time protect the vulnerable
core of his person. Hence, it is no coincidence that other norms of
this kind, (for example, those that oblige us to respect normative
validity claims) are formulated in positive terms. Gert's ten rules also
include such positive duties :
7 . You should keep your promises.
1 0. You should do your duty.

The fundamental status of the kinds of commands that Gert includes
in his catalog of duties cannot be explained in terms of the deter
minateness of norms prescribing omissions but only in terms of their
self-referential character. These duties regulate precisely the neces
sary pragmatic presuppositions of communicative action from whose
normative content discourse ethics derives the basic substance of
morality by analyzing the universal and necessary communicative
presuppositions of the practice of argumentation-that is, the reflec
tive form of communicative action. Gert's ten rules are concerned
immediately with the integrity of the person himself as a symbolic
structure that is produced and reproduced through relations of re
ciprocal recognition; they are concerned mediately with the preser
vation and development of the bodily existence of the respect-worthy
person. The constitutional susceptibility of a personality structure
that is at the mercy of interpersonal relations is of even greater
moment than the more tangible vulnerability of the integrity of body
and life: the symbolic structure can disintegrate while the physical
substrate remains intact. For this reason, we sometimes risk death
rather than live a life devoid of freedom. This insight is indeed open
to ideological misuse, but its truth is not thereby denied as such.
In regard to (c) , because a person can be individuated only through
socialization and because an individual's inner center develops only
as he engages in communicative structures of reciprocity, morality,
which counteracts this vulnerability, reflects the intersubjective struc
ture of this vulnerable identity itself. Hence the integrity of the
individual person, which calls for equal respect for all, cannot be
safeguarded without simultaneously safeguarding the social fabric of
68
relations of reciprocal recognition. Just as justice and solidarity are

simply two sides of the same coin, so too negative and positive duties
spring from the same source. If rights and duties are to foster the
integrity of individuals who are by their very nature socialized, then
the constitutive social context of interaction is not something second
ary for those whose lives and identity are made possible and sustained
by it. Omissions are no less a potential threat to personal integrity than
injuries actively inflicted.
In complex societies, of course, claims to a fair share of scarce
social resources-positive rights to basic goods such as food and
housing, health care, education, and opportunities for work--can be
satisfied only through organizations. In this context individual rights
and duties are transformed into institutional rights and duties: the
organized society as a whole is the subject of obligations--claims to
positive rights are raised against it.
Society as a whole is an anonymous addressee, a fact that generates
complications over and above the problems already mentioned. In
the case of positive rights, problems of what can be reasonably ex
pected and of accountability, which call for a moral "division of labor,"
are particularly pressing.54 These problems become even more acute
when the individual, as a member of an organized state, raises claims
against institutions to benefits resulting from institutionally mediated
positive duties, while at the same time fulfilling the role of an ad
dressee of such claims. These problems cannot be solved at the level
of moral rules; they call for mechanisms of coordination such as law
and political power. What interests me here from the moral point of
view is solely the fact that the relation between negative and positive
duties is represented differently from the liberal tradition once we
abandon the narrow, individualistic concept of the person.
The deontological force of normative validity claims cannot be
explicated in terms of an abstract primacy of negative rights of free
dom over positive rights of participation, for, as far as the integrity
of essentially socialized individuals is concerned, these categories of
rights are coeval. Fried too recognizes this: "Respect for our common
humanity has provided the basis for the account of the categorical
wrongs and the corresponding negative rights. Now, the same com
mon humanity is the source of positive rights as well . . . . Common
human nature is not merely something which each of us possesses
69
singly. . . . It is also like a single thing we all share, like the common
thread that runs through each bead in a string. There are aspects of
common humanity which we share because of the efforts of others
to produce them : the fruits of common labour, the security of civil
society, the riches of culture and civilization, the fact of language. "55
If we take these reflections seriously and elaborate them through a
theory of communication, we will no longer accord negative rights
and duties privilege over positive rights and duties on account of
their allegedly categorical character, the determinateness of their
content, or the lack of ambiguity concerning their addressee and his
undifferentiated duty. It is quite sufficient to appeal to a common
process of discursive will formation that subjects all norms to the
same standard-the capacity to command general assent-and se
cures the deontological force of their validity against an unlimited
orientation to consequences by admitting only regulations that are
equally in the interest of all. Only the universalistic privileging of
what is equally good for all brings the moral point of view to bear in
the justification of norms.
9. Thus far I have assumed that the deontological distinction between
the right and the good, as well as the primacy of what is morally
right or obligatory over what is ethically desirable or preferable, are
unproblematic. But this involves a prior determination against the
possibility of construing morality as one aspect of a more compre
hensive ethics of the good and of grounding it within this framework.
This decision may reflect a specifically modern prejudice-thus, at
any rate, argue defenders of classical approaches. Charles Taylor, for
example, claims that the modern morality of justice is based on a
selective understanding of modern identity. Once we grasp the latter
in its full complexity, the underlying ethical strata of competing
fundamental goods, in which the modern principles of justice and
solidarity are rooted in turn, will be uncovered. 56 From this perspec
tive the right loses its primacy over the good ; indeed, the deontolog
ical distinction between questions of justice and questions of the good
itself becomes problematic. By endeavoring to revive the claims of
an ethics of the good under modern conditions, Taylor is disputing
not the universalistic claim of the morality ofjustice but the autonomy
of rational morality. At the same time, he raises the issues of the
70
place of moral theory in philosophy as a whole and of the tasks to

be accomplished by philosophy as such, for if it were open to philos
ophy to uphold its classical claim to make universally valid statements
about the meaning of the good, or not unsuccessful, life, then it
would also have to accept the task of privileging one particular way
of life over others-for example, the classical project of living an
examined life. It would not have to limit itself to a formal account
of the moral point of view or of the ethical problematic but would
clarify the concrete totality of the existential orientations within which
the limited role, but also the existential significance, of morality first
become comprehensible.57
Taylor's critique does in fact make explicit certain background
assumptions of the modern morality of reason. Deontological ap
proaches that take the normative validity of moral commands and
norms as the phenomenon in need of explanation at the outset do
indeed presuppose that moral prescriptions are oriented to the con
sensual resolution of conflicts of action. This involves at least four
distinct assumptions. (1) Morality is tailored to interactive relations
between subjects capable of action and not to goods-that is, things
in the world that can acquire value and importance for individual
agents. In this way morality is set apart from the beginning as a social
phenomenon and is delimited from individual aspirations to happi
ness, existential problems, and sensuous needs. (2) Morality relates
to interactions regulated by norms. These bind their addressees to
fulfill reciprocally interconnected behavioral expectations by selecting
a few possibilities from a broad spectrum of value orientations and
making them generally binding. Morality is thereby restricted from
the outset to what admits of normative validity and invalidity, and
thus to rights and duties, and is demarcated from the axiological
sphere of what is to be preferred and optimized. (3) Morality is
understood as a peaceful alternative to violent resolution of action
conflicts for which there is no equivalent. Moral commands must be
capable of gaining the assent of the parties to the conflict and must
themselves be shown to be legitimate or worthy of recognition in
respect of what they impose as a duty. In this way morality is secured
against all purely naturalistic interpretations. (4) From this assump
tion there follows, where recourse to collectively binding religious or
metaphysical worldviews is no longer available, the more far-reaching
71
assumption that a consensual justification of rights and duties can be

generated only through argument, that is, through the cogency of
good reasons.
Taylor's critique is sparked by the question of the rationality of
rational morality. Kant sought to ground morality in a postmeta
physical fashion, not by appeal to a totality higher than the human
mind but from reason itself as the source of world-constitutive ideas.
Taylor too does not want to fall back on traditional natural law; he
wants to eschew any direct recourse to an objective realm of essences
or values. At the same time, however, he regards an autonomous
grounding of morality in reason to be impossible. A morality that
rests solely on a rational procedure of impartial judgment must, in
his view, alienate itself in a subjectivistic manner from sources of
motivation and relinquish the possibility of illuminating the back
ground in terms of which alone the existential meaning of moral life
becomes comprehensible: "It has no easy way of capturing the back
ground understanding surrounding any conviction that we ought to
act in this or that way-the understanding of the strong good in
volved. And in particular, it cannot capture the peculiar background
sense, central to much of our moral life, that something incomparably
important is involved" (87). But if we cannot justify moral commands
without first having experienced the relevance of the whole dimension
through acquaintance with "higher-order goods," the differentiation
between the right and the good, and above all the primacy of the
right over the good, become problematic. Then the justice of the
morality of j ustice and the rationality of rational morality are not
ultimate but derive their inspiration, impulse, and pathos from an
tecedent commitments and affective states:
The more one examines the motives-what Nietzsche would call the "gen
ealogy"---o f these theories of obligatory action, the stranger they appear. It
seems that they are motivated by the strongest moral ideals, such as freedom,
altruism and universalism . These are among the central moral aspirations of
modern culture, the hypergoods which are distinctive to it. And yet what
these ideals drive the theorists towards is a denial of all such goods. . . .
Impelled by the strongest metaphysical, epistemological, and moral ideas of
the modern age, these theories narrow our focus to the determinants of
action, and then restrict our understanding of these determinants still further
by defining practical reason as exclusively procedural. They utterly mystify
the priority of the moral by identifying it not with substance but with a form
72
of reasoning, around which they draw a firm boundary. They then are led
to defend this boundary all the more fiercely in that it is their only way of
doing justice to the hypergoods which move them although they cannot
acknowledge them. (88.)
In what follows I will present the alternative Taylor proposes and

examine whether an ethics of the good consonant with the modern
identity, though universalistic in aspiration, can indeed be defended
with the cognitive resources available to postmetaphysical reflection. 58
Taylor is neither a metaphysician in his theoretical orientation nor
an antimodernist in the conclusions he arrives at, but his Catholic
skepticism toward the potential self-sufficiency of a purely secular,
proceduralist ethics leads him to cling to the classical claim of philos
ophy. He wants to explicate the meaning of the fulfilled life at least
to the degree that it sheds light on the "modern identity." Taylor is
interested not merely in a descriptive cultural history of the config
urations of values that have attained preeminence in the modern era
but in the justification of the self-understanding that has become
ineluctable for us in the modern age. This analysis is by no means
value neutral; on the contrary, it makes fundamental value orienta
tions explicit and understands itself as an ethics of the contemporary era.
This accounts for the theoretical orientation to an ethics of the good :
goods are an objective matter and not merely a reflection of subjective
values and preferences. The systematic results of Taylor's historically
informed investigation can be summed up in four theses :
(a) The different ethical systems we encounter provide general
orientation concerning the meaning of life, the integrity of one's
fellow human beings, and one's own dignity; they provide standards
for strong evaluations, that is, an orientation to what has its end or
purpose within itself. We judge our own goals and purposes in the
light of sud) ultimate ends. This objectivity justifies us in speaking in
an ontologizing fashion of "higher-order goods" : "these ends or
goods stand independent of our own desires, inclinations or choices"
(20). Strong evaluations belong to the ontological constitution of a
being who finds himself situated in a force-field of attractions and
repulsions and must seek the meaning of life through his orientation
to the highest goods.
(b) The ethics of the good presupposes that the good is something
objective and independent of the wills of particular subjects. How-
73
ever, the horizon of goods and values is disclosed only through a

language in which subjects can articulate their ethical feelings and
judgments in an appropriate fashion. Of course, the "vision of the
good" can be made accessible to experience only through linguistic
articulation and cannot be justified in a strict sense. We must be able
to say what motivates us most deeply; we must be able to articulate
what constitutes a fulfilled life as such. Taylor falls into an ontological
mode of thought, posing the question of the Good underlying indi
vidual goods, of the "constitutive Good" that first confers on goods
the motivating quality of what is binding and worthy of allegiance
that infuses subjects with passion and moral pathos. The Platonic
Idea of the Good, which we grasp through contemplation, the agape
of the benign Christian God, or the wellspring of autonomy, the
Kantian capacity of the subject to bind his own will through insight:
these ideas are "moral sources" of which modern moral theories have
lost sight.
(c) Modern identity draws on three different moral sources simul
taneously: the Christian notion of the love of God, of whose goodness
all creation partakes ; the Enlightenment notion of the self-responsi
bility of the subject, who in virtue of his reason is capable of acting
autonomously; and, finally, the romantic belief in the goodness of
nature that finds expression in the creative accomplishments of the
human imagination : in the work of art and in the self-realization of
the individual (understood on the model of artistic production).
There is indeed broad consensus concerning the fundamental values
of freedom, justice, welfare, and the eradication of suffering, "but
under this general agreement, there are profound rifts when it comes
to the constitutive goods, and hence moral sources, which underpin
these standards" (495) . Taylor thinks that modern identity is prey to
a shallow, unreflective, and, to that extent, false self-understanding
that obstructs the sources of enthusiasm for the good. Today selective
and deficient views of modernity compete against each another. This
war of gods and demons can be resolved only by an ethic that makes
explicit in a different fashion the existential tension between God's
word, human reason, and the creative capacities of inner nature. just
this is what Taylor seeks to accomplish.
(d) It is one thing to make a diagnosis and elucidate it historically
and quite another to administer the therapy. Taylor is caught in a
74
dilemma and knows it. His goal is not merely to describe modern
identity by exploiting the resources of the history of ideas but to
justify the identity outlined as a formation of fundamental orienta
tions that is ineluctable and authoritative for us (and for all other
moderns). An ethics of the good that discloses the order of consti
tutive goods as a publicly accessible reality can realize this ambitious
argumentative goal. But under the premises of postmetaphysical
thinking, this route is closed to us. Arguments will not suffice to open
the eyes of the "value-blind" children of modernity to the efficacy of
the highest goods; that requires the world-disclosing power of an
evaluative language that first lifts the scales from our eyes: "There is
no coherent place left for an exploration of the order in which we
are set as a locus of moral sources, what Rilke, Pound, Lawrence and
Mann were doing in their radically different ways . . . . The order is
only accessible through personal, hence 'subjective', resonance" (5 1 0).
Thus philosophy must pin its hopes on art. Only in aesthetic expe
rience, freed from the spell of anthropocentric thinking, do we en
counter something objective capable of awakening our sense for the
good : "The great epiphanic work actually can put us in contact with
the sources it taps" (5 1 2).
The reflections summarized under (a) through (d) point to a divi
sion of labor among philosophical ethics, art, and aesthetics. But this
evasion reveals the epistemological impasse in which a metaphysical
ethics of the good finds itself. The aesthetic considerations that Taylor
adduces are themselves problematic. In the wake of Adorno and
Derrida, what can be accomplished by modern art can scarcely any
longer be construed as "epiphany. "59 Moreover, Friedrich Schlegel
already saw that what is distinctive of modernity is that the "aesthetic"
tears itself loose from the good and the true. Modern art can no
longer be tapped as a source of the moral. But even if we could
accept an aesthetics that still believes in the ethical relevance of the
world-disclosing power of modernism , its implications for philosophy
would be of a renunciatory nature: it would either have to resign itself
to the role of aesthetic criticism or itself become aesthetic. At any
rate, it would have to abandon any pretension to convince on the
basis of its own arguments. Adorno faced a similar conclusion and
responded to it by developing a negative dialectics. But that does not
75
accord with Taylor's philosophical goal of bringing a modern ethics

of the good to bear within the discourse of experts.
Regardless of how we assess Taylor's proposed alternative, the fact
that an ethics of justice has nothing to say concerning questions of
motivation raises a problem. Thus far I have understood the question
"Why be moral?" in such a way that it admits of a trivial answer: it
makes no sense "to try to prove a 'good will' to someone through
rational argumentation, that is, to conclusively compel him, as it were,
to translate his cognitive grasp of moral obligation into a correspond
ing act of will. "60 But Taylor has given the question a different, less
trivial meaning. Over and above the problem of weakness of will, the
willingness to look at a matter from the moral point of view depends
on both one's awareness of the moral dimension as such and one's
taking it seriously. Hence philosophy should protect us from becom
ing blind or cynical toward moral phenomena. It should, Taylor
believes, persuade us of the preeminent importance of the orientation
to the good ; it should sensitize us to the hidden dimension of the good
and infuse us with the strength for passionate engagement in the
cause of the good. But a postmetaphysical philosophy is too belated
to perform the one task, the awakening of moral sensibility, and is
overtaxed by the other, that of overcoming moral cynicism.
Philosophy is overtaxed by what Apel terms "the existential ques
tion concerning the meaning of being moral."61 For moral despair
demands an answer to the fundamental ethical question of the mean
ing of life as such , of personal or collective identity. But the ethical
existential process of reaching an individual self-understanding and
the ethical-political clarification of a collective self-understanding are
the concern of those involved, not of philosophy. In view of the
morally justified pluralism of life projects and life-forms, philoso
phers can no longer provide on their own account generally binding
directives concerning the meaning of life. In their capacity as philos
ophers, their only recourse is to reflective analysis of the procedure
through which ethical questions in general can be answered. As to the
propaedeutic task of awakening the faculty of moral perception,
philosophy is not in a privileged position when it competes with the
rhetorically moving, exemplary representations of the novelist or the
quietly insistent intuitions of common sense . We learn what moral,
and in particular immoral, action involves prior to all philosophizing;
76
it impresses itself upon us no less insistently in feelings of sympathy

with the violated integrity of others than in the experience of violation
or fear of violation of our own integrity. The inarticulate, socially
integrating experiences of considerateness, solidarity, and fairness
shape our intuitions and provide us with better instruction about
morality than arguments ever could .
Theories are directed against other theories, but at least they can
correct bad theories that tend to alienate us from our better moral
intuitions. The demonstration that the moral point of view, as ex
pressed by (U), is generally valid and does not merely express a
culture-specific or class-specific evaluative orientation prevents us
from succumbing to a relativism that robs moral commands of their
meaning and moral obligations of their peculiar force. Hence moral
theory is competent to clarify the moral point of view and justify its
universality, but it can contribute nothing to answering the question
"Why be moral?" whether this be understood in a trivial, an existen
tial, or a pedagogical sense. While Apel shares this view, at the same
time he interprets the question in yet another sense, as "the question
of the ultimate rational ground of 'moral existence."' Understood in
this sense it supposedly calls for a philosophical answer and receive
one in the form of an "ultimate justification" (Letztbegrilndung) .
1 0. A pel proposes the following intuitive elucidation of the idea of

an ultimate justification :
It is mistaken to assume that questions such as "Why be moral?" or "Why be
logical?" or "Why be rational?" must be answered either by a deductive
justification or an irrational decision. In reality the situation in which the
problem is supposed to arise does not even exist, that is, the situation in
which we are faced with a decision for or against reason, logic, or morality
and could nevertheless already engage in argument, or at least pose the why
question . . . . One who poses the why-question in earnest has thereby already
entered the terrain of argumentative discourse, which means that he can
assure himself through reflection on the meaning of his act that he has
already necessarily recognized the rules of cooperative argumentation and
hence also the ethical norms of a communication community. 62
I have deployed a similar argument to justify (U) as a rule of argu

mentation. 53 The goal is to establish two nontrivial results: that ques
tions of justice do admit of rational answers and how-that is, by
77
means of which rules of argumentation or which moral principle

they can be answered. Transcendental-pragmatic arguments are de
signed to remind anyone who so much as enters into argumentation
that he is already participating in a normatively structured practice.
We must be careful not to read too much into this statement. The
normative content of the general presuppositions of argumentation
is appealed to in this way solely in response to the epistemic question
of how moral judgments are possible and not the existential question
of what it means to be moral. Contrary assertions notwithstanding,
A pel appears to saddle the demonstration of the moral principle with
the latter more ambitious claim. He regards the issue of whether we
can view ourselves as capable of moral insight as depending on
whether we are already in a certain sense moral. This is reminiscent
of a figure of thought familiar from religion and metaphysics: we
may regard ourselves as capable of "the truth" because we are already
"within the truth." This thought could be understood in an unob
jectionable sense in the present context. Our argumentative praxis
belongs to a form of life that is completely structured by communi
cation, and everyday language use oriented to reaching understand
ing is already tied to unavoidable pragmatic presuppositions that are
by no means morally neutral. In communicative action as such, we
always already move within the boundaries of a lifeworld saturated
with ethical value. However, what Apel has in mind cannot be moral
existence in this sense. Demonstrating the substantive normative pre
suppositions of argumentation is by no means sufficient to reveal the
foundational interrelation of discourse, action oriented to reaching
understanding, and the lifeworld ; that is a matter for a formal prag
matic analysis of language, not for moral theory. Moreover, Apel's
notion of ''moral existence" has the connotation of obligatoriness that
transcends the particularity of the concrete ethical substance of a
lifeworld-the universalistic standard of morality "cuts into the eth
ical substance of the lifeworld. " In addition to an epistemic claim,
Apel invests practical reason with a more far-reaching existential
obligatoriness and elevates the justification of the moral principle to
the status of an "ultimate justification ." Although Kant's practical
reason is not restored to its position of preeminence in the intelligible
realm, it nonetheless retains something of its omnipotence.
78
Let me first clarify what is involved in this existential claim. A

cognitivist ethical theory understands the operation of practical rea
son in purely epistemic terms. It is content to point out that, in order
to provide answers to moral questions of the type "What should I
do?" that take the form of singular judgments, participants engage
in discourses concerning application in particular cases and, when
necessary, must engage in justificatory discourses. And it belongs to
the normative sense of every moral judgment that it makes a duty of
the action expressed in its propositional content. But a supernorm
that made a duty of acting in accordance with duty could say no
more than what is already contained in the normative import of an
individual moral judgment. A supernorm that made the realization
of justice in general a duty in a reflexive manner would be either
redundant or meaningless and hence would not need to be justified.
I suspect that Apel's "ultimate justification" would amount to just
that: the justification of a supernorm that states that justice as such
ought to exist. This is how we must understand Apel when he criti
cizes Rawls "for failing to offer a prior justification for why justice as
such should exist."64 Again he criticizes Kohlberg for failing to pro
vide a justification for the claim that "postconventional conscience
. . . should make a moral rather than, for example, an egoistic-stra
tegic use of the capacity for 'role-taking."'65 By insisting on this de
mand, Apel ultimately invests the question concerning morality with
existential significance.
What more could this existential obligatoriness add to the epistemic
function of practical reason ? In everyday practice, we can indeed
suspend the presuppositions of communicative action in a concealed
or open manner; at any time we can switch from an orientation to
reaching understanding to that of a strategically acting subject con
cerned with his own success. All agents to whom freedom of will is
attributed must be capable of making such decisions. But it is quite
another issue whether strategic action is permitted or should be
forsworn in a given situation, and thus whether it is morally admis
sible. That morally inadmissible actions should be forsworn is already
expressed by the relevant norms. There is no need for a supernorm
that makes the observance of moral commands a duty in a self
referential fashion. It would not augment in the least the binding
force of elementary duties. Since this is so obvious, those who seek
79
an ultimate justification must have some other end in view. I suspect

that the meaning of an "ultimate justification" cannot be explicated
solely within the framework of moral-theoretical reflections but must
be understood in terms of a lingering (though unacknowledged)
foundationalism that still pervades the architectonic disposition of
Apel's philosophy. Only when philosophy takes upon itself problems
it can solve through neither self-reflection nor self-referential argu
mentation does the philosopher find himself constrained to offer
"ultimate justifications."
For Taylor, as we have seen, the proper task of philosophy is an
ethical one-specifically, the traditional task of clarifying the meaning
of life. A pel, by contrast, views the deontological question concerning
the principle of justice as the pivotal issue of philosophy, though he
is aware that a justification of the moral does not of itself provide a
solution to the problem of meaning and of realizing individual or
collective happiness.66 In his view there is a relation of complemen
tarity between the standard of morality first articulated by Kant "and
the problem of personal or communal ethical life in relation to the
concrete universal of a whole form of life."67 The general statements
of philosophy, while they express quite well what is essential to mo
rality, do not extend to the substantive ethical life embodied in tra
ditional or utopian forms of life. But this declaration of abstinence
is ambiguous, for philosophy is supposed to clarify the meaning of
the moral not just in the sense that it makes intelligible what it means
to judge something from the moral point of view. It is supposed to
explain what it means to be moral-the unique significance of morality
in life as a whole-and thereby provide the will with a rational incen
tive to justice as such. But Apel thereby seeks to ')ustify" philosoph
ically something that cannot be achieved through argumentation. A
good will is awakened and fostered not through argumentation but
through socialization into a form of life that complements the moral
principle. A comparable effect may also be produced by the world
disclosing power of prophetic speech and in general by those forms
of innovative discourse that initiate better forms of life and more
reflective ways of life-and also the kind of eloquent critique that
enables us to discern these indirectly in works of literature and art.
World-disclosing arguments that induce us to see things in a radically
different light are not essentially philosophical arguments and a for
tiori not ultimate justifications.
80
There is, of course, an internal relation between moral conscious

ness and ego identity. Postconventional moral consciousness needs to
be supplemented by an enlightened existential self-understanding
that entails that I can respect myself only as someone who as a general
rule performs the actions he takes to be morally right. But while
moral consciousness can "foster" this complementary structure in a
weak sense, it cannot guarantee it. Whether I have sufficient resolve
to act in accordance with my moral insights even when they are
opposed by strong interests of a different kind does not depend
primarily on my capacity for moral judgment or on the level of
justification of moral judgments but on my personality structure and
form of life. This problem of weakness of will cannot be solved
through moral cognition.
Apel might cite this fact in support of the thesis that more must
be accomplished by philosophy than is allowed by a crude cognitivist
ethical theory; it must aim at philosophical enlightenment and trans
form the existential self-understanding and worldview of members
of modern societies through its claim to universal validity. At first
sight, this thesis is incompatible with the goal of providing only
answers that do not depend on metaphysically privileging one of the
irreducible plurality of interpretations of life that today have equal
claim to our allegiance. I suspect that Apel hopes to avoid this con
tradiction, by construing communicative reason as in essence moral
practical reason and then, in virtue of this Fichtean primacy of prac
tical reason, according the philosophical-explanatory discourse distin
guished by self-referentiality a preeminent position in the hierarchy
of scientific discourses. Discourse ethics is supposed to remain neutral
over against the plurality of belief systems yet not pay the price of
renouncing substantive sources of motivation entailed by its proce
duralism. Apel believes that he can find an Archimedean point of
self-reflection for the sphere of philosophical reflection dealing with
the discourse principle. From this vantage point, philosophy is sup
posedly able to provide an ultimate justification that encompasses
both theory and practice through the self-clarification of argumen
tation as such. ss
Apel's theoretical framework, as I see it, rests on two problematic
assumptions. The first premise-the privileging of discourse ethics on
the basis of fundamental-philosophical considerations-Ape} adopts
81
at the cost of assimilating communicative to practical reason. By

drawing on the resources of the theory of argumentation, the concept
of communicative reason can be developed in the direction of an
analysis of validity claims and the conditions of their discursive re
demption. This involves in part the analysis of general presupposi
tions of argumentation. The more ambitious task of investigating
different forms, and corresponding rules, of argumentation leads to
specialized theories. In this way, too, a moral principle can be
grounded and what it means to regard something from the moral
point of view explicated. This addresses a specific kind of question,
a specific aspect of validity in general: one moment of an encompas
sing communicative reason.
The latter has normative import only in the broader sense that a
communicatively acting subject must accept pragmatic presupposi
tions of a counterfactual kind. He must make certain idealizations
for example, ascribe identical meanings to linguistic expressions,
make context-transcending validity claims for his utterances, or re
gard his addressees as accountable subjects. The communicative
agent thereby becomes the subject of a "must" in the sense of weak
transcendental necessitation, without encountering the prescriptive
"must" of a rule of action, whether the latter is understood deonto
logically in terms of the normative validity of a moral command,
axiologically in terms of a constellation of preferred values, or em
pirically in terms of the efficacy of a technical imperative. Commu
nicative reason, unlike practical reason, is not itself a source of norms
of right action. It spans the full spectrum of validity claims (of asser
toric truth, subjective truthfulness, and normative rightness) and
hence extends beyond the sphere of moral-practical questions. Nor
mativity in the more restricted sense of a binding practical orientation
is not identical with the rationality of action oriented to reaching
understanding as a whole. Normativity and rationality overlap in the
field of justification of moral insights that are attained in a hypo
thetical attitude and, as we have seen, generate only weak rational
motivations and at any rate cannot sustain an existential understand
ing of self and world.
The second premise-the privileging of a foundational discourse in
general-is directly related to Apel's project of reformulating the
basic problems of the first two paradigms of first philosophy-ontol-
82
ogy and the philosophy of consciousness-in terms of a third, that

of a pragmatics of language. Underlying this "transformation of
philosophy"69 is the idea that the substantive contents of metaphysics
can be preserved only in the form of global scientific hypotheses that
have a fallible status, whereas transcendental reflection on the con
ditions of objectively valid experience and argument as such uncovers
a realm of genuine philosophical knowledge, which, itself infallible,
accounts for the presuppositions of fallibilism and thereby satisfies
the conditions of an ultimate grounding.70 It is supposed to be pos
sible to perform this justification of validity "from above" in one fell
swoop, so to speak, "in strict reflection" on the general presupposi
tions of argumentation, and to secure the presuppositionlessness of
original philosophical reflection for a philosophy that clarifies the
presuppositions of all possible knowledge. The notion of ultimate
justification suggests a residual foundationalism limited to the analysis
of transcendental conditions, which A pel defends by appealing to the
unique status enjoyed by the philosophical determination of presup
positions of argumentation in general. On this account, philosophy
goes beyond the autonomous, differentiated spheres of methodically
secured knowledge and enjoys a special immunity; its preeminent
position in the hierarchy of forms of knowledge makes it invulnerable
to externally produced cognitive dissonances. Philosophy thereby be
comes the final haven of certainties that are otherwise unavailable on
the assumptions of a generalized fallibilism.
If this is the import of the claim to provide ultimate justifications,
two objections immediately suggest themselves. First, the abstract
opposition of philosophy and science is a carry-over from the theo
retical framework of a philosophy of consciousness that posits the
basic questions of epistemology as fundamental. After the turn to the
pragmatics of language, the conditions of possible agreement instead
become the focus of interest. But the objectifying sciences-as the
reflective form of our knowledge of the objective world-thereby lose
their special status. They take their place among the diverse forms
of knowledge operative in everyday life and in expert cultures among
which philosophy mediates in the role of an interpreter without
needing to "ground" any of them. At the same time, philosophy is
involved in diverse forms of cooperation within the system of scien-
83
tific knowledge, not just as a contingent matter but for methodolog

ical reasons. It has already forfeited its role as judge and director for
the simple reason that there is no hierarchical gradation between
discourses and corresponding metadiscourses.7 1 The metatheoretical
interconnection between theoretical results of the now-autonomous
disciplines and spheres of knowledge is henceforth assured only by
coherence, not by "grounding."
Moreover, both the hermeneutical-interpretive approach to sym
bolically prestructured objects and the rational reconstruction of the
pretheoretical knowledge of competent judging, speaking, and acting
subjects suggest a close kinship between philosophy, on the one side,
and the humanities and the social sciences, on the other. The abstract
opposition between philosophy and "science" (conceived on the
model of the nomological empirical sciences) can no longer be sus
tained. 72 The distinctiveness of philosophy consists in neither its
preoccupation with conceptual analysis nor the universalistic char
acter of its questions. What sets philosophy apart is the self-referen
tiality of some of its arguments. But the self-referentiality of the
undoubtedly central analysis of general presuppositions of argumen
tation that can be made only by participants in argument does not
secure the philosophical enterprise the independence and infallibility
A pel associates with the notion of an ultimate justification.
This second objection relates to the status and meaning of tran
scendental arguments, a topic I cannot here go into in detail. I will
simply recall that thus far no equivalent has been found for Kant's
transcendental deduction of the categories of the understanding, and
there is little prospect that one will be found. Without it we are left
with weak transcendental arguments in Strawson's sense. 73 These can
establish only the factual inescapability of the general and necessary
presuppositions of a nonreplaceable discursive order, such as the
language game of argumentation. Because we do not in fact have a
functional equivalent for rational discourses, we are left with no
choice: we unavoidably accept the pragmatic presuppositions of this
demanding form of communication because there is no alternative.
This demonstration of the factual inescapability of substantive nor
mative presuppositions of a practice internally interwoven with our
sociocultural form of life is indeed conditional on the constancy of
this life-form. We cannot exclude a priori the possibility of its under-
84
going changes. But that remains an empty possibility since-science

fiction scenarios that transform human beings into zombies aside
we cannot even imagine a fundamental alteration in our form of life.
The weak transcendental proof, if we can indeed carry it out as I
assume in agreement with Apel, suffices to ground the universalistic
validity claim (valid for all subjects capable of speech and action) of
a moral principle conceived in procedural terms. If this principle can
be deduced from the normative content of factually inescapable pre
suppositions, it would be shown that it cannot be meaningfully placed
in question as such but only in some particular interpretation or other.
But we do not need an ultimate justification to show this. An expla
nation of the moral point of view that rests on fallible reconstructions
of the normative content of factually unavoidable presuppositions of
argumentation by no means signifies renouncing moral universalism
and tying the moral principle to elements of the background knowl
edge of a particular lifeworld. 74
An ultimate justification of ethics is neither possible nor necessary.
One can, of course, speculate whether a philosophy developed from
such an Archimedean point could provide ready answers to questions
that a rigorously cognitivist version of disq:mrse ethics must leave
unanswered. I am thinking, for example, of problems resulting from
the fact that a rational morality must reflect on the limits beyond
which adherence to its demands cannot be reasonably required or
on the legal-institutional arrangements through which the imposition
of special duties is mediated. A hierarchical philosophy that posited
a privileged reflective standpoint would offer better preconditions
for moral theory's once again overtaking and reflectively transcend
ing itself.
Apel presents what he calls "Part B" of his ethics as a process of
reflective self-transcendence. There he adopts the perspective of a
politician acting on the basis of an ethics of responsibility and takes
his orientation from cases where strict adherence to moral require
ments would lead to irresponsible consequences because a strategi
cally acting adversary does not accept the preconditions for the
consensual resolution of an acute practical conflict. To prevent this
solitary representative who shoulders political responsibility for the
interests of the state from falling back into the sphere of unrestrained
power politics, Apel supplies him with a principle by which choices
85
between alternative strategies can be justified in accordance with an

ethics of responsibility. He must orient himself to the more long
trm goal of producing the conditions under which general observ
ance of moral rules and consensual principles first becomes a reason
able demand and then the current dilemma of power politics
disappears. Apel defines the goal of a cosmopolitan order, which
Kant outlined in terms of a philosophy of history, as a morally grounded
political goal of individuals who act in a representative capacity, and
summarily elevates this goal of realizing the indispensable conditions
of life under which moral action would also be a viable political duty
to the status of an auxiliary principle to be justified by moral theory. He
postulates the duty "to strive for moral progress and to regard its
historical realization as conceivable in relation to the historical pre
conditions and-through ever renewed hypothetical reconstructions
of history-to show it to be such. "75 I will not here go into A pel's
paradoxical subordination of thought to the imperatives of morality;
at present I am concerned with the undesirable consequences that
result from Apel's preempting issues within moral theory that can be
meaningfully addressed only after the transition to the normative
problematic of political and legal theory.
Apel's auxiliary principle introduces a teleological perspective into
deontological moral theory, as he himself remarks: the realization of
morality itself is elevated to the highest good. But he thereby explodes
the conceptual framework of a deontological ethics. In the first place,
it must be asked on what level the auxiliary principle is to be justified.
If on the same level as the moral principle itself, then one would
have to be able to offer some analogue of the analysis of the content
of general presuppositions of argumentation, which is not the case.
Again, the fact that the auxiliary principle is formulated as a special
duty addressed to holders of specific positions creates problems. An
analogous translation of the moral principle into a concrete duty
would take the form : "Always act in accordance with correctly applied
valid norms," which amounts to nothing more than the redundant
demand, "Act morally." If the auxiliary principle is to be viewed as
a special duty, it must, like all other norms, be treated on a different
level; that is, it must be subjected to generalization by participants in
action and those affected themselves in justificatory discourses. How
ever, that means that the principle would be tested under precisely
86
the presuppositions of argumentation whose sufficient fulfillment it

implicitly disputes. I cannot see how a form of argumentation that
thematizes itself in such a contradictory manner could be possible.
Aside from this, the auxiliary principle calls for a structurally in
coherent mode of action, one that is both morally required and at the
same time to be pursued in a purposive manner. Although an ethics
of the good can enjoin a course of action oriented to values in a
normative fashion, deontological positions cannot make a duty of
action oriented to success. With his auxiliary principle Apel wants to
introduce something like a moral competence-competence and
preempt in abstracto something that can be judged situationally only
at the level of political will formation in the light of concrete possi
bilities of action. Apel clearly allows himself to be misled by the
hierarchical structure of his theory into immediately addressing ques
tions of political ethics "directly from above" through a superprinci
ple, although these questions are not situated on the same level as
the justification of the moral principle. Behind the solitary politician
of Apel's imagination there lurks the philosopher king who wants to
put the world in order, not the citizen of a democratic state. In a
democratic state, political will formation--even reforms with far
reaching political ramifications-is always pursued within the institu
tions of a legally constituted social and political order. Politicians are
holders of democratically sanctioned positions of power whose deci
sions are contingent on legally institutionalized processes of opinion
and will formation.
Relations between states too are so closely regulated by suprana
tional institutions, treaties, and norms of international law that clas
sical foreign politics is gradually becoming transformed into a
worldwide internal politics that (fortunately) leaves increasingly little
room for the heroic initiatives of solitary politicians. Questions of
political ethics constitute an especially complicated application prob
lem for which considerations in legal and perhaps even social theory
are indispensable.
With the initial differentiation between "Part A" and "Part B" of
his ethical theory, Apel already moves in a false direction by contrast
ing an ethics of conviction indifferent to consequences with an ethics
of responsibility oriented to consequences. For within moral dis
courses, the principle of universalization necessitates the weighing of
87
consequences and demands in general that all relevant features of

the given situation be taken into account already in the justification
of norms, but most especially in their application, as is explicitly
required by the principle of appropriateness. And self-referential
reflections on the reasonableness of moral demands already play a
role within moral discourses.
In extreme cases, a tension may develop between moral insight
and ethical self-understanding that cannot be uniformly resolved in
favor of what is morally required, even when viewed morally, as
would be the case if it were merely a problem of weakness of wi11.76
Thus, in the case of an existential conflict, the single applicable norm
might demand an action that the individual concerned indeed ac
knowledged to be morally required but could not perform without
ceasing to be the person he is and wants to be-the burden imposed
by this obligation might be too much for him to bear. In an excep
tional case such as this, the preconditions of moral self-respect cannot
be harmonized with one's relation to oneself as a person who en
deavors to pursue his life project in an authentic fashion. A moral
theory that insists on the primacy of the tight over the good will not
dispute this phenomenon as such but will describe it differently. One
proposal is the familiar move of introducing duties toward oneself
that can compete with other duties. But the concept of duties toward
oneself cannot be reconciled with the requirement of symmetry be
tween rights and duties. Hence a more plausible solution is to admit
the possibility of a reflective turn in discourses of application; once
we have established which norm is the only appropriate one in the
particular case, it may be necessary to examine whether the singular
judgment that follows from it requires an action that cannot be
reasonably expected from an existential point of view.
Generally the problem of which moral actions can be reasonably
enjoined first arises with the transition from moral to legal theory.
Modern rational law makes the transition from morality to law by
way of specific refections on the reasonableness of imposing obliga
tions. Norms are judged to be valid in the light of the moral principle
only under the presupposition-explicitly stated in (U)--of general
observance of norms. I f this precondition is not met, norms cannot
be reasonably imposed, regardless of whether they are valid. Kant
uses this fact to justify the state monopoly on the exercise of legal
88
Remarks
on
Discourse Ethics
coercion. Legal institutionalization alone can ensure general adher

ence to morally valid norms. That amounts to a moral reason for law
in general. Conversely, there are moral considerations relevant to the
justification of particular legal norms that the democratic legislative
process must take into account. Since problems of application must
be distinguished from problems of justification also in the case of
legal norms, there arises the additional demand for institutionaliza
tion of discourses of application in the shape of an independent
judiciary. In this way, the contours of a constitutional state, which
Kant consistent with these determinations treated in his doctrine of
law, become discernible; here the moral point of view is no longer
applied directly to modes of action but to legal and political institu
tions. Finally, this normative approach can be pursued to the point
where the classic problem of the limits of obedience to the law, and
hence the preconditions for civil disobedience, legitimate resistance,
and a "right to revolt," is raised. Then it is a question of the conditions
under which a politically motivated violation of positive law would
be morally justified or legitimate. 77 These remarks are intended only
to lend plausibility to the view that the normative relations between
morality, law, and politics are better explained by discourse theory in
a step-by-step manner, without recourse to a false hierarchy of forms
of discourse.
1 1 . In laying the moral-theoretical foundations of a philosophical

theory of justice, it is important to undertake the normative justifi
cation of the transition from morality to law in the proper manner.
I propose (a) to take an objection of Thomas McCarthy as an occasion
to point at least to some new perspectives in the theory of law and
politics opened up by discourse ethics. Since underlying this objection
is a doubt concerning the priority of the right over the good, I will
also discuss (b) the concept of an "overlapping consensus" developed
by John Rawls in his more recent writings.
(a) McCarthy grounds the need for legal regulation of beh<.. vioral
expectations, and in particular for legal institutionalization of the
opinion and will formation of a political legislator, in the indetermi
nacy of discursive procedures where different worldviews and a mul
tiplicity of evaluative languages are competing with one another:
"Disagreements of these sorts are likely to be a permanent feature of
democratic public life. They are in general not resolvable by strategic
89
compromise, rational consensus or ethical self-clarification, in Ha

bermas's sense of these terms. All that remains in his scheme are
more or less subtle forms of coercion, e.g. majority rule and the
threat of legal sanctions."78 Interestingly enough, McCarthy intro
duces the claim that morality must be supplemented by law with an
argument that, were it sound, would undermine the very core of the
discourse-ethical concept of morality. On his account, legal compul
sion is supposed to compensate for the motivational weaknesses and
cognitive indeterminacy of a moral discourse that rests on abstraction
and idealization ; however, he conceives the legally institutionalized
processes of opinion and will formation through which legislation
must legitimate itself not merely as a more complex network of
discourses and negotiations but as an alternative to the discourse
model: "We have to modulate the idea of rationally motivated agree
ment beyond Habermas's basic distinction between a strategically
motivated compromise of interests and an argumentatively achieved
consensus on validity. "'9
By contrast, a discourse theory of law grounded in discourse ethics
holds fast to the discourse model. It explains the difference between
the normative validity claim of moral rules and the claim to legitimacy
of legal norms in terms of the fact that not just moral (and ethical)'
considerations enter into a presumptively rational process of political
will formation of a legislator, but, in the first instance, programs and
goals that are the result of more or less fair compromises, as well as
information and predictions resulting from more or less controversial
discussions among experts. A process of collective will formation
along these lines would have the presumption of rationality in its
favor to the extent that it could be realized in general in forms of
discourse. The idea of the constitutional state can then be understood
in terms of the endeavor to institutionalize the demanding commu
nicative presuppositions and procedures of a network of forms of
argumentation and negotiation differentiated in terms of the prob
lem areas, for the political power employed in executing the require
ments of the principle of discourse must itself be tamed through
discourse.
I will address this theoretical proposal in detail in another place;
here I am interested only in why McCarthy regards the alternative
between force and rationally motivated agreement as incomplete. In
90
his view, a third element comes into play with the will formation of
a pluralistic society, the legitimate compulsion that he illustrates only
by reference to the legal form of an institutionally regulated major
itarian process of decision making. As grounds, he cites primarily the
incommensurability of standards of evaluation, worldviews, evalua
tive languages, and traditions that set narrow limits to the general
ization of interests undertaken in justificatory moral discourses:
The success of Habermas's universalization principle in getting from multi
farious "I want's" to a unified "we will" depends on finding "universally
accepted needs." The argument just sketched suggests that this may not be
possible when there are fundamental divergences in value orientations. The
separation of formal procedure from substantive content is never absolute:
we cannot agree on what is just without some measure of agreement on what
is good. But practical discourse is conceived by Habermas to deal precisely
with situations in which there is an absence of such agreement, that is, when
there is a need to regulate matters concerning which there are conflicting
interests and values, competing conceptions of the good.80
McCarthy knows how this objection could be met within the frame
work of discourse ethics. Need interpretations are not an ultimate
given but depend on intersubjectively shared evaluative languages
and traditions that are not anybody's personal property. Hence, crit
ical revision of the vocabulary in terms of which we interpret our
needs is a public and, where appropriate, discursively negotiable
affair. In modern societies, we are confronted with a pluralism of
forms of life-and a progressively greater individualization of life
projects-that is not only unavoidable but even desirable; and this
makes it ever more improbable, as McCarthy emphasizes, that we will
agree on shared interpretations in such disputes. We can draw less
and less on experiences and straightforward examples that have the
same significance for different groups and individuals. We can count
less and less on the same reasons having the same weight for different
individuals and groups within different systems of relevance.
This fact, however, supports rather than undermines the univer
salistic approach of discourse ethics. The more that principles of
equality gain a foothold in social practice, the more sharply do forms
of life and life projects become differentiated from one another. And
the greater this diversity is, the more abstract are the rules and
91
principles that protect the integrity and egalitarian coexistence of

subjects who are becoming increasingly unfamiliar with one another
in their difference and otherness. To be sure, the sphere of questions
that can be answered rationally from the moral point of view shrinks
in the course of the development toward multiculturalism within
particular societies and toward a world society at the international
level. But finding a solution to these few more sharply focused ques
tions becomes all the more critical to coexistence, and even survival,
in a more populous world. It remains an empirical question how far
the sphere of strictly generalizable interests extends. Only if it could
be shown in principle that moral discourses must prove unfruitful
despite the growing consensus concerning human rights and democ
racy-for example, because common interests can no longer even be
identified in incommensurable languages-would the deontological
endeavor to uncouple questions of justice from context-dependent
questions of the good life have failed.
These reflections bring us back to a theme already broached in
discussing Taylor's universalistic ethics of the good. The priority of
the right over the good is also the main bone of contention in the
disputes between communitarians and liberals. 81 The socio-ontological
aspects of this involved debate need not concern us here, since the
objections raised by communitarians against individualistic concepts
of the person or instrumentalist concepts of society do not apply to
the basic concepts of discourse ethics derived from the theory of
communicative action. Discourse ethics occupies an intermediate po
sition, sharing with the "liberals" a deontological understanding of
freedom, morality, and law that stems from the Kantian tradition
and with the communitarians an intersubjective understanding of
individuation as a product of socialization that stems from the He
gelian tradition.82 But the priority of the right over the good is
disputed primarily on epistemological grounds. In this respect dis
course ethics is as vulnerable as liberal theories ofjustice to objections
that appeal to the contextuality and the rootedness in tradition of all
conceptions of justice and practical reason, including the procedural.
The contextualists insist that behind the allegedly general and neutral
explanations of the moral point of view and the perspective of justice
there are always concaled particular world interpretations informed
92
by specific evaluative languages and traditions. This context depen

dency contradicts the asserted independence of the general from the
particular and the consequent priority of the right over the good.
(b) Rawls responded in the course of the 1 980s-most concisely in
the brief "Restatement" of his theory83-to the contextualist criticism
in both a defensive and an offensive manner. He has, on the one
hand, retracted the claim to justify a universally valid concept of
justice that had come under attack as foundationalist. He now pres
ents the theory of justice merely as a systematic reconstruction of the
best normative intuitions of the Western tradition in political thought.
Rawls leaves open the question of whether the reconstructively
grounded principle of justice should be regarded as valid only for
societies shaped by our political-cultural traditions and not for mod
ern conditions of life in general, and thus for all modern societies
irrespective of their cultural orientation and tradition. On the other
hand, Rawls accords the context of our traditions only the status of
an informative background. This intuitive background determines
the hermeneutic point of departure for the description of a hypo
thetical original position under whose normative constraints the par
ties can justify principles of justice neutral between different forms
of life. These constraints compel them to abstract from all compre
hensive views of self and the world in which the different conceptions
of the good are articulated. Through the specification of the views
of justice built into the conditions of the original position, our tra
dition remains effective in an unthematized fashion in the conscious
ness of the parties. But neither our own nor alien traditions can be
thematized within the original position, unless it be in terms of the
anonymous concept of "comprehensive doctrines" in general. This is
the substantively indeterminate, complementary concept to the con
cept of justice, which Rawls terms "political, not metaphysical" be
cause of its indifference over against all such worldviews.
In a further step Rawls examines whether this abstraction is pos
sible and whether it proves to be fruitful. The principles of justice
justified in the original position should not be left hanging in the air
but must also be capable of being acted upon, of falling under "the
art of the possible." It belongs to the design of a "well-ordered
society" that its ideal citizens are also loyal to the institutions that they,
by hypothesis, recognize as legitimate, especially as just practices
93
constitute an experiential context in which the sense of loyalty is

renewed with each successive generation. But the real citizens of
co ntemporary liberal societies are flesh and blood individuals who
have grown up in different traditions and forms of life and owe their
self-understanding to competing worldviews. The political public
sphere in which they come together to form a public body of citizens
is characterized by a plurality of belief systems and interest structures
and the coexistence and confrontation of cultural life-forms and
individual life projects on which McCarthy bases his objection. Thus,
the theory of justice first takes on a realistic shape by reflecting on
itself as a whole and clarifying the conditions of its own acceptability.
It must render plausible through a self-reflexive turn in argumen
tation the belief that its justification of principles of justice can expect
to meet with agreement in the public arena of an existing body of
citizens: "Justice as fairness is not reasonable unless in a suitable way
it generates its own support by redressing each citizen's reason, as
explained in its own framework" ( 1 4 1 ) .
After all that has been said about the abstraction from comprehen
sive views of self and the world imposed by the structure of the
original position, Rawls's response will come as no surprise: a concept
of justice developed in ignorance of the content of competing world
views will be an intersection set comprising normative statements on
which the divergent value orientations of the comprehensive doc
trines agree. An "overlapping consensus" in regard to the theoreti
cally justified principles of justice will emerge in the public discourse
of citizens who try to convince each other. of the rightness of their
respective political views.
To this the contextualist could object that Rawls prejudices the
point at issue between them by making two assumptions specifically
connected with the concept of a "modern" society. One of these
assumptions is empirical in nature. In modern societies not just a
political order has emerged that is perceptibly independent of other
social action systems; value spheres specifically concerned with ques
tions of political justice have also emerged. This assumption is em
pirically well confirmed, though it also has controversial implications:
"Taking the political as a distinctive domain, let's say that a political
conception formulating its basic characteristic values is a freestanding
view. This means two things: first, that it is framed to apply in the
94
first instance to the basic structures of society alone; and second, that
it formulates the characteristic political values independent of non
political values and indeed of any specific relationship to them" ( 1 33).
The other presupposition is of a conceptual nature and in this context
deserves special consideration. A postmetaphysical concept of justice
is not compatible with all comprehensive doctrines, only with non
fundamentalistic worldviews : "While an overlapping consensus may
obtain between not unreasonable comprehensive doctrines, when is
a comprehensive doctrine not unreasonable or simply reasonable? A
reasonable doctrine must recognize the burdens of reason" ( 1 39).
Modern world views must accept the conditions of postmetaphysical
thought to the extent that they recognize that they are competing
with other interpretations of the world within the same universe of
validity claims. This reflective knowledge concerning the competition
between equally valid warring "gods and demons" creates an aware
ness of their fallibility and shatters the naivete of dogmatic modes of
belief founded on absolute truth claims. Recognizing the "burdens
of reason" entails knowing that proponents and opponents in the
contest between substantive worldviews may (for the time being) have
equally good grounds for their inability to reach a consensus and for
leaving contentious validity claims undecided. This fallibilism is
grounded in the indeterminacy of discursive procedures, in local
limitations on available information and reasons, and, in general, in
the provinciality of our finite minds regarding the future. Under
these conditions there is no guarantee that a rationally motivated
consensus could always be attained. The idea of "reasonable disagree
ment" permits us to leave truth claims undecided while simulta
neously upholding their unconditional character. One who, with this
in mind, accepts the coexistence of competing worldviews by no
means resigns himself to a mere modv.s vivendi; while upholding his
own validity claims, he simply postpones the possibility of consensus,
kept open in principle, to an indefinite future.
The model that Rawls has in mind is not that of competing para
digms governed by the standards of scientific rationality, and thus
not the fallibility of the empirical sciences, but the uncoupling of
religious creeds from the sanctioning power of the state and their
coexistence under conditions of religious tolerance. But although this
principle can indeed be justified in the original position, it is not the
95
one that Rawls must presuppose. What Rawls in fact prejudges with
the concept of an "overlapping consensus" is the distinction between
modern and premodern forms of consciousness, between "reasonable"
and "dogmatic" world interpretations. Modern forms of conscious
ness do not involve renunciation of interpretations of the world as a
whole, and thus of "comprehensive doctrines" as such. They are set
apart by a form of reflexivity that enables us to adopt an external
perspective toward our own traditions and bring them into relation
to other traditions. Whether we can in this way transcend the tradition
in which our identity has been formed depends on a strong premise
that is far from trivial and affords the contextualist an occasion to
make his objection more pointedly. There must be a common basis
on which mutual understanding among alien cultures, belief systems,
paradigms, and life-forms is possible-that is, a translation between
different evaluative languages and not merely communication among
members of the same language group relying on reciprocal observation
of alien cultures. The languages and vocabularies in which we inter
pret our needs and communicate our feelings must be mutually
permeable; they should not be so deeply rooted in monadically self
enclosed contexts that cannot be transcended from within that they
imprison the subjects who have been born and socialized into them.
Rawls can defend the thesis of the primacy of the right over the
good with the concept of an overlapping consensus only if it is true
that postmetaphysical worldviews that have become reflexive under
modern conditions are epistemically superior to dogmatically fixed,
fundamentalistic worldviews-indeed, only if such a distinction can
be made with absolute clarity. Otherwise the disqualification of "un
reasonable" doctrines that cannot be brought into harmony with the
proposed "political" concept of justice is inadmissible. This differ
entiation between a modern and a traditional understanding of the
world is possible only if competing interpretations of the world are
not utterly incommensurable, if we at least intuitively accept context
transcending assumptions concerning rationality that first make pos
sible translations from one context into another. But this is precisely
what is disputed by a strong contextualism for which there is no
single "rationality." On this conception, individual "rationalities" are
correlated with different cultures, worldviews, traditions, or forms of
96
life. Each of them is viewed as internally interwoven with a particular

understanding of the world.
This touches on a debate concerning rationality that has been
carried on for the past two decades by anthropologists, sociologists,
and philosophers and to which Rorty, Davidson, and Putnam have
also made contributions. I will limit myself to a discussion of the most
recent book of Alasdair Macintyre who draws upon different con
cepts of justice developed in the philosophical tradition to argue for
his version of strong contextualism.
1 2 . Macintyre examines the concepts of justice of the three philo
sophical traditions that have their origins respectively in Aristotle, in
Augustine and Aquinas, and in Scottish moral philosophy and con
trasts these classical concepts with the specifically modern conceptions
of morality and politics he associates, somewhat globally, with "lib
eralism."84 Liberalism adopts a critical attitude not just toward rival
traditions but toward tradition as such in an attempt to represent its
own concepts of rationality and its own principles as generally valid.
In Macintyre's view, this abstract universalism obscures the fact that
liberalism itself has by now become a tradition, though it refuses to
acknowledge this. Each of these traditions is so intimately interwoven
with a particular language and local form of life that practical reason
and political justice can be understood only from the perspective that
structures one of the different traditions. There is no standpoint
outside the context of any particular tradition from which a rational
evaluation of these theories and explanatory schemas could be
undertaken.
Obliviousness of this fact is constitutive of modern forms of con
sciousness, as Macintyre attempts to show in the case of liberalism.
He attributes this bias to the structure of a language-game that gen
erates the transcendental illusion of a universalistic self-misunder
standing on the part of a philosophical tradition that repudiates
tradition as such . One of the so-called international languages forms
the grammatical ground of this boundless universalism. These lan
guages have established themselves-after the fashion of American
scientific prose-as lingua franca and create the impression of a me
dium into which any utterances whatever, regardless of how alien the
traditions from which they derive, can be translated without further
97
ado. International languages of this kind exert grammatical pressure

on their users to appropriate alien traditions in a context-neutralizing
manner. Macintyre also regards as a grammatical illusion the histor
ical consciousness that Nietzsche denounced in his second "Untimely
Meditation" as antiquarian historiography: "When texts from tradi
tions with their own strong, substantive criteria of truth and ration
ality, as well as with a strong historical dimension, are translated into
such languages, they are presented in a way that neutralizes the
conceptions of truth and rationality and the historical context. . . .
The distortion by translation out of context-from the standpoint of
those who inhabit the traditions from which the distorted texts are
taken-is of course apt to be invisible to those whose first language
is one of the internationalized languages of modernity" (384.). In
ternational languages prejudice the self-understanding of the forms
of life of which they are constitutive. Hence, the reflexivity of modern
worldviews does not establish their superiority over the traditional
understanding of the world, as Rawls assumes in accordance with the
liberal tradition. Because they are parasitic on unacknowledged tradi
tions, they are themselves degenerate forms of world interpretation.
They impose a false ontological preunderstanding on the roodess
heirs of modernity, a "boundless cosmopolitanism" that leads them
to absolutize their own standards of rationality and assimilate every
thing alien without regard to context. The liberal inhabitants of the
modern world have no sense for what is ultimately inaccessible in
alien traditions and precisely for this reason fail to recognize the
possibility of actually learning something from alien cultures. Abstract
universalism is merely the obverse side of a globally objectifying
historicism.
But Macintyre fails to mention that this diagnosis of the contem
porary scene follows in the footsteps of the cultural criticism of the
German mandarins. If the diagnosis is to express something more
than the mere distaste of the educated European for "Americanism"
and "shallow" Enlightenment culture, it stands in need of systematic
grounding. Macintyre offers an interesting proposal to this effect in
the last three chapters of his book. His argument proceeds on two
fronts. Against the alleged abstract universalism of the Enlighten
ment, Macintyre defends (a) the thesis that there is no such thing as
a context-transcending rationality, only different forms of rationality
98
rooted in traditions. And against a performatively self-contradictory

relativism, he upholds (b) the thesis that productive communication
between such self-centered traditions is no less feasible than learning
from alien traditions. From the perspective of our own culture , we
maintain steadfastly the unconditionality of validity claims; we take
our forms of argumentation and investigation seriously and do not
have to regard our practices of justification and criticism as merely
arbitrary social conventions, as Rorty would have it. In what follows
I will . examine whether the contextualist thesis can be reconciled with
the antirelativist one.
Macintyre's argument involves two steps. First, he shows how we
can speak of the rationality of a particular tradition without having
to renounce the moment of unconditionality we unavoidably associate
with fundamental concepts in every tradition. Then he seeks to dem
onstrate how communication between alien cultures is possible with
out postulating an underlying context-transcending rationality.
Macintyre portrays the competition between comprehensive views of
self and world in a different light from Rawls. The coexistence of
equally strong doctrines, with conceptions of the good that can over
lap in a postmetaphysical concept of justice on condition that they
submit themselves, as products of modern culture, to the conditions
of reflexivity-this idea is now superseded by the model of a dispute,
oriented to truth, concerning the hegemony of the most cogent tra
dition, the one that proves to be superior in its concept of practical
reason, its ideals of life, and its interrelated conceptions of the good
and the just. A tradition will prove to be rationally superior only if,
like all historically rooted traditions, it possesses the power of sub
stantive interpretations of the world and adopts a critical stance to
ward modern forms of consciousness.
In regard to (a) , the rationality embodied in an inherited context
is articulated in an endogenous tradition of inquiry of the kind in
vestigated by Macintyre, who takes as his examples the four philo
sophical movements mentioned. It is only at the third stage in the
process of a tradition's becoming reflexive, however, that traditions
of inquiry first arise. During the first stage, the authority of learned
opinions and texts is not yet placed in question, while at the second
stage, the processes of dogmatic articulation and systematization have
advanced to a level where contradictions and unsolved problems
99
come to light but cannot yet be methodically resolved. Only at the

third stage are problem solutions construed as learning processes, so
that one's own earlier conceptions can now be criticized as errors in
the light of the corrected conceptions. From the perspective of the
clarification of one's own errors there emerges a coherence concept
of truth : that is true which serves within the context of our traditions
to correct errors. Although this concept of truth reflects only the
experience of tradition-immanent learning processes, from within
our respective traditions we raise truth claims that have absolute
validity for us; they confirm the dissolution of the only form of
cognitive dissonance to which we are subject. In this way standards
of truth have determinate meaning only within the context of a
particular tradition of investigation, but for their adherents they do
not (as a result) have merely local significance. Like the basic concepts
of the language game of argumentation itself, they preserve a con
text-transcending sense: "The concept of truth is timeless" (363).
The concept of truth is crucial to the interesting limit cases in
which epistemological crises can no longer be resolved by a renewal
of the basic conceptual vocabulary that draws on the resources of
one's own tradition but only by learning another tradition recognized
as superior. If the claim concerning the dependency of forms of
rationality on traditions is to hold good in this case too, the instruction
derived from alien traditions must be described as both an internally
motivated paradigm shift and at the same time as a paradigm leap
from one tradition to the other.
The endogenous resolution of an epistemological crisis must satisfy
three conditions. First, it makes possible an explanation in terms of
the new paradigm of the problems that could not be solved in the
old vocabulary; second, it will make clear retrospectively the obstacles
that hindered the tradition from finding a productive solution to
problems prior to the innovation; and, finally, it will ensure certain
continuities within the tradition through the paradigm shift and be
yond : "these first two tasks must be carried out in a way which exhibits
some fundamental continuities of the new conceptual and theoretical
structures with the shared beliefs in terms of which the tradition of
inquiry had been defined up to this point" (362) But precisely this
third condition cannot be fulfilled by an exogenously induced learning
process : "Derived as it is from a genuinely alien tradition, the new
1 00
explanation does not stand in any sort of substantive continuity with

the preceding history of the tradition in crisis. In this kind of situation
the rationality of tradition requires an acknowledgment by those who
have hitherto inhabited and given their allegiance to the tradition in
crisis that the alien tradition is superior in rationality and in respect
of its claims to truth to their own" (365). This account turns on the
fact that the rational discrediting of one's own tradition still proceeds
according to its own standards of rationality, whereas learning a
rationally superior alien tradition presupposes a process of conver
sion, namely, the adoption of new standards of rationality. If different
forms of rationality inhere in different traditions, there can be no
bridge between them. The shift between mutually exclusive totalities
requires a transformation in the identities of learning subjects. The
latter must alienate themselves from themselves in the moment of
transition and in the course of a conversion in their understanding
of self and world must learn to understand their own past in the
light of another tradition acknowledged to be superior.
But this model of what is involved in learning alien traditions is
open to at least three objections. In the first place, the description is
too selective. The model holds at best for learning processes in which
the transition to a new level of justification signifies global, categorial
devaluation of the sort of grounds valid up to that point-as happens,
for example, in the transition from mythical narratives to monoth
eistic or cosmological explanations from first principles, or in the
dissolution of religious and metaphysical worldviews in postmeta
physical thought. But all of the ethical systems Macintyre has in mind
compete within the same universe of discourse of Western philosophy.
They interpenetrate to a sufficient degree that they can learn some
thing from each other without compromising their identity. Further
more, Macintyre cannot uphold his own learning theory-modeled
on the ideas of recent postempiricist philosophy of science-without
falling into a performative self-contradiction. Either his metatheo
retical claims about stages of increasing reflexivity hold for any tra
dition whatsoever, in which case they could not have been developed
from within the context of a particular research tradition, which
contradicts the presupposition on which they rest; or these claims
lose their context-transcending meaning and have only local validity,
but then Macintyre becomes entangled in precisely the relativism he
101
tries to avoid by means of his learning theory. Finally, the proposed

concept of learning through conversion is not internally consistent.
The recognition of the rational superiority of an alien tradition can
be sufficiently motivated from the perspective of one's own tradition
only if the learning subject can compare the explanatory power of
both traditions in relation to the same problems. But precisely this is
denied him, because in the absence of a zone of rational overlap the
two traditions are incommensurable.
In regard to (b), Macintyre attempts to counter this last objection
with a theory of "thicker" though less than complete translation. If
learning from an alien tradition is to be possible, the two traditions
must be capable of communicating with one another without assum
ing common standards of rationality. Now "good" translations cannot
limit themselves to literal correlation of expressions of the one lan
guage with expressions of the other; rather, they must conform to
the model of acquisition of a "second first language" to the extent
that they take into account the internal intermeshing of communi
cative expressions with the totality of a life practice. Understanding
a language presupposes at the very least virtual participation in the
native language games, if not a secondary socialization into the in
digenous forms of life. The meaning of a communicative utterance
is revealed from the contexts in which it is uttered. Hence, the un
derstanding of a communicative utterance becomes deeper with ac
cess to the context-constitutive horizons of the linguistically
structured lifeworld in which the utterance is situated.
This way of distinguishing between good and bad translations
draws on well-worn hermeneutic insights. It loses its triviality only
with the additional assertion that the better the interpreter learns to
master his second language, the more clearly he recognizes the un
translatability of its central components: "The characteristic mark . . .
is to be able to recognize where and in what respect utterances in the
one [language] are untranslatable into the other" (375) . With the
acquisition of at least two "first," that is, actually spoken, languages
and by becoming accustomed to the corresponding practices of an
extended form of life, the "good" translator gradually gains a stereo
scopic perspective from which he is supposed to be able to perceive
"zones of untranslatability." He learns to move back and forth be
tween both contexts and experiences the failure of attempts to trans-
1 02
pose adequately constitutive expressions of the one language into the

other: "You cannot express some of Plato's key thoughts in the He
brew of Jeremiah or even of the Wisdom literature, but you also
cannot express them in Homeric Greek" (375). With this intuitive
appeal to "zones of untranslatability," Macintyre clarifies, on the one
hand, how those who belong to different traditions, each crystallized
around its own rational core, can at all communicate with one other;
and, on the other, he wishes to explain thereby why, if they enter such
zones, they can recognize the rational superiority of another tradi
tion. When an interpreter experiences the failure of attempts at
translation, he must be capable of adopting both irreconcilable per
spectives and yet of understanding the meanings that cannot be
mutually transformed into one another.
The inconsistency of this account becomes apparent once we reflect
on what Macintyre must think he can accomplish if he is even to find
examples to illustrate his thesis of untranslatability. Equal mastery of
two essentially incommensurable language worlds-let us say, biblical
Hebrew and classical Greek-enables him to paraphrase examples
from both languages with the help of English in such a way that the
difficulty of an exact or literal translation becomes plausible for the
reader. Of course, Macintyre cannot understand such paraphrasing
as translation, since the circumscribing explications of inaccessible
meanings are supposed instead to demonstrate the untranslatability
of central expressions. Under the precondition of genuine bilingual
ism, Macintyre presumably experiences essential inaccessibility,
which he nevertheless communicates to readers in a third language.
But in doing this, the interpreter has assuredly found a language for
his experience of the "absolutely other" in zones where the mutually
exclusive forms of rationality inhering in the different traditions no
longer overlap. The fact that the experience of untranslatability is
not inexpressible should raise our suspicions.
Macintyre himself anticipates the obvious objection "that only in
sofar as we can come to understand what it is that is allegedly inac
cessible to us could we have grounds for believing in such
inaccessibility, and that the acquisition of such understanding is of
itself sufficient to show what we alleged to be inaccessible is in fact
not so." He responds to his own objection in the following manner:
"If it is the case . . . that a condition of discovering the inaccessible
103
is in fact a matter of two stages, in the first of which we acquire a

second language-in-use as a second first language and only in the
second of which can we learn that we are unable to translate what
we are now able to say in our second first language into our first first
language, then this argument loses all its force" (387). The ad hoc
irreversibility built into this description has an air of arbitrariness
about it; the fact, contingent in itself, that Macintyre in his own
example makes use of a third language-English-to paraphrase the
presumably untranslatable meanings of expressions of the first and
second languages already betrays the conceptual necessity of a tertium
comparationis that enables us bring the two language worlds into rela
tion to each other.
How would a subject have to be constituted who comes upon
something absolutely untranslatable in communicating with the ad
herents of an alien tradition? Let us suppose to begin with that the
speaker's identity, which has been constituted through primary so
cialization and becoming habituated to the traditions and life-forms
of his mother tongue, remains essentially unaltered by the acquisition
of a second first language. His identity, we might say, is inflexible. As
soon as a speaker of this kind advances into regions where the ra
tionalities of different languages and traditions prove to be mutually
impervious, he is faced with an alternative: either he changes his
identity by converting to an avowedly rationally superior language,
tradition, and life-form, or he acquires a second identity but one he
cannot bring into relation with his first identity rooted in the past.
In short, he suffers a split personality. In either case there would no
longer be a subject to whom the experience of change of identity or
of fragmentation of identity could be ascribed. Something analogous
holds for the experience of incommensurability of language worlds
and the untranslatability of their constitutive expressions; we could
not attribute such an experience to a speaker with an inflexible iden
tity essentially rooted in a tradition. For in the course of the experi
ence ascribed to him, this referent would change; hence, if the
referent is to remain constant, we must abandon the assumption of
a rigid identity. In its place we make the more realistic assumption that
a subject who learns to master two languages ideally well emerges
from the process of acquiring the second first language with a bilin
gually extended identity. The speaker has so broadened his understand-
1 04
ing of self and world that he is sufficiently flexible to remain identical

with himself when he makes the transition from one language world
to the other-and to escape the fate of conversion or even of a split
personality. A flexibly maintained identity enables him to bring the
languages and rationalities of both of his past worlds, the primary and
the secondary, into relation in such a way that both are fused into a
broadened horizon of possible understanding. And after this fusion
of horizons in his own person, the bilingual speaker can express the
difficulties of translating from one tradition into another by reciprocal
paraphrasings of each language in the other-and also in a third
language. The problem-or, better, apparent problem-of complete
untranslatablility must have already been solved before we can even
identify zones of more or less radical difficulties of translation. This
has been shown in a different way by Davidson in his critique of the
notion of a conceptual scheme. Macintyre's rehabilitation of a strong
concept of tradition is inspired by culture-critical motives. He wants
to make students-who live betwixt and between traditions-aware
that they cannot live an authentic life without the critical appropri
ation of the substantive contents of a tradition with which they iden
tify. This pedagogical intention does not account for his radically
contextualist position, for which the insights Gadamer articulated in
his philosophical hermeneutics, for example, would be sufficient.
Against this Macintyre not only emphasizes the ties of an interpreter
to the conditions of his hermeneutic point of departure but maintains
in addition the peculiar asymmetry of a communicative situation in
which one side is viewed as capable of learning from the other only
from a perspective that remains bound to its own context (or, alter
natively, through subjection to an alien perspective). Macintyre seems
to restrict all interpretive procedures to processes of self-understanding.
He is misled -by his pedagogical intention into assimilating the un
derstanding of arbitrary symbolic expressions to the clarification of
an existential self-understanding. This ethical curtailment of his con
cept of interpretation explains the ethnocentric self-centeredness
from which alone traditions that satisfy Macintyre's description can
encounter one another.
Even in exemplary instances of intercultural understanding where
not just rival conceptions but conflicting standards of rationality clash,
unprejudiced communication between "us" and "them" nevertheless
1 05
necessitates a symmetrical relation. The fusion of interpretive horizons

at which, on Gadamer's account, every communicative process aims
should not be understood in terms of the false alternative between
an assimilation "to us" and a conversion "to them." It is more properly
described as a convergence between "our" perspective and "theirs"
guided by learning processes, regardless of whether "they" or "we"
or both sides must reform the practices of justification thus far ac
cepted as valid. Concepts such as truth, rationality, and justification
play the same role in every language community, even if they are
interpreted differently and applied in accordance with different cri
teria. And this fact is sufficient to anchor the same universalistic
concepts of morality and justice in different, even competing, forms
of life and show that they are compatible with different conceptions
of the good--on the assumption that the "comprehensive doctrines"
and "strong traditions" enter into unrestricted dialogue with one
another instead of persisting in their claims to exclusivity in a fun
damentalistic manner. In this sense Rawls's concept of an overlapping
consensus represents one possible interpretation of the principle that
the general and the particular reciprocally presuppose each other.
1 3 . A deontological moral theory that concentrates on questions of

justice pays for the presumptive general validity of moral judgments
with a narrow concept of morality. But it does not thereby banish
questions of the good or unfailed life from the pale of rational
treatment. It need only maintain that ethical discussions, in contrast
to moral arguments, are always already embedded in the traditional
context of a hitherto accepted, identity-constituting form of life.
Moral judgments differ from ethical judgments only in their degree
of contextuality.
The implications of a narrow concept of the moral for issues in
ecological ethics are of greater moment. The anthropocentric profile
of theories of the Kantian type seems to render them blind to ques
tions of the moral responsibility of human beings for their nonhuman
environment. These theories proceed on the assumption that moral
problems arise only within the circle of subjects capable of speech
and action because we, in Gunther Patzig's words, "as members of a
community of human beings count and depend on the cooperation
and consensus of others."85 But our moral feelings, judgments, and
1 06
actions are directed not only to subjects capable of speech and action
but also to animals. Patzig has the merit of confronting pressing
ecological questions of the protection of animals and the preservation
of species without overstepping the limits of an "ethics without meta
physics." As with the extension of the spectrum of issues of justice to
include questions of the good life, here it is also a matter of supple
menting moral questions in the narrow sense-limited to rational, or
at least potentially rational, subjects-with questions of another kind.
Is there such a thing as responsibility toward nature independent of
responsibility for present and future humanity? What, in particular,
is the nature of our obligation to protect animals?86
Patzig describes the predicament in which an anthropocentric
moral theory finds itself when it attempts to justify the protection of
animals, for example : "Is it possible to extend the sphere of validity

of moral obligations beyond the human realm to encompass all living
creatures who are capable of experiencing pain and suffering but
also pleasure? Here we run up against a clear barrier. . . . For animals
cannot enter into relations of principled reciprocity with us of the
kind that govern our conduct toward other human beings."87 On the
other hand, our moral intuitions speak an unambiguous language.
We have an unmistakable sense that the avoidance of cruelty toward
all creatures capable of suffering is a moral duty and is not simply
recommended by prudential considerations or even considerations
of the good life. And we cannot settle the matter, following Kant, by
saying that whereas we have duties in relation to animals, we do not
have duties toward them. Animals confront us as vulnerable creatures
whose physical integrity we must protect for its own sake.
The founder of utilitarianism, jeremy Bentham, had little difficulty
in answering the question of the scope and sphere of application of
moral duties: "The question is not, Can they reason? nor, Can they
talk? but, Can they suffer?"88 Yet how is this utilitarian response to
be reconciled with an anthropocentric approach if duties in the strict
sense can only follow from rules that rational beings impose upon
themselves through insight? Valid norms deserve the intersubjective
recognition of all potentially affected because the latter, were they to
engage in argumentation, could convince themselves that these
norms are equally in the interest of all. Now animals are not creatures
of a kind that "could enter into a relation of principled reciprocity
107
with us." But what, then, is the status of duties that impose a deter
minate responsibility on us as addressees of valid norms, not only for
or in respect of animals but toward these animals themselves, if animals
for their part do not belong to the domain of possible addressees of
norms?
Patzig summarily dispenses with the symmetry between duties and
rights that is a conceptually necessary consequence of the reciprocal
recognition of at least potentially free and equal subjects and defines
the status of duties toward animals in an asymmetrical manner: "An
imals do not have duties toward humans, but humans have duties
toward animals. "89 At the same time he opens up the possibility of a
gradation of duties incompatible with the binary encoding of the
mode of validity of strict duties. The greater the sensitivity to pain
of the animals, the more onerous duties should be. But this amounts
to a transformation of the concept of duty; it is no longer a question
of duties in the deontological sense but of the relative preferability
of goods. This suggests that questions of environmental ethics, as
well as ethical questions of how to live one's life, should be treated
from a different perspective-a teleological one-and that only ques
tions of the well-ordered coexistence of persons should be treated in
a deontological manner.
Now the exclusion of ethical questions from the sphere of moral
questions or questions of justice no doubt makes sense because the
question of what is good for me or for us, all things considered, is
already formulated in such a way that it invites an answer whose
claim to validity is relativized to prior life projects and forms of life.
Identity-constitutive values and ideals cannot obligate us in the same
sense as moral norms; they lack the unconditionality of a categorical
ought. But we have a sense of being under categorical obligations
toward animals. The horror inspired by the torment of animals is, at
any rate, more closely related to outrage at the violation of moral
demands than to the pitying or condescending attitude toward people
who, as we are wont to say, have made nothing of their lives or are
failures by their own standards of authenticity. We "ought" not to
neglect animals callously, much less cruelly torment them.
In an attempt to do justice to these intuitions, Patzig tries to connect
the injunction to protect animals, grounded initially in utilitarian
terms, to the deontological theory: "Our theory of rational morality
1 08
suggests a different explanation of the norm forbidding the infliction

of unnecessary suffering on animals. Each of us knows what pain
and suffering are and we expect all other human beings to respect
our vital interest in avoiding them as much as possible. But it would
be irrational to make a radical distinction between human and non
human creatures where the latter behave in such a way that we are
compelled to assume that they are capable of experiencing pain and
suffering. Thus the prohibition on arbitrarily inflicting suffering and
callous neglect is extended beyond the human domain to include
non-human creatures."90 Regrettably, the "thus" here masks a non
sequitur, for the principle of generalization does not entail that the
norm of refraining from inflicting pain on others should be extended
to encompass all creatures sensitive to pain, so long as this principle
demands onl y equal consideration of the interests of all those we may
assume are capable of adopting the perspective of all others. Apart
from their potential "yes" or "no" responses it cannot be determined
whether such a norm could meet with the agreement of all and hence
is valid. Animals, as Patzig correctly emphasizes, simply do not belong
"to the parties to the contractual reciprocity that underlies human
morality."91
Otherwise we would be at a loss to explain why the command to
refrain from inflicting suffering on animals does not encompass-as
in the case of human beings-the further duty to respect animal life,
that is, not to kill animals. If animals had a moral claim of the same
kind as human beings to be protected from suffering inflicted on
them by human beings, it would unavoidably appear paradoxical that
they should not hae an even more emphatic claim not to be killed
by human beings. But the discussion concerning the qualified per
missibility of animal experiments with fatal results has already shown
us that we do not regard the rationally justified and painless killing
of animals as an act of murder. Moreover, nonvegetarians seem to
have few scruples about eating meat. This paradox can at least be
brought closer to resolution if we take the discourse-ethical interpre
tation of rational morality as our starting point and recall that the
feeling of duty has its roots in the fundamental relations of recog
nition we always already presuppose in communicative action.
Social interactions mediated by the use of language oriented to
mutual understanding are constitutive for sociocultural forms of life.
1 09
This kind of communicative socialization through which persons are

simultaneously individuated generates a deep-seated vulnerability,
because the identity of socialized individuals develops only through
integration into ever more extensive relations of social dependency.
The person develops an inner life and achieves a stable identity only
to the extent that he also externalizes himself in communicatively
generated interpersonal relations and implicates himself in an ever
denser and more differentiated network of reciprocal vulnerabilities,
thereby rendering himself in need of protection. From this anthro
pological viewpoint, morality can be conceived as the protective in
stitution that compensates for a constitutional precariousness implicit
in the sociocultural form of life itself. Moral institutions tell us how
we should behave toward one another to counteract the extreme
vulnerability of the individual through protection and considerate
ness. Nobody can preserve his integrity by himself alone. The integ
rity of individual persons requires the stabilization of a network of
symmetrical relations of recognition in which nonreplaceable individ
uals can secure their fragile identities in a reciprocal fashion only as
members of a community. Morality is aimed at the chronic suscepti
bility of personal integrity implicit in the structure of linguistically
mediated interactions, which is more deep-seated than the tangible
vulnerability of bodily integrity, though connected with it.
Clearly this difference between personal and bodily integrity dis
appears in the animal world because we do not attribute personality
to creatures with whom we cannot speak and cannot come to an
understanding about something in the world. Nevertheless, we com
municate with animals in a different way once we involve them in
our social interactions, in however asymmetrical a fashion. Such in
teractions take on a measure of continuity in our association with
domestic animals : our conscience is particularly insistent "concerning
species with which we can communicate with particular ease. ; '92 Like
moral obligations generally, our quasi-moral responsibility toward
animals is related to and grounded in the potential for harm inherent
in all social interactions. To the extent that creatures participate in
our social interactions, we encounter them in the role of an alter ego
as an other in need of pro tection; this grounds the expectation that
we will assume a fiduciary responsibility for their claims. There exists
a quasi-moral responsibili ty toward animals that we encounter in the
1 10
role (if not completely filled) of a second person, one whom we look
upon as if it were an alter ego. We can adopt a performative attitude
toward many animals, though not toward plants. In that case, they
are no longer objects to be observed by us, or even just objects of
our empathy, but beings who, in their interaction with us, make their
distinctive mode of being felt in a manner different than a rock does
its mineral hardness or a plant does the osmotic interaction of an
organism with its environment. To the extent that animals participate
in our interactions, we enter into a form of contact that goes beyond
one-sided or reciprocal observation because it is of the same kind as an
intersubjective relation.
In making such statements, we must be wary of mystifications.
Interactions between humans and animals are mediated by nonlin
guistic gestures, and the injuries that human beings are capable of
inflicting on an animal do not affect anything resembling a personal
identity-rather, they are a direct assault on its psychophysical integ
rity. Again, an animal does not experience its pain reflexively like a
human being, who in suffering is cognizant of the fact that he i in
pain. These and similar asymmetries characterize the way in which
animals participate in our interactions. At the same time, the inter
actions must satisfy the condition that we should not confront animals
in the objectifying attitude of a third person, nor just communicate
about animals but with them. We must be able to ascribe characteristics
of agents to animals, among others the ability to initiate utterances
and to address them to us. Then we have duties that are analogous
to our moral duties, because like the latter they are rooted in the
presuppositions of communicative action. Of course, they are anal
ogous to moral duties only to the extent that the asymmetries in the
interactions admit comparison with relations of recognition between
persons.
If this justification of interactive duties toward animals by appeal
to a theory of intersubjectivity is not completely wide of the mark,
we can also explain why animals, on the one hand , are rendered
particularly dependent on human beings and in need of protection
by the asymmetrical structure of possible interactions, but why, on
the other, they can enjoy this moral protection only within the inter
subjective horizon of our modes of interaction. Human beings always
find themselves already within this horizon and as persons can never
Ill
leave it, whereas animals belong to other species and other forms of
life and are integrated into our forms of life only through partici
pation in our interactions. The limits of our quasi-moral responsibility
toward animals are reached once humans, in their role as members
of one species, confront animals as exemplars of another. However,
it is a tricky moral question to determine in which situations this is
permissible. I do not wish to preclude a priori that some vegetarians
exhibit a moral sensibility that may prove to be the correct moral
intuition under more auspicious social circumstances. In that case
animals would come to be recognized in all situations as possible
participants in interaction, and the protection to which we feel obli
gated in our interactions with animals would be extended to include
protecting their existence.
Human responsibility for plants and for the preservation of whole
species cannot be derived from duties of interaction, and thus cannot
be morally justified.95 Nevertheless, I am in accord with Patzig in
believing that, aside from prudential considerations, there are good
ethical reasons that speak in favor of the protection of plants and
species, reasons that become apparent once we ask ourselves seriously
how, as members of a civilized global society, we want to live on this
planet and how, as members of our own species, we want to treat
other species. In certain respects, aesthetic reasons have here even
greater force than ethical, for in the aesthetic experience of nature,
things withdraw into an unapproachable autonomy and inaccessibil
ity ; they then exhibit their fragile integrity so clearly that they strike
us as inviolable in their own right and not merely as desirable ele
ments of a preferred form of life.
3
Lawrence Kohlberg and Neo-Aristotelianism
Among my American friends Larry Kohlberg was always the one

who for me embodied most convincingly the distinctively American
tradition that stretches back from Mead and Dewey via Royce and
Peirce to the transcendentalists-from Chicago to Concord, so to
speak. It is hardly a coincidence that Kohlberg came of age academ
ically in the place where Charles Morris, among others, kept alive
the spirit of pragmatism. I recall one evening at his house in Cape
Cod, as we sat beside the fireplace with a view of the ocean shortly
after sunset. Larry took a battered volume from the shelf and began
to recite an ode of Walt Whitman, rather haltingly, running the
syllables together, with a strange mixture of understatement and deep
involvement. At that time I scarcely understood a word of the poem,
but the sound of his voice, the rhythmic flow of the free verse, the
whole scene spoke eloquently enough: here we were witnessing one
of the roots from which this man's life drew its nourishment. During
my last visit to Harvard, he took me to William james Hall and
pointed out the inscription that expresses the essential intuition of
pragmatism :
The commu n ity stagnates without the impulse of the individual,
the impulse dies away without the sympathy of the community.
-William james
l l4
These words are deeply embedded in the American tradition. But is

this also true of the proposition the sentence expresses, of the insight
the maxim encapsulates? To put it more pointedly: Is this statement
true for Americans alone? Or does it also embody a truth for those
who are not the fortunate heirs to the political thought of a Thomas
Paine and a Thomas Jefferson? It might be objected that it was easy
for a German of the younger generation like myself, who graduated
from high school after World War II under the influence of "reed
ucation," to distance himself from the elitism of the German Gym
nasium and to open himself unreservedly to the universalistic,
egalitarian values of the radical democratic tradition. Yet the crucial
issue is to what extent the genesis of such convictions compromises
their validity.
In what follows I will defend a thesis that does not sit well with the
spirit of the times : that anyone who has grown up in a reasonably
functional family, who has formed his identity in relations of mutual
recognition, who maintains himself in the network of reciprocal ex
pectations and perspectives built into the pragmatics of the speech
situation and communicative action, cannot fail to have acquired
moral intuitions of the kind articulated in propositions such as that
of James. The maxim asserts the reciprocal dependence of sociali
zation and individuation, the interrelation between personal auton
omy and social solidarity, that is part of the implicit knowledge of all
communicatively acting subjects; it does not merely express a more
or less subjective opinion concerning what some person believes is
the good life. 1 That proposition articulates an intuition we acquire in
various different contexts on the condition that we grow up in sur
roundings that are not completely undermined by systematically dis
torted communication.
Kohlberg defends the thesis that moral judgments exhibit the same
structure in all cultures and societies in the following way: "We claim
that there is a universally valid form of rational moral thought process
which all persons could articulate, assuming social and cultural con
ditions suitable to cognitive-moral stage development. We claim that
the ontogenesis toward this form of rational moral thinking occurs
in all cultures in the same stepwise, invariant stage sequence."2 Each
of these sentences contains two theses that need to be defended
independently-the first with the resources of moral philosophy, the
1 15
second in the field of developmental psychology. However, the in

validity of the one assertion would not be without implications for
the truth of the other.
Philosophers engage in the controversy surrounding the question
of moral universalism on normative grounds without drawing on
empirical theories. But the relevant empirical theories must at least
be in accord with the rational reconstructions of the intuitive knowl
edge of competent speaking, acting, and judging subjects proposed
by philosophy. Einstein's theory of relativity was a challenge to Kant's
"Metaphysical Foundations of Natural Science." In another sense
Kohlberg's theory, as we shall see, represents a touchstone against
which Kant's "Foundations of the Metaphysics of Morals" must be
tested. A developmental theory of the Piagetian kind can function at
best as an indirect test of the theories the psychologist has relied upon
in his description of the cognitive operations of the highest stage of
moral-psychological development. This amounts only to a test of the
coherence of different theories that cannot be brought into direct
relation to one another but must not contradict one another at the
metatheoretical level where we piece together the different theoret
ical elements. On the other hand, cognitive psychology is dearly
dependent on the philosophical theories it employs in constructing
its empirical theory. Kohlberg's theory furnishes a good example of
this interdependence. He uses arguments from Hare, Frankena, and
Rawls, among others, to delimit the domain of 'udgments ofjustice."
I n addition, he relies on moral theories of this sort to provide a
structural description of the postconventional judgments of Stage 6,
which are not adequately supported by empirical evidence, for con
siderations of logical completeness of the sequence of developmental
stages demand that a highest stage be postulated.5
In what follows, I will concentrate on the philosophical aspects of
the universality thesis. A moral-philosophical defense of the univer
salistic position is crucial for Kohlberg's research program for at least
two reasons :
( 1 ) Kohl berg himself was well aware o f the fact that his "metaethical
assumptions" stand in need of justification within the framework of
philosophical discourse. Were there already compelling moral philo
sophical reasons against universalistic approaches as such, a univer
salistically oriented theory of moral development would also be
1 16
subverted . Yet Kohlberg's use of the term "metaethical" to charac

terize his categorial framework is misleading,4 for we cannot delimit
the object domain of moral judgments independently of the nor
mative theories that already inform the description of the highest
stage of moral judgment, nor can we make clear what is to be under
stood by problems ofjustice and their corresponding solutions. More
over, the fundamental philosophical assumptions of psychological
theories have to be clarified if we are to be able to counter objections
against reconstructive research projects based on an erroneous phi
losophy of science.
(2) In addition, differentiations and alterations in moral philo
sophical foundations have implications for the construction of the
empirical theory extending to the details of experimental design. At
the metatheoretical level, moral-philosophical and moral-psycholog
ical programs are implicated in a hermeneutic circle in which they
must complement each other. Thus, changes in the philosophical
explanation of the moral point of view shed new light on how to
conduct empirical surveys of moral judgments and how to select
instruments and evaluate data.
In recent years universalistic theories such as those of Rawls, Dwor
kin, Apel, and myself have been sharply criticized by philosophers
such as Alasdair Macintyre, Bernard Williams, Charles Taylor, Mi
chael Sandel, and David Wiggins. From these involved discussions I
will first extract three serious objections. Then I will reverse the
direction of criticism and discuss some difficulties in the neo-Aristo
telian position on which these objections are based. This discussion
should put us in a position, finally, to revise the fundamental premises
of Kantian moral theory in such a way that neo-Aristotelian objections
can be accommodated within a deontological framework.
II
Classical ethics, which achieved its mature form in the work of Ar

istotle, is inspired by the ambition to find a philosophical answer to
one fundamental question: "How should I live?" or "How should one
live?" In virtue of this prior orientation, practical questions are in
vested with teleological significance. The question "What ought I to
do?" or "What is right for me?" is subordinated to a more encom-
1 17
passing question, "What is the good life?" A consequence of this turn

toward an ethics of the good is the uncoupling of practical reason
from theoretical cognition. Practical reason is identified with pru
dence, prudentia or phronesis, and thus as a faculty that always operates
within the horizon of established customs and practices and conse
quently renounces the claim to knowledge reserved by Aristotle ex
clusively for science or theory in the strict sense. With the emergence
of the modern empirical sciences, however, philosophical theory also
had to give up the strong claim to knowledge grounded ultimately
in metaphysics. But phronesis, as the Aristotelian counterpart of epis
te'TIU!, was also affected by the latter's altered status and weakened
claim. With the unsettling of theoretical reason, the status of practical
reason also becomes problematic. 5
Practical reason needed to be reconsidered in the light of the
modern ideas of theory and science. The path of ethics now branched
out in three directions. The capacity for moral judgment was either
excluded from the sphere of reason altogether in an empiricist man
ner, thereby giving rise to a long series of noncognitivist positions;
or moral reflection was reduced to instrumental calculations of con
sequences of action, as in the case of utilitarianism. Kant alone re
served a place for moral judgment in the realm of practical reason
and thereby upheld its status as genuine knowledge. In Kant's view
all judgments--empirical, normative, even aesthetic-raise a claim to
validity that can be supported or criticized by appeal to good reasons.
These theoretical determinations entail at the same time that the
concept of morality henceforth stands in need of mediation by a
theory of knowledge. What one understands by morality becomes a
matter of how one answers the question concerning the possibility of
rationally resolving practical questions in general.
Aristotle could still construe the claim to knowledge of his ethics,
which was embedded in a metaphysical worldview, in the sense of a
weak cognitivism. Because he took his orientation from the situation
of the citizen in the polis, understood as the proper telos of human
existence, his ethics could still provide orientation concerning the
ontological conditions and the institutional framework of the good
life. Without such a metaphysical background on which to draw, Kant
confronted the task of first explaining the possibility of moral insight
as such and then of situating practical reason within the theoretical
1 18
framework of a tripartite, though only formally unified, reason. The

categorical imperative can be understood as an explication of the
viewpoint of impartial judgment formation.
The moral point of view is the perspective from which we can
decide among controversial normative claims impartially, solely on
the basis of reasons. Kant had to pay a price for this move to a
postmetaphysical concept of autonomous morality, and this is the
issue on which the dispute between philosophers still turns today.
The proponents of Aristotelian ethics are convinced that modern
directions in moral theory have come at too high a price; these
theories have renounced the proper goal of ethics-that of answering
the existential question of the wherefore of our lives-but have
gained nothing worth speaking of in return. Even the categorical
imperative demands an exercise of abstraction with far-reaching con
sequences-disregarding everyday practices and hence the very con
text in which moral judgment and action takes place. Here I set aside
questions of interpretation and accept for the sake of argument the
concept of morality the neo-Aristotelians ascribe to Kant. They ad
monish deontological positions in the Kantian tradition for detaching
practical reason from the context of ethical life and restrict it to a
narrow moral outlook defined by one or other formulation of the
principle of universalization. The intuitive meaning of the categorical
imperative is clear: moral questions should be decided with regard
to what all could will. Only those norms and modes of action are
morally valid or obligatory that could meet with the justified assent
of all involved (or all affected). In fact, the moral point of view
requires a threefold abstraction : ( 1 ) abstraction from the motives
required of those involved , (2) abstraction from the particular situa
tion, and (3) abstraction from existing institutions and forms of life.
( 1 ) The moral point of view involves a shift in the focus of practical
reason from questions of the form, "What is good for me/for us?" to
questions of justice of the form "What should one do?" With this
change in perspective, the significance of the once-canonical orien
tation to happiness and well-being also changes. The question con
cerning eudaimonia originally encompassed the sphere of all possible
goods, including justice, as well as all of the virtues, including the
sense of justice. But under the deontological aspect of the question
of what all could will, justice and autonomy (that is, the capacity to
1 19
act in accordance with self-imposed laws) are singled out as the sole
morally relevant determinations. The moral point of view necessitates
differentiating between the right and the good. All goods, including
the highest good of my life project or of the form of life to which
we collectively aspire, are deprived of their moral status and are
lumped together with things designed to satisfy contingent needs and
wants ; henceforth, goods fulfill only subjective preferences. In con
trast to the ethics of the good, the morality of justice sets duty and
inclination in conflict with one another. From this opposition there
follow two disturbing consequences:
(a) Viewed from the perspective of classical ethics, practical reason
completely loses its point if moral judgment is restricted to interper
sonal obligations, so that the question of how we should live-as
individuals and members of a community-must be left up to blind
decision or sheer impulse. It was precisely these issues of individual
life projects and collective ways of life that philosophy once promised
to illuminate through reflection. What was held to be essential to
philosophy was that it should help us to lead a reflective life. Must
philosophy now renounce this preeminent task in favor of that of
submitting just one subset of practical questions, questions of justice,
to a more clear-cut standard of judgment?
(b) Second, once the bond between the right and the good is
broken, the question of why one should act morally at all can no
longer be answered satisfactorily. Classical ethics sought to demon
strate that striving for the highest good is compatible with both our
obligations and our true interests. But in Kant's view, free will is
constituted by moral insight alone ; he conceives of autonomy as a
subject's capacity to bind his will and to be guided in action by moral
judgment alone. However, motivation by insight into what is morally
required extends no further than the motivating power of good
reasons. Hence Kant must invest practical reason with the higher
authority of an intelligible world. But if we reject Kant's metaphysical
background assumptions concerning the noumenal and the phenom
enal realms, the question of why we should make the moral point of
view our own, even when it is not in our interest, becomes unavoid
able. The question "Why be moral?" remains open: "The I that stands
back in rational reflection from my desires is still the I that has those
1 20
desires . . . . It cannot, just by taking this step, acquire the motivations

of justice. "6
(2) From the moral point of view, practical reason focuses its atten
tion on the justification of norms while problems of application re
cede into the background. Practical reason is thereby transformed
from a context-dependent faculty of prudent deliberation that op
erates within the horizon of an established form of life into a faculty
of pure reason operating independently of particular contexts. While
Aristotelian phronesis focuses on specific alternative courses of action
and links normative considerations with empirical observations and
questions of rational choice in a nonperspicuous manner, practical
reason in the Kantian sense, by contrast, is restricted to questions of
normative justification in such a radical fashion that no tradition or
way of life, however deeply rooted in our everyday praxis, can escape
problematization. No normative validity claim raised in the lifeworld
is immune to challenge ; everything counts as a hypothesis until it has
regained its validity through the authority of good reasons.
This focus on questions of justification in abstraction from prob
lems of application is characteristic of the posttraditional level of
moral judgment. But the way in which Kant makes the break with
conventional thought leaves him vulnerable to the charge of context
insensitivity. Because he explicates the moral point of view in terms
of the "moral law" and construes the question "What should I do?"
exclusively in terms of the justification of maxims, he seems to restrict
the exercise of practical reason to assessing alternative courses of
action in the light of the moral law. Moral justification seems to
amount to nothing more than the deductive application of an abstract
basic principle to particular cases, with the result that the specific
context of the given situation loses its peculiar relevance. Abstraction
from the particular constellation of the individual case also has the
effect of excluding the consequences of action from normative con
sideration. For this Kant pays the price of the rigidity of an ethics of
conviction. The model of moral judgment tailored to questions of
justification overlooks the fact that the choice between competing
norms of action and their application gives rise to problems of its
own. As a general rule, problems of applying implicitly presupposed
norms are even more exigent than those of justification .7
121
(3) Seen from the moral point of view, issues of moral cognition
take precedence over questions of practical orientation; problems are
shifted from the existential plane of the concrete practical concern
of socialized individuals regarding their individual and collective lives
to the abstract level of the reflection of isolated subjects on what all
could accept as a law governing their actions. Kantian practical reason
transcends the sphere of validity of traditions and institutions that
are constitutive of the particular life-form of a given collectivity with
its specific practices and customs and concrete virtues and duties.
From this abstraction there follow in turn two unfortunate
consequences :
(a) First, an atomistic conception of the person is presupposed,
since each individual who examines his maxims from the moral point
of view finds himself compelled to go beyond the context of his own
form of life in foro interno if he is to encounter all others as Others,
that is, as isolated individuals. Free will operates in a vacuum, dis
engaged from the social bonds that first invest ethical life with mean
ing. To the atomistic concept of the person who relates to himself as
his own private property there corresponds a contractualist concept
of society that denies any inherent moral quality to a life reduced to
calculations of self-interest: "For a society to be a community in this
strong sense, community must be constitutive of the shared self
understandings of the participants and embodied in their institu
tional arrangements, not simply an attribute of certain of the partic
ipants' plans of life."s
(b) Furthermore, from a contextualist perspective, it is doubtful
whether any concept ofjustice can claim universal validity. Lifeworlds
are totalities that exist only in plural form. But if ethical life is tied
to the specific characteristics of a certain form of life, then every
concept of good and evil, however abstract, is affected and shaped
by the intu itive preunderstanding embodied in the concrete totality
of moral conceptions dominant in a particular place. Concepts of
justice cannot be separated from the complex totality of a concrete
ethical life and a particular idea of the good life. This is why in
Macintyre's view "the Enlightenment project of justifying morality"
had to fail. 9 The conceptions of an autonomous morality developed
by Rousseau and Kant in the eighteenth century also seem to remain
tied to the period from which they arose. Already in their basic core
122
we can discern traces of the possessive individualism that would

become dominant in the course of capitalist modernization.
III
The objections leveled against moral theories of the Kantian type

may be summarized as follows. The deontological privileging of what
one ought to do as the basic moral phenomenon makes the abstract
separation between the right and the good unavoidable. This leads
to abstraction from the requisite motivations, with the result that the
question of why we should act morally at all no longer admits of a
plausible answer. The cognitivist privileging of the posttraditional level
of moral judgment gives prominence to questions of normative jus
tification. This leads to abstraction from particular situations and
neglect of questions of the application of norms. The formalist privi
leging of the general over the particular, finally, goes hand in hand
with an atomistic concept of the person and a contractualist concept
of society. This leads to abstraction from an ethical life that can
assume concrete form only within particular forms of life. This in
terrelation raises doubts about the possibility of a strict separation
between form and content and of a context-independent concep
tualization of justice as such.
The outcome of our reflections leaves us with two alternatives:
either we return to the Aristotelianism underlying these criticisms,
or we modify the Kantian approach to take account of legitimate
objections. Only those who are ready to restore a metaphysical mode
of thinking could unhesitatingly embrace the first alternative. Con
temporary neo-Aristotelians are not willing to do this. But their at
tempts to develop a practical philosophy in an Aristotelian spirit
without drawing on metaphysical premises run up against difficulties
that are, in my view, insurmountable.
( 1 ) In modern societies we encounter a pluralism of individual life
styles and collective forms of life and a corresponding multiplicity of
ideas of the good life. As a consequence we must give up one of two
things : the claim of classical philosophy to be able to place competing
ways of life in a hierarchy and establish at its acme one privileged
way of life over against all others or the modern principle of tolerance
according to which one view of life is as good as any other, or at least
1 23
Lawrence Kohlberg and Neo-Aristotelia nism
has equal right to exist and be recognized. Macintyre wants to have

his cake and eat it. On the one hand, he wants to uphold the classical
cognitive claim : "In what does the unity of an individual life consist?
. . . To ask 'What is the good for me?' is to ask how best I might live
out that unity and bring it to completion. To ask 'What is the good
for man ?' is to ask what all answers to the former question must have
in common. " 1 0 On the other hand, given the premises of postmeta
physical thought, Macintyre can no longer appeal directly to what
he calls Aristotle's "metaphysical biology." On this was founded the
thesis that "man" can actualize his nature and attain "the good" only
within the privileged context of one particular form of life : the polis.
Macintyre, by contrast, has to work his way through a proliferatir:tg
multiplicity of practices, traditions, and biographical patterns of life
without being able to reach any single privileged complex of distin
guishing characteristics of "the" good life. As Richard Bernstein re
marks: "While there are constraints on how I can answer the question
'What is good for me?' this does not prevent a wide variety of answers,
each of which may clash, and be incompatible with others-may
require participation in practices incompatible with others, and even
may require incompatible virtues. This (modern) conception of a
human good does not limit 'the subversive arbitrariness' that can
invade 'the moral life. " ' 1 1
(2) If we take modern pluralism seriously, we must renounce the
classical philosophical claim to defend one uniquely privileged mode
of life-for example, the vita contemplativa in opposition to the dif
ferent forms of the vita activa. Bernard Williams nevertheless wants
to secure the cognitive status of moral consciousness under these
premises. But practical reason then begins to look suspiciously like
phronesis. Once its metaphysical underpinning has been removed, the
latter must either be assimilated to everyday knowledge or elevated
to the status of reflective knowledge. In moral consciousness Williams
sees the possibility of a form of ethical cognition that qualifies as
knowledge and yet remains below the level of reflection of objecti
fying science and philosophy. Thus, on the one hand he stresses the
limits of philosophy: " How truthfulness to an existing self or society
is to be combined with reflection, self-understanding, and criticism is
a question that philosophy, itself, cannot answer. It is the kind of
question that has to be answered through reflective living. " 12 But then
1 24
he must, on the other hand , specify criteria in terms of which a

reflective form of life can be distinguished from an unreflective one:
"Ethical reflection becomes part of the practice it considers, and
inherently modifies it." But it is thereby faced with a paradoxical
task. Although practical reason eschews the objectifying procedure
of empirical inquiry into moral behavior no less than the theoretical
constructions of philosophy, yet it is supposed to produce a form of
practical knowledge "that would help us to find our way around in
a social world which . . . was shown to be the best social world for
human beings . " 1 3 Williams is compelled to attribute to practical rea
son a form of rationality that goes beyond sheer common sense but
whose difference from scientific rationality remains to be determined.
If practical reason can inform members of a social lifeworld con
cerning what is best for them and how they should regulate their
common life, it makes possible a form of practical knowledge that is
undoubtedly secured from within the lifeworld and yet points beyond
its horizon. The horizon of every form of life is fluid, its boundaries
permeable. There is no absolute barrier to the "desire for as much
intersubjective agreement as possible."14 Practical knowledge can all
the more readily claim to be knowledge the more radically we open
ourselves to others and expand our local knowledge and ethnocentric
outlook-indeed, extend our community in a virtual manner such
that our discourse ultimately includes all subjects capable of speech
and action. But this process would yield the perspective Kantians call
the moral point of view. Certainly, "if the agreement were to be
uncoerced, it would have to grow from inside human life."15 How
ever, this qualification also holds for an intersubjective construal of
the concept of practical reason that preserves intact its universalistic
core: what "we" agree about coincides with what all could possibly
assent to, provided we respond in a sufficiently radical fashion to
Rorty's pragmatist call for the most comprehensive possible extension
of the sphere of application of the word "we." If each individual
community can achieve "knowledge" (in not just a metaphorical
sense) concerning what is good for it, it is far from obvious why this
practical knowledge should not be extended in an intercultural di
rection and become so thoroughly emancipated from provincial lim
itations that it orients itself to what is equally good for all. Without
metaphysical backing, what Aristotle called phronesis must either dis-
1 25
solve into mere co mmon sense or be developed into a concept of

practical reason that satisfies the criteria of procedural rationality. 16
(3) By contrast, if we wish to remain faithful to the Aristotelian
conviction that mo ral judgment is bound to the ethos of a particular
place, we must be prepared to renounce the emancipatory potential
of moral universal ism and deny so much as the possibility of sub
jecting the structural violence inherent in social conditions character
ized by latent exploitation and repression to an unstinting moral
critique. For only the transition to the posttraditional level of moral
judgment liberates us from the structural constraints of familiar dis
courses and established practices. 1 7 Charles Taylor counters this
charge with a universalistic ethics of the good that appeals to supreme
goods transcending all particular forms of life. The examples he
proposes, however, derive from Plato, the Stoics, and Christianity,
that is, from traditions that appeal to the authority of reason, uni
versal natural law, or a transcendent deity. 18 These are the forerun
ners of moral universalism but are grounded in cosmological and
religious worldviews that are even more difficult to reconcile with
postmetaphysical thought than the teleological world view of Aristotle.
, Contemporary neo-Aristotelianism has played quite different po
litical roles in Germany and the United States. Similar lines of ar
gument take on a different significance in the two political cultures
in Germany a conservative, in America a critical. But as Herbert
Schnadelbach has shown, the conservative bent of neo-Aristotelian
ism is not altogether accidental:
For the neo-Aristotelian turned conservative, autonomy is the same as funda
mental opposition. . . The neo-Aristotelian ethic of prudence, with its weak
foundationalism and its refined relativization of the principle of autonomy,
is in its implications an eminently political ethic. The hermeneutic linking of
ethics with ethos engenders not just a habitual bias in favour of the existing
state of affairs, it also generates a systematic mistrust of the individual who
can only exemplify, but never himself be the bearer of, an ethos . . . . The fact
that the ethos already embodied in the political must always take precedence
over the moral individual accounts for the tendency of neo-Aristotelians to
collapse the difference between politics and morality from the perspective of politics,
while at the same time warning against the moralization of politics . t 9
.
In sum, all attempts a t a historicist revival o f Aristotelian ethics on

a postmetaphysical footing are beset with insuperable difficulties.
1 26
Hence I propose to explore the other alternative left open to us and

examine whether the misgivings concerning the deontological, cog
nitivist, and formalist abstractions can be accounted for within the
framework of a moral theory of the Kantian kind, though reinter
preted in intersubjective terms.
IV
I will now take up once again with particular reference to Kohlberg's

theory the objections discussed in the second section.
Abstraction from the Requisite Motives
Kohlberg consistently made a sharp distinction between moral dilem
mas and questions of the good life. He illustrates what I have called
the deontological abstraction with the responses of a young woman
talking about a critical decision in her life: "I've had a personal
decision, my decision to divorce . . . but I didn't view it as a moral
problem. It wasn't. There weren't any moral issues involved really.
The issues involved were-was it the right thing for us? . . I knew
that I'd be giving up my entire life and I'd have to begin again
because it involved a geographical change and a change in work."
When asked to say what would constitute a moral problem, she imag
ined the following situation : "Usually where two principles that I
consider valuable look as though they may be clashing, then it's very
hard to make a decision about things. When I think about things like
child abuse, for example, there is the principle of family unity and
the principle of the welfare of the child . . . although in this case I
would always look out for the welfare of the child." Concerning this
Kohlberg remarks : "Her resolution of the dilemma of child abuse vs.
family unity can claim to be universalizable, impartial and agreed
upon as right by all human beings. The decision of her own divorce
. . . cannot be resolved from the 'moral point of view', whereas the
dilemma of child abuse can be."20
Ethical questions of the good life can be distinguished from moral
questions by a certain self-referentiality. They refer to what is good
for me or for us-in this case, to what was best for the woman, her
husband, and others directly affected by the divorce. This egocen.
1 27
tric--or, as in the case of political questions, ethnocentric-reference

is a sign of the internal relation between ethical questions and prob
lems of self-understanding, of how I should understand myself (or
we ourselves as members of a family, community, nation, and so
forth). The question "What is the best thing for me (or us) in this
situation?" must be answered in the light of the underlying question:
"Who am I, and who would I like to be?" ("Who are we, and who do
we want to be?") It is one's own identity that is here at stake. The
fact that ethical questions are implicitly informed by the issues of
identity and self-understanding may explain why they do not admit
of an answer valid for everyone. But the logic of such questions does
not completely exclude the possibility of rational answers in this
dimension. Ethical problems need not be abandoned to subjective
decisions or preferences. Contrary to what is often alleged, the latter
is not a consequence of the deontological approach : "Personal deci
sions are understood to be culturally, historically and individually
relative, though some degree of empathy, sensitivity and communi
cation is required to resolve such problems."2 1
The hermeneutical clarification of one's identity also appeals to
reasons; a self-referential interpretation must satisfy the precondition
that it admits of assessment in terms of authenticity and inauthentic
ity. This kind of interpretation depends on a descriptive grasp of
one's own processes of development-"Who am I?"-as much as on
life projects and ego ideals-"Who would I like to be?" Hence, an
understanding of self interpreted in this way can claim validity only
for persons with determinate histories and specific life projects. The
relativization of the validity of ethical statements does not constitute
a deficiency but is the result of the logic of a question directed to me
(or us) alone and ultimately can be answered only by me (or by us),
though authentic interpretations must, of course, be compatible with
valid moral norms.
This differentiation does not yet solve the problem of how the
deontological chasm between moral judgment and factual conduct
can be bridged. But this motivational problem can be posed only
within the framework of a philosophy that is still confident of pro
viding general answers to questions of the good life. A moral theory
that no longer claims to know the telos of "the" good life must leave
the question "Why be moral?" unanswered. On the premises of post-
1 28
metaphysical thought, there is no reason why theories should have

the binding power to motivate people to act in accordance with their
insights when what is morally required conflicts with their interests.
The disposition to act responsibly is contingent on processes of so
cialization and the degree of success in identity formation. But an
identity cannot be produced by arguments. Thus, the fact that moral
theory does not aspire to anything over and above the task of recon
structing the moral point of view and justifying its general validity
does not amount to a deficiency. It can only show the participants
the procedure they must follow if they want to solve moral problems
and must leave all concrete decisions up to them. To think that one
has the right answer is to know that one does not have good reason
to act otherwise. The capacity of moral judgments to motivate agents
to act stands in direct proportion to the rationally motivating force
of the reasons on which they are grounded. The extent to which
rational motives actually influence action depends on the individuals,
the circumstances, the interest positions, and the institutions involved.
Abstraction from the Given Situation
Kant did not effect the transition to autonomous morality in a suffi
ciently consistent fashion. He presented the categorical imperative as a
response to the concrete question "What ought I to do?" and did not
fully realize that the shift in focus to problems of justification also
entailed a strict separation of issues of justification of norms from
those of their application. The categorical imperative must not be
understood as a moral law that can be applied directly to maxims and
actions. The categorical imperative is, rather, a proposal concerning
how the viewpoint of impartiality, from which the validity of norms
can be judged, should be understood. We cannot justify norms and
defend concrete actions in the same breath. In justificatory discourses
we appeal to the constellation of individual situations only by way of
illustrating the conditions of application of a norm by means of
examples. This is not a matter of correctly deciding a concrete case
that permits different courses of action in a particular situation but
of establishing the validity of norms that could be appealed to in such
decisions. Knowledge of valid norms does not extend to knowing
how one ought to decide in a particular situation. Discourses of ap-
1 29
plication require different data and principles from discourses of

justification. Kohlberg's substages A and B may have something to
do with this differentiation, at least at the postconventional level.
If the employment of practical reason is no longer conceived as an
operation in the mind of a solitary individual but is instead opened
up to public argumentation, then the empirical consequences of nor
matively regulated action can also be submitted to the tribunal of
norm-testing reason. Hence an ethics of conviction must yield to an
ethics of responsibility. The notion of ideal role taking, which Kohl
berg borrowed from G. H. Mead, is realized in the form of a shared
practice of argumentation or deliberation that compels each partici
pant to take the perspectives of all others. This idea has only recently
attracted attention in analytic philosophy. David Wiggins, for exam
ple, interprets the procedure of universalization in terms of uniting
the perspectives of the agent, the object of action, and the observer
in a specific form of communication: "What is envisaged here is a
public scene . . . in which moral agents are at once actors and spec
tators . . . . Actors here are persons doing things and persons having
things done to them. Spectators are not strangers to these roles. Nor
are actors strangers to the role of spectators. For everyone plays each
of these three roles at some point, and his direct and indirect knowl
edge of the other roles constantly informs his playing of each."22 This
description at least comes close to the interweaving of perspectives
produced under the communicative presuppositions of moral argu
mentation. Here what concerns me is that a communication-theoret
ical interpretation of the moral point of view should liberate the
intuition expressed in the categorical imperative from the burden of
a form of moral rigorism that is deaf to the consequences of actions.
A norm can command the rationally motivated assent of all only if
everyone involved or potentially affected has taken into consideration
the consequences and side effects of the general observance of the
norm for himself and others.
The same liberating consequence follows from the logic of dis
courses of application. 23 The principle of universalization that regu
lates discourses of justification does not exhaust the normative sense
of the impartiality of a just judgment. A further principle must be
adduced to guarantee the correctness of singular judgments. An
impartial judge must assess which of the competing norms of action-
1 30
whose validity has been established in advance-is most appropriate

to a given concrete case once all of the relevant features of the given
constellation of circumstances have been accorded due weight in the
situational description. Thus, principles of appropriateness and the
exhaustion of all relevant contextual features come into play here.
And practical reason thereby also comes into play in discourses of
application.
Abstraction from Concrete Ethical Life
Neo-Aristotelians have a valid point when they criticize the atomistic
concept of the person and the contractualist concept of society that
have informed virtually every modern approach in moral and legal
theory. But Hegel, who was among the first to spell out this criticism,
was equally aware that there is also an element of truth in these
modern abstractions. Individualistic approaches give expression to
the notions of autonomy and freedom of conscience that are indis
sociable from the self-understanding of the modern period. Thus,
Kant too grounded moral consciousness in the intelligible ego of the
individual subject. What he had in view were the limit situations in
which moral insights have to be defended against the prejudices of
the majority or even of a prejudiced society as a whole. Idealist
thought was influenced by Luther's Protestant maxim, "Here I stand,
I can do no other." American Pragmatism, also a child of the Prot
estant spirit, must be credited with the achievement of overcoming
the individualism of contract theories without sacrificing the moral
contents that had previously been situated in the conscience of the
individual. Peirce, Royce, and Mead developed the ideas of an unlim
ited communication community and a universal discourse that rep
resent an alternative to abstract internality because they transcend all
existing states of affairs while retaining the character of a public court
of appeal. Already the notion of "ideal role taking" preserves the
characteristics of a transcendental socialization and invokes the social
bond that unites humanity as a whole.
With the notion of an ideally extended communication community
as its point of reference, moral theory also leaves behind all presocial
concepts of the person. Individuation is merely the reverse side of
socialization. Only in relations of reciprocal recognition can a person
131
constitute and reproduce his identity. Even the innermost essence of

a person is internally connected with the outermost periphery of a
far-flung network of communicative relations. Only in the aggregate
of his communicative expressions does a person become identical
with himself. The ego is also rendered vulnerable by the social inter
actions through which it is formed-by the dependencies in which it
becomes implicated and the contingencies to which it is exposed.
Morality serves to counterbalance this susceptibility implicit in the
very process of socialization.
But in that case it is not possible to deduce a substantive normative
concept of autonomy from the necessary presuppositions of the te
leological action of isolated subjects.24 To achieve this goal we must
take the model of action oriented to reaching understanding as our
point of departure. In communicative action speaker and hearer
assume that their perspectives are interchangeable. By entering into
an interpersonal relation in the performative attitude, they commit
themselves to recognizing each other symmetrically as responsible
subjects capable of orienting their actions to validity claims. And their
behavioral expectations remain reciprocally interconnected in the
normative context of the lifeworld. In this way the necessary presup
positions of communicative action constitute an infrastructure of pos
sible communication constructed around a moral core-the idea of
unforced intersubjectivity. Since argumentative praxis is merely a
reflexive form of communicative action, it transmits the normative
force of its presuppositions to the communicative presuppositions of
argumentation. Only at this level do the perspectives, relations of
recognition, and normative expectations built into communicative
action become completely reversible in all relevant respects, for partici
pants in argumentation are credited with the ability to distance them
selves temporarily from the normative spectrum of all existing forms
of life.
The foregoing prefigures an ethics of discourse that can explain
how Kohlberg can postulate an internal connection between co gnitive
abilities and moral consciousness. Robert Selman, for example, has
extended research on perspective taking into the field of interper
sonal negotiation strategies. 25 These and similar investigations26 pro
vide empirical confirmation of the underlying intuition that an
intersubjective interpretation of the moral point of view can be pur-
1 32
sued as far as the presuppositional analysis of action oriented to

reaching understanding. This coheres in turn with the Aristotelian
insight that we acquire our moral intuitions not through philosophical
instruction or other explicit communications but in an implicit man
ner through socialization. Because they must take place in the me
dium of communicative action, all socialization processes are shaped
by the structure of this type of action, but the presuppositions of
communicative action already carry within themselves the germ of
morality.
v
From his dissertation onward , Kohlberg held steadfastly to his the

oretical program. He repeatedly analyzed the same data and elabo
rated in an ever more precise fashion the same fascinating ideas.
This attitude contrasts impressively with the latent opportunism of
institutionalized research. His bearing made Kohlberg a philosopher
among scientists, though one who never deviated from the standards
of a true researcher. He was, on the other hand, so much a philos
opher that his conception of how his theory should be implemented
in educational praxis bordered on the Platonic. He walked a stretch
of the road that led a disillusioned Piaget from philosophy to psy
chology, but in the opposite direction. In the role of psychologist, he
embodied more completely the classical self-understanding of phi
losophy than the majority of his philosopher colleagues who are
intimidated by the authority of the sciences.
4
To Seek to Salvage an Unconditional Meaning
Without God Is a Futile Undertaking:
Reflections on a Remark of Max Horkheimer
Max Horkheimer's late philosophy, scattered throughout various

notes and essays, takes the form of reflections from a damaged life.
Alfred Schmidt has deciphered in them the outline of a systematic
intention. His proof is an indirect one, using Horkheimer's tools as
a key to unlock the door to Schopenhauer's philosophy of religion. 1
These illuminating reconstructions have impressed upo n me the rea
sons and motives that induced Horkheimer to look to Schopenhauer
in his quest for a religion that could satisfy the longing for perfect
justice. Horkheimer's interest in the doctrines of judaism and Chris
tianity was spurred less by a concern with God as such than with the
redemptive power of God's will. The inj u stice that comes to pass in
a suffering creature should not be permitted to have the last word.
At times it seems as if Horkheimer wanted to put the religious prom
ise of redemption directly at the service of morality. At one point he
explained the prohibition of images in terms of the notion that "in
the jewish religion what is important is not how things stand with
God, but how they stand with men. "2 Schopenhauer's metaphysics
seemed to offer a resolution of an aporia in which Horkheimer had
become involved in consequence of two equally strong convictions.
For him too, the critical task of philosophy consisted essentially in
salvaging the truth in religion in the spirit of the Enlightenment;
nevertheless, it was dear to him that "one cannot secularize religion
without giving it up."S
This aporia has haunted Greek philosophy like a shadow from the
moment of its initial encoun ter with the jewish and Christian tradi-
1 34
tion onward. In Horkheimer's case, it is made even more acute by

his profound skepticism concerning reason. What for him is the
essential substance of religion-morality-is no longer tied to reason.
Horkheimer praises the dark writers of the bourgeoisie for having
"trumpeted far and wide the impossibility of deriving from reason
any fundamental argument against murder."4 I have to admit that
this remark irritates me now no less than it did almost four decades
ago when I first read it. I have never been altogether convinced of
the cogency of the skepticism concerning reason underlying Hork
heimer's ambivalence toward religion. The idea that it is vain to strive
for unconditional meaning without God betrays not just a metaphys
ical need; the remark is itself an instance of the metaphysics that not
only philosophers but even theologians themselves must today get
along without.
Before I attempt to back up this objection, I want to clarify the
fundamental moral intuition that guided Horkheimer throughout
his life ; I will then turn to the kinship between religion and philos
ophy that Horkheimer never lost from sight and, finally, reveal the
premises on which he based his appropriation of Schopenhauer's
negative metaphysics. In what follows I draw on notes and essays that
Alfred Schmidt made available to the public5 and to whose systematic
import he was first to draw attention.6
I
Once the rationality of the remorse experienced by a religiously

tutored conscience is rejected by a secularized world, its place is taken
by the moral sentiment of compassion. When Horkheimer expressly
defines the good tautologically as the attempt to abolish evil, he has
in view a solidarity with the suffering of vulnerable and forsaken
creatures provoked by outrage against concrete injustices. The re
conciling power of compassion does not stand in opposition to the
galvanizing power of rebellion against a world devoid of atonement
and reparation for injustice. Solidarity and justice are two sides of
the same coin ; hence, the ethics of compassion does not dispute the
legitimacy of the morality of justice but merely frees it from the
rigidity of the ethics of conscience. Otherwise the Kantian pathos
expressed in Horkheimer's injunction "to proceed into the desert in
1 35
spite of everything, even if hope were lost" would be incomprehen

sible.7 And under the banner "necessary futility," Horkheimer does
not shrink from drawing the almost Protestant conclusion : "It is true
that the individual cannot change the course of the world. But if his
whole life is not a gesture of wild despair that revolts against it, he
will fail to realize that infinitely small, insignificant, futile, nugatory
modicum of good of which he is capable as an individual."8 The
shared fate of exposure to the infinitude of an indifferent universe
may awaken a feeling of solidarity in human beings, but among the
community of the forsaken, the hope of solidarity and pity for one's
neighbor must not undermine equal respect for everyone. Moral
feelings imbued with a sense of justice are not just spontaneous
impulses; they are more intuitions than impulses. In them a correct
insight, in an emphatic sense of "correct," comes to expression. The
positivists "have not the faintest inkling that hatred of the decent and
admiration for the depraved are inverted impulses not just before
the tribunal of custom, but of truth, and that they are not merely
reprehensible in an ideological sense, but are objectively debased
experiences and reactions."9
Horkheimer is so secure in his fundamental moral intuitions that
he can qualify them only as "correct insights." This moral cognitivism
seems to place him firmly in the Kantian camp. Nevertheless, he is
so profoundly influenced by the dialectic of enlightenment that he
repeatedly disputes the role Kant still accorded practical reason. What
remains is only a "formalistic reason" that is no "more closely allied
to morality than to immorality."10 Material investigations alone could
overcome this sterile formalism, though indeed only in a paradoxical
manner. Unable to specify the good, a critical theory of society should
reveal specific injustices in given cases. Because this theory, in its
skepticism toward reason, no longer maintains a positive relation to
the normative contents it uncovers step by step in the criticism of
unjust condition s, it must borrow its normative orie ntations from a
cultural ethos that has already been superseded-that of a meta
physically grounded theology. The latter preserves the legacy of a
substantive reason that has since been rendered impotent.
Horkheimer is under no illusions about the vertiginous nature of
this theoretical undertaking. Social theory "has superseded theology
but has no new heaven to which it can point, not even a mundane
1 36
one. Of course, social theory cannot completely efface [heaven's]

traces and hence is repeatedly questioned about how it is to be at
tained-as though it were not precisely the discovery of social theory
that the heaven to which one can point the way is no heaven."1 1 No
theory could possibly accommodate itself to this Kafkaesque figure
of thought, at least not without embracing an aesthetic mode of
expression and becoming literature. Hence, the thoughts of the late
Horkheimer circle around the idea of a theology that must be "dis
placed" by the critical and self-critical activity of reason, yet which,
in its capacity as justifying morality's claim to unconditionality, cannot
be replaced by reason . Horkheimer's late philosophy may be under
stood as wrestling with this dilemma and his interpretation of Scho
penhauerian metaphysics as a proposal for resolving it.
In the essay entitled "Theism-Atheism," Horkheimer traces the
development of the Hellenistic notion of an interrelation between
theology and metaphysics up to the great metaphysical systems in
which theology and natural philosophy converged. He is interested
above all in the militant atheism of the eighteenth century that "was
able to promote rather than to stifle interest in religion." 12 Even the
materialistic antithesis to Christianity that substituted "Nature" for
"God" and merely readjusted the fundamental concepts accordingly
still remained caught up in the metaphysical framework of world
views. Kant's critique of metaphysics opened the door to the mystical
and messianic currents that, from Baader and Schelling to Hegel and
Marx, found their way into philosophy. Horkheimer was aware of
the theological current in Marxist theory from the beginning: with
the idea of a just society, the Enlightenment opened up the prospect
of a new beyond in the here and now; the spirit of the Gospel was
now to reach worldly fulfillment through the march of history.
The secular sublation (Aufhebung) of ontotheology by the philoso
phy of history has profoundly equivocal implications. On the one
hand, philosophy becomes disguised theology and salvages the latter's
essential content. The very meaning of atheism itself ensures the
enduring relevance of theism : "Only those who employed the word
in a derogatory sense understood it as the opposite of religion. Those
who professed atheism at a time when religion still had power were
wont to identify themselves more sincerely with the theistic precept
of devotion to one's neighbor and other creatures as such than most
137
Reflections o n a Remark o f Max Horkheimer
of the adherents and fellow travelers of the various religious confes

sions." 1 3 On the other hand, philosophy can recover the idea of the
unconditioned only in the medium of a reason that has in the interim
sacrificed the infinite on the altar of historical contingency and has
abjured the unconditioned. A reason that can appeal to no authority
higher than that of the sciences is a naturalized faculty that has
regressed to intelligence in the service of pure self-affirmation; it
measures itself by the yardstick of functional contributions and tech
nical successes, and not by a mode of validity that transcends space
and time: "With God dies eternal truth." 14 In the wake of the Enlight
enment, the truth in religion can be salvaged only in a way that
annihilates truth. A critical theory that sees itself as the "successor"
to theology finds itself in this unhappy predicament because every
thing to do with morality ultimately derives from theology.
II
The rational sublation of theology and its essential contents : how can
this still be accomplished in the present day, in the light of the
irreversible critique of metaphysics, without destroying the import of
religious doctrines or of reason itself ? With this question Hork
heimer, the pessimistic materialist, appeals to Schopenhauer, the
pessimistic idealist. On Horkheimer's surprising interpretation, Scho
penhauer's enduring importance lies in the fact that his thorough
going negativism salvages the "spirit of the Gospels." According to
Horkheimer, Schopenhauer accomplished the improbable feat of
providing an atheistic justification of the morality underlying theol
ogy, and thus of preserving religion in the absence of God.
In the world as will and representation, Horkheimer discerns, first,
the sterile Darwinian operation of instrumental reason degraded to
a tool of self-preservation, which-up to and including a globally
objectifying scientific intellect-is dominated by a blind and indefa
tigable will to life that pits one subject against another. On the other
hand, precisely this reflection on the abysmally negative ground of
being is supposed to awaken in subjects who seek remorselessly to
dominate one another some inkling of their common fate and an
awareness that all manifestations of life are pervaded by an identical
will : " If the realm of appearances, sensible reality, is not the work of
1 38
positive divine power, an expression of inherently good, eternal

Being, but of a will that affirms itself in everything finite, that is
mirrored in a distorted fashion in multiplicity, and yet that remains
at a profound level identical, then everyone has reason to view him
self as one with all others, not with their specific motives, but with
their entanglement in delusion and guilt, their drivenness, joy and
decline. The life and fate of the founder of Christianity becomes a
model, no longer based on commands but on insight into the inner
constitution of the world." 15
What fascinated Horkheimer in Schopenhauer is the prospect of
a metaphysical justification of morality through insight into the con
stitution of the world as a whole, yet in such a way that this insight is
at the same time directed against central assumptions of metaphysics
and coheres with postmetaphysical skepticism concerning reason.
Negative metaphysics upholds the distinction between essence and
appearance only with an inversion of the order of priority between
them-inverted Platonism . This in turn grounds the expectation that
insight into the "pitiless structure of infinitude" could produce "a
community of the forsaken." However, Horkheimer is aware of the
shadow of performative self-contradiction that has haunted all neg
ative metaphysics since Schopenhauer and Nietzsche. Even if we
prescind from epistemological misgivings about intuitive, bodily ac
cess to the thing-in-itself, it remains mysterious how the turning of
the irrational world-will against itself that constrains it to continual
reflection will come about: "The metaphysics of the irrational will as
the essence of the world must lead to reflection on the problematic
of truth." 1 6 In Schmidt's formulation of the dilemma, "If the essence
of the world is irrational, then that cannot remain incidental to the
truth claim of precisely this thesis." 1 7 In the light of this result, the
statement that it is futile to seek to salvage unconditional meaning
without God can also be understood as a criticism of Schopenhauer,
as a critique of the "last great philosophical attempt to rescue the
essential core of Christianity." 18
In the final analysis, Horkheimer's ambiguous formulations vacil
late between Schopenhauer's negative-metaphysical justification of
morality and a return to the faith of his forefathers. This unresolved
argumentative impasse leads me to rexamine the premise from
which Horkheimer's late philosophy begins : that "formalistic" reason,
1 39
or the procedural reason that remains under conditions of postme

taphysical thought, is equally indifferent to morality and immorality.
As far as I can discern, Horkheimer's skeptical assertion rests pri
marily on the contemporary experience of Stalinism and on a con
ceptual argument that presupposes the ontological concept of truth.
III
Horkheimer's thought is influenced even more than Adorno's by the

harrowing historical fact that the ideals of freedom, solidarity, and
justice deriving from practical reason, which inspired the French
Revolution and were reappropriated in Marx's critique of society, led
not to socialism but to barbarism under the guise of socialism: "The
vision of instituting justice and freedom in the world which underlay
Kant's thought has been transformed into the mobilization of nations.
With each revolt that followed in the wake of the great revolution in
France, it seems, the humanistic elements atrophied while nationalism
thrived. In this century it was socialism itself that orchestrated the
supreme farce of perverting the pledge to humanity into an intran
sigent cult of the state. . . . What Lenin and the majority of his
comrades aspired to before assuming power was a free and just
society. In reality they prepared the way for a totalitarian bureaucracy
under whose sway there was no more freedom than in the tsarist
empire. That the new China is entering on a phase of barbarism is
plain to see."19 From this experience Horkheimer drew consequences
for the reconstruction of the architectonic of reason announced in
the concept of "instrumental reason." There is no longer any differ
ence between the operation of the understanding in the service of
subjective self-assertion, which imposes its categories on everything
and transforms it into an object, and reason as the faculty of ideas
whose place understanding has usurped. Indeed, the ideas them
selves have been caught up in the dynamic of reification; elevated to
absolute ends, they retain merely a functional significance for other
ends. But by exhausting the supply of ideas in this way, every claim
that points beyond instrumental rationality loses its transcending
power; truth and morality forfeit their unconditional meaning.
Thinking that is sensitive to historical changes, even down to its
fundamental concepts, submits itself to the tribunal of new experi-
140
ence. Thus, it is not inappropriate to ask whether the bankruptcy of

state socialism that has in the interim become apparent does not offer
other lessons, for this bankruptcy is partly the doing of ideas that the
regime, while distancing itself from them ever further, misused for
the purposes of its own legitimation because-which is more impor
tant-it had to appeal to them. A system that collapsed despite its
brutal Orwellian apparatus of oppression because social conditions
eloquently contradicted everything prefigured by its legitimating
ideas, manifestly cannot dispose of the inner logic of these ideas as it wishes.
In the ideas of the constitutionally embodied republican tradition,
however egregiously abused, there persists the element of existing
reason that resisted the "dialectic of enlightenment" by eluding the
leveling gaze of the negative philosophy of history.
The controversy surrounding this thesis could be resolved only by
recourse to material analyses. As a consequence, I will limit myself
to the conceptual argument that Horkheimer develops from the
critique of instrumental reason.
Horkheimer's assertion that the difference between reason and
understanding has become obsolete in the course of the world-historical
process still presupposed, in contrast to contemporary poststructur
alism, that we can recall the emphatic concept of reason. The critical
import of "instrumental reason" is first thrown into relief by this act
of recollection. And through anamnetic retrieval of the subst<;t ntive
reason of religious and metaphysical worldviews, we can reassure
ourselves of the unconditionality that concepts such as truth and
morality once carried with them before they succumbed to positivistic
and functionalistic disintegration. An Absolute or Unconditional be
comes accessible to philosophy only together with justification of the
world as a whole, and hence only as metaphysics. But philosophy
remains true to its metaphysical beginnings only as long as it attempts
"to imitate positive theology" and proceeds on the assumption that
cognizing reason rediscovers itself in the rationally structured world
or itself actually confers a rational structure on nature and history.
As soon as the world "in its essence, by contrast, does not of necessity
cohere with the spirit, philosophical confidence in the being of truth
dissipates completely. Then truth is henceforth sublated only in tran
sient human beings themselves and becomes as transient as they
are."2 0
141
Reflections o n a Remark o f Max Horkh eimer
It never occurred to Horkheimer that there might be a difference

between "instrumental" and "formal" reason. Moreover, he uncere
moniously assimilated procedural reason-which no longer makes
the validity of its results dependent on the rational organization of
the world but on the rationality of procedures through which it solves
its problems-to instrumental reason. Horkheimer assumes that
there cannot be truth without an Absolute, without a world-tran
scending power "in which truth is sublated." Without ontological
anchoring, the concept of truth is exposed to the inner-worldly con
tingencies of mortal men and their changing situations; without it,
truth is no longer an idea but merely a weapon in the struggle of
life. Human knowledge, including moral insight, can lay claim to
truth, he believes, only if it judges itself in terms of relations between
it and what is as these relations are manifested to the divine intelli
gence alone. In contrast to this strangely traditional conception, in
the final section I will argue for a modern alternative-a concept of
communicative reason that enables us to recover the meaning of the
unconditioned without recourse to metaphysics. But first we must
clarify the true motive that causes Horkheimer to hold fast to the
classical concept of truth as adaequatio intellectus ad rem.
Decisive for Horkheimer's persistence in maintaining an ontologi
cal anchoring of truth are the ethical reflections he attributes to
Schopenhauer: only insight into the identity of all life, into a unitary
ground of being, even if it be irrational, in which all individual
appearances are brought into harmony with one another, "can
ground solidarity with all creatures long before death. "2 1 The unified
thought of metaphysics renders plausible why it should be that the
effort to overcome egoism would find a sympathetic response in the
constitution of the world. For this reason alone, unity takes prece
dence over multiplicity for philosophers, the unconditional occurs
only in the singular, and the one God has greater importance for
Jews and Christians than the multiple deities of antiquity. It is the
peculiar fate of bourgeois culture that individuals entrench them
selves in their particularity and thereby reduce individualism to a
falsehood. Horkheimer so emphatically regards this societal state of
nature of competitive society as the fundamental problem of morality
that for him justice and solidarity become synonymous with "ren
ouncing the self-assertion of the isolated ego." Egoism has congealed
1 42
to such an extent into an inverted condition of things that the tran

sition from self-love to devotion to others is unthinkable without
metaphysical assurance of the prior unity of an unfathomable world
will that provokes us to insight into a possible solidarity of the des
titute: "Schopenhauer drew the necessary consequences : the insight
into the baseness of one's own life which cannot be separated from
the suffering of other creatures is correct; the identification with
those who suffer, with man and animal, is correct, the renunciation
of self-love, of the drive to individual well-being as the ultimate goal;
and the induction after death into the general, the non-personal, the
nothing is desirable."22 The individuated will is base only when it
turns itself against others ; it becomes good when, through compas
sion, it recognizes its true identity with all other beings.
IV
Already in The Dialectic of Enlightenment Horkheimer credits de Sade

and Nietzsche with the recognition that "after the formalization of
reason, compassion still remained, so to speak, as the sensual con
sciousness of the identity of the general and the particular, as natu
ralized mediation."23 On the Schopenhauerian interpretation, of
course, compassion cannot assume the role of dialectical mediation
between individual and society, between equal respect for all and the
solidarity of each with all. Here it is solely a matter of the abstract
self-overcoming of individuality, of the dissolution of the individual
in an all-encompassing oneness. But with this the very idea in which
the moral substance of Christianity consists is abrogated. Those who
at the Last Judgment come, one after the other, before the eyes of
God as unrepresentable individuals stripped of the mantle of worldly
goods and honors-and hence as equals-in the expectation of re
ceiving a fair judgment, experience themselves as fully individuated
beings who must give an account of their life histories in full respon
sibility for their actions. Together with this idea, however, the pro
fond intuition that the bond between solidarity and justice must not
be severed must also be given up.
Admittedly, in this respect Horkheimer does not follow Schopen
hauer without misgivings. His interpretation of the Ninety-first Psalm
reveals his struggle to overcome a certain dissonance. The doctrine
143
of the individual soul, he writes, has an additional significance in

Judaism, one unadulterated by the expectation of an afterlife: "The
idea of continued existence signifies in the first instance, not the
after-life, but the identification with the nation so crassly distorted
by modern nationalism, which has its prehistory in the bible. By
conducting his life in accordance with the Torah, by spending days,
months, years in obedience to the Law, the individual becomes so
much one with others despite personal differences that after his death
he continues to exist through those who survive him, in their observ
ance of tradition, of love for the family and the tribe, in the expec
tation that at some time things may still become better in the world .
. . . Not unlike the figure of Jesus in Christianity, Judaism as a whole
bore witness to redemption."24 Horkheimer tries to circumvent the
problem of superseding the individual, of repudiating inalienable
individuality, by changing the question. The issue is not whether the
kingdom of the Messiah is of this world but whether the fundamental
moral intuition of Judaism and Christianity to which Horkheimer
unwaveringly adheres can ultimately be adequately explained without
reference to the unrestricted individuation possible in a universal
confederation.
The moral impulse of unwillingness to resign oneself to the force
of circumstances that have the effect of isolating the individual and
to secure the happiness and power of one person only at the cost of
the misfortune and powerlessness of another-this impulse confirms
Horkheimer in the view that the reconciling potential of solidarity
with those who suffer can be realized only if individuals renounce
themselves as individuals. He fails to see that the danger of a nation
alistic distortion of the identificatory bond with the nation arises
precisely at the moment when false solidarity permits individuals to
be subsumed into the collectivity. Unified metaphysical thought
however negatively accented-transposes solidarity, which has its
proper place in linguistic intersubjectivity, communication, and in
dividuating socialization, into the identity of an underlying essence,
the undifferentiated negativity of the world-will.
Quite a different, dialectical unity is produced in communication
in which the structure of language inscribes the gap between I and
Thou. The structure of linguistic intersubjectivity makes harmony
between the integration of autonomy and devotion to others possible
1 44
for us-in other words, a reconciliation that does not efface differ
ences. Horkheimer is by no means deaf to this promise of reconcili
ation inherent in language itself. At one point he puts it trenchantly:
"Language, whether it wants to or not, must lay claim to truth."25 He
also recognizes that we have to take into account the pragmatic di
mension of language use, for the context-transcending truth-claim
of speech cannot be grasped from the blinkered perspective of a
semantics that reduces utterances to propositions : "Truth in speech
is not properly predicated of detached, naked judgments, as though
printed on a piece of paper, but of the conduct of the speaker toward
the world that is expressed in the judgment and concentrates itself
in this place."26 What Horkheimer has in mind is clearly the theolog
ical tradition, extending from Augustine through logos mysticism to
radical Protestantism, that appeals to the originary character of the
divine Word and to language as the medium of the divine message
to man : "But theological metaphysics is in the right against positivism,
because no proposition can avoid raising the impossible claim, not
merely to an anticipated result, to success, as positivism maintains,
but to truth in the proper sense, regardless of whether the speaker
reflects on it or not."27 Prayer, in which the believer seeks contact
with God, would lose its categorial difference from incantation and
regress to the level of magic if we confused the illocutionary force of
our assertions with their perlocutionary effects, as does the unreal
izable program of linguistic nominalism.
But these insights remain sporadic. Horkheimer fails to treat them
as clues to a language-pragmatical explanation of the unconditional
meaning associated with unavoidable truth claims. His skepticism to
ward reason is so thoroughgoing that he can no longer see room for
communicative action in the world as it is now constituted : "Today
talk has become stale and those who do not want to listen are not
altogether wrong . . . . Speaking has had its day. Indeed so has action,
at least insofar as it was once related to speech. "2 8
v
His pessimistic diagnosis of the times is not the only reason that
Horkheimer refrains from seriously entertaining the question of how
something we accomplish on a daily basis-orienting our action to
1 45
context-transcending validity claims-is possible. In fact, a profane

answer to this question, such as the one proposed by Peirce, for
instance, could not have sufficiently satisfied Horkheimer's metaphys
ical need for religion.
Horkheimer equated Kant's formalistic reason with instrumental
reason. But Peirce reinterprets Kantian formalism in the direction of
a pragmatics of language and construes reason in procedural terms.
The process of sign interpretation achieves self-awareness at the level
of argumentation. Peirce now shows how this nonquotidian form of
communication is commensurate with the "unconditional meaning"
of truth and of context-transcending validity claims in general. He
conceives of truth as the redeemability of a truth claim under the
communicative conditions of an ideal community of interpreters
that is, one extended ideally in social space and historical time. The
counterfactual appeal to an unlimited communication community of
this kind replaces the moment of infinitude or the supratemporal
character of "unconditionality" with the idea of an open yet goal
directed process of interpretation that transcends the boundaries of
social space from within from the perspective of an existence situated
in the world. In time, learning processes are to form an arch bridging
all temporal distance; in the world, the conditions we assume are at
least sufficiently fulfilled in every argument are to be realized. We
are intuitively aware that we cannot rationally convince anyone, not
even ourselves, of something if we do not accept as our common
point of departure that all voices that are at all relevant should be
heard, that the best arguments available given the current state of
our knowledge should be expressed, and that only the unforced force
of the better argument should determine the "yes" and "no" re
sponses of participants.
The tension between the intelligible realm and the realm of phe
nomena is thereby shifted to general presuppositions of communi
cation, which-despite their ideal and only approximately realizable
content-participants must in every case actually accept if they wish
to thematize a controversial truth claim. The idealizing force of these
transcending anticipations penetrates into the very heart of everyday
communicative praxis, for even the most fleeting speech-act-offer,
the most conventional "yes" or "no," point to potential reasons, and
hence to the ideally extended audience they must convince if they
1 46
are valid. The ideal moment of unconditionality is deeply rooted in

factual processes of communication because validity claims are Janus
faced : as universal, they outstrip every given context; at the same
time, they must be raised and gain acceptance here and now if they
are to sustain an agreement capable of coordinating action. In com
municative action, we orient ourselves to validity claims that we can
raise only as a matter of fact in the context of our language, of our
form of life, whereas the redeemability implicitly co-posited points
beyond the provinciality of the given historical context. Whoever
employs a language with a view to reaching understanding lays him
self open to a transcendence from within. He is left without any
choice because he masters the structure of language through the
intentionality of the spoken word. Linguistic intentionality outstrips
subjects but without subjugating them.
Postmetaphysical thought differs from religion in that it recovers
the meaning of the unconditional without recourse to God or an
Absolute. Horkheimer's dictum would have been justified only if by
"unconditional meaning" he had meant something different from
the notion of unconditionality that also belongs to the meaning of
truth as one of its moments. The significance of unconditionality is
not to be confused with an unconditional meaning that offers con
solation. On the premises of postmetaphysical thought, philosophy
cannot provide a substitute for the consolation whereby religion in
vests unavoidable suffering and unrecompensed injustice, the contin
gencies of need, loneliness, sickness, and death, with new significance
and teaches us to bear them. But even today philosophy can explicate
the moral point of view from which we can judge something impar
tially as just or unjust; to this extent, communicative reason is by no
means equally indifferent to morality and immorality. However, it is
altogether a different matter to provide a motivating response to the
question of why we should follow our moral insights or why we should
be moral ar all. In this respect, it may perhaps be said that to seek to
salvage an unconditional meaning without God is a futile undertak
ing, for it belongs to the peculiar dignity of philosophy to maintain
adamantly that no validity claim can have cognitive import unless it
is vindicated before the tribunal of justificatory discourse.
5
Morality, Soc iety, and Ethics : An Interview
with Torben Hviid Nielsen
Torben Hviid Nielsen: The main topic of discussion will be your views
on moral theory and ethics, particularly in the form they have as
sumed since the publication of Theorie des kommunikativen Handelns in
1 98 1 . 1 In the first part we will concentrate on the concept of morality
and the relation between justice, law, and care. The second part will
deal with questions concerning the universal-pragmatic justification
of discourse ethics. The focus will be on the validity of norms, the
status of the so-called ideal speech situation, and the differentiation
between the justification of norms and the procedure of democratic
will formation. And in the third part we will discuss morality and
ethics in relation to the notions of system and lifeworld.
I
THN: How should we understand the development that has led you
from the sociological critique of the pathologies of modernity in The
Theory of Communicative Action to the moral theory developed in Moral
Consciousness and Communicative Action2 and the subsequent series of
articles and lectures? Can discourse ethics be seen as a philosophical
answer, from the perspective of the individual, to the sociological
question of the proper, nonpathological relation between system and
lifeworld in the modern world? Why have you concentrated since
1 98 1 on issues in philosophical ethics rather than on this sociological
question left open in Theory of Communicative Action ?
148
Morality, Society, and Ethics
]urgen Habermas : I see this differently. What was of primary impor

tance for the philosophical foundations of The Theory of Communicative
Action was the introduction of the linguistic-pragmatic concept of
communicative rationality. Granted, I did deal with legal and moral
development following Weber and Durkheim, but the two theoretical
approaches I drew upon in my discussion, discourse ethics and Kohl
berg's theory of stages of moral consciousness, remained in the back
ground at that time. Only in subsequent years did I turn my attention
directly to these matters that had been left to one side. The title essay
of Moral Consciousness and Communicative Action dates back to my time
at the Max Planck Institute at Starnberg and reflects the kind of
research undertaken there. The essay on discourse ethics is the result
of seminars I held directly after my return to Frankfurt, and hence
in a philosophy department. Since 1 983 I have been working in
different professional surroundings, which tends to influence the
direction of research.
Your assumption is also mistaken in that this concern with questions
of moral theory simply takes up again a number of problems I had
already discussed in 1 973 in the final part of Legitimationsprobleme im
Spatkapitalismus. 5 At that time I proposed a model of "suppressed
generalizable interests" in order to clarify the sense in which it is
possible to distinguish between "general" and "particular" interests.
Later, in The Theory of Communicative Action I did not return to this;
rather, as you correctly point out, I attempted to describe social
pathologies by means of a two-level concept of society, that is, as
deformations that can be traced back to disruptions in the reproduc
tion of the lifeworld (2 : 1 42ff.). There I was particularly interested in
pathologies that arise when systemic imbalances that are symptomatic
of crisis in the economy or the state apparatus are displaced onto the
lifeworld and interfere with its symbolic reproduction (2 : 385ff.). Had
I wanted to analyze more closely this phenomenon of reification of
communicative relationships through monetarization and bureaucra
tization-that is, the phenomenon to which Marx attached the global
term alienation-moral-theoretical considerations would have been
completely out of place. Rather, a more careful definition of the
concept of systematically distorted communication would have been
needed. This I considered-on the basis of empirical research on
pathologies within the family-as the interpersonal counterpart of
149
the intrapsychical disturbances that psychoanalysis traces back to the

unconscious avoidance of conflict and explains in terms of corre
sponding defense mechanisms. But since 1 974 I have not returned
to these ideas on communication pathologies that arise at the level of
elementary interactions. 4 I still consider my suggestions relevant, by
the way, and I think that this is confirmed by the interesting work of
Jim Bohman and Martin Low-Beer.
THN: You present discourse ethics as both the continuation and

completion of your earlier theoretical work on ethics and a response
to the political agenda of the public discussions of the 1 980s. Do you
feel a certain tension between these two concerns, between your own
theoretical development, which points back to questions raised in the
1 960s, and the political themes of the 1 980s? Have these themes
contributed to your shift of emphasis from a more social, " Hegelian"
ethics to a more individual, " Kantian" conception of morality?
JH: Actually, my research program has remained the same since
about 1 970, since the reflections on formal pragmatics and the dis
course theory of truth first presented in the Christian Gauss Lec
tures.5 On the other hand, anyone who is at all sensitive to politics
and the political impact of theories is bound to react to changed
contexts. In the 1 960s it was necessary to engage the theories of
technocracy of one group and in the early 1 970s the crisis theories
of another. And since the mid- 1 970s I have felt the pressure of the
neoconservative and the poststructuralist critiques of reason, to which
I responded with the concept of communicative rationality. This
constellation remained unchanged in the 1 980s, and it was for this
reason that I continued to work on a critique of the philosophy of
consciousness and sought to lend it greater philosophical precision.
In Der Philosophische Diskurs der Modeme ( 1 985) I tried to show that
"representational thinking" can be replaced by something other than
the defeatism of the deconstructionists or the contextualism of the
neo-Aristotelians. 6
It was in connection with this intersubjective self-criticism of reason
that I also reacted to the rather suspect popularity enjoyed by philo
sophical ethics at present and proceeded to work out the issues raised
by Mead's communicative ethics, which had always interested me. 7
Hence discourse ethics, like Mead's work, takes its lead from intui-
1 50
tions underlying Kantian moral theory while eschewing the latter's

individualistic premises.
THN: Discourse ethics refers specifically to modern conditions of

life, as do The Theory of Communicative Action and The Philosophical
Discourse of Modernity. They defend the Enlightenment and modernity
against traditionalism, on the one hand, and postmodernism, on the
other. Thus, both you and one of your main opponents, the neo
Aristotelian Alasdair Macintyre, agree that the concept of virtue, for
instance, is incompatible with modern life. Why is it that all tradi
tional, substantive moralities have become obsolete? And what is your
response-a nonsubstantive morality that is better warranted than
Macintyre's answer, better, that is, than the proposed return to tra
ditional virtues?
]H: In my opinion, After Virtue has two principal weaknesses. First,
Macintyre makes things too easy for himself from a critical perspec
tive by selecting in Alan Gewirth an untypical and rather easily crit
icizable example of a universalistic position instead of dealing with
Rawls or Dworkin or Apel. Second, his appeal to the Aristotelian
concept of praxis gets him into trouble as soon as he tries to extract
a universal core from the pluralism of equally legitimate forms of
life that is unavoidable in modernity. Where can he find an equivalent
for what Aristotle could still fall back on-1 mean a substitute for the
metaphysical preeminence of the polis as the exemplary form of life
in which people (all those, that is, who do not remain barbarians)
could realize the goal of a good life? Under modern conditions,
philosophy can no longer stand in judgment over the multiplicity of
individual life projects and collective forms of life, and how one lives
one's life becomes the sole responsibility of socialized individuals
themselves and must be judged from the participant perspective.
Hence, what is capable of commanding universal assent becomes
restricted to the procedure of rational will formation.
THN: Discourse ethics offers a view of ethics that is narrow or
minimal in two respects : its approach is deontological, cognitivistic,
formalistic, and universalistic, and it focuses on issues of justice as its
primary subject matter. Thus, the traditional concern with the good
or happiness (or a combination of both) is excluded. Why this exclu
sive focus on justice? Do you regard this as a necessary feature of all
modern ethical theories?
151
JH: Under modern conditions of life none of the various rival

traditions can claim prima facie general validity any longer. Even in
answering questions of direct practical relevance, convincing reasons
can no longer appeal to the authority of unquestioned traditions. If
we do not want to settle questions concerning the normative regula
tion of our everyday coexistence by open or covert force-by coer
cion, influence, or the power of the stronger interest-but by the
unforced conviction of a rationally motivated agreement, then we
must concentrate on those questions that are amenable to impartial
judgment. We can't expect to find a generally binding answer when
we ask what is good for me or for us or for them; instead, we must
ask what is equally good for all. This "moral point of view" throws a
sharp, but narrow, spotlight that picks out from the mass of evaluative
questions practical conflicts that can be resolved by appeal to a gen
eralizable interest; in other words, questions of justice.
In saying this, I do not mean that questions of justice are the only
relevant questions. Usually ethical-existential questions are of far
more pressing concern for us-problems, that is, that force the in
dividual or group to clarify who they are and who they would like to
be. Such problems of self-clarification may well be of greater concern
to us than questions of justice. But only the latter are so structured
that they can be resolved equitably in the equal interest of all. Moral
judgments must meet with agreement from the perspective of all
those possibly affected and not, as with ethical questions, merely from
the perspective of some individual's or group's self-understanding or
world view. Hence moral theories, if they adopt a cognitivist approach,
are essentially theories of justice.
THN: Why is it that the sphere of ustice" cannot be further dif
ferentiated ? Why should the fragmentation of modernity come to a
halt before the three Kantian critiques and the corresponding spheres
of value, so that questions ofjustice can be treated only under a single
aspect? Michael Walzer's book Spheres of justice may be read as an
extended argument for dividing justice into different spheres (of
membership, welfare, the economy, education, and so on) ; in this
way he seeks to defend pluralism and equality: "the principles of
justice are themselves pluralistic in form . . . different social goods
ought to be distributed for different reasons, in accordance with
different procedures, by different agents." 8
1 52
I agree entirely with this statement though not with the con
sequences Walzer wishes to draw from it.
That a norm is just or in the general interest means nothing more
than that it is worthy of recognition or is valid. Justice is not some
thing material, not a determinate "value," but a dimension of validity.
Just as descriptive statements can be true, and thus express what is
the case, so too normative statements can be right and express what
has to be done. Individual principles or norms that have a specific
content are situated on a different level, regardless of whether they
are actually valid.
For example, different principles of distributive justice exist. There
are material principles of justice, such as "To each according to his
needs" or "To each according to his merits" or "Equal shares for all."
Principles of equal rights, such as the precepts of equal respect for
all, of equal treatment, or of equity in the application of the law,
address a different kind of problem. What is at issue here is not the
distribution of goods or opportunities but the protection of freedom
and inviolability. Now all of these principles of justice can be justified
from the perspective of universalizability and can claim prima facie
validity. But only in their application to particular concrete cases will
it transpire which of the competing principles is the most appropriate
in the given context.
This is the task of discourses of application. Within the family, for
instance, conflicts of distribution will tend to be decided on the prin
ciple of need rather than on the principle of merit, whereas the
situation may well be the reverse in cases of conflicts of distribution
at the level of society as a whole. It depends on which principle best
fits a given situation in the light of the most exhaustive possible
description of its relevant features. But I find the idea of a universal
correlation of principles of justice with spheres of action highly prob
lematic. The kinds of considerations Walzer entertains could be ac
commodated in discourses of application, but then they would have
to prove themselves in each particular instance in its own right.
JH:
THN: The limitation of moral theory to questions of justice leads

you to make a sharp distinction between "moral questions" (which
can in principle be decided rationally in terms of a principle of
universalization) and "evaluative questions" (which are questions of
1 53
the good life and as such admit of rational treatment only within the
horizon of a concrete historical form of life or an individual life
history). But do you in fact exclude the possibility of congruence
between justice and the good life? John Rawls, who also maintains
the priority of the right over the good, presupposes the possibility of
such congruence at least under the conditions of a "well-ordered
society." He takes the view that a moral theory should specify how
the right and the good are related to each other.9
JH: Yes, in a society that possessed all of the resources of a modern

society and was in addition well ordered-that is, just and emanci
pated-socialized individuals would enjoy not only autonomy and a
high degree of participation but also relatively broad scope for self
realization, that is, for the conscious projection and pursuit of indi
vidual life plans.
THN: While you separate justice from the good life, you also in
dude elements of care and responsibility in your concept of justice.
But how can Carol Gilligan's notions of "care" and "responsibility"
be integrated into your discourse-theoretical conception of justice? 10
The ethics of care is oriented to the concrete other, not the gener
alized other; it calls for a contextual, rather than an abstract, formal,
mode of thought; it focuses on social relationships, not fixed roles,
and traces moral questions back to conflicting interests instead of
competing rights. How can these differences be subsumed under a
formal concept of justice?
JH: Let me address the first two points together and then turn to
the second two.
The impression that deontological ethical theories such as Kantian
ethics compel us to overlook the concrete other and his or her par
ticular situation is created only by an avoidable one-sided preoccu
pation with questions of justification. Kant conceived of morality in
general from the Rousseauean perspective of a legislator who reflects
on how some matter can be regulated in the common interest of all
citizens, hence from the perspective of universalizability. In this way,
the problem of application is lost sight of. The unique disposition of
a particular case that calls for regulation, and the concrete charac
teristics of the people involved, come into view only after problems
1 54
of justification have been resolved. It is only when it has to be estab

lished which of the prima facie valid norms is the most appropriate to
the given situation and the associated conflict that a maximally com
plete description of all the relevant features of the particular context
must be given. Thus Klaus Gunther entitled his excellent study of
discourses of application Der Sinn fur Angemessenheit (The sense of
appropriateness) (Frankfurt, 1 988). Practical reason is not fully re
alized in discourses of justification. Whereas in justifying norms prac
tical reason finds expression in the principle of universalization, in
the application of norms, it takes the form of a principle of appro
priateness. Once we grasp the complementarity of justification
and application, it becomes clear how discourse ethics can address
the misgivings you share with Carol Gilligan and also with Seyla
Benhabib. 1 1
Let us now turn to the other objection: that deontological ethics
concentrates on rights to the exclusion of needs and in addition
neglects the significance of membership relations in favor of institu
tionally defined roles. Viewed historically, in the light of the individ
ualism of the Kantian tradition, this objection is justified, but it does
not apply to discourse ethics. The latter adopts the intersubjective
approach of pragmatism and conceives of practical discourse as a
public practice of shared, reciprocal perspective taking: each individual
finds himself compelled to adopt the perspective of everyone else in
order to test whether a proposed regulation is also acceptable from
the perspective of every other person's understanding of himself and
the world. Justice and solidarity are two sides of the same coin because
practical discourse is, on the one hand, a procedure that affords
everyone the opportunity to influence the outcome with his "yes" or
"no" responses and thereby takes account of an individualistic un
derstanding of equality; on the other hand, practical discourse leaves
intact the social bond that induces participants in argumentation to
become aware of their membership in an unlimited communication
community. It is only when the continued existence of this commu
nication community, which demands of all its members an act of
selfless empathy through ideal role taking, is assured that the net
works of reciprocal recognition, without which the identity of each
individual would necessarily disintegrate, can reproduce themselves.
155
THN: How should we understand the division between morality

and law? According to Durkheim and Weber, they are two distinct
spheres that result from the dissolution of traditional ethics, but
somehow they also remain bound to one a nother through a common
center. Must morality and law be conceived in the modern period
merely as different ways of institutionalizing procedures that serve
the same purpose?
Positive law and postconventional morality complement each

other and together overlay traditional ethical life. From a normative
point of view, it is easy to see that universalistically grounded moral
norms are in need of supplementation. A precondition of the general
acceptance of any norm that passes the generalization test is that it
should also be actually adhered to by everyone. It is precisely this
condition that a reflexive morality which breaks with the certainties
of a concrete ethical life cannot itself guarantee. Thus, the very
premises of an ambitious postconventional mode of justification gen
erate a problem of what can be reasonably expected (Problem der Zumutbar
keit) : adherence to a valid norm is to be expected only of an individual
who can be sure that all others will also follow the norm. Kant already
justified the transition from morality to state-sanctioned law in this
way. But Kant also recognized the problem generated by the recourse
to the medium of state power. Political power is not a neutral me
dium ; its use and organization must themselves be subjected to moral
constraints. The idea of a state governed by the rule of law is the
response to this demand.
Kant and early liberalism share a conception of the rule of law to
the effect that the legal order is itself exclusively moral in nature or
at least is one form that the implementation of morality assumes. But
this assimilation of law to morality is misleading. The political aspect
of law brings quite different factors into play. Not all matters that
call for and admit of legal regulation are of a moral nature. Even if
legislation were to approximate sufficiently to the ideal conditionsof
discursive belief formation and will formation, the decisions of the
legislator could not be based solely on moral grounds-and most
certainly not those of a legislator in a welfare state. Pragmatic reasons
always play a considerable role in reaching a (more or less) fair
balance of nongeneralizable interests, as do ethical grounds in the
]H:
1 56
accepted self-understanding and preferred form of life of a group

in which different traditions, grounding different identities, come
together and have to be harmonized with one another. For this
reason, the claim to legitimacy of positive law could not be assimilated
to the claim to moral validity, even if it were based on rational will
formation. Together with the introduction of pragmatic and ethical
reasons, other factors come into play in the question of the legitimacy
of law: legitimacy rests on a broader range of aspects of validity than
the normative validity of moral norms of action.
Moreover, legal validity has two distinct components : the empirical
component of the enforcement of law and the rational component
of the claim to legitimacy. Legal validity requires that both compo
nents be justifiable simultaneously to the addressees: the cognitive
expectation that, if necessary, force may be brought to bear to ensure
that everyone adheres to the individual legal norms (for which reason
legality of behavior, that is, behavior that merely conforms to the law,
is sufficient for the law). At the same time, it must warrant the nor
mative expectation that the legal system merits recognition on good
grounds (for which reason the law must at all times make possible more
than mere legality, namely, obedience based on insight into the legit
i m acy of the legal order) .
II
THN: Let us turn to the topic of your linguistic-pragmatic justifi

cation of discourse ethics, in particular your development of Toul
min's analyses in The Uses of Argument, Wittgenstein's "language
games," and Chomsky's "universal grammar" into a formal pragmat
ics. Already at this methodological level, I would like to know how
you would respond to a new variation on the old objection to the
alleged Eurocentrism of your defense of the Enlightenment and your
concept of evolution.
One might wonder whether the whole notion of a formal prag
matics is merely the result of a dubious generalization from Indo
European languages. Whorf compared "Standard Average Euro
pean" with non-European languages and found that such central
features as the function of verbs, the tense structure, and the gram
matical relation between subject and predicate in the latter languages
1 57
were fundamen tally different from the characteristics you take to be

universal. 12 I ca nnot go into the details here, but there seem to be
abundant lingu istic data that raise serious doubts about the whole
program of a u niversal pragmatics or that even refute it. Perhaps
you could argue that the non-European languages are less developed,
but then you would have to show how development of grammatical
deep structures is even possible.
]H: The Sapir-Whorf hypothesis was discussed at length in the

1 950s, by and large with negative results. Clearly the surface struc
tures of individual languages may differ radically, without the agree
ment between the underlying semantic structure of simple assertoric
statements or the pragmatic structure of the speech situation (such
as personal pronouns, spatial and temporal indices) being thereby
diminished. What Whorf had in mind were, rather, the differences
between linguistic worldviews, which had already interested Hum
boldt though it did not lead him to draw conclusions in favor of
linguistic relativism. Avoiding relativism by no means forces us to fall
back on the notion of an evolution of language systems. In the case
of natural languages, an evolutionary hypothesis is manifestly inap
propriate, since the grammatical complexity of languages scarcely
changes over time.
Whorf's intuition has recently found resonance again on a differ
ent level, in the rationality debate initiated by anthropologists, which
has developed in a number of different directions. I think the crucial
point in this debate is whether we must take account of an asymmetry
that arises between the interpretive capacities of different cultures in
virtue of the fact that some have introduced "second-order concepts"
whereas others have not. These second-order concepts fulfill neces
sary cognitive conditions for a culture's becoming self-reflective, that
is, for its members' adopting a hypothetical stance toward their own
traditions and on this basis grasping their own cultural relativity. This
kind of decentered understanding of the world is characteristic of
modern societies. What the argument is about, therefore, is whether
such cognitive structures represent a threshold that demands similar
processes of learning and adaptation of any culture that crosses it.
According to the contextualists, the transition to postmetaphysical
concepts of nature and posttraditional conceptions of law and mo-
158
rality is characteristic of just one tradition among others and by no

means signifies that tradition as such becomes reflexive. I don't see
how this thesis could be seriously defended. I think that Max Weber
was fundamentally right, especially in the careful universalistic inter
pretation that Schluchter has given his thesis of the universal cultural
significance of Occidental rationalism. 1s
THN: Your moral theory takes the form of an examination of moral

argumentation. The only moral principle you lay down is a principle
of universalization, which is supposed to play a role in moral argu
mentation similar to that played by the principle of induction in
deciding empirical-theoretical questions. It specifies that a norm is
valid only if it could be accepted by all potentially affected in a real
process of argumentation, meaning that the norm must be capable
of satisfying the interests of each participant. But why should the
participants agree to the consequences of the general observance of
the norm? Often they merely arrive at a consensual determination
of their disagreement. That would be similar to the procedure of
political will formation resulting in a consensus that certain issues
and controversies must be left (o other forms of discussion.
JH: Argumentation is not a decision procedure resulting in collective
decisions but a problem-solving procedure that generates convictions.
Of course, argumentative disputes concerning the truth claims of
assertoric statements or the rightness claim of normative statements
may remain undecided and no agreement be reached; then the ques
tions are left open for the time being, though on the assumption that
only one side can be right. In practical discourses, however, it may
transpire that the conflict at issue is not a moral one at all. It may be
an ethical-existential question affecting the self-understanding of a
given individual or group, in which case any answer, however ra
tional, will be valid only relative to the goal of my or our good, or
not unsuccessful, life and cannot claim to be universally binding. Or
perhaps it is a pragmatic question of balancing opposed but nonge
neralizable interests, in which case the participants can at best reach
a fair or good compromise. Thus, the breakdown of attempts at
argumentation in the sphere of practice may also reflect the realiza
tion that self-interpretive discourses or negotiations are called for
rather than moral discourses.
1 59
Parliamentary processes of will formation also have a rational core,

for political questions admit of discursive treatment either from em
pirical and pragmatic or moral and ethical points of view. There is a
definite time frame within which decisions must be reached in legally
institutionalized processes of belief formation such as these. The
standard procedures combine elements of belief formation oriented
to truth with majoritarian will formation. From the perspective of a
theory of discourse that seeks to sound out the normative potential
of such procedures, majority rule must retain an internal relation to
the cooperative search for truth. Ideally, then, a majority decision
must be reached under discursive conditions that lend their results
the presumption of rationality: the content of a decision reached in
accordance with due procedure must be such as can count as the
rationally motivated but fallible result of a discussion provisionally
brought to a close under the pressure of time. Hence, one should
not confuse discourse as a procedure for making moral or ethical
judgments with the legally institutionalized procedures of political will
formation (however much the latter are mediated by discourse) .
THN: You defend cognitivism i n moral theory b y maintaining an

analogy between truth claims and normative validity claims. However,
this analogy can be upheld only by identifying the norms that un
derlie the principle of universalization with normative validity as
such. How can, and why must, moral theory neglect norms that have
de facto social currency but are not valid in the strict sense? And is
this exclusion possible without severing the dialectical connection
between abstract morality and social ethics?
JH: Viewed from the performative perspective of their addressees,
norms claim to be valid in a manner analogous to truth. The term
"analogous" indicates, of course, that the mode of validity of norms
should not be assimilated to the truth of propositions. The differ
ences can be seen even prior to divergences at the level of rules of
argumentation and in the specific nature of the arguments admissible
in either case; they already begin with the fact that normative validity
claims are embodied in norms, that is, in structures that are on a
higher level than individual moral actions and regulative speech acts,
whereas truth values are ascribed only to individual assertoric state
ments and not to theories. In the latter case, the higher-level struc-
1 60
tures-the theories-owe their validity to the set of true propositions

derivable from them, whereas particular commands and prohibitions
obtain their validity from the norms underlying them.
One interesting difference, is that taking propositions to be true
has no impact on what is essential to the truth of sentences, the
existence of states of affairs. By contrast, the fact that norms are
taken to be right has immediate consequences for the regulation of
action essential to norms. As soon as a norm governing action gains
sufficient recognition among, and is adhered to by, its addressees, a
corresponding practice is generated, regardless of whether the norm
can be justified and deserves recognition or whether it is merely de
facto recognized for the wrong reasons, for example, or is adhered
to out of sheer habit. For this reason, it is important to distinguish
between the validity of a norm and its social currency, the fact that
it is generally held to be valid. I can agree with you thus far.
However, I am not sure that I understand the meaning of your
question. The moral theorist adopts a normative point of view ; he
shares the attitude of an addressee of the norm who participates in
discourses of justification and application. From this perspective, we
must begin by abstracting from existing traditions, customary prac
tices, and present motives, in short, from the established ethical life
of a society. On the other hand, it is primarily this ethical life that
must interest the sociologist. But the latter adopts the objectifying
attitude of a participant-observer. We cannot simultaneously adopt
the second-person attitude of an addressee of a norm and the third
person attitude of a sociological observer. What you have in mind is
presumably the complicated case where knowledge acquired in the
one attitude is interpreted in the other. That is the situation of the
sociologist who assesses a descriptively formulated belief in the legit
imacy of an observed, socially valid order in terms of the reasons that
could be adduced in support of its legitimacy from the perspective
of potential addressees. The participant in argumentation (or the
moral theorist, as his philosophical alter ego) switches his role ac
cordingly the moment he views the empirical aspects of matters in
need of regulation through the spectacles of a legislator and takes
account in his reflections of the reasonableness or acceptability of
regulations. These different ways of viewing things and their differ
ent objects must be clearly distinguished from one another, but such
161
differences yield no arguments in support of a fallacious sociologi

zation of moral theory.
You spoke of empirically valid norms that do not have (nonnative)
validity. Strictly speaking, this formulation applies only to conventions
such as table manners, that is, to customary rules that are followed
for the most part without needing, or being amenable to, rational
justification.
THN: The preconditions of an ethics of discourse are fulfilled at

the ontogenetic level only at Kohlberg's final, postconventional stages
of moral-psychological development. But only a minority of the adult
population ever reaches these stages (if we can rely on the relevant
longitudinal studies) . 14 Then we are faced with the paradox of a
society with postconventional social institutions while a majority of
the population languish at the preconventional or conventional stages
of moral consciousness. How is this possible? And how can it be
reconciled with your claim that normative structures function as the
pacemaker of social evolution? 15 If the answer is that postconven
tional morality is embedded in legal structures, then you must give a
plausible explanation of how such a situation can be stabilized.
JH: Social innovations are often initiated by marginal minorities,
even if they are later generalized to the whole of society at the
institutional level. This may explain why positive law in modern
societies must be understood as the embodiment of postconventional
structures of consciousness, although many members are found to
have reached only a conventional stage of moral consciousness. Nor
does a conventional understanding of a postconventional legal system
necessarily lead to instabilities; for example, it sometimes prevents
radical interpretations that lead to civil disobedience.
What's more, findings concerning the moral consciousness of the
general population are problematic; it is a matter of considerable
controversy whether Kohlberg's methods of collecting data do not in
fact lead to artificial results in the definition of stages. For instance,
children master the moral judgments of a given stage long before
they have the verbal resources to articulate this knowledge in re
sponse to the familiar dilemmas.
THN: The question about the analogy between normative validity
and truth is merely a variation on a question that was originally put
162
to you in an interview with the New Left Review and that has remained
unanswered: "How do you conceive the relation between philosoph
ical and scientific truth-claims? Are philosophical truth-claims cog
nitive claims, and would a rational consensus ultimately guarantee
the truth of a consensus theory of truth itself ?"16
JH: I think that philosophy today plays two roles simultaneously:

the role of an interpreter who mediates between the lifeworld and
expert cultures and a more specialized role within the scientific system
in which it cooperates with various reconstructive sciences. In so
doing, it generates statements that claim to be true in the same way
as other scientific statements. The discourse theory of truth also
contains assertions that must be defended against rival theories of
truth within the relevant universe of discourse.
But your question expresses another doubt. You seem to suggest
that the self-referentiality of philosophical statements-in this case,
the assertions of a theory of truth-necessarily land the discourse
theory of truth in a vicious regress. I disagree. Of course, the recon
struction of our intuitive understanding of truth in terms of a theory
of discourse that I propose might well turn out to be wrong, or at
least inadequate. But the practice, whether in everyday life or in
science, which depends on the correct use of this intuitive knowledge
remains unaffected by these attempts at philosophical reconstruction
and by their possible revisions. It is impossible to refute this practical
knowledge itself, only an incorrect description of it.
THN: The concept of a performative contradiction, which you have
appropriated from Karl-Otto Apel (though stripped of its transcen
dental connotations), plays a decisive role in your justification of
morality. I' The argument underlying your employment of perfor
mative contradictions seems convincing in the narrow sense that no
body can systematically dispute the necessary presuppositions of
communicative action while engaging in it without placing his own
rationality or responsibility in question. But how can one establish in
this way that one type of ethical theory is better than another?
JH: The demonstration of performative contradictions in particular
cases serves to refute skeptical counterarguments. This casuistic prac
tice can also be developed into a method that serves, as in Strawson,
to identify unavoidable presuppositions of a practice for which there
163
is no functional equivalent in our form of life. Apel and I employ

this method to discover universal pragmatic presuppositions of ar
gumentation and to analyze their normative content. In this way I
attempt to justify a principle of universalization as a moral principle.
The initial intention is simply to demonstrate that moral-practical
questions can indeed be decided on the basis of reasons. These gen
eral presuppositions of argumentation have the same status in dis
course ethics as the construction of the original position in Rawls's
theory of justice. It must then be left to discussion between such
theoretical positions to show which version of Kantian ethics is the
best. This professional dispute will involve many different aspects
and certainly cannot be resolved by direct appeal to performative
self-contradictions.
THN: What is the status of the "ideal speech situation"? Is it partially

counterfactual, or is it part of a society fictionally represented as a lifeworld,
or is it a hypostatization ? Or are all three theses somehow interrelated?
The first you adopt explicitly in Moral Consciousness and Communicative
Action. 18 The second follows from an understanding of discourse
ethics as a further development of the third of the fictions posited in
The Theory of Communicative Action as necessary to conceptualize society
as lifeworld-I mean the assumption of complete transparency of
communication. 19 The third is attributed to you by Wolfgang
Schluchter, who claims that the logic of your argumentation compels
you to transform the ideal speech situation from a necessary presup
position of communication into an ideal of reality, and hence to
hypostatize it as such.2
JH: For present purposes, we can ignore the second position be
cause the passage cited deals with a concept of the lifeworld that I
reject as idealistic. The first position implies only that the unlimited
communication community (unlimited, that is, in social space and
historical time) is an idea that we can approximate in real contexts
of argumentation. At any given moment we orient ourselves by this
idea when we endeavor to ensure that ( 1 ) all voices in any way
relevant get a hearing, (2) the best arguments available to us given
our present state of knowledge are brought to bear, and (3) only the
unforced force of the better argument determines the "yes" and "no"
responses of the participants. Unfortunately, I once dubbed the
1 64
conditions under which these idealizing presuppositions would be

fulfilled the "ideal speech situation," a term whose concretistic con
notations are misleading. It suggests the kind of hypostatization that
Schluchter, albeit with reservations, mistakenly attributes to me.
Schluchter bases his argument, among other things, on the formula
"Vorschein einer Lebensform" ("prefiguration of a form of life") , from
which I distanced myself a decade ago. 2 1 However, at no time did I
"hypostatize the unlimited communication community from a nec
essary presupposition into an ideal of reality," as Schluchter, citing
Wellmer, would have it.
As a matter of fact, I hesitate to call the communication community
a regulative idea in the Kantian sense, because the notion of an
"unavoidable idealizing presupposition of a pragmatic kind" cannot
be subsumed under the classical opposition between the "regulative"
and the "constitutive."
From the participant perspective, the idea of the truth of fallible
statements we make here and now is regulative. On the one hand,
all reasons available here and now warrant us in claiming truth for
'p' ; on the other, we have no assurance that 'p' will be able to with
stand all future objections-we have no way of knowing whether it
will be among the valid statements that would repeatedly command
assent in the unlimited communication community ad infinitum. But
the general pragmatic presuppositions of argumentation as such are
by no means merely regulative, for these conditions must be satisfied
to a sufficient degree here and now if we want to engage in argument
at all, where that degree of satisfaction counts as "sufficient" which
qualifies our actual argumentation as a spatiotemporally localizable
component of the universal discourse of the unlimited communica
tion community. But this does not mean that the latter is thereby
transformed into an idea that constitutes a reality. The conceptual
schema of world constitution is inapplicable here. The point is, rather,
that if we want to enter into argumentation, we must make these
presuppositions of argumentation as a matter of fact, despite the fact
that they have an ideal content to which we can only approximate in
reality.
With the validity claims raised in communicative action, an ideal
tension is imported into social reality itself, which comes to conscious
awareness in participating subjects as a force that explodes the limits
1 65
of the given context and transcends all merely provincial standards.

To put it paradoxically, the regulative idea of the validity of utter
ances is constitutive for the social facts produced through commu
nicative action. To this extent I go beyond Kant's figures of thought,
as Schluchter remarks, though without embracing the totalizing view
point of Hegel. Already in Peirce the idea of the unlimited commu
nication community serves to replace the moment of infinitude or
timelessness of the unconditionality of truth with the idea of a pro
cedure of interpretation and communication that transcends the lim
its of social space and historical time from within the world. The learning
processes of the unlimited communication community should form
an arch in time bridging all temporal distances, and in the world they
should realize the conditions whose fulfillment is a necessary presup
position of the unconditionality of context-transcending validity
claims.
While it is introduced in the context of a theory of truth, this
notion also structures a concept of society grounded in communica
tive action, since communicative interactions are regulated by inter
subjectively recognized validity claims. These unconditional validity
claims introduce into the lifeworld a moment of transcendence that
permeates its symbolic structures. For this reason, even the counter
factual assumptions of communicatively acting subjects can expect to
meet with support from the side of social reality; every factually
raised claim to validity that transcends the limits of a given lifeworld
generates a new fact with the "yes" and "no" responses of its addres
sees. Mediated by this cognitive-linguistic infrastructure of society,
the results of the interplay between inner-worldly learning processes
and world-disclosing innovations become sedimented [in social real
ity] . This is the Hegelian elemen t detected by Schluchter, which he
can see from his Kantian perspective (in my opinion erroneously)
only as the illicit objectificatio n of a regulative idea.
III
THN: You understand discourses as a reflective form of commu

nicative action, which for its part is situated in the lifeworld. By
contrast, all normative elements disappear at the level of the social
subsystems steered by the media of money and power. In another
1 66
place you have explained how your expression norm{ree sociality has
led to misunderstandings. Even after the modern uncoupling of
system and lifeworld, system integration remains indirectly connected
with the lifeworld through the legal institutionalization of the steering
media. You claim only that the integration of the subsystems is not
"in the final instance" dependent on the socially integrative accom
plishments of communicative action. You write: "It is not the illocu
tionary binding effects [of speech acts] but steering media that hold
the economic and administrative action systems together."22 This an
swer makes your position more flexible, but you still maintain that
the media of money and power demand that the agent adopt a
strategic attitude. I have my doubts about this.
Your picture of the economic actor shares important traits with
the model of neoclassical economic theory. Why do you neglect the
arguments developed by institutionalist economic theory to demon
strate that the model of purely strategic and utilitarian action died
out at the latest with Adam Smith's "invisible hand"? Amitai Etzioni's
most recent book offers numerous arguments and copious evidence
for the claim that "the most important bases of choices [also in the
market] are affective and normative. That is, people often make non
or subrational choices, first because they build on their normative
affective foundations, and only secondly because they have weak and
limited intellectual capabilities."23
]H: I think that this is a misunderstanding. I use "system" and
"lifeworld" as concepts for social orders that differ in their mecha
nisms of societal integration, that is, in the intermeshing of interac
tions. In "socially integrated" spheres of action, this interlinking or
sequential ordering is achieved either through the intentions of the
agents themselves or through their intuitive background understand
ing of the lifeworld ; in "systemically integrated" spheres of action,
order is generated objectively, "over the heads of the participants" as
it were, through the functional interlocking and reciprocal stabiliza
tion of consequences of action of which the agents need not be aware.
The concept of the lifeworld must be introduced in the context of a
theory of action. But only if the concept of system were introduced
in a similar manner could one establish the dear, reversible relation
167
between systemically integrated spheres of action and forms of pur

posive-rational action you attribute to me.
In fact, I introduce the concept of systemically differentiated, self
steering, recursively closed spheres of action in terms of mechanisms
of functional integration, and specifically the steering media of
money and power. The latter undoubtedly have correlates at the level
of social action in the form of interactions steered by media. But this
fact in no way compromises the rationality of the choices of partici
pants in interaction. In each case, the medium specifies the standards
by which conflicts are ultimately resolved. To this extent the structural
limitations to which media-guided interactions are subject occasion
more or less rational plans of action, but they neither necessitate that
action be rationally orientated nor can they obligate actors to behave
rationally. Hence, the empirical evidence you allude to is compatible
with a media-theoretical description of economic and administrative
behavior.
THN: You have taken over the concepts of the state as a system
and power as a medium from Talcott Parsons, both of which entail
the separation of politics and administration. Thomas McCarthy has
criticized this separation as contrary to both empirical investigations
and your own concept of democracy: "If self-determination, political
equality, and the participation of citizens in decision-making pro
cesses are the hallmarks of true democracy, then a democratic gov
ernment could not be a political system in Habermas's sense."24 You
have emphasized that the democratic state cannot be reduced to
positive law. I n the case of civil disobedience, legality must even be
abrogated in favor of those who must ultimately safeguard the legit
imacy of government-the citizens.25 But how can civil disobedience
be interpreted in this way without abandoning the separation of
politics and administration underlying the concepts of the state as
system and of power as a medium?
]H: I do not regard processes of legitimation per se as part of the
administrative system regulated by power; they unfold in the political
public sphere. Here, two opposed tendencies meet and intersect:
communicatively generated power proceeding from democratic pro
cesses of opinion formation and will formation (Hannah Arendt)
runs up against the production of legitimation by (and for) the ad-
1 68
ministrative system. How these two processes-more or less sponta

neous opinion formation and will formation through public channels
of communication, on the one hand, and organized production of
mass loyalty, on the other-affect each other, and which gains dom
inance over the other, are empirical matters. A similar form of inter
action occurs in institutionalized forms of political will formation
for instance, in parliamentary bodies. Only with the complete inte
gration of political parties into the state apparatus would these insti
tutionalized processes of will formation be fully absorbed into an
administrative system that was self-regulating (albeit within the limits
of existing laws) .
But to return to your question : the boundaries between commu
nicatively regulated political opinion formation and will formation,
on the one hand, and administration steered by power, on the other,
could be blurred under modern conditions of life only at the cost of
reversing differentiation within the system of public administration.
The production of power by communicative means obeys a different
logic from the assertion and exercise of administrative power.
By contrast, civil disobedience-in the sense of nonviolent
transgression of rules intended as a symbolic appeal to a differently
minded majority-is merely an extreme case that exhibits the inter
play between noninstitutionalized public communication and con
stitutional processes of democratic will formation. The one can
influence the other because institutionalized will formation embodies
the idea just mentioned : in parliaments, opinion formation oriented
to truth should act as a kind of filter for majority decisions, thereby
lending them the presumption of reasonableness.
THN: How can power be compared with money, even if the former
is understood as a steering medium? In The Theory of Communicative
Action you enumerate (again following Parsons) differences in mea
surement, circulation, and storage that exist between the two media
but then maintain that both would equally make action coordination
independent of the resources of the lifeworld. Yet the mode of in
stitutionalization of power in the lifeworld exhibits significant dissim
ilarities from that of money. Thus, obedience is the appropriate
attitude toward administrative power, whereas the market calls for
an orientation to enlightened self-interest. The respective attitudes
1 69
Morality , Society, and Ethics
must, for example, be correlated with different stages of moral con

sciousness. How can such differences be explained when both steer
ing media occupy the same or a paralle l place in the architectonic of
your theory?
JH: The contradictio n you have just set forth can be resolved as
follows. The two media, money and power, function in a symmetrical
fashion insofar as they serve to hold together differentiated, self
regulating action systems independent of any intentional effort, and
thus of the coordinating activity, of actors. They behave asymmetri
cally in respect of their mode of dependency on the lifeworld, though
both are legally institutionalized and hence anchored in the lifeworld.
Whereas the capitalist economy also subsumes the production pro
cess, including labor power (as the substrate that generates exchange
values), the democratic state apparatus remains dependent on the
repeated provision of legitimation over which it can never gain com
plete control through the exercise of administrative power. Here com
municatively produced power constitutes a substrate that can never
be cut off from the roots of discursive-and to that extent, admin
istratively nonregulable-processes of public opinion formation and
will formation to the same extent that production steered by market
forces can be severed from lifeworld contexts of active labor.
On the other hand, this asymmetry should not mislead us into
thinking that the administrative system can be subsumed under life
world categories. It is indeed a necessary condition of the possibility
of making demands on the administrative system in the name of
imperatives originating in the lifeworld; and the latter, in contrast to
consumer decisions, do not have to be formulated from the outset in
the language of the steering medium in question, that is, in prices
and institutional directions, in order to become "comprehe nsible" to
the corresponding system. This can be seen from the different atti
tudes that politics and admini stration take toward law-respectively,
a normative and a more instrum ental one. 26 The a dministrative sys
tem treats law primarily in an instrumental fashion; what counts from
the perspective of administrative power is not the practical reason
employed in the justification or application of norms but rather ef
ficiency in the implementation of a program in part laid down and
in part developed by the ad ministrative system itself. The normative
1 70
reasons, which in the language of politics serve to justify positive

norms, function in the language of administration as constraints and
post hoc rationalizations for decisions occasioned by other consider
ations. At the same time, normative reasons remain the sole currency
in which communicative power becomes operative. It can affect the
administrative system by cultivating the pool of reasons on which
administrative decisions, which are subject to the rule of law, must
draw. Not everything 11ULJ be done that could be done by the admin
istrative system if the preceding political communication and will
formation had discursively devalued the reasons required within a
given legal framework.
THN: The last three questions can be drawn together in an argu

ment to the effect that your analysis of the pathologies of modernity
needs to be supplemented in ways that pull in opposing directions.
You maintain that the system colonizes the lifeworld. But there also
exist countervailing tendencies. Normative expectations and demo
cratic processes of will formation can so influence the two subsystems
that they can no longer be integrated exclusively through their sys
temic mechanisms. Thus, your analysis of processes of legal regula
tion, for example, could be supplemented by an analysis of social
movements that have as their goal the democratization of the econ
omy, user participation, and so forth. Why do you ignore this possi
bility? Would the results of such investigations undermine the
architectonic of your theory?
JH: When I was writing The Theory of Communicative Action my main
concern was to develop a theoretical apparatus with which the phe
nomena of "reification" (Lukacs) could be addressed. But this ap
proach to systemically induced disruptions in communicatively
rationalized lifeworlds was one-sided ; it failed to exhaust the analytic
resources provided by the theory of communicative action. The ques
tion of which side imposes limitations on whose imperatives, and to
what extent, must be treated as an empirical one that cannot be
decided beforehand on the analytical level in favor of the systems. I
have already emphasized in the preface to the third edition of the
book, in response to similar objections by Johannes Berger, that the
colonization of the lifeworld and democratic constraints on the dy
namics of systems that remain unresponsive to the "externalities"
171
they produce represent two equally legitimate analytic perspectives.

A one-sided diagnosis of the contemporary scene is by no means
implicit in the architectonic of the theory.
THN: In a series of articles that have appeared since the original

publication of Moral Consciowness and Communicative Action, you have
discussed the Hegelian concept of "ethical life" (Sittlichkeit) or, alter
natively, of a "pragmatic ethics," in an effort to mediate between
discourse ethics and social reality. You gauge the rationality of a form
of life by the degree to which it enables and encourages its members
to develop a moral consciousness guided by principles and to convert
it into practice.27 But can rationality be identified with morality?
Ethical life or practical ethics would seem to be reducible to the
existing norms of a given society. Then the question becomes whether
the socially current norms are also valid or whether they at least
promote such valid norms. You seem to want to preserve normative
validity, or the pure "ought," exclusively within the framework of
Kant's individual morality, while abandoning Hegel's project of un
iting "is" and "ought" as untenable.
]H: The concept of communicative rationality comprises a number
of different aspects of validity, not just the moral aspect of the nor
mative validity of commands or actions. Hence the rationality of a
form of life is not measured solely by the normative contexts or the
potential sources of motivation that facilitate the translation of po
stconventional moral judgments into practice. Nevertheless, what
seems to me to be essential to the degree of liberality of a society is
the extent to which its patterns of socialization and its institutions, its
political culture, and in general its identity-guaranteeing traditions
and everyday practices, express a noncoercive, nonauthoritarian
form of ethical life in which an autonomous morality can be embod
ied and take on concrete shape. Intuitively we recognize fairly
quickly-like ethnologists who have been integrated into a foreign
society-how emancipated, responsive, and egalitarian our surround
ings really are, how minorities, marginal social groups, the handi
capped, children, and the elderly are treated, the social significance
of illness, loneliness, and death, how much tolerance there is toward
eccentricity and deviant behavior, the innovative and the dangerous,
and so on.
172
However, your question seems to confuse two issues. When I draw

the distinction between morality and ethical life in the normative
attitude of a moral theorist (or indeed of a participant in argumen
tation), I have something different in mind than when I compare, in
the role of sociologist, the moral beliefs of individuals I observe, or
the moral content of their legal principles, with the established prac
tices and the concrete manifestations of ethical life in this society. But
even from the sociological perspective, it is not as if the complete
normative substance resides in the heads of those making moral
judgments (or in the wording of their juridical texts) and thus would
be exhausted by universalistic morality. The ethical practice actually
obtaining in society also belongs to this normative substance, however
much it may deviate from the socially accepted morality.
THN: In the Howison Lecture of 1 988 you make a further attempt
to mediate between discourse ethics and society. There you assert
that "the application calls for argumentative clarification in its own
right. . . . the impartiality of judgment cannot again be secured
through a principle of universalization."28 But how can a new form
of relativism be avoided when a so-called principle of appropriateness
serves as a substitute for the context-sensitive application of norms
in all situations?
]H: The logic of discourses of application can be investigated from
the normative point of view of the philosopher or the legal theorist.
Ronald Dworkin provides examples of this and develops a corre
sponding theory, and Klaus Gunther formulates this approach con
vincingly within a discourse-theoretical framework. 29 The latter
shows that the principle of appropriateness, no less than the principle
of universalization, brings impartiality to bear in judging practical
questions and thereby makes rationally motivated agreement possi
ble. In discourses of application, we also rely on reasons that are in
principle valid for everyone, not just for you and me. One must be
wary of making the fallacious inference that an analytic procedure
that calls for sensitivity to context must itself be context dependent
and lead to context-dependent results.
THN: In the Howison Lecture you emphasize that ethical ques
tions-in contrast to moral questions-do not require a complete
break with the egocentric perspective: . . . "in each instance they take
1 73
their orientatio n from the telos of one's own life."30 You also intro
duce maxims of action as a kind of bridge between morality and
ethics "because they can be judged alternately from ethical and moral
points of view."3 1 How are maxims related to normative validity
claims? Don't maxims somehow claim empirical and normative valid
ity simultaneously?
]H: Yes, ethical questions, questions of self-interpretation, have as

their goal your or my good life, or better, not unsuccessful life. We
survey our life history or our traditions and ask ourselves, with the
ambiguity that typifies strong preferences, who we are and who we
would like to be. The answers must therefore refer to a particular
perspective on life assumed to be valid for particular individuals or
groups. Answers such as this cannot claim to privilege one form of
life as exemplary and binding on all, as Aristotle privileged the polis.
But relative to a given context, ethical questions can be answered
rationally, that is, in such a way that they make sense to everybody
and not just to those immediately affected, from whose perspective
the question was posed.
You touched on another point: What are maxims? Following Kant,
I understand by "maxims" rules of action or customs that constitute
practices or even a whole way of life, in that they relieve agents of
the burden of continually making decisions in daily life. Kant had in
mind primarily the maxims of a professionally stratified, early capi
talist society. In the lecture I said that maxims can be judged from
either an ethical or a moral point of view. What may be good for me,
given my view of myself and of how I would like to be seen, may not
be equally good for all. But the fact that maxims can be judged from
a twofold perspective does not confer on them a twofold nature.
Again, the normative discussion we have just been conducting must
be distinguished from a sociological discussion. From the viewpoint
of a sociological observer, maxims may recommend themselves as a
class of phenomena with reference to which the concrete ethical life
of a group can be productively studied. Maxims enjoy social currency;
as such, they are normatively binding for the agents themselves, as
least insofar as they are not mere conventions. Hence, we can change
perspective and move from observation to judgment by considering
whether the reasons for which they have chosen their maxims are also
good reasons for us.
1 74
THN: You defend ethical cognitivism against skeptics but leave the
consideration of moral feelings to one side. But these come into play
once more, at the very latest, in the application of norms. What is
the status of moral feelings? Don't feelings and "habits of the heart"
have intrinsic worth? Or do they merely have a catalytic function in
the development of moral consciousness and become superfluous
once a certain level of moral competence has been reached?
Moral feelings are both an important topic and a broad field

in their own right, so I will limit myself to a few remarks.
First, moral feelings play an important role in the constitution of
moral phenomena. We would not experience certain conflicts of ac
tion ;;ts morally relevant at all unless we felt that the integrity of a
person is threatened or violated. Feelings form the basis of our
perception of something as moral. Someone who is blind to moral
phenomena is blind to feeling. He lacks a sense, as we say, for the
suffering of a vulnerable creature who has a claim to have its integrity,
both personal and bodily, protected. And this sense is manifestly
closely related to sympathy or compassion.
Second, and more important, moral feelings, as you correctly ob
serve, provide us with orientation in judging morally relevant particular
cases. Feelings are the experiential basis of our initial intuitive judg
ments: feelings of guilt and shame are the basis of self-reproach,
pain and resentment of reproaches against another person who of
fends us, anger and indignation of the condemnation of a third
person who offends someone else. Moral feelings are reactions to
disruptions of relations of intersubjective recognition or of interper
sonal relations in which agents are involved in the role of first, second,
or third persons. Hence, moral feelings are structured in such a way
that they reflect the system of personal pronouns.
Third, moral feelings clearly play an important role not only in
the application of moral norms but also in their justification. At the
very least, empathy-the ability to project oneself across cultural
distances into alien and at first sight incomprehensible conditions of
life, behavioral predispositions, and interpretive perspectives-is an
emotional prerequisite for ideal role taking, which requires everyone
to take the perspective of all others. To view something from the
moral point of view means that we do not elevate our own self-
JH:
1 75
understanding and worldview to the standard by which we univer

salize a mode of action but instead test its generalizability also from
the perspectives of all others. It is unlikely that one would be able to
perform this demanding cognitive feat without the generalized com
passion, sublimated into the capacity to empathize with others, that
points beyond affective ties to immediate reference persons and
opens our eyes to "difference," to the uniqueness and inalienable
otherness of the other.
Of course, however indispensable the cognitive function of moral
feelings, they do not have a monopoly on the truth. In the final
analysis, moral judgment must bridge a gap that cannot be filled
emotionally. Ultimately we must rely on moral insiglw if all of hu
mankind is to have a claim to moral protection. The counterfactual
idea that all human beings are brothers and sisters is already difficult
enough to encompass, but the notion of a horizon including all
human beings proves to be even more fragile if it is supposed to be
held together by spontaneous feelings. So your question is not at all
easy to answer. Certainly moral feelings initially sensitize us to moral
phenomena, and in questions of the justification and application of
norms, they have in addition an invaluable heuristic function. But
they cannot be the final reference point for judging the phenomena
they bring to light.
THN: You have often emphasized that a narrow concept of morality

demands a correspondingly modest self-assessment of moral theory.
On your conception, "the philosopher ought to explain the moral
point of view, and-as far as possible-justify the claim to universality
of this explanation. . . . Anything further than that is a matter for
moral discourse between participants. " S2 However, it seems to me
that this modesty and separation of spheres is being superseded in
your more recent writings by a new tripartite division in which a neo
Kantian moral theory (discourse ethics) is mediated with a curtailed
form of Hegelian ethical life (a pragmatic ethics) through a broad
ened concept of practical reason or even a Kierkegaardian notion of
"radical choice." How do you view this?
JH: I think that the task of philosophy is to clarify the conditions
under which moral and ethical questions alike can be decided ration
ally by the participants themselves. The moral point of view enables
us to grasp generalizable interests in common; an ethical decision to
1 76
adopt a conscious plan of life first puts a person or group in the

proper frame of mind to appropriate critically their own life history,
or identity-constituting traditions, in the light of an authentic life
project. But philosophy cannot arrogate to itself the task of finding
answers to substantive questions ofjustice or of an authentic, unfailed
life, for it properly belongs to the participants. It can help prevent
confusions; for example, it can insist that moral and ethical questions
not be confused with one another and thereby be addressed from an
inappropriate perspective. But when it offers material contributions
to the theory of justice-as Rawls does in some sections of his book
or if it becomes engaged in drawing up normative blueprints for an
emancipated society-as do Ernst Bloch and Agnes Heller-then the
philosophical author steps back into the role of an expert who makes
proposals from the perspective of a citizen participating in the polit
ical process.
Anyone who goes beyond procedural questions of a discourse the
ory of morality and ethics and, in a normative attitude, immediately
embarks on a theory of the well-ordered, or even emancipated, so
ciety will very quickly run up against the limits of his own historical
situation and his failure to take into account the context in which his
own development has taken place. Hence, I advocate an ascetic con
strual of moral theory and even of ethics-indeed, of philosophy in
general-so as to make room for a critical social theory. Critical theory
can contribute to the scientific mediation and objectivation of pro
cesses of self-interpretation in quite a different way; the latter should
neither succumb to a hermeneutic idealism nor fall between the twin
stools of philosophical normativism and sociological empiricism. This
is more or less the architectonic I have in view-from the negative
perspective of what is to be avoided.
Chapter l
l . Ursula Wolf, Das Problem des moralischnl Sollens (B erli n , 1984).
2. Hans Albert, Treatise on Critical Reason, trans. M. V. Rorty (Prince ton , N.j., 1985).
3. Charles Taylor, "The Concept of a Person," in Philosophical Papers (New
1:97ff.
York, 1985)
4. Hans-Georg Gadamer, "Hermeneutics as Practical Philosophy," in Reason in the Age

of Science, trans. F. Lawrence (Cambridge, Mass . , 1 982), pp. 88ff.
5. Michael Sandel, Liberalism and
the Limits ofJustice (New York , 1982).
6. Hobbes, De Give, I I I , 14.

7. Ernst Tugendhat, "Antike und moderne Ethik,"
1 984), pp. 33ff.
in Probleme
der Ethill (Stuttgart,
8. Kant, Foundations of the Metaphysics of Morals, trans. L. W. Beck (New York, 1959),
p. 42 [425]. [Numbers in brackets in quotations from Kant refer throughout to the
pagination of the Prussian Academy edition-Trans.]
9. Cf. Tullio Maranhao, Therapeutic Discourse and Socratic DitJWgue (M adison , Wis.,
1986).
10. Karl-Otto A pe!, "The a priori of the Communication Community and the Foun
dations of Ethics," in Towards a Transformation of Philosophy, tra ns . G. Adey &: D. Frisby
(London, 1980), pp. 225ff.
1 1 . Klaus Gunther, Der Sinn fur Angemessenheit (Frankfurt , 1988).

12. Ernst Tugendhat, Probleme der Ethill (Stuttgart, 198 4), pp. 87ff.
1 78
Notes
13. Here I part company with Karl-Otto Apel and his "principle of complementarity"
(cf. below, chapter 2, pp. 84ff.); see also Jiirgen Habermas, "Volkssouveriinitat als
Verfahren," in Forum fUr Philosophie (ed.), Die ldeen von 1 798 in der deutschen Rezeption
(Frankfurt, 1 989), pp. 7-36.
1 4. This objection is raised by Martin See) in Die Kunst der Entzweiung (Frankfurt,
1 976).
Chapter 2
I . Kant, Foundations of the Metaphysics of Morals, trans. L. W. Beck (New York, 1959),
p. 20 [403-404] .
2. Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, Mass., 1 985).
3. Williams, Ethics, p. 163.
5 . See the remarks on the type of discussion pursued on the basis on mutual trust in
Williams, Ethics, pp. 1 70f.
7. Williams, Ethics, p. 1 52. For criticism of Williams's conception of science, cf. Hilary
Putnam, Realism with a Human Face (Cambridge, Mass., 1 990), pp. 1 63- 1 78.
8. John Rawls, A Theory ofJustice (Cambridge, Mass., 1971).
9. john Rawls, "Kantian Constructivism in Moral Theory," Journal of Philosophy (Sep
tember 1 980), p. 5 1 9.
1 0. john Rawls, ''justice
as
Fairness: Political Not Metaphysical," Philosophy and Public
Affairs 14 (Summer 1985), p. 230.
11. "[I]t was an error in Theory [of Justice] and a very misleading one, to describe a
theory of justice as part of the theory of rational choice, as on pp. 16 and 583. What
I should have said is that the conception of justice as fairness uses an account of
rational choice subject to reasonable conditions to characterize the deliberation of the
parties as representatives of free and equal persons." Rawls, ':Justice as Fairness,"
p. 237, n. 20.
1 2. Rawls, "justice as Fairness," p. 223. With the shifting of normative premises from
the procedure to the concept of the person already undertaken in the Dewey Lectures,
Rawls presents an unprotected flank to the familiar neo-Aristotelian objections. Cf.
Michael I. Perry, Morality, Politics, and Law (Oxford, 1 988), pp. 57ff.
1 3 . For a discourse-theoretical interpretation of Rawlsian constructivism, see Kenneth
Baynes, The Normative Grounds of Social Criticism: Kant, Rawls and Habermas (New York,
1 99 1 ) , pp. 68-76.
1 4. Gadamer makes the connection between philosophical hermeneutics and a histor
ically informed Aristotelian ethics : " U ber die Moglichkeit einer philosophischen
1 79
Notes
Ethik," in Kleine Schriften I (Tubingen, 1 967), pp. 1 79ff. ; cf. Ernst Tugendhat, "Antike
und Modeme Ethik," in Probleme der Ethik (Stuttgart, 1 984), pp. 33ff.
1 5 . Albrecht Wellmer, The Persistence of Modernity, trans. D. Midgley (Cambridge, Mass.,
1 99 1 ), p. 1 85.
1 6. For a critical discussion of important aspects of Wellmer's study, see Lutz Wingert,
"Moral und Gemeinsinn" (Dissertation, Frankfurt, 1 99 1 ).
1 7. In the original edition of Moralbewufttsein und lwmmuniMtives Handeln (Frankfurt,
1 983), pp. 1 02f., I employed an overly strong notion of normative justification. This
error has been corrected in subsequent German editions and in the English edition :
Moral Consciousness and Communicative Action, trans. C. Lenhardt and S. W. Nicholsen
(Cambridge, Mass., 1 990), pp. 92f.
1 8. A detailed proposal for carrying through this justification is outlined by William
Rehg, "Discourse and the Moral Point of View: Deriving a Dialogical Principle of
Universalization," Inquiry 34 ( 1 99 1 ) , pp. 27-48.
1 9. Wellmer, Persistence, p. 225. The same holds of the well-founded objection that
Gunther Patzig raises against the more comprehensive claims of K. 0. Ape!; cf. Patzig,
"Principium diiudicationis und Principium executionis," in Gerold Prauss, Handlung
stheorie und Transzendentalphilosophie (Frankfurt, 1 986), pp. 204ff., especially pp. 2 1 4ff.
20. Wellmer, Persistence, p. 1 55.
2 1 . Cf. Klaus Gunther, Der Sinn fur Angemessenheit (Frankfurt, 1 988), pp. 23-1 00.
22. Gunther, Der Sinn, p. 62.
25. Gunther, Der Sinn, pp. 55ff.
26. Cf. also Gunther, "Ein normativer Begriff der Koharenz fUr eine Theorie der
juristischen Argumentation," Rechtstheorie 20 ( 1 989), pp. 1 63-1 90.
27. Gunther, "Ein normativer Begriff," p. 182.
28. On Durkheim, cf. Jurgen Habermas, Theory of Communicative Action, trans. T.
McCarthy (Boston , 1 987}, 2 :43ff.
29. I have discussed Tugendhat's Gauss Lectures in another context (Habermas [ 1 990] ,
pp. 68ff.). Here I refer t o the "Retractions" in Tugendhat, Probleme der E thik, pp. 1 321 76, and to his contribution to the Fahrenbach Festschrift, "Zum Begriff und zur
Begrundu ng von Moral," in C. Bellut and Muller-SchOll (eds.), Mensch und Moderne
(Wurzburg, 1 989}, pp. 1 45- 1 63. The transitional view between the earlier and later
positions is represented by an interesting manuscript, "A New Conception of Moral
Philosophy" ( 1 987).
30. Tugendhat, "Zum Begriff und zur Begrundung von Moral," p. 1 46.
1 80
Notes
3 1 . Tugendhat, "Zum Begriff und zur Begrii ndung von Moral," p. 147.
33. Tugendhat, "Retractions," pp. 1 56ff.
34. Cf. the thoroughgoing critique developed by Wingert, "Moral und Gemeinseinn."
35. Cf. Tugendhat, "A New Conception of Moral Philosophy."
36. Tugendhat, "Zum Begriff und zur Begriindung von Moral," pp. 1 58f.
39. The first-person plural perspective cannot be replaced, or even circumscribed, by
an alternation between the first- and third-person perspectives. Thomas Nagel, whose
thought is situated within the theoretical framework of the philosophy of the subject,
confuses the standpoint of impartial judgment of moral questions with the "external
standpoint" of an observer in The View from Nowhere (Oxford, 1 986), pp. 1 38ff. This
premise generates a dilemma between an objectivistic neutralization of the spheres of
value-" If we push the claims of objective detachment to their logical conclusion, and
survey the world from a standpoint completely detached from all interests, we discover
that there is nothing-no values left of any kind : things can be said to matter at all
only to individuals within the world" (p. 1 46)-and the subjectivistic dissolution of
commands and values into preferences. Nagel believes that he can escape this dilemma
through a procedure of expanding the subjective perspective by reflective integration
of objective knowledge: "We simply aim to reorder our motives in a direction that will
make them more acceptable from an external standpoint. Instead of bringing our
thoughts into accord with an external reality, we try to bring an external view into the
determination of our conduct" (p. 1 39). But preserving the connection to the first
person perspective prejudices the view in favor of ethical issues and closes off the
dimension of moral judgment. Nagel cannot deal with the "obscure topic of deonto
logical constraints" (p. 1 75), as can be seen from his counterintuitive treatment of the
example on pp. 1 76ff.
40. See, for example, Foundations of the Metaphysics of Morals, p. 48, n. 1 4 [430].
4 1 . Cf. Karl-Otto Apel, Charles S. Peirce: From Pragmatism to Pragmaticism, trans. J. M .
Krois (Amherst, Mass., 1 98 1 ), and "Sprachliche Bedeutung, Wahrheit und normative
GUitigkeit," Archivio di Filosofia 55 ( 1 987), pp. 5 1-88; also C. J. Misak, Truth and the
End of Inquiry (Oxford, 1 99 1 ).
42. Cf. Robert Alexy, "Probleme der Diskurstheorie," Zeitschrift fur Philosophische For
schung 43 ( 1 989), pp. 8 1-93.
43. Stephen Lukes, "Of Gods and Demons," in John B. Thompson and David Held
(eds.), Habermas-Critical Debates (Cambridge, Mass., 1 982), p. 1 4 1 .
44. Williams, Ethics, p . 1 97.
45. Cf. B. Peters, Rationalitiit, Recht und Gesellschaft (Frankfurt, 1 99 1 ), pp. 227ff.
181
Notes
46. Cf. R. Alexy, "Nachwort" to the new edition of Theorie der juristischen Argumentation
(Frankfurt, 1 99 1 ), pp. 399ff. [In English: A Theory of Legal Argumentation, trans. P.
Adler and N. McCormick (New York, 1 989).]
47. On the communication-theoretical concept of the subject, cf. Jiirgen Habermas,

"Individuation Through Socialization," in Habennas, Postmetaphysical Thinking (Cam
bridge, Mass. , 1 992), pp. 1 49-204.
48. Cf. Marcus G. Singer, Generaliz.ation in Ethics (New York, 1 97 1 ), p. 240; Wellmer,
Persistence, pp. 1 23ff.
49. Wellmer, Persistence, p. 1 93.
50. See section 4.
5 1 . Cf. Jiirgen Habermas, "Labor and Interaction: Remarks on Hegel's jena Philosophy
of Mind," in Theory and Practice, trans. J . Viertel (Boston, 1 983), pp. 1 42ff.
52. Bernard Gert, The Moral Rules (New York, 1 970), p. 1 25; Morality-A New jwtifi
cation of the Moral Rules (New York, 1 988), p. 1 5 7 .
53. Charles Fried, Right and Wrong (Cambridge, Mass., 1 978), p. 29.
54. Cf. Henry Shue, "Mediating Duties," Ethics 98 ( 1 988), pp. 687-704.
55. Fried, Right and Wrong, p. 1 1 8 .
56. Cf. Charles Taylor, Sources of the Self: The Making of the Modem Identity (Cambridge,
Mass., 1 989). Page references to this book are noted parenthetically in the text.
57. The same claim is made by Dieter Henrich, Fluchtlinien (Frankfurt, 1982); see also
the appendix to his Ethik zum nuklearen Frieden (Frankfurt, 1 990), pp. 274ff.
58. Other neo-Aristotelians dispute the neutrality of the principle of justice regarding
conceptions of the good life rooted in concrete forms of life and traditions but without
taking into account, as does Taylor, the fact that the priority of the right over the
good first emerges at a postmetaphysical or nonfoundationalist level of thought. Only
in modern societies do cultural traditions become reflective in the sense that competing
worldviews no longer simply assert themselves against one another in noncommuni
cative coexistence but are compelled to justify their claims to validity self-critically in
the light of argumentative confrontations with the competing validity claims of all
others. Modern conditions of life leave us without alternatives; we are not free to alter
them at will, and hence they do not stand in need of retrospective nonnative justifi
cation. This fact was dramatically highlighted by the Rushdie affair: no culture that
has, with the transition to modernity, integrated secod-order concepts into its every
day practices and its collective identity can accod claims that are merely asserted in a
fundamentalistic manner the same argumentative role as reflectively jwtified claims.
Perry's critique (cf. Perry, Morality, pp. 57ff ) o the so-called liberal project of Ack
ermann, Dworkin, and Rawls fails to accord th1s f sufficient weight. Perry's own
concept is inconsistent in, on the one hand, s_ubordmating the demands of an auton
_ usly or metaphysically grounded forms
omous morality ofjustice to the claims of rehg.
of life while, on the other, linking the coeXIStence of these life-forms to forms of
communication in which it is not difficult to discern the existence of a moral principle,
whose neutrality Perry nevertheless denies: "If one can participate in politics and law
.
1 82
Notes
. . . only as a partisan of particular moral (religious) convictions about the human, . . .

then we who want to participate, whether as theorists or activists or both, must examine
our convictions self-critically. We must be willing to let our convictions be tested in
ecumenical dialogue with others who do not share them . . . . If necessary, we must
revise our convictions until they are credible to ourselves, if not always or even often
to our interlocutors" (p. 1 83). But how should the liberation theologians, with whom
Perry sympathizes, deal with their pope. for instance, who does not even want to enter
into ecumenical dialogue? How are they to defend the view that he should expose
himself to such arguments when the appeal to authenticity has priority over deonto
logical notions of impartiality, equal respect for all, autonomy, and so forth? "The
relevant ontological category is not autonomy but authenticity. To have achieved some
degree of (political) freedom is not necessarily to have achieved any degree of (human)
authenticity" (p. 1 82).
59. Cf. Christoph Menke, Die Souveriinitiit der Kunst (Frankfurt, 1 988). [English trans
lation forthcoming, MIT Press.]
60. Karl-Otto Ape!, Diskurs und Verantwortung (Frankfurt, 1 988), p. 348.
6.J . Ape!, Diskurs, p. 347.
62. Apel, Diskurs, p. 352.
63. Jiirgen Habermas, Moral Consciousness and Communicative Action ( 1 990), pp. 86-98.
65. Ape!, Diskurs, p. 352.
66. Ape!, Diskurs, p. 69, and "Is the Ethics of an Ideal Communication Community a
Utopia?" in S. Benhabib and F. Dallmayr (eds.), The Communicative Ethics Controversy
(Cambridge, Mass., 1 99 1 ), pp. 23ff.
67. Apel, Diskurs, p. 1 65.
68. Karl-Otto Ape!, "Normatively Grounding 'Critical Theory' Through Recourse to
the Lifeworld ?" in A. Honneth, T. McCarthy, C. Offe, and A. Wellmer (eds.), Philo
sophical Interventions in the Unfinished Project of Enlightenment (Cambridge, Mass., 1 992),
pp. 1 25- 1 70.
69. Karl-Otto Apel, Transformation der Philosophie, 2 vols. (Frankfurt, 1 973) [English
translation of selected essays: Towards a Transformation of Philosophy, trans. G. Adey and
D. Frisby (London, 1 980).]
70. Karl-Otto Ape!, "Kann es in der Gegenwart ein postmetaphysisches Paradigma
der Ersten Philosophic geben?" (ms., 1 99 1 ).
7 1 . See Martin Seel, "The Two Meanings of Communicative Rationality," in A. Hon
neth and H. Joas (eds.), Communicative Action (Cambridge, Mass., 1 99 1 ), pp. 36-48,
and my response, pp. 222-228.
72. Habermas, Moral Consciousness, pp. 1-20.
73. See Marcel Niquet, Transundentale Argumente: Kant, Strawson und die Sinnkritische
Aporetik der Detransundentalisierung (Frankfurt, 1 99 1 ) .
1 83
Notes
74. Apel, "Normatively Grou nding. "

76. Cf. Wingert, "Moral und Gemei nsinn."
77. Cf. my remarks on the issue of civil disobedience in Habermas, Die Neue Uniiber
sichtlichkeit (Frankfurt, 1 985) , pp . 79 - 1 20.
78. Thomas McCarthy, "Practical Discourse: On the Relation of Morality to Politics,"
in Ideals and Illusions: On Reconstruction and Deconstruction in Contemporary Critical Theory
(Cambridge, Mass., 1 99 1 ), p. 1 98.
79. McCarthy, "Practical Discourse On the Relation of Morality to Politics," p. 1 96.
80. McCarthy, "Practical Discourse On the Relation of Morality to Politics," pp. 1 9 l f.
8 1 . Cf. Charles Taylor, "The Liberal-Communitarian Debate," in N. Rosenblum (ed.),
Liberalism and the Moral Life (Cambridge, Mass., 1 989).
82. Seyla Benhabib, "Autonomy, Modernity and Co mm unity, in A. Honneth, T.
McCarthy, C. Offe, and A. Wellmer (eds.), Cultural-Political lnterventiuns in the Unfinished
Project of Enlightenment (Cambridge, Mass., 1 992), pp. 39-59, and "In the Shadow of
Aristotle and Hegel: Communicative Ethics and the Current Controversy in Practical
Philosophy," Philosophical Forum 2 1 ( 1 989/90), pp. 1-3 1 ; Kenneth Baynes, "The. I..ib
eral-Communitarian Controversy and Communicative Ethics," Philosophy and Social
Criticism 1 4 ( 1 988), pp. 293-3 1 5 .
"
83. John Rawls, "Justice a s Fairness : A Briefer Restatement" (ms 1 989) (page refer
ences in parentheses are to this text). Cf. also Rawls "justice as Fairness."
.,
84. Alasdair Macintyre, Whose justice? Which RationalityP (Notre Dame, Ind., 1 988) .
Page references to this book are noted parenthetically in the text.
85. Gunther Patzig, Ethik ohne Metaphysik, 2d ed. (GOtti ngen 1 983), p. 132.
,
86. D. Birnbacher, "Sind wir fii r die Natur verantwortlich?" in Okologie und Ethik
(Stuttgart, 1 980), p. 72.
87. Gunther Patzig, " O kologische Ethik-innerhalb der Grenzen blo8er Vernunft" in
H. J . Elster (ed.), Umweltschutz-Herausforderun unser Generation ( Stdienzentru m
Weikersheim, 1 984), 67. In Ethik ohne Metaphysik Pawg defends a Kanban approach
in ethics that incorporates utilitarian arguments.
88. Jeremy Bentham, An Introduction to the Principles of Morals and Government (Oxford,
1 789), chap. 1 7, I, sec. 4, footnote.
89. Patzig, " O kologische Ethik," p. 74.
90. Patzig, " O kologische Ethik," p. 7 3.
9 1 . Patzig, " Okologische Ethik." p. 7.
92. Gunther Patzig, "Der wissenscha ftliche Tierversuch unte r ethischen Aspekten" in
W. Hardegg and G. Preiser (eds.), Tierversuch und medizinische Ethik (Hildesheim, 1 986),
p. 77.
1 84
Notes
93. Cf. Joel Feinberg, "The Rights of Animals and Unborn Generations," in William
T. Blackstone (ed.), Philosophy and Environmental Crisis (Athens, Ga., 1 974), pp. 43-68.
Chapter 3
l . Jiirgen Habermas, "Individuation Through Socialization," in Habermas, Postmeta

physical Thinking (Cambridge, Mass., 1 992).
2. Lawrence Kohlberg, "The Current Formulation of the Theory," in Kohlberg, The
Psychology of Moral Development (San Francisco, 1 984), p. 286.
3. I have discussed how empirical assumptions and rational reconstruction are inter
related in Kohlberg's theory of moral development in "Reconstruction and Interpre
tation in the Social Sciences," in Habermas, Moral Consciousness and Communicative
Action, trans. C. Lenhardt and S. W. Nicholsen (Cambridge, Mass., 1 990), pp. 2 l ff.
4. Kohlberg, "Current Formulation," pp. 277ff.
5. Herbert Schnadelbach, "What is Neo-Aristotelianism?" Praxis International 7,
pp. 225-237.
6. Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, Mass., 1 985), p. 69.
7. Albrecht Wellmer, The Persistence of Modernity, trans. D. Midgley (Cambridge, Mass.,
1 99 1 ), pp. 200ff.
8. Michael Sandel, Liberalism and the Limits ofjustice (Cambridge, Mass., I 983), p. 1 73.
9. Alasdair Macintyre, After Virtue (Notre Dame, Ind., 1 984), pp. 5 l ff.
10. Macintyre, After Virtue, pp. 2 1 8-2 19.
1 1 . Richard J . Bernstein, "Nietzsche or Aristotle," in Philosophical Profiles (Philadelphia,
1 986), p. 1 29.
12. Williams, Ethics, p. 200 (my emphasis).
1 3 . Williams, Ethics, p. 1 55.
14. Richard Rorty, "Solidarity or Objectivity," in J. Rajman and C. West (eds.), Post
Analytic Philosophy (New York, 1 985), p. 5.
15. Williams, Ethics, p. 1 72 .
1 6 . Karl-Otto Apel, "Der postkantische Universalismus i m Lichte seiner aktuellen
Mi6verstii.n dnisse," in Diskurs und Veranlworlung (Frankfurt, 1 988), pp. 2 1 7ff.
1 7 . William E. Connolly, Politics and Ambiguity (Madison, Wis., 1 987).
18. Charles Taylor, "Die Motive einer Verfahrensethik," in Kuhlmann, Moralitlit,
pp. 1 0 l ff.
1 85
Notes
19. Schnadelbach, "What is Neo- Aristotelianism ? " pp. 2 34-2 3 5 .

20. Kohlberg, "Current Formulation," pp. 230.
2 1 . Kohlberg, "Current Formulation," pp. 2 3 2.
22. David Wiggins, Needs, Values, Truth (Oxford, 1 987), p. 8 2
23. Klaus Gunther, Der Sinn fur Angemessenheit (Fra nkfurt, 1 988), pp. 255ff.
24. Alan Gewirth, Reason and Morality (Chicago, 1 978).
25 . Robert Selman, The Growth of Interpersonal Understanding (New York, 1 980) ; "ln
terpersonale Verhandlungen," in W. Edelstein and J . Habermas (eds.), Soziale lnter
aktion und soziales Verstehen (Frankfurt, 1 984), pp. 1 1 3ff.
26. Fritz Oser, Moralisches Urteil in Gruppen (Frankfurt, 1 98 1 ).

Chapter 4
I . Alfred Schmidt, Die Wahrheit im Gewande der Lilge (Munich, 1 986); "Religion as Trug
und als metaphysisches Bediirfnis," Quatuor Coronali ( 1 988), pp. 87ff. ; "Aufklarung
und Mythos im Werk Max Horkheimers," in A. Schmidt and N. Altwicker (eds.), Max
Horkheimer heute (Frankfurt, 1 986), pp. 1 80ff.
2. Max Horkheimer, "Gesprach mit Helmut Gumnior," Gesammelte Schriften [henceforth
cited as GS] (Frankfurt, 1 985- 1 99 1), 7 : 387.
3. GS, 7 :393.
4. Max Horkheimer and Theodor W. Adorno, Dialectic of Enlightenment, trans. J.
Cumming (New York, 1 972), p. 1 1 8.
5. Max Horkheimer, Notizen 1 950 bis 1 969 (Frankfurt, 1 974).

6. This holds above all for the philosophical articles that Alfred Schmidt already
included in the appendix to the German edition of The Critique of Instrumental Reason
(Zur Kri.tilr. der instrumentellen Vemunft [Frankfurt, 1 967], pp. 1 7 7ff.).
7. Horkheimer, Notiun, p. 93.
8. Horkheimer, Notiun, p. 1 84.
1 0. Horkheimer and Adorno, Dialectic, p. 1 1 8.
I I . Horkheimer, Notiun, p. 6 1 .
1 2 . GS, 7 : 1 78.
1 3. "Theismus-Atheismus," GS, 7 : 1 85ff.
Notes
1 4 . GS, 7 : 1 84.
1 5. "Religion und Philosophie," GS, 7 : 193.
16. "Die Aktualitiit Schopenhauers," GS, 7 : 1 36.
17. Schmidt, Die Wahrheit im Gewande der Liige, p. 1 2 1 .
1 8 . "Religion und Philosophie," GS, 7 : 1 9 1 .
1 9 . "Die Aktualitat Schopenhauers," GS, 7 : 1 38f.
20. GS, 7 : 1 35f.
2 1 . "Schopenhauers Denken," GS, 7 : 252.
22. "Pessimismus Heute," GS, 7 :227f.
23. Horkheimer and Adorno, Dialectic, p. 1 0 1 .
24. "Psalm 9 1 ," GS, 7 : 2 1 0.
26. Horkheimer, Notiun, p. 1 72.
27. "Die Aktualitiit Schopenhauers," GS, 7 : 1 38.
28. Horkheimer, Notizen, p. 26.
Chapter 5
1 . English version : The Theory of Communicative Action, 2 vols. trans. T. McCarthy

(Boston, 1 984 , 1 987)
2. Moral Consciousness and Communicative Action, trans. C. Lenhardt and S. W. Nicholsen
(Cambridge, Mass., 1 990). Original version: MoralbewujJtsein und kommunikatives Han
deln (Frankfurt, 1 983).
3 . English version : Legitimation Crisis, trans. T. McCarthy (Boston, 1 975).
4. J iirgen Habermas, Vorstudien und Ergiinzungen zur Theone des kommunikativen Handelns
(Frankfurt, 1 984) , pp. 226-270.
5. "Vorlesungen zu einer sprachtheoretischen Grundlegung der Soziologie," in Ha
bermas Vorstudien, pp. 1 1-1 26.
6. English version: Habermas, The Philosophical Discourse ofModernity, trans. F. Lawrence
(Cambridge, Mass., 1 987).
7. Cf. Habermas, "Theory of Communicative Action," 2 : 92 ff.
8. Michael Walzer, Spheres ofJustice (New York, 1 983), p. 6.
1 87
Notes
9. Cf. John Rawls, A Theory ofJustice (Cambridge, Mass. , 1 97 1 ), pp. 446, 567ff.
1 0. Cf. Carol Gilligan, In a Different Voice (Cambridge, Mass. , I 982) .
I I . Seyla Benhabib, "The Generalized and the Concrete Other," in Seyla Benhabib
and Drucilla Cornell (eds.), Feminism and Critique (Minneapolis, I 987), pp. 77-96.
12. Cf. Benjamin Lee Whorf, Language, Thought and Reality (Cambridge, Mass. , I 973).
I3. Wolfgang Schluchter, The Rise of Western Rationalism (Berkeley, I 98 I ) , pp. 6ff.
I 4 . Cf. Anne Colby and Lawrence Kohlberg, The Measurement of Moral Judgment (Cam
bridge, 1 987), vol. I .
I 5 . Cf. Habermas, Communication and the Evolution of Society, trans. T. McCarthy (Bos
ton , I 979), p. I 20.
I 6. Cf. Habermas, Autonomy and Solidarity: Interviews with Jilrgen Habermas, ed. Peter
Dews (London, I 986), p. 1 60.
17. See Habermas, Moral Consciousness, p. 80.
1 8. Habermas, Moral Consciousness, p. 92.
1 9. Habermas, "Theory of Communicative Action," 2 : I 49-I 50.
20. Cf. Wolfang Schluchter, Religion und Lehensfilhrung. Studien zu Max Wehers Kultur
und Werttheorie (Frankfurt, I 988), I : 322-323.
2 1 . Cf. Habermas), Vorstudien, p . I 26 n. 94 ; also "A Reply to My Critics," in John B.
Thompson and David Held (eds.), Habermas-Critical Debates (Cambridge, Mass., I 982),
pp. 26 1 .
22. "A Reply," in Axel Honneth and Hans Joas (eds.), Communicative Action (Cambridge,
Mass., I 9 9 I ) , p. 257.
23. Cf. Amitai Etzioni, The Moral Dimension (New York, I 988), p. 90.
24. Thomas McCarthy, "Complexity and Democracy, or the Seducements of Systems
Theory," New German Critique (Spring, I 985), p. 44.
25. Cf. Habermas, "Civil Disobedience : Litmus Test for the Democratic Constitutional
State," Berkeley Journal of Sociology 30 ( I 985), p. I 06.
26. Cf. Jiirgen Habermas, "Volkssouveranitiit als Verfahren," in Forum fiir Philosophie
(ed.), Die ldeen von 1 798 (Frankfurt, I 989), pp. 28f.
27. Cf. Habermas : " U ber Moralitat und Sittlichkeit," in H. Schnadelbach (ed.), Ration
alitii.t (Frankfurt, I 984), p. 228; also Moral Consciousness, pp. 207ff.
28. Habermas, "On the Concept of 'Practical Reason,"' Howison Lecture, ms. (Uni
versity of California, Berkeley, I 988), p. I 9. This lecture appears as the first chapter
of this book (cf. above p. I 3).
1 88
Notes
29. Cf. Klaus Gunther, "Ein normativer Begriff der Koharenz," Rechtstheorie 20 ( 1 989):
163-1 90.
30. Cf. p. 6.
3 1 . Cf. p. 7.
32. Habermas, Autonomy and Solidarity, p. 1 60.
Action, 1 4 , 64-66, 70, 1 20, 1 29. See also

Communicative action
autonomous, 1 4
moral, 33 -34 , 48
moral judgment of, 9
norm-regulated, 9
possibilities of, 23
strategic, 6, 3 1
Adorno, Theodor W. , 74, 1 39
Agent, 63-64, 66
American Pragmatism, 1 30
Animals, duties toward, 1 06-1 1 1
Apel, Karl-Otto 1 9, 50, 75-80, 82-86,
1 1 6, 1 50, 1 62- 1 63
and an ethics of responsibility, 84
foundationalism of, 79, 82
on moral existence, 77
and the pragmatics of language, 8 182
Appropriateness, principle of, 14, 37,
87, 1 30, 1 7 1
Aquinas, St. Thomas, 96
Arendt, Hannah, 1 6 7
A rgume n tatio n , 1 1 , 4 9 , 53, 55, 9 8 , 1 06,
1 29, 1 45, 1 54
and universalization, 35-36
as a problem-solving p rocedure, 1 58
commu nicative presuppositions of, 16,
19, 29, 48
definition of, 58
forms of, 1 6
ideali z ing assumptions of, 50
and imagination, 1 5
institutional guarantees of, 3 1
intersubjective process of, 5 1
Jack o f absolute, 59
language game of, 83 , 99
logic of, 29
modeJ of, 27
moriD, 1 6, 24, 33, 1 29, 158
network of, 89
normative content of, 30
normative presuppositions of, 33
participants in, 5 7
practice of, 67
presuppositions of, 32, 59, 77, 82, 8486, 1 3 1 , 1 64
procedure of, 58
rationality assumptions of, 3 1
rules of, 32, 76, 1 59
self-clarification of, 80
self-referential, 79, 83
self-reflexive, 93
shifting terrain of, 30
and social roles, 1 60
spontaneous, 57
transcendental constraints on, 3 1
transcendental-pragmatic, 77, 83
and truth claims, 59
universal presuppositions of, 50, 1 63
world-disclosing, 79
Argumentative praxis, 1, 77, 1 3 1
Arguments, deductive, 59
Aristotelian ethics, 1 1 8, 1 25
Aristotle, 1 0, 1 7 , 2 1 , 24, 96, 1 1 6- 1 1 7,
1 32, 1 50, 1 73
on the good life, 4-5

teleologiciD worldview of, 1 23- 1 25
Art, 73-74
Assertoric pro position s , 32, 40

Atheism, 1 36
Augustine, Sain t, 96, 1 44
Authenticity, 9, 1 2
Autonomous will, 42
1 90
Index
Autonomy, 14, 34, 43, 1 1 4, 1 1 9, 1 25,

1 30-1 3 1 , 1 53
Baader, Franz Xavier von, 1 36
Benhabib, Seyla, 1 54
Bentham, jeremy, 1 06
Berger, Johannes, 1 70
Bernstein, Richard, 1 23
Bilingualism, 1 02, 1 04
Bloch, Ernst, 1 76
Bohman, Jim, 149
Categorical imperative, 1, 5, 8, 13, 40,
60, 63, 1 1 8 , 1 28. See also Commands
Choice, 9, 6 1 , 66
Chomsky, Noam, 1 56
Christianity, 73, 1 25, 1 33, 1 36, 1 38,
1 42- 1 43
Civil disobedience, 88, 1 6 1 , 1 67-1 68
Cognitivism, 20, 1 1 7
and morality, 33, 78, 80, 1 22, 1 35, 1 74
Commands, 3, 9, 1 6, 62, 67, 70. See also
Categorical imperative
Communal life, 7, 16, 27, 33, 59-60
Communication, 23, 49, 77, 83, 1 29,
1 43, 1 45. See also Discourse, Language
and alien cultures, 98
counterfactual presuppositions of, 55
distorted, 1 48
and idealizations, 54-55
presuppositions of, 3 1 , 1 45
public forms of, 1 7
theory of, 69
transparency of, 1 63
Communication community, 1 3 , 76, 1 30,
1 54. See also Community
ideal, 57
idealizing presuppositions of, 5 1
as a regulative idea, 1 64
unlimited, 53-54, 1 63- 1 65
Communicative action, 55, 77-78, 1 08,
1 1 0, 1 3 1 - 1 32, 1 44, 1 46, 1 62
and context-transcendence, 1 64- 1 66
and form of life, 48-49
idealizing presuppositions of, 50
pragmatic presuppositions of, 66-67
pragmatics of, 1 1 4
reflective form of, I
theory of, 9 1 , 1 70
Communicative ethics. See Discourse
ethics
Communicative practice, 28, 55, 145
Communicative reason, 80-8 1 , 1 4 1 , 1 46,
1 48- 1 49
Communitarianism, 1 , 9 1
Community, 23-24, 30, 40-4 1 , 46, 48,

60, 1 1 9, 1 2 1 , 1 24, 1 45. See also
Communication community
Compassion, 1 34, 1 42
Conflicts, 6, 9
Conscience, 14, 1 30
Consciousness, 80, 95, 97-98, 1 23, 1 6 1
transcendental, 1 6
Consensus, 32, 57, 7 1 , 84, 1 05, 1 58
argumentatively achieved, 89
hypothetical, 93
overlapping, 93-95
rational, 29, 162
Constructivism, 26, 30
Contextualism, 95, 1 04, 1 2 1 , 149
Contract theory, 25, 1 30
Contractualism, 27, 48
Cosmopolitanism, 85, 97
Critical theory, 137, 1 76
Darwin, Charles, 1 37
Davidson, Donald , 96, 1 04
Decisions, 3-4, 9, 22, 1 58
Deconstructionism, 1 49
Deliberation, 7, 1 0, 23, 1 20, 1 29
Democracy, 86, 9 1 , 1 1 4
Deontological ethics, 1 9, 39, 63, 85-86,
1 1 8, 1 22 , 1 54
Derrida, Jacques, 74
Dewey, john, 1 1 3
Dialectic of enlightenment, 1 35, 1 40
Dignity, 1 4
Discourse. See also Communication,
Language
empirical, 1 0
ethical, 1 1- 1 2, 1 6, 23-24
foundational, 8 1
moral-practical, 9 , 1 2- 1 3 , 1 5 , 8 7 , 89,
9 1 , 1 58
practical, 50, 58, 90, 1 54, 1 58
pragmatic, 1 0-1 1 , 1 4 , 1 6
public, 1 5
rational, 1 , 50
self-interpretive, 1 58
universal, 1 2, 49, 53
Discourse ethics, 1-2, 30-3 1 , 50, 60, 67,
80, 8 1 3 1 , 1 5 1 56, 1 6 1 , 1 72
and concept of morality, 88-9 1
and Kantian moral theory, 147- 1 50
objections against, 1 9
and two-stage process o f justification,
34-36
Discourse theory, 1 6, 88-89, 1 49
Discourses of application, 1 4 , 37-38,
1 28- 1 30, 1 52, 1 60, 1 72
191
I ndex
Discourses of justification, 35, 37, 1 281 29, 1 55 , 1 60

Discursive procedures, 88, 94
Durkheim , Emile, 42, 1 49, 1 56
Duty, 14, 40-4 1 , 6 1 , 63-69, 78, 85, 87,
1 07, 1 1 9
Dworkin, Ronald, 1 1 6, 1 5 1 , 1 7 2
Ecological ethics, 1 05
Egocentric perspective 6, 24, 28
Einstein, Albert, 1 1 5
Empirical sciences, 25, 30
Empiricism, 1 0, 20
Enlightenment, 73, 1 33 , 1 36- 1 37, 1 50,
1 56
Environment, 1 05 , 1 07
Episteme, 2 1 , 25
Equality, principles of, 64, 90, 1 5 1- 1 52
Ethical life, 1 , 1 2 , 22, 25, 47, 5 1 , 1 1 8,
1 2 1- 1 22 , 1 60, 1 7 1-1 72, 1 75
Ethical questions, 6, 24
Ethics, 7, 1 1 7- 1 1 8, 1 25. See also
Discourse ethics
Aristotelian, 1
of care, 1 53
classical, 2
of compassion, 1 34
of conviction, 1 , 64, 86, 1 20, 1 29
of the good, 72, 74-75, 86, 1 1 7, 1 1 9,
i25
of justice, 75
political, 29, 86
pragmatic, 172
procedural, 72
of prudence, 1 25
of responsibility, 84-86, 1 29
sphere of, 4
Ethnocentrism, 1 04
Ethnology, 1 7 2
Ethos, 6, 1 0, 1 25
Etzioni, Amitai, 1 66
Euclid, 54
Eudaimonia, 1 1 8
Evaluation, 62, 1 53
Everyday life, 25, 32
Expert cultures, 82, 1 62
Fallibilism, 38-39, 59, 82, 94
Fichte, Immanuel Hermann, 80
Form of life, 1 1- 1 2, 80, 83, 90, 1 0 1 ,
1 05 , 1 08, I l l , 1 25, 1 46, 1 50, 1 7 1 . See
also Lifeworld, World, Worldview
and argumentative praxis, 77
and egocentric detachment, 24
historical, 6
incommensurable, 1 63
intersubjective, 6, 23, 47
and normative validity, 1 1 9- 1 22
plurality of, 22
prefiguration of, 1 64
of real individuals, 92-93, 95-97
and universalism, 1 5
universally valid , 8
utopian, 79
Formalism, 1 22 , 145
Frankena, William, 1 1 5
Free will, 27, 40, 1 2 1
Freedom, 1 , 1 5 , 42-43 , 6 1 , 66-68, 1 39
French revolution, 1 39
Fried, Charles, 1 9 , 65-66, 68
Gadamer, Hans-Georg, 1 04- 1 05
Gehlen, Arnold, 22
Genealogy, 7 1
Genetic structuralism, 30
Gert, Bernard, 62, 65, 67
Gewirth, Alan, 1 50
Gilligan, Carol, 1 53-1 54
Goals, 3, 1 1
God , 73, 1 3 7
Go od , 2, 4, 2 1 , 73, 1 1 9
Good life, 2, 4, 1 1 7 , 1 2 1 , 1 26
Guilt, 1 4 , 40, 47
Gunther, Klaus, 19, 36-38, 1 72
coherence concept of, 39
on impartiality, 58
Der Sinn fur Angemessenheit, 1 54
Habermas, Jiirgen, 89, 90
Howison Lecture of 1 988, 1 72
Legitimation.sprobleme im Spatkapitalismus,
1 48
Moral Consciousness and Communicative
Action, 1 47-1 48, 1 63, 1 7 1
The Philosophical Discourse of Modernity,
1 49-1 50
on suppressed generalizable interests,
1 48
on system, 167
The Tlieory of Communicative Action, 14 71 48, 1 50, 1 63 , 1 68, 1 70
Hare, R. M., 1 1 5
Hegel, G. W. F., 1 , 1 30, 1 36, 149, 1 65 ,
171
Heidegger, Martin, 30
Heller, Agnes, 1 76
History, 39, 85
Hobbes, Thomas, 7 , 27
Horkheimer, Max, 1 33- 1 46
The Dialectic of Enlightenment (with
Theodor Adorno), 142
on the good, 1 34
1 92
Index
Horkheimer, Max (cont.)

on language, f44
on religion, 133
on Schopenhauer, 1 38
"Theism-Atheism," 1 36
Humanity, 1 5, 69
Humboldt, Wilhelm, 1 5 7
Ideal role-taking, 5 2 , 54, 1 30, 1 54, 1 74
Ideal speech situation, 56-57, 1 63
Idealism, 1 76
Idealization, 54-55, 57, 8 1 , 89
Identity, 6, 1 6, 69, 73, 1 03
Illocutionary force, 9, 1 44
Imagination, 73
Impartiality. See Moral point of view
Imperatives. See Commands
Inclination, 4, 1 1 9
Individualism, 64, 1 22, 1 4 1
Individuation, 9 1 , 1 1 4, 1 30
Instrumental reason, 1 37, 1 39- 14 1 , 1 45
Intersubjectivity, 1 3 1 , 1 43
theory of, 43, 1 1 0
James, William, 1 1 3- 1 1 4
Jefferson, Thomas, 1 1 4
Jeremiah, 1 02
Jesus, 1 43
Judaism, 1 33, 1 43
Judgment, 1 0, 1 7
Justice, 26, 1 1 9, 1 2 1- 1 22, 1 39, 142
as dimension of validity, 1 52
as fairness, 93
contextualists on, 9 1
contractualist model of, 28
discourse-theoretical conception of, 1 53
distributive, 1 52
Kantian theory of, 29
philosophical theory of, 88
politica1, 96
postmetaphysical concept of, 94, 98
principle of, 27, 79, 92
questions of, 2, 76, 105
and solidarity, 1 , 50, 68, 1 34, 1 54
sphere of, 1 5 1
theory of, 9 1 , 93
Justification. See also Discourses of
justification
moral, 1 3, 2 1
and norm-application, 36
normative, 32
Kant, Immanuel, 1, 1 7, 30, 40, 48, 64,
73, 79, 83, 1 20, 1 30, 1 34, 1 39, 145,
1 5 1 , 1 54, 1 64-1 65, 1 7 1 . See also
Categorical imperative, Kingdom of

Ends
and anthropocentrism, 1 05
on autonomy, 33-34, 42, 1 1 8, 1 28
critique of metaphysics, 1 36
his doctrine of law, 88
on duty, 14, 34, 1 06
Foundations of the Metaphysics of Morals,
1 15
on free will, 27, 1 1 9
on general validity, 52
on the general will, 5 1
on hypothetical imperatives, 3
on the Kingdom of Ends, 57
on maxims, 6-7, 63, 1 73
Metaphysical Foundations of Natural
Science, 1 1 5
on moral knowledge, 20
and the moral point of view, 24, 1 22,
1 24
moral theory of, 1 49
and the neglect of rule application, 35
on obligation, 42
on practical reason, 10, 1 3, 77, 1 1 7,
1 2 1 , 1 35
and principle of rights, 6 1
on reason and morality, 7 1
and Rousseau, 1 53
on state monopolies, 87, 1 55
on transcendental consciousness, 1 6, 48
on the understanding, 20
Kantian ethics, 1 , 34, 39, 1 53, 1 63
cognitivism of, 20
Kierkegaard, Sj/Jren, 1 75
Kingdom of Ends (Kantian) , 40, 5 1 , 57,
64
Knowledge
clinical, 1 1
discursive, 24-25
empirical, 1 1 , 22
ethical, 2 1-24
everyday, 24-25
intuitive, 25
moral, 2 1 , 38
nomological, 25
philosophical, 23, 82
postmetaphysical conception of, 2 1
practical, 25, 38, 1 24
and reflection, 23
scientific, 2 1
theoretical, 2 1
therapeutic, 23
Kohlberg, Lawrence, 78, 1 1 4- 1 16, 1 26,
1 29, 1 3 1-1 32, 1 6 1
on deontological abstraction, 1 26
on moral development, 1 1 5, 1 48
1 93
I ndex
Language, 96-97, 103, 1 44-1 45 See also

Communication, Discourse, Language
games
Language community, 1 05
Language-games, 32, 96, 1 0 1 , 1 56
Law, 16, 87-89, 1 55, 1 5 7
international, 8 6
moral, 9
philosophy of, 1 7
rational, 87
self-given, 9
Lawrence, D. H . , 74
Legal theory, 87, 1 30
Lenin, V. 1 . , 1 39
Liberalism, 9 1 , 96, 1 55
Life history, 6, 9-1 2, 1 5 , 23, 142, 1 73
Life-forms, 93, 95, 1 2 1
Life-plans, 90, 93
Lifeworld, 1 0 1 , 1 20, 1 3 1 , 1 47-1 48 , 165,
1 68-1 70. See also Form of life, World,
World view
boundaries of, 77
concrete, 50, 84, 1 2 1
conflict and, 58
and expert cultures, 1 62
intuitive knowledge of, 1 3
reflective, 1 24
shared, 1 1 , 52, 66
and system, 1 63, 166
transcendence of, 24, 1 65
LOw-Beer, Martin, 1 49
Lukacs, rg, 1 70
Lukes, Stephen, 1 9 , 57
Luther, Martin, 1 30
.
McCarthy, Thomas, 1 9 88-90

on constraints limiting universalization
principle, 90
on democracy and system, 1 67
on incommensurability of discursive
procedures, 88
and plurality of belief systems, 93
Macintyre, Alasdair, 1 9 , 97-98, 1 021 04, 1 1 6, 1 2 1 , 1 23, 1 5 1
o n abstract universalism, 97
After Virtue, 1 50
contextualism of, 96
on cultural translation, 1 0 1
learning theory of, 1 00
untranslatability thesis of, 102
Mackie, J . L., 43
Mann, Thomas, 74
Marx, Karl, 30, 1 36, 1 39, 1 48
Marxism, 1 36
Maxims, 6, 7, 8, 9, I I , 24, 5 1 , 6 1 , 6364, 1 20- 1 2 1 , 1 73
,
Mead, George Herbert, 1 1 3 , 1 30

communicative ethics, 1 49
on ideal role-taking, 1 29
on universal discourse, 49
Meaning of life, 72, 75, 79
Merleau-Ponty, Maurice, 30
Messiah, the, 143
Metaphysics, 82, 1 36, 1 40- 1 4 1
critique of, 1 37
negative, 1 38
theological, 144
Modernism, 22, 74
Modernity, 73-74, 1 50- 1 5 1
internationalized languages of, 97
pathologies of, 1 70
Modernization, 1 22
Money, 1 68
Moral cognition, 27, 1 2 1
Moral commands, 8 , 1 4 , 1 5 , 25-26, 3334, 76, 8 1 . See also Categorical
imperative
illocutionary force of, 40
validity of, 34
Moral facts, 26-27
Moral feelings, 40, 59, 1 35, 1 74-1 75
Moral insight, 1 4 , 20, 33, 42
Moral intuition, 20, 25, 28, 40, 76, 1 06,
1 1 4, 1 3 5
Moral judgment, 14, 3 0 , 34, 4 8 , 77-78,
1 25
and action, 62, 1 1 9, 1 28
and cognitive claims, 20-22
and discourse, 1 59
and intuitive knowledge, 25
justification of, 80
and moral principles, 33
validity of, 1 05
Moral law, 27, 1 20
Moral life, 1 5 , 7 1
Moral point o f view, the, 6 , 27, 39, 5860, 66, 68-69, 75, 79, 88, 9 1 , 1 2 1 ,
1 24, 1 3 1 , 1 60, 1 75- 1 76
and abstraction, 1 1 8
and argumentation, 1 9, 84
differentiating rights and goods, 1 1 81 20
and impartiality, I , 5, 36-37, 48, 1 46,
151
and justification, 70, 1 28- 1 29
and maxims, 7, 24
and practical discourse, 50
and universal validity, 24, 76
Moral principles, 1 4 , 33
Moral questions, 6, 24, 32, 152
Moral theory, 2, 1 7 , 2 1 , 24-25, 70, 7677, 87, 1 1 8, 1 2 7-1 28, 1 48, 1 60
.
1 94
Index
Moral theory (cont.)

anthropocentric, 1 06
cognitivism in, 1 59
cognitivist approaches in, 30
deontologica1, 1 05, 1 1 6
Neo-Aristotelian critique of, 1 30
and the right and the good , 1 5 3
sociologization of, 1 6 1
Moral truth, 25-26, 34
Morality, 2, 1 3 , 16, 6 1 , 69-70, 76, 79,
87-89, 1 05, 1 09, 1 1 7, 1 3 1 , 1 39,
1 55
autonomous, 1 2 1
boundaries of, 1 5
cognitivist interpretation of, 26
cognitivist understanding of, 34
discourse-ethical concept of, 89
dissolution of, 1
and ethics, 7, 1 7 1 - 1 7 2
imperatives of, 8 5
o f justice, 1 1 9
language-game of, 43
and life, 24
noncognitivist conception of, 43
postmetaphysical concept of, 1 1 8
and practical reason, 1 0
principle of, 1
procedural, 27
rational, 34, 69-70, 84
unconditionality of, 140
universal conditions of, 47
Morris, Charles, 1 1 3
Multiculturalism, 9 1
Mutual respect, 44, 46
Nationalism, 1 39
Natural law, 1 , 7, 16, 27, 125
Naturalistic fallacy, 50
Nature, 73, 1 58
Negotiation, 1 6
Neo-Aristotelianism, 28-29, 1 30
New Left Review, 163
Nielsen, Torben Hviid, 1 47
Nietzsche, Friedrich, 22, 30, 7 1 , 1 38 ,
142
Untimely Meditations, 97
Nominalism , linguistic, 144
Normative validity, 1 3 , 4 1 , 49, 54, 64,
161
Normativism, 1 76
Normativity, 8 1
Norms
abstract universality of, 1 3
o f action, 63
application of, 1 3, 37
determinateness of, 67
legal, 88
moral, 47
noncoercive binding force of, 4 1
obligatory character of, 3 3
Objectification, 22, 25
Objectivity, 13, 26
Obligation, 4 1
Ontotheology, 1 36
Orwell, George, 1 40
Paine, Thomas, 1 1 4
Paradigm shift, 99
Parsons, Talcott, 1 67- 1 68
Participant perspective, 22, 49
Participation, 68
Patzig, Gunther 1 9, 1 05, 1 08, 1 1 1
on duties towards animals, 1 07
on ecological questions, 1 06
Peirce, Charles Sanders, 1 7 , 29-30, 50,
52-53, 58, 1 1 3, 1 30, 1 45, 1 65
on pragmatics of language, 1 45
on truth, 53, 1 45
Perelman, Chaim, 54
Performative attitude, 1 3 , 25, 49, 55, 66,
131
Performative self-contradiction, 33, 56,
1 00, 1 38, 1 62
Perlocutionary effects, 1 44
Person, 23, 60, 62, 66-68
atomistic conception of, 1 2 1 , 1 30
concept of the, 28, 60, 1 30
Philosophy, 23, 74, 1 1 9, 1 37, 1 46, 1 50
classical claims of, 70, 72, 1 23
of consciousness, 82
critical task of, 1 33, 1 76
ethical task of, 79-80
and existential concerns, 75
of history, 85, 1 36, 1 40
limits of, 2 1
moral, 1 1 4
postmetaphysical, 75
and religion, 1 34
of science, 1 00, 1 1 6
and self-referentiality, 83
self-understanding of, 1 32
transformation of, 82
Phromsis, 2 1 , 25, 1 1 7, 1 20, 1 23- 1 24
Piaget, Jean, 30, 1 1 5, 1 3 2
Plato, 1 02 , 1 25, 1 32, 1 3 8
Platonism, 34
Pluralism, 1 5 1
Polis, 6
Politics, 88, 93
Positivism, 1 35, 144
1 95
Index
Postmetaphysical thought, 74, 94, 1 00,

1 23 , 1 25, 1 28, 1 46
Posmetaphysical worldviews, 95
Postmodernism, 1 50
Poststructuralism, 140
Pound, Ezra, 74
Power, 1 6 , 86, 1 69
communicative, 1 70
institutionalization of, 1 68
of labor, 1 69
political, 68, 89, 1 5 5
production of, 1 68
Practical deliberation, 2 1 , 28
Practical reason, 2-3, 1 0, 20, 30, 42, 78,
80-8 1 , 98, 1 1 8- 1 20, 1 30, 1 35 , 1 39,
1 54. See also Reason
and argumentation, 1 29
Aristotelian view of, 2 1
and contextualism, 96
exercise of, 4, 7
intersubjectivity of, 43
and justification, 9
Kantian concept of, 1 3-14, 28, 77
moral employment of, 1 5
normative content of, 28
and phronesis, 1 1 7, 1 23- 1 24
unity of, 1 6- 1 7
Pragmatic point o f view, 5-6
Pragmatics, 3, 82, 1 49, 1 56, 1 58
Pragmatism, 1 7 , 30, 1 1 3, 1 54
Praxis, 32, 1 50
Problematization, 22, 1 20
Prohibitions, 40, 63-64
Protestantism, 1 44
Public discourse, 49, 93
Public life, 26, 89
Public sphere, 60, 93
Purposiveness, 2, 43
Putnam, Hilary, 96
Rational choice, 3
Rational reconstruction, 25
Rationalism, 1 58
Rationality, 30, 8 1 , 97, 1 00, 1 7 1 . See also
Reason
communicative, 1 7 1
context-transcending, 95 , 98
debate over, 96
forms of, 1 02
and nonagreement, 94
procedural, 1 25
purposive, 3
scientific, 1 24
standards of, 1 0 1 , 1 04
suppositions of, 3 1-32
Rawls, John, 1 9, 29, 78, 97-98, 1 1 51 1 6, 1 50, 1 76

constructivist account of justice, 27
on contextualism, 92
contract theory of, 25
and impartial judgment, 48
Kantianism of, 28
on the original position, 28, 57, 92-93,
1 05 , 1 63
on overlapping consensus, 88, 93-95
Restatement, 92
on the right and the good, 1 53
social contract of, 27
Theory ofjwtice, 27-28
on truth, 26
Realism, 25
Reason, 2, 1 0-13, 34, 42, 1 34, 1 36, 1 39.
See also Practical reason, Rationality
architectonic of, 1 39
argumentative, 1 3
burdens of, 94
norm-testing, 1 5
passivity of, 2 7
poststructuralist critiques of, 1 49
procedural, 1 39, 1 4 1 , 1 45
public, 27
pure, 1 20
reflective, 3
theoretical, 30, 1 1 7
Reciprocal perspective-taking, 64, 1 54
Reciprocal recognition, 60, 67, 68
Reciprocity, communicative structures
of, 67
Recognition, relations of, 1 3 1
Redemption, religious, 133, 143
Reflection, 2 1-22, 1 24
Reflectiveness, 22
Reflexivity, conditi?ns of, 98
Relativism, 22, 76, 98, 1 00, 1 72
linguistic, 1 57
Religion, 47, 1 34, 136- 1 37
Respect, 1 5, 40, 66-68, 80, 1 35
and admiration, 45
and esteem, 45
reciprocal, 44
symmetrical, 6
Responsibility, 66, 153
Right, 6 1 , 1 1 9
Rightness, 29, 32
Rights, 40, 6 1 -63, 65, 68, 87, 1 1 9
Rilke, Rainer Maria, 74
Rorty, Richard, 96, 98, 1 24
Rousseau, Jean-Jacques, 1 2 1 , 153
Royce, Josiah, 1 1 3, 1 30
Rules, 3, 36
1 96
Index
Sade, Marquis de, 142

Sanctions, 4 1-42, 44
Sandel, Michael, 1 1 6
Sapir-Whorf hypothesis, 1 5 7
Schelling, Friedrich W. , l 36
Schiller, Friedrich, 34
Schlegel, Friedrich, 74
Schluchter, Wolfgang, 1 58, 1 63-165
Schmidt, Alfred, 1 33- 1 34, 1 38
Schnlidelbach, Herbert, 1 25
Schopenhauer, Arthur, 1 34, 1 38, 1 4 11 42
atheistic justification of morality of, 1 3 7
interpretation of compassion, 1 42
metaphysics of, 1 36
negativism of, 1 37
and philosophy of religion, 1 33
Science, 2 1-22 , 82, 1 37
Self, social constitution of, 46
Self-clarification, 1 1 , 1 2
Self-reflection, ethical, 23
Self-respect, 6, 40, 43-46, 87
Self-understanding, 4-5, 7, 12, 23, 26,
72
ethical, 2 1 , 24
existential, 4, 1 04
hermeneutic, 5
Selman, Robert, 1 3 1
Shame, 40, 47
Skepticism, 1 34-1 35, 1 44
postmetaphysical, 1 38
Smith, Adam, 1 66
Social point of view, 26
Social sanction, 43
Social sciences, 83
Social theory, 1 35-1 36, 1 76
Social world, 26, 39
Socialism, 1 39- 1 40
Socialization, 25, 33, 46, 67, 79, 1 1 4,
1 28, 1 30- 1 32 , 1 43
Society, 68, 1 48
contractualist concept of, 1 2 1 , 1 30
critical theory of, 1 35
Solidarity, 1 , 1 5, 50, 68, 76, 1 1 4, 1 341 35, 1 39, 1 42- 1 43, 154
Speech acts, 29, 1 59
Stalinism, 1 39
Steering media, 1 66- 1 67, 1 69
Stoics, 1 25
Strawson, Peter, 40, 83, 1 62
Subject, philosophy of the, 16
Subjectivism, 25
System, 1 66, 1 70
Taylor, Charles, 4, 1 9 , 70-75, 79, 1 1 6,
1 25
on the constitutive good, 73

on justice and identity, 69
on universalistic ethics, 9 1
Theology, 1 35- 1 37, 1 40
Torah, the, 1 43
Toulmin, Stephen, 1 56
Traditionalism, 1 5 1
Transcendentalists, 1 1 3
Truth, 97, 99, 1 40, 1 6 1
and coherence, 99
concept of, 29, 99, 1 39, 1 4 1
consensus theory of, 1 62
context-transcending sense of, 99
correspondence theory of, 26, 29
discourse theory of, 1 62
discursive redemption of, 29
ontological anchoring of, 1 4 1
problematic of, 1 38
theory of, 1 62, 1 65
unconditionality of, 1 40, 1 65
Tugendhat, Ernst, 19, 43, 45-48
Unconditionality, 1 4 1 , 145
Understanding, 5 , 10
Universal discourse, 1 2, 49, 53
Universal pragmatics, 50
Universalism, 35, 96
abstract, 97
boundless, 96
moral, 1 5, 84, 1 1 4-1 1 5 , 1 2 5
Universality, 8, 4 0 , 1 1 5, 1 53 , 1 56
Universalizability test, 7
Universalization, principle of, 1 3, 5 1 , 86,
1 29, 1 58- 1 59, 1 72
and ideal role taking, 52
formulation of, 37, 1 1 8
justification of, 1 63
and reason, 1 53- 1 54
Untranslatability, 1 02
Utilitarianism, 1-2, 1 06, 1 1 7
Validity, 14, 29, 38, 53. See also Validity
claims
categorical, 14, 34
deontological force of, 69
local, 1 00
moral, 63
sphere of, 9
strategic, 1 1
universal, 24-25
Validity claims, 30, 33, 50, 66, 8 1 , 89,
94, 1 20, 1 3 1 , 1 65 . See also Validity
assertoric, 38
context-transcending, 59, 145- 1 46
discursive redemption of, 29
generality of, 29, 34
197
Index
hypothetical stance toward, 1 2

and maxims, 1 73
and moral commands, 1 4
and moral judgments, 36
normative, 68, 1 59
and range of application, 36
and rationally motivated assent, 36
theory of, 29
and truth claims, 1 59
unconditional, 34
unconditionality of, 98
universalistic, 84, 1 46
universalization test of, 37
Volition, 2, 1 0
Walzer, Michael, 1 52
Spheres ofJwtice, 1 5 1
Weber, Max, 1 48, 1 55 , 1 58
Wellmer, Albrecht, 1 9 , 33, 36, 6 1 , 1 64
on affective preconditions of moral
argument, 34
on argumentation, 3 1
critique of moral universalism, 35
on moral obligations, 30
Whitman, Walt, 1 1 3
Whorf, Benjamin, 1 56- 1 5 7
Wiggins, David, 1 1 6, 1 29
Will, 3, 1 0- 1 6, 34, 6 1 , 79, 1 1 9, 1 37. See
also Will-formation
activity of, 27
of the agent, I 0
alien , 2, 1 6
autonomous, 1 0, 1 3- 1 4 , 33-34
autonomy of, 1 4, 42
catharsis of a, 34
common, 6 1 , 62
concept of, 9, 1 6
division of, 1 4
empirical, 1 4
free, 9- 1 0
good, 1 0
and guilt, 1 4
individual, 1 6
omnipotent, 1 0
positive, 6 1
principles of, 1 4
rational, I 0
weakness of, 87
Will-formation, 1 5, 1 50, 1 55- 1 56. See
also Will
collective, 1 6- 1 7
democratic, 1 68- 1 70
discursive, 69
individual, 1 6
political, 6 1 , 86, 88-89, 1 58- 1 59, 1 68

procedure of, 28
Williams, Bernard, 1 9 , 2 1-25, 57, 1 1 6,
1 23-124
Wittgenstein, Ludwig, 1 56
World, 8, 9 1 , 1 38, 1 64 . See also Form of
life, Lifeworld, Worldview
objective, 26, 38-39
Worldview, 60, 94, 97. See also Form of
life, Lifeworld, World
Studies in Contemporary German Social Thought
Thomas McCarthy, General Editor

Theodor W. Adorno, Against Epistemology: A Metacritique
Theodor W. Adorno, Hegel: Three Studies
Theodor W. Adorno, Prisms
Karl-Otto A pel, Understanding and Explanation: A Transcendental-Pragmatic
Perspective
Seyla Benhabib and Fred Dallmayr, editors, The Communicative Ethics Debate
Richard J. Bernstein, editor, Habermas and Modernity
Ernst Bloch, Natural Law and Human Dignity
Ernst Bloch, The Principle of Hope
Ernst Bloch, The Utopian Function of Art and Literature: Selected Essays
Hans Blumenberg, The Genesis of the Copernican World
Hans Blumenberg, The Legitimacy of the Modem Age
Hans Blumenberg, Work on Myth
Susan Buck-Morss, The Dialectics of Seeing: Walter Benjamin and the Arcades
Project
Jean Cohen and Andrew Arato, Civil Society and Political Theory
Helmut Dubiel, Theory and Politics: Studies in the Development of Critical
Theory
John Forester, editor, Critical Theory and Public Life
David Frisby, Fragments of Modernity: Theories of Modernity in the Work of
Simmel, Kracauer and Benjamin
Hans-Georg Gadamer, Philosophical Apprenticeships
Hans-Georg Gadamer, Reason in the Age of Science
Jiirgen Habermas, justification and Application: Remarks on Discourse Ethics

Jiirgen Habermas, On the Logic of the Social Sciences
Jiirgen Habermas, Moral Consciousness and Communicative Action
Jiirgen Habermas, The New Conservatism: Cultural CriticUm and the Historians'
Debate
Jiirgen Habermas, The Philosophical Discourse of Modernity: Twelve Lectures
Jiirgen Habermas, Philosophical-Political Profiles
Jiirgen Habermas, Postmetaphysical Thinking: Philosophical Essays
Jiirgen Habermas, The Structural Transformation of the Public Sphere: An
Inquiry into a Category of Bourgeois Society
Jiirgen Habermas, editor, Observations on " The Spiritual Situation of the Age"
Axel Honneth, The Critique of Power: Reflective Stages in a Critical Social
Theory
Axel Honneth and Hans Joas, editors, Communicative Action: Essays on
Jurgen Habermas's The Theory of Communicative Action
Axel Honneth, Thomas McCarthy, Claus Offe, and Albrecht Wellmer,
editors, Cultural-Polztical Interventions in the Unfinished Project of Enlightenment
Axel Honneth, Thomas McCarthy, Claus Offe, and Albrecht Wellmer,

editors, Philosophical Interventions in the Unfinished Project of Enlightenment
Max Horkheimer, Between Philosophy and Social Science: Selected Early
Writings
Hans Joas, G. H. Mead: A Contemporary Re-examination of His Thought

Reinhart Koselleck, Critique and Crisis: Enlightenment and the Pathogenesis of
Motkm Society
Reinhart Koselleck, Futures Past: On the Semantics of Historical Time

Harry Liebersohn, Fate and Utopia in G1!1'man Sociology, 1 887-1 923
Herbert Marcuse, Hegel's Ontology and the Theory of Historicity
Gil G. Noam and Thomas Wren, editors, The Moral Self: Building a Better
Paradigm
Guy Oakes, Weber and Rickert: Concept Formation in the Cultural Sciences
Claus Offe, Contradictions of the Welfare State
Claus Offe, Disorganized Capitalism: Contemporary Transformations of Work and
Politics
Helmut Peukert, Science, Action, and Fundamental Theology: Toward a Theology

of Communicative Action
Joachim Ritter, Hegel and the French Revolution: Essays on the Philosophy of
Right
Alfred Schmidt, History and Structure: An Essay on Hegelian-Marxist and
Structuralist Theories of History
Dennis Schmidt, The Ubiquity of the Finite: Hegel, Heidegger, and the
Entitlements of Philosophy
Carl Schmitt, The Crisis of Parliamentary Democracy
Carl Schmitt, Political Romanticism
Carl Schmitt, Political Theology: Four Chapters on the Concept of Sovereignty
Gary Smith, editor, On Walter B enjamin: Critical Essays and Recollections
Michael Theunissen, The Other: Studies in the Social Ontology of Husser[,
Heidegger, Sartre, and Buber
Ernst Tugendhat, Self-Consciousness and Self-Determination
Georgia Warnke, Justice and Interpretation
Mark Warren, Nietzsche and Political Thought
Albrecht Wellmer, The Persistence of Modernity: Essays on Aesthetics, Ethics and
Postmodernism
Thomas E. Wren, editor, The Moral Domain: Essays in the Ongoing Discussion
between Philosophy and the Social Sciences
Lambert Zuidervaart, Adorno 's Aesthetic Theory: The Redemption of Illusion

Habermas - Justification and Application

Uploaded by

Copyright:

Available Formats

Habermas - Justification and Application

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Habermas - Justification and Application

Uploaded by

Copyright:

Available Formats

Justification and Application

Remarks on Discourse Ethics

The MIT Press, Cambridge, Massachusetts, and London, England

Third printing, 2001

This edition 1 993 Massachusetts Institute of Technology. This English version

Ethics. I. Habermas, Jiirgen. Erlauterungen zur Diskursethik.

I On the Pragmatic, the Ethical, and the Moral

2 Remarks on Discourse Ethics

4 To Seek to Salvage an Unconditional Meaning

5 Morality, Society, and Ethics: An Interview with

Torben Hviid Nielsen

The discussions of the working group on legal theory that took

This book is a partial translation of Jiirgen Habermas's book Erliiu

Habermas's discourse theory of morality represents one of the most

While self-consciously Kantian in its cognitivism and its commitment

course ethics represents a sustained critique of the central role Kan

the center of the moral universe. 5 In light of this ineluctable historical

relation between meaning and standards of validity than is possible

rightness of norms, consensual interaction is disrupted and can be

(ii) For Habermas, as for Kant, the goal of moral theory is to

or practice demonstrably unavoidable for us, in the sense that we

with different employments of practical reason. 22 Pragmatic questions

In order to situate Habermas's approach within the context of con

contemporary moral debate, he suggests, is nothing but the debased

horizon of possibility, and hence that neo-Aristotelian appeals to

potentially damaging attacks on moral theories in the Kantian tra

of the individual deliberating subject. On this account, thinkers such

norm : on the one hand, it should be capable of commanding the

of how discourse ethics can provide concrete guidance for action,

evaluative commitments must be entered into voluntarily (in the sense

in such a stark fashion, he seems to lend substance to the view that

Notes on a Program of Philosophical Justification" (henceforth "DE") in Mural Con

3. Habermas's repeated criticisms of the "monological" character of the reflective

procedure enjoined by the categorical imperative-in contrast with the dialogical

providing essential empirical confirmation of his discourse theory of ethics. For a

of Habermas's formal pragmatics in relation to analytic theories of meaning, see

the Aristotelian Society 79 ( 1 978-1979), pp. 151-165.

28. Phronesis involves a kind of situational appreciation which Aristotle assimilates to

32. Cf. Gunther, "Impartial Application," p. 200.

free actualization of the personality of one individual depends on the actualization of

Contemporary discussions in practical philosophy draw, now as be

has met from its inception. Admittedly, it adopts a narrowly circum

Practical problems beset us in a variety of situations. They "have to

On the Employments of Practical Reason

we must make a career decision. Whether one travels to Scandinavia

In the case of complex decisions-for example, choosing a career

are interwoven in it: the descriptive component of the ontogenesis

status of means or limiting conditions for the realization of one's own

On the Employments of Practical Reason

which an individual customarily regulates his actions. They relieve

if my identity and my life project reflected a universally valid form

Thus, the question "What should I do?" takes on a pragmatic, an

On the Employments of Practical Reason

cation of an individual's self-understanding and clinical questions of

Pragmatic discourses in which we justify technical and strategic rec

serve to relate empirical knowledge to hypothetical goal determina

In ethical-existential discourses, reason and the will condition one

the individual participant's perspective without becoming discon

tive application, practical reason must be informed by a principle of

entrenched interest that is susceptible only of being pushed through.