Inquiry

ISSN: 0020-174X (Print) 1502-3923 (Online) Journal homepage: http://www.tandfonline.com/loi/sinq20

On Being Neither Post- Nor Anti-Kantian: A Reply

to Breazeale and Larmore Concerning The Fate of
Autonomy

Karl Ameriks

To cite this article: Karl Ameriks (2003) On Being Neither Post- Nor Anti-Kantian: A Reply
to Breazeale and Larmore Concerning The Fate of Autonomy , Inquiry, 46:2, 272-292, DOI:
10.1080/00201740310001236

To link to this article: http://dx.doi.org/10.1080/00201740310001236

Published online: 05 Nov 2010.

Submit your article to this journal

Article views: 69

View related articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=sinq20

Download by: [University of the Arts London] Date: 19 March 2017, At: 09:53
 Inquiry, 46, 272292

On Being Neither Post- Nor Anti-Kantian:

A Reply to Breazeale and Larmore
Concerning The Fate of Autonomy
Karl Ameriks
University of Notre Dame

Daniel Breazeale and Charles Larmore offer readings of my Kant and the
Fate of Autonomy that implicitly agree on several exegetical points but differ
strikingly in their basic systematic perspectives. They both contend that
historically Kant deserves to be linked more closely to his Idealist successors,
especially Fichte, than on my interpretation. Larmore makes this claim as part
of a plea that the whole Idealist tradition, so construed, should be rejected
because of a too subjective conception of autonomy. Breazeale, in contrast,
argues that appreciating closer connections between Kant and Fichte and
consciousness centered German Idealism in general is the best way to find
grounds for praising rather than deserting the Critical philosophy. Breazeale
and Larmore thus instantiate two long-standing traditions of reaction to Kant.
The first is attracted to him because of a supposed broadly Cartesian Idealist
connection, while the second is repelled by it whereas my aim is to explore
ways that Kant can be understood, and to some extent saved, precisely by
freeing his thought from the supposition common to both these traditions.

I
Before replying to specific points, it may help to review some of the most
basic themes of my treatment of Kant and his successors. On my reading, the
structural kernel of Kants Critical philosophy consists in four fundamental
features:
(1) A starting point in common experience (E);
(2) A transcendental derivation (TD) from this of various categories and pure
principles;
(3) An ultimate metaphysical account of all this in turn as making sense only
on the basis of transcendental idealism (TI); and, finally,
(4) A guiding idea that these first three steps are the essential theoretical
prerequisites for vindicating the ultimate goal of our autonomy (AUT), in

DOI 10.1080/00201740310001236 # 2003 Taylor & Francis

 On Being Neither Post- Nor Anti-Kantian 273

various key practical and methodological as well as theoretical senses. (In

schematic form: E only if TD, and this only if TI; and then, given E and TD,
AUT also only if TI.)

In most general terms, this means that I read Kant as offering a meta-
physics of experience, but one in which both key terms of this well-known
phrase need to be properly understood. First, Kantian experience is defined
not in terms of private so-called Cartesian representations, but instead
designates a cognitive situation occurring, roughly speaking, at a level no
lower than that of the core perceptual judgments of common sense. Second,
the Critical metaphysics, even while it criticizes its scholastic predecessors,
remains fundamentally rationalist in its epistemology, insofar as it requires
certain and pure principles for empirical knowledge that take us beyond a
completely naturalistic framework, and also in its transcendental idealist
ontology, insofar as this implies a kind of immaterialism, i.e., an insistence on
a non-spatio-temporal character for things in themselves.
Kants versions of rationalism and immaterialism are relatively modest
doctrines that can be overlooked because sometimes it is assumed that the
only options to empiricism and materialism are the more radical alternatives
of a dogmatic rationalism, i.e., an epistemology relying on an inflated
presumption that theoretical reason can positively determine substantive
features of things all on its own, entirely apart from sensible considerations,
and a spiritualistic immaterialism, i.e., an ontology that ascribes all sorts of
specific and independent powers to what is non-material.
For these reasons and others, including an explicit self-characterization of
his philosophy in these terms in one of Kants metaphysics lectures, I have
characterized his kind of rationalist system in general as a relatively moderate
or modest one (p. 37). By this term I do not at all mean to deny that Kant is
extremely interested in what would seem to us nowadays to be a much too
ambitious and elaborate system; obviously, like many traditional philoso-
phers, Kant is notorious for all sorts of bold claims to have solved
philosophys basic problems in a certain and complete way. My point has
been simply that, despite these claims, the system that Kant actually sets out
can be easily read as having several very significant limitations that
distinguish it as relatively modest in comparison to that of most of his major
predecessors and successors. Looking back, there are many obvious ways in
which Kants epistemology and ontology is not as radical as that of earlier
philosophers such as Aquinas, Descartes, Hume, Berkeley, Spinoza, and
Leibniz. Quite clearly, he does not, as they do, contrast our manifest image
with a system of alleged sure and detailed purely theological, mechanistic,
psychologistic, spiritualistic, monistic, or monadological theoretical claims
about ultimate reality. But, even more importantly for us, I believe Kant just
as surely does not make the kind of incredibly strong methodological and
274 Karl Ameriks

metaphysical claims found in his immediate Idealist successors: Reinhold,

Fichte, Schelling, and Hegel. He is of course responsible for suggesting to
them the very idea of a fundamentally autonomous rational system one that
would put autonomy at the heart of its notion of reality in general, as well as at
the ground of its view of philosophical methodology and of the role that
philosophy should play in culture in general. But there is an awful lot that
depends on exactly how the notion of autonomy is developed in detail and
once the details are attended to, I believe it can be shown that there are many
specific ways in which Kants conception of our autonomous rationality is
both more modest and more appealing than that of the later Idealists, even if it
no doubt has flaws of its own.
This modesty can be specified by comparing the nature of the first three
basic components in Kants system with that of his successors, that is to say,
its (1) basis (E), (2) its development (TD), and (3) its scope (TI). First, since
the Kantian basis is experience in the sense of putative empirical knowledge
Erfahrung it has what I would call a relative rather than an absolute
certainty, that is, it does not have the kind of certainty that a Cartesian skeptic
finds in the mere existence of representations. Second, since Kants
arguments from this basis to the specific categories and pure principles
depend on given claims that require pure forms of intuition and judgment
that are not themselves deduced from an absolute basis, such as the mere
notion of representation, the development of his system inherits the merely
relative certainty of its basis as well as a dependence on the irreducible posits
of the additional given presumptions just noted. Third, the scope of Kants
metaphysical claims is sharply limited by his doctrine of transcendental
idealism; what reality is specifically like in itself theoretical philosophy is
largely forbidden to say in any positive way, although one thing we
supposedly can assert with certainty is that it is not spatio-temporal as such.
I have argued that the base, development, and scope of later Idealist
systems differ strikingly from Kants: they all insist on a much less relative
basis, a much more tightly connected and unified development of principles,
and an in principle denial that the scope of their system leaves any major
features of reality undetermined. Thus Reinhold, Fichte, and Hegel each insist
on an absolute basis (that no rational being at all can consistently deny); they
each insist on a tightly structured deduction from this basis that does not allow
independent Kantian presuppositions about specific forms of intuition and
understanding, and they each develop a complete system that in one way or
the other leaves no sense, as Kants transcendental idealism does, in enter-
taining anything wholly beyond the spatio-temporal realm of experience.
Given the extremely strong resistance that has been offered to the
especially strong claims that these Idealists made about the base, devel-
opment, and scope of each of their systems, I take it to be obvious that Kants
more modest approach has at least an initial advantage. (Whereas I do not see
 On Being Neither Post- Nor Anti-Kantian 275

anything in Breazeales remarks that gives support to what I have called the
specific especially strong claims of the later idealists whatever other
virtues they may have on his account.) Hence, if they all got into trouble by
claiming too much, it seems worthwhile to explore the possibility that, in
going back to Kants less ambitious claims, we may save ourselves at least
some trouble without giving up on autonomy altogether. Moreover, this is
not at all to say that Kants own claims are modest to the point of triviality,
that they are so limited that they themselves say nothing very significant or
controversial.

II
It is remarkable how often the moderate but still significant nature of Kants
approach in the first two steps of his system tends to this day to be missed by
readers of the Critique of Pure Reason, and of my reconstruction of it, simply
because they presume that there are only two basic alternatives for the
Kantian transcendental arguer. These are: either to pursue a strongly pro-
gressive program, i.e., one that begins with a so-called Cartesian foundation
of pre-judgmental states such as sense data, unclear ideas, or what
Breazeale calls mere private representations in order then to ascend to a
proof that there is objectivity and an external world at all; or, in a direct
reversal, to pursue a strongly regressive program, i.e., one that simply
assumes the specific principles of the Newtonian science of Kants day and
then merely adds to this an abstract account of the faculties involved in this
knowledge and an explanation of how the ultimate metaphysical ground of
such science requires the doctrine of transcendental idealism.
I am driven to suppose that there is some kind of fatal attraction to think
simply in terms of these two extremes, for otherwise I cannot understand how
it is that so often when I write that Kant should be read as assuming
experience in a relatively thick rather than thin, or so-called Cartesian
sense, it still happens that some readers immediately assume that this must be
tantamount to saying that Kant is always already starting with something
much more than the mere fact that there is commonsense experience, i.e.,
putatively warranted judgments about distinct objects in a common spatio-
temporal domain. Despite using the term regressive ever since my initial
characterization of Kants basic argument structure, I have never maintained
that the regression in the Critique starts originally from any higher-level
strong and contentious premises such as the truth of objective a priori
principles or presumptions about any specific physical system. Nonetheless,
Breazeale claims, for example, that on my reconstruction Kants argument
must from the very start presuppose the objective validity of natural
science (p. 246), the reality of synthetic judgments a priori, and the
276 Karl Ameriks

necessity of universal and necessary objective judgments of experience (p.

246). In fact, however, all that is presupposed with my view of Kants
commonsense starting point is that there is some objectivity to our
experience, that we have something that allows us to provide some sort
of justification for saying that some of our states are true rather than false. (I
point out that this informal use of common sense, or common knowledge,
is not meant to be equivalent to any sympathy for the methodology of the
popular philosophy of Kants day or similar movements; the common or
ordinary is an admittedly vague notion, but starting from it does not imply
abdicating systematic philosophys distinctive reflective duties and critical
capabilities.)
Note that this is obviously still a long way from even trying to argue as
Kant does later that we need specific pure principles (going far beyond
mere common sense) to fill out and give a substantive, informative backing to
our common presumption of objectivity, to argue, rather than presuppose, for
example, that the common notion that there are objective and not merely
subjective time sequences ultimately does require an appeal to universal
principles such as a general law of causality. To be sure, the Transcendental
Analytics arguments for such principles are not entirely disconnected from
Kants interest in Newtonianism, but they do not themselves presuppose any
specific scientific principles and are instead explicitly aimed at defending
various much more general principles that any developed physics of spatio-
temporal experience would have to use, be it like Newtons or not. And this
means that there is after all, right in the overt structure of the Critique, a
distinct and highly relevant moderate alternative to the strongly progressive
and the strongly regressive approaches.
Note also that this moderate alternative is consistent with my view that
Kant is not aiming to defeat radical skepticism as such. Such an interpretation
is often met with incredulity by those who suppose, as Breazeale may (p.
246), that the Critique must be concerned with an answer to radical Humean
skepticism. This supposition fails to note that one can as I believe Kant
obviously does bracket radical Humean doubts about whether there are
sensible objects at all, or even any substantive validity to pure mathematics,
while still working out a position that is very much a response to worries
specifically about the necessary objective claims of our higher faculties and of
pure reason which is after all the topic of the title of Kants book. The
moderate regressive interpretation still leaves plenty of room to see the
Critique as trying to respond to numerous higher level skeptical worries,
generated by Hume among others, about what pure thought can establish;
these worries are precisely what is addressed by the Transcendental
Deduction and the arguments for specific pure principles in the second or
main development part of Kants system. (Here it is, of course, also possible
to note, as some have, following on Maimons lead, that even once Kants
 On Being Neither Post- Nor Anti-Kantian 277

system is charitably understood in these sections, a host of serious skeptical

problems may remain concerning how it can be applied to particular cases
and how it can account for itself in general.1)
Sometimes the tendency to overlook Kants moderate regressive approach
may arise from a more general presumption that if philosophers are not
making a strong theoretical argument, something that would directly defeat
skepticism from an extremely thin or so-called Cartesian foundation, then
their whole project is not very interesting anyway, not genuinely foundational
or idealistic in a good sense, and so it doesnt matter much what else is going
on. The desire to defeat skepticism in this way from the ground up is
remarkably widespread; it recurs not only in Breazeales pro-Fichtean
remarks but also in broad strands of both English and German philosophy,
and in our own time it has been reinforced by influential reconstructions such
as Strawsons and Rortys.2 Whatever its immediate sources, I have tried to
rebut the presumptions of the strong reading in a number of ways. In addition
to repeating that a moderately regressive form of argument clearly matches
the headings of Kants own Transcendental Deduction, which explicitly
center on establishing not objectivity as such but rather the objectivity of the
categories, given that there is experience (see 26), I have tried to explain
how there are a number of understandable reasons why readers could
nonetheless have been confused on this basic point.
The sources of confusion here are at once deeply textual, systematic, and
historical. One basic confusion has to do with idea, central also to
Breazeales interpretation, that Kant is starting simply from private
representations (p. 245). There is no doubt that Kant speaks about
representation as basic in some sense, simply as the most general term for
all components of mental life (A320/B376),3 and that his extremely
influential immediate successor Reinhold made this notion the explicit and
supposedly sufficient foundation for his best known and most ambitious
system (the theory of the Vorstellungsvermogen). The term also plays a
central and very different role in some of Kants expressions of
transcendental idealism, as when he says that all we know are representations
but here the term is meant to contrast in a global way with thing in
themselves and does not mean anything like individual Cartesian sense
data. Unfortunately, interpreters from the eighteenth through the twentieth
century have often conflated Kants transcendental use of the term
representation, i.e., its use as an abbreviation for what the whole sphere
of our theoretical knowledge can reach, with an empirical use, i.e., as
standing for a set of primitive psychological acts.
When Kant himself attempts to characterize the popular notion of
representation (A320/B376) he makes quite clear that it is such a rudimentary
term, possibly signifying something as simple as either a bare sensation or
idea on its own, that it is immediately evident from the most basic and best-
278 Karl Ameriks

known principles of his epistemology (intuitions without concepts are blind,

A51/B75, and the only use which the understanding can make of these
concepts is to judge by them, A68/B93) that, strictly speaking, representation
by itself cannot mean a genuine cognitive state. For Kant a state can become a
cognition, something that can be justified or unjustified, true or false, only
once it has passed beyond the situation of being a mere psychological
representation to that of being a cognitive aspect of something that has the
logical form of a judgment, i.e., the complex and distinctive epistemic
synthesis that Kant believes cannot be found in the many other species of
representing beings that are lower than us.
All this is a reason, I believe, for also not saying simply that Kants starting
point is mere self-consciousness or even the unity of our self-conscious-
ness. The problem here is that there are many non-cognitive ways of taking
these terms, for example, as designating a collection of mere states of passive
inner sense, or reflexivity, or, in English, of indeterminate social unease, or, in
German, an attitude of immediate confidence, and in none of these cases does
there have to be any judgment or claim that such and such is the case. From
such a meagre basis, Kant neither can nor does make a transcendental
argument. On the other hand, there surely can exist thick states of self-
consciousness that happen to include a specific kind of cognitive
representation, a claiming to know that something is true about ones self
and from such states Kant can and does go on to make arguments, for
example, that this may require categories or pure principles. Note that in that
event, however, the focus specifically on the self does not do the essential
work; the conclusions can follow simply from the general conditions of the
cognitive unity of an act.
Nonetheless, it might be asked, what about the indisputable fact that Kant
does frequently discuss self-consciousness, especially near the beginning of
the Deduction; doesnt that mean that some kind of Cartesian basis is present
after all? I would argue in just an opposite way that Kants discussion of
self-consciousness there is most naturally understood as rather a key step
precisely in his anti-Cartesian strategy, which is strongly emphasized
throughout the B edition (think of the wholesale revision of the Paralogisms),
and which consists in arguing that self-knowledge is not privileged but is
subject to the same basic structures as knowledge in general (see B 160, thus
the unity of the synthesis of the manifold, without or within us). What the B
Deduction is repeatedly saying unlike what Descartes, and Leibniz, and
even the pre-Critical Kant argue is that even the self really knows (as
opposed to merely senses or intuits or has a general concept of) itself only via
the general rules of synthesis that govern all experience, all putative
knowledge claims.4
The crucial feature of Kants starting point of experience is that it is not a
completely inner or Cartesian episode but rather a state that makes some kind
 On Being Neither Post- Nor Anti-Kantian 279

of genuine cognitive claim. This is what constitutes the bare meaning of the
experience at the Kantian basis, in contrast to the buzzing confusion of
mere impressions or ideas that other philosophers start from. The mere
meaning of a thick representation that is something for me, and hence can
be part of an apperceptive experience, is not as specific as a claim about
physical or spatial or non-solipsistic objects, although instances of such
representations might well still involve such claims. But, whether or not they
do, they are still quite distinguishable from the mere existence of a fuzzy
Cartesian image or a bare Humean idea, or any merely clarified or associated
concatenation of these. That we also in fact (supposedly) must have
knowledge of outer spatial things, and never of merely pure inner things, is
something Kant is eventually very concerned with, but the key argument for
this in the Critique does not come until long after the Deduction, in the B
edition Refutation of Idealism. Confusion may arise here simply because,
even though neither of the mere terms representation and apperception
means specifically spatio-temporal putative knowledge, Kant can and does go
on eventually to argue that we have no actual states of apperception or even
representation, that we can examine that do not already bring with them this
kind of knowledge in some way.
I turn now to some points about how these features of the first steps of
Kants system are related to his third step, his transcendental idealism, and to
the immediate reception of his work that I have summarized under the
heading of the short argument to idealism. The lack of an absolute Cartesian
basis for the Kantian system is precisely what first came to bother Kants
immediate successor, Reinhold. Reinholds Cartesian move to found
philosophy explicitly on a principle of representation (or Satz des
Bewusstseins) was extremely influential, but, as I note in my book, important
historical work has shown that there was a significant alternative to
Reinholds highly influential Cartesian project that was developed even in
his own time. Followers of Kant in the Jena Herbert/Niethammer circle
explicitly demanded a return to the standpoint of experience, which they
directly linked to a commonsense hypothetical orientation, one that explores
what happens if we allow, as we in fact do, that there is experience in Kants
thick sense.5 In view of this sensible alternative, I have described Reinholds
original Cartesian demands as resulting not from an irreversible insight but
from an understandable but mistaken extreme conception of autonomy and its
cultural prerequisites: Reinhold was so interested in preserving the Critical
philosophy as a secure foundation for the Enlightenment in general, that he
wanted its effectiveness guaranteed by an irreversible basis that could be
found in the simplest circumstance imaginable, namely, the having of a
representation at all. When Reinholds program encountered problems, the
history of philosophy entered a moment when in both its interpretative and
its systematic work it might have (but unfortunately did not, in its main line)
280 Karl Ameriks

turned back away from methodological Cartesianism in order to reconsider

the more moderate Kantian manner of defending a rationalist philosophy of
autonomy. Instead, the most influential idealist philosophers went on to
contend that there were other and better ways than Reinholds to try to realize
imperial rather than not merely moderate autonomous aspirations of reason.
A crucial and very understandable tactic along this line was Fichtes shift to
speaking of apperception not as a fact of experience or as a Tatsache des
Bewusstseins, but as a Tathandlung, a unique absolutely free act whereby I
recognize myself as responsible for taking my representations as mine.
Here a kind of self-determination is worked right into the content of the
first truth of the rationalist system. Put extremely briefly, the weakness of
Fichtes approach, from a Kantian perspective, is that it seems much too
dogmatically to assert an absolute freedom of consciousness that Kant had
taken great pains to show that we cannot ever theoretically warrant for how
can we ever with certainty exclude hidden external causes? This is an issue
that Kant was very familiar with, since until shortly before the Critique he
himself was inclined to affirm the old rationalist thesis that the mere
representation of the I does prove our absolute spontaneity and this is why it
is so striking that the Critique itself dramatically backs off from any such
proof.6 My own hypothesis is that Fichte himself came to appreciate this
problem over time, and that this is precisely why, in very perceptive works
such as the Vocation of Man, he was willing to stress that it is not as a
theoretical insight but fundamentally as a pure practical voice of conscience
that as Breazeale also says we first come into supposedly certain contact
with our absolute freedom. Without a moral perspective, we might still worry
that, in the realm of theoretical possibilities not refutable from a traditional
Cartesian basis, no matter how much it seems that ones thoughts literally
issue from the absolute activity of ones I, there might ultimately be
neither any such activity nor any such I (in this way the remarkable beginning
of the third part of the Vocation of Man constitutes nothing less than a
dramatic deconstruction of the classical cogito). From the perspective of
traditional theory alone there could be, as Hume and Jacobi suggested earlier,
nothing but a stream of representations passively flowing along.
Consistently enough, to hold on to a strong sense of autonomy, Fichte then
went on as Breazeale reminds us to argue that all the content that was
traditionally wanted from theoretical philosophy could be derived after all
once the demands of the conscientious I are granted as legitimate. To accept
that there is a genuine moral I, an I that has genuine oughts and therefore
genuine cans, is at the same time to accept after all the external world that
seems needed to fulfill that I and also, Fichte contends in his Divine
Government essay, to reject as meaningless anything beyond the world so
disclosed. In this way, the base, development, and scope of the Fichtean
system might seem to construct a fully autonomous home for our rationality.
 On Being Neither Post- Nor Anti-Kantian 281

But precisely in calling for an absolute basis while turning ultimately to a

practical ground for the claim of our activity, Fichte still leaves himself open
to deep problems known since Hobbes and Hume and not only the Critical
Kant. (For this reason to anticipate an objection of Larmores there is a
sense in which the problem of dualism seems to me more of a difficulty for
the post-Kantians than for Kant.) Rather than truly unifying all knowledge
and philosophy and life, Fichte introduces claims of absolute freedom that
threaten to sunder philosophy in two; we are called on at first to ignore the
apparent ultimacy of the determinist empirical patterns of ordinary life,
modern science, and philosophical arguments such as Kants Second Analogy
which I take to be the core of his theoretical philosophy, the core that it still
seems (even after Breazeales defense) that Fichte neglects. Sensing this
problem from the very beginning, the third step of Kants system built in a
metaphysics of transcendental idealism that would preserve the crucial claim
of autonomy in its fourth and final step by creating some theoretical room for
at least the possibility of the moralists assertion of our absolute freedom
(which is only one part of Kants notion of autonomy, but an indispensable
one).
It is true that Kants arguments here are highly controversial, but it is very
significant that he at least appreciated the kind of difficult argument that
would be needed to reconcile our conflicting modern longings for freedom as
well as theoretical and scientific responsibility. Rather than shore up anything
like these arguments, Fichtes position, following Jacobi, appears too quickly
to assume that Kants own idealist metaphysics is not only lacking in
evidence but is straightforwardly contradictory, and so he seems to leave no
room for any argument that would give any purely theoretical warrant for the
space of possibility needed to preserve the rationality of an insistence on
absolute freedom. The weakness of this Fichtean approach is that, on its own
admission, it claims autonomy in behalf of a libertarian-moralistic absolutism
that can fly in the face of all our merely theoretical impressions. For this
reason, even though Fichtes system was clearly constructed in a sincere
attempt to deliver the most rigorous form of autonomy, in the end it can look
like a presumption of self-determination that evades complete rationality; the
determinist voice within the self is squelched all too abruptly.
To appreciate the significance of this point, it may help to return to some
basic issues that come up in Breazeales remarks. In contrast to my reading,
Breazeale holds not only that Fichte is not to be criticized here, but also that I
have misread Kants position on freedom. Breazeale reads a crucial passage I
refer to in the B Preface (Bxxix)7 as saying merely that Kant thinks that his
Critical philosophy is the only way to avoid a contradiction between freedom-
oriented Sittenlehre (doctrine of ethics) and determinism-oriented Naturlehre
(doctrine of nature). On this reading, theoretical philosophy and practical
philosophy are in a balanced, mutually canceling situation, and our actual,
282 Karl Ameriks

practical belief in freedom stands in no need of any theoretical justification or

propaedeutic whatsoever (p. 247). In contrast, I see asymmetry rather than
balance in this text, for what Kant clearly goes on to say here is that without
his Critical philosophy, morality would have to yield to the mechanism of
nature (Bxxix). Moreover, on my reading Kant is saying that this is a
condition not merely of philosophy about morality but of morality itself,
because he says that it is a condition not only of Sittenlehre but also of what
he repeatedly calls morality itself, die Moral. Surely it is Kants view that it
is not merely philosophers but rather ordinary rational persons as such who
presume that morality requires genuine freedom. Hence it is no wonder that,
given Kants view that nothing other than the standpoint of his Critical
metaphysics can allow this freedom to be thought consistently, he insists that
we must have previously established our unavoidable ignorance of things in
themselves and limited all that we can theoretically know to mere
appearances (Bxxix). We do after all need a theoretical argument for
transcendental idealism first, and thus a first Critique of theoretical reason as a
propaedeutic to defend the very possibility of the moral life systematized in
the second Critique.
Breazeale suggests that Kant believes this point is relevant only to
philosophers, not ordinary persons, since their actual belief in freedom will
continue in any case. This may or may not be true at a psychological level, but
I take it to be clear from Kants remarks that he thinks that anybody now
relying simply on that belief would not be proceeding rationally (even if the
freedom they believe in is truly there). For better or worse, Kant clearly
presumed that by his era any rational person would have to be prepared to
acknowledge a mechanistic understanding of all empirical nature, and so any
such person could properly hold on to an attachment to freedom only by
following Kants lead in restricting the determinism of mechanism to a
doctrine of appearances, a doctrine defined by a contrast with the place left
open for absolute freedom by and only by the Critiques theoretical
argument for things in themselves that are not to be identified with
appearances. Hence, even though I certainly agree with Breazeale that it is
often very important to keep distinct what may be demanded at the immediate
and the reflective levels of experience, in the end the main question in this
case cannot be simply divided into two separate issues, what ordinary people
may assert and what philosophers may assert. The crucial issue is what
anyone can any longer say rationally, and here Kants own position appears
clear enough: without his specific doctrine of transcendental idealism, there is
no longer any responsible way to hold on to asserting absolute freedom and
the morality that presupposes it.
For this reason alone, ones understanding of transcendental idealism is
crucial to interpreting Kants project and the innovations of his successors.
On this issue as well there are deep differences between my interpretation and
 On Being Neither Post- Nor Anti-Kantian 283

the position Breazeale reiterates. Put most simply, I see Kants idealism as
essentially defined by his long and complex treatment of our specific forms of
intuition: space and time. Kant argues that space and time must be ideal, and
then he argues that, since all our specific theoretical knowledge is determined
by space and time, all that knowledge must also be ideal. Understanding the
essential relation between ideality and space and time is crucial for
understanding not only the logic of Kants argument but also the content of
its conclusion: since ideality follows not from representation or sense or
thought as such, but only from the specific features of our forms of intuition,
there is a realm of the non-spatio-temporal left over that Kant can and does
exploit as meaningful in contexts that go beyond specific theoretical
determinations. In contrast, if one sets out what I have called a short
argument to idealism (because of a remark by Reinhold about how his
argument takes a shorter way) that abstracts from these specific forms, then
one will have not only a method that directly contrasts with Kants own texts
but also a content for idealism that is not restricted by claims about the forms
of space and time and what follows from our reliance on them.
Following this shorter way, one can end up with an idealism of the kind
that Breazeale finds in Kants immediate followers, i.e., one that I take it is
no longer Kants own transcendental idealism, insofar as this is an ontological
thesis rooted in specific claims about space and time, but is instead, in
Breazeales words, simply a general project of proceeding without
appealing [in order to ground experience] to anything transcending con-
sciousness (p. 248). In seeing this generic Cartesian project as at the heart of
the post-Kantian enterprise, and in agreeing that it differs from the space-time
specific argument in Kant, Breazeale, despite all else, seems to be agreeing at
the interpretive level with one of my most basic points. At the systematic
level, however, he still finds himself very attracted to the Cartesian project,
whereas I see it is as a questionable detour that needs to be diagnosed rather
than propped up.
Breazeale also remains unconvinced by one of my key interpretative
claims, namely that Reinhold was chiefly responsible for a Cartesian turn
away from Kant because of his very influential and unfortunate attachment to
an extreme ideal of autonomy, an absolute demand for a philosophy that
would have a wholly self-evident foundation. Breazeale disputes my claim
that Reinhold ever held the thesis that a strict form of philosophical autonomy
is a condition of autonomy in ordinary life. Here I can only reiterate that I give
several references to Reinholds having actually linked moral and political
autonomy to prior autonomy established at a philosophical level, and the
historical details offered about the social orientation of Reinholds early
writings are meant to be one of the main points of my study.8 We might agree
that the tight linkage Reinhold insists on can sound absurd, but for me it at
least fits in very smoothly with Reinholds extensive popular enlightenment
284 Karl Ameriks

activities, his extraordinarily strong demands for a self-determining system

within philosophy, and his clear belief that properly understood philosophy is
not an ancillary discipline but has become a necessary basis for all truly
rational life.
Reinholds philosophical enthusiasm is important, I believe, because it was
carried over into many influential features of philosophies such as Fichtes,
which also insist on a strongly unified systematic philosophy and a basis in a
very strong claim about freedom that is not clearly ontologically grounded.
On Fichte, I do agree at the interpretive level with Breazeales main points
that no theoretical argument could establish the truth of Fichtes first
principle (p. 253), and that this is consistent with allowing that there is a
theoretical series of arguments worked out on its own within Fichtes system,
once the ultimate practical basis is set out. I may also agree with Breazeales
proposal that Fichtes system is most defensible (if defensible at all) when
understood as a kind of experiment, as the thought that if we start with a
practical posit of our absolute freedom and can derive all we want
theoretically after all, as conditions of this very freedom, then in the end
this might give us enough to feel entitled to our original starting point. This
intriguing interpretive proposal reminds me very much of the main tactic of
Reinholds second Jena system (and a similar idea is picked up in Larmores
remarks), and it certainly sounds less foundational and Cartesian than the
Fichte I was criticizing. However, like the work of the later Reinhold, this is
also an approach that at the time had much less main line influence (Hegel
despised it) and, as far as I can tell, it has left us with hardly any specific
arguments taken up directly by well-known philosophers.
Instead of pursuing that proposal further here, I will conclude this section
by simply reviewing my response to the three main claims Breazeale made
against my interpretation of Kant (pp. 245246). His first claim was that Kant
should be read as beginning with subjective representations. Whereas
Breazeale gives no extensive textual support for this claim, I have argued
against it (while also acknowledging and giving an explanation for its earlier
widespread influence) not only in my books but also in detailed studies
elsewhere.9 This is a matter of Kant scholarship that seems to be one of the
few positions shared by myself and such different interpreters as Paul Guyer,
Henry Allison, and Michael Friedman. Breazeale claimed, second, that Kant
should not be read as assuming the judgments of common sense or specific
objective sciences. But he does not distinguish these two very different levels,
and takes my interpretation to presume the second, higher level, whereas it
really presupposes only the first, much lower level. Breazeale claimed, third,
that Kant must be read as responding to Humean skepticism, but here he does
not distinguish two very different issues, radical skepticism about objectivity
at all, and a mere skepticism of reason about universal and necessary
judgments. The point of my interpretation is to stress how Kants main
 On Being Neither Post- Nor Anti-Kantian 285

arguments are a response to the second, not the first kind of skepticism. Once
all these distinctions are made, I believe the core of the original Kantian
project remains clearly more viable than the very different kinds of
interesting, but overly ambitious, projects of the post-Kantians.

III
Charles Larmore elegantly recapitulates the main points of my work and then
forcefully draws attention to some very difficult problems that I can only
begin to address here. Larmores main challenge comes from his belief that
Kants notion of autonomy remains too subjective to do justice to our
fundamentally realistic experience of value. Ironically, my own still largely
undeveloped views on value overlap considerably with Larmores realism.
A background motivation of my book was precisely to make room for a view
of Kant that would reveal how genuine autonomy can be consistent with a
sensible form of value realism so that we are not forced to make a choice
between the two notions, as Larmore and many others have presumed. It is for
this reason that I still aim to find a place for both Larmore and myself within a
camp of fans of autonomy but only once the term is liberated from the
subjectivistic distortions that popular (mis)readings of Kant have imposed
upon it.
Before focusing on this fundamental problem of practical philosophy,
some other issues should be clarified first. Larmore suggests that I may not
have done justice to the concern for unity that can be found in Kants own
system, and that underlies the therefore excusable attacks on Kantian
dualism that motivate much of post-Kantianism. The passage he cites to
bolster his case has a long history of influence, but I believe it points in quite
another direction than his remark implies. Kant does mention the idea that
sensibility and understanding spring from a common, but to us unknown
root (A 15/ B 29; p. 264), but whereas Larmore says that the idea here is that
there must be such a root, Kants statement is only that perhaps (vielleicht)
there is one. Moreover, ever since Dieter Henrichs work,10 there have been
good reasons for believing contra Heidegger and others that the main
point of the passage is precisely to indicate that the root remains unknown to
us. Rationalist and empiricist reductions of sensibility to understanding or
vice versa are all left behind with the Critical turn, and with good reason. All
sorts of underlying roots can still be imagined, of course, and one might
even think that God is there as the ultimate source of unity but for the
Kantian all this remains nothing that should be straightforwardly asserted or
assumed at the outset.
Similarly, Larmore suggests that the post-Kantians were correct to say
Kant did not have his own key dualisms under firm control (p. 264), and he
286 Karl Ameriks

reiterates Jacobis charge that the categories of the understanding, especially

causality, cannot consistently concern both things in themselves and also
action upon our sensibility. But this influential charge cannot be sustained.
Nothing in Kants distinction of the realms of understanding and sensibility,
or (a different distinction) of things in themselves and appearances rules out
the possibility of terms holding in common for both realms (for example, each
may be considered singular, or plural), and even our having true and basic
beliefs about them. Kants Critical turn inhibits only claims of theoretical
knowledge here. Therefore, although Larmore is certainly right to note that
Niethammer and others believed Kants system was in trouble and some kind
of new systematic method should still be sought for achieving its aims, I have
yet to see any significant inconsistencies that they truly perceived in the
Critique.
A milder objection that Larmore offers is that it is wrong to for me to
suggest that the original post-Kantian project was driven simply by a quest for
certainty. But although this quest may have been the first point that I stressed,
I did not mean it to be the main one; on the contrary, I thought it very
important also to stress the goal that the idealists were striving after, the
project of achieving and securing autonomy. According to my reading, it just
happened that for a while this goal seemed to some to require a certain base.
A related consideration that Larmore raises is that my reading does not say
enough about Kants own thought that reason has a natural tendency to go
beyond experience and, in particular, to seek the Unconditioned. This is
definitely an important theme about which more could have been said. But I
do not read the passage that Larmore cites about reason demanding a system
on the basis of one principle (A645/B673) as in tension with my
interpretation. Among other things, there is evidence that this principle is
nothing other than the notion of judgment and its forms, which guides all of
Kants Critical investigations (see Axx, on the common principle). A
passage that Larmore himself cites says reason must itself show the way
with principles of judgment (Bxiii), and this perfectly fits the passage I cite in
which Kant calls for a modest system.11 Moreover, I believe this issue is not
exactly the same one that Kant has in view when he speaks in the Dialectic
about the demand of reason (A332/ B389). Kant is perfectly willing to
stress that in some contexts there are ambitious demands of Reason in a
capital R sense that a Critical thinker should precisely not expect to find
satisfied (just as in the search for one root of our faculties) and should regard
as the unrealistic demands of an abstract faculty. Matters are complicated by
the fact that sometimes Kant speaks generically about how we should follow
reasons own devising (Bxiii), but in that case there is no ground to believe
he does or should mean Reason in the strong Idealist sense that contrasts
sharply with the (pure) Understanding. A more complicated consideration
that Larmore raises concerns the issue of considering theoretical reason to be
 On Being Neither Post- Nor Anti-Kantian 287

self-legislative. Larmore argues that within his own Critical system (by his
own lights, p. 268), Kant should have allowed absolute theoretical, and not
only practical, claims of self-legislation, since in its theoretical activity too,
the mind is bound by categorical norms by the rules of logic, for example
(p. 268), and, supposedly, the only possible Kantian account of categorical
norms must be in terms of self-legislation. Larmore takes such an account to
be absurd, on the ground that ultimate norms cannot ever be explained as a
matter of our creativity but Larmore thinks Kant had no other internally
consistent alternative to pursue, despite the reservations that the second
Critiques doctrine of the fact of reason expressed about proving our
freedom theoretically.
There are a number of different issues to be sorted out here. I also
emphasize that ultimately Kant restricts himself to practical and not
theoretical arguments for our absolute activity; and, like Larmore, I see
this as an unstable position that calls out for attempts at a stronger backing.
However, I believe that this restriction has to do with whether we can have
theoretical certainty about being absolutely free in individual acts in an
efficient causal sense, and this issue should be distinguished from that of
justifiably claiming that reason as a formal justificatory ground can be in some
sense self-legislative.12 (Readers can miss this point, since Kant thinks that
we in fact have and need both efficient and formal self-determination, and so
he does not always clearly stress the distinction between these two
components of the selfs absolute activity.) One might question efficient
self-determination while not at the same time disputing formal self-
determination (or vice-versa). Larmore seems to believe that self-legislation
always must mean something like mere imposition, as in arbitrary reflexive
and efficient causality, a matter of making something authoritative merely by
taking it to be authoritative. But from my perspective this is precisely not the
way to understand genuine Kantian self-legislation in its formal sense. Kant
repeatedly distinguishes the concepts of legislative and executive determina-
tion,13 making it obvious that the self that is acting in pure legislation is
nothing like the arbitrary imposing individual who is the villain of countless
contemporary tracts against autonomy worship.14 Kant thus can regard the
principle of contradiction, for example (or the table of the forms of
judgment), as a categorical principle of transcendental theoretical reason
without in any way implying that its validity as opposed to its application
is a matter of any of our empirical impositions.
In other words, I agree with Larmore that if autonomy means anything it
means self-legislation, (p. 269), but I do not believe that for Kant any of these
terms has to be understood as a matter of giving principles validity simply by
imposing them on ourselves in any ordinary sense as an efficient cause.
Larmores way of speaking about autonomy and I would certainly grant
that, given the very history of post-Kantian influence that I have charted, this
288 Karl Ameriks

has become its most popular meaning uses paradigms such as voting
procedures, which are clearly settled by individuals working together in
ordinary empirical self-legislative ways. And his critical point is that this is
an activity that takes place against a background of reasons that we must
antecedently recognize (p. 269). Suppose we grant this point. What follows, I
think, is simply that mere efficient self-legislation cannot account for
categorical norms fully or all the way down. This still leaves open the idea
that there might be a kind of self-legislation in a formal sense that is a source
of such norms and it is precisely this kind of autonomy that I believe is
Kants prime concern (even if he is also very interested in autonomy in many
of its empirical, e.g., political, dimensions). Just as in general with Kant we
must distinguish the transcendental work of the necessary forms of
judgmental representation (which determines the necessary conditions for
all particular empirical cognitive claims) from an individual selfs empirical
theoretical acts, so too in Kantian practical philosophy we must distinguish
the legislative work of pure practical reason, and the necessary rules it reveals
for proper willing, from questions about particular acts of will. Larmore
contends that the most one might get along this line is self-governance,
which is a matter of seeing and doing what is right regardless of the threats
and rewards that others might offer (p. 270). But, as he also notes, I have
proposed that we understand self-legislation for Kant as a matter of following
the sheer rationality in our essential nature. I take this to be a positive
notion that goes beyond mere self-governance, since one might be self-
governed, as defined so far, and simply go off on a whim.15 The first crucial
point about Kants notion of reason as self-legislative is that it involves laws
in a strong sense. His moral rules are not simply generalized intentions but
principles whose generalization is supported by reason in general (e.g., the
principle respect rational agency has an inherent positive ground that makes
it unlike the general but pointless maxim dont anybody whistle while you
eat). This makes them self-legislated in the strong sense that that for each
rational being they are from and for its own reason as such, and not merely in
fact in accord with it.
Note that the self here is obviously tied to the nature or structure of
reason itself. It is not a self in any mere empirical or psychological or physical
sense, although such an empirical self might live in actual accord with the
laws of pure practical reason, just as individual physical objects may act in
accord with pure concepts that have a meaning that transcends merely
empirical contexts. The self-legislation of reason is literally reasons
determination of itself; it is what the idea of reason determines should be
done in order to realize reason. This might sound like a mysterious
hypostatization, but it need not be understood in any mystical way because
the determining here, although it is reflexive, is originally formal rather than
efficient. The claim is not that reason literally does something, as a person
 On Being Neither Post- Nor Anti-Kantian 289

might cross a street, but simply that the nature of reason itself primarily
determines (as a matter of essence) what is for reason (after all, wouldnt it be
the most appropriate determinant for such a thing?). In theoretical contexts,
this is no more mysterious than saying that consistency is required by reason
itself which of course leaves open the empirical issue of whether particular
empirical beings will assert consistent sentences. In practical contexts, Kants
notion does get more controversial because it is the thought that pure practical
reason can generate norms ultimately out of itself, and also that these are
norms that otherwise rational beings can understandably disobey. This notion
might be a non-starter for a skeptic or radical empiricist, but it should be at
least a possibility for Larmore, who allows the notion of categorical practical
norms.
Another way to put this interpretation is to say that it takes the autonomy
issue to be literally what Kant says it fundamentally is, namely a matter of the
autonomy of reason; we might not immediately catch this simply because he
calls it Wille (IV, p. 433), but for Kant this just is equivalent to pure practical
reason. This is nothing like the willful autonomy of me in the me
generation sense of our times (a derogatory sense with predecessors even in
Greece, where the Sophoclean chorus chastized Antigone for being
autonomous16). At the same time, Kants libertarianism and other features
of his thought insure that, unlike other universalistic theories such as
utilitarianism, his theory does not ignore the common belief in the ultimate
reality and significance of the individual human agent. Just as in his
discussion of the transcendental features of knowledge, there is no reason to
presume that Kants discussion of pure practical reason does not apply to,
express, and fulfill concrete individuals, even if it is no way built up from an
investigation of their mere natural and empirical peculiarities.
A major stumbling block for many interpreters here has no doubt been
Kants talk in the Groundwork about a moral being as subject only to a law
given by himself (IV, p. 432). But Kant immediately makes clear that the self
of this being is generic and its lawgiving must be universal in a
transcendental rather than merely empirical sense, for it has to do with the
reference of all action by which alone a kingdom of ends is possible (IV, p.
434). There is nothing in the notion of this kingdom that forces a contingent
and arbitrary active as opposed to realistic and appreciative reading of the
moral rules, since they are dictated by the general conditions of what
necessarily makes a universal harmony of wills possible. They do involve
activity but in a harmless way insofar as they entail principles that every
rational will is actually supposed to approve and act from in order to fulfill
rational activity in general. Readers may also be misled by the fact that Kant
speaks of maxims as being imposed (IV, p. 434), but this can be literally
true at an individual level (since the incorporation of a maxim in ones life
is a matter of contingent self-imposition) even while the laws that the
290 Karl Ameriks

maxims observe are not themselves imposed in any troublesome empirical

sense.17
There obviously remain many questions here, such as whether the idea of
pure practical reason is rich enough to be norm generating in anything like the
way that logic and mathematics may be. Also, once reason is understood, as it
is here, as something more like a pure and general faculty than an individual
act or orientation, one can wonder how particular sensible agents can come to
have an interest, let alone a sense of overriding obligation, in respect to it.
After all, an underground man might well say, yes, I understand now that
that is what the faculty of reason (even the voice of reason in me) says but I
am not going to follow that voice, I am going to follow my own natural,
egocentric voice! Not surprisingly, this brings us back to some of the most
basic problems of Kantian ethics how can pure practical reason provide
content, and how can it motivate?18 These are challenging questions, but they
are not the same as the question that has been our main topic, namely, whether
there is a so far at least intelligible notion of autonomy that is truly close to
Kants own texts (especially his rejections of all empiricist ethics) and yet
does not have the self-defeating characteristics that Larmore fears. To the
extent that my considerations point to a way that such a notion can make
sense after all, autonomy can have not only a fate but also a future.19

NOTES

1 See Paul Franks, Does Post-Kantian Skepticism Exist? Internationales Jahrbuch des
Deutschen Idealismus International Yearbook of German Idealism, ed. Karl Ameriks and
Jurgen Stolzenberg, vol. 1, 2003 (Berlin: Walter de Gruyter), pp. 14163.
2 See my Problems from Van Cleves Kant: Experience and Objects, Philosophy and
Phenomenological Research 66(2003), pp. 196202; and Idealism from Kant to Berkeley,
in S. Gersh and D. Moran (eds), Eriugena, Berkeley and the Idealist Tradition (Notre Dame:
University of Notre Dame Press, 2003), pp. 00*.
3 In referring to Kants Critique of Pure Reason I use the standard references to the A (1781)
and B (1787) editions, and the Norman Kemp Smith translation (New York: St. Martins
Press, 1965). References to Kants Groundwork of the Metaphysics of Morals will be to the
volume (IV) and page of the Akademie Ausgabe (Berlin: de Gruyter, 1900 ) and the Mary
Gregor translation (Cambridge: Cambridge University Press, 1996).
4 This is a major theme throughout my Kants Theory of Mind: An Analysis of the Paralogisms
of Pure Reason (Oxford: Clarendon Press, 1982; second, revised ed., Oxford University
Press, 2000).
5 For more background on this point, see my Introduction: Interpreting German Idealism, in
K. Ameriks (ed.), The Cambridge Companion to German Idealism (Cambridge: Cambridge
University Press, 2000), pp. 117.
6 This point is argued in my Kants Theory of Mind and Kant and Short Arguments to
Humility, in P. Cicovacki (ed.) Kants Legacy: Essays in Honor of L. W. Beck (Rochester:
Rochester University Press, 2001), pp. 16794; and supported by new data in Lectures on
Metaphysics/ Immanuel Kant, ed. and tr. Karl Ameriks and Steve Naragon (Cambridge:
Cambridge University Press, 1997).
7 Breazeale (n. 6) helpfully points out that when I refer to this passage a printers error gives an
incorrect reference to Bxix. A very similar passage can be found in Groundwork, IV, p. 456.
 On Being Neither Post- Nor Anti-Kantian 291

8 See Kant and the Fate of Autonomy, p. 123 (e.g., rights can be recognized by states only
when philosophers are clear about them), and my Reinholds Challenge: Systematic
Philosophy for the Public, in M. Bondeli and A. Lazzari (eds), Fichte-Studien Supplementa
(Amsterdam/Atlanta: Rodopi, 2002), pp. 88113; and Reinhold uber Systematik,
Popularitat und die Historische Wende , Philosophie ohne Beinamen, ed. M. Bondeli
and A. Lazzari (Basel: Schwalbe, 2003), pp. 30336.
9 See e.g., my Kants Transcendental Deduction as a Regressive Argument, Kant-Studien 69
(1978), pp. 27385; Recent Work on Kants Theoretical Philosophy, American
Philosophical Quarterly 19 (1982), pp. 124; Kant and Guyer on Apperception, Archiv
fur Geschichte der Philosophie 65 (1983), pp. 17486; Contemporary German
Epistemology, Inquiry 25 (1982), pp. 12538; The First Edition Paralogisms of Pure
Reason, in G. Mohr and M. Willaschek (eds), Immanuel Kant: Kritik der reinen Vernunft,
(Berlin: Akademie Verlag, 1998), pp. 36988.
10 See especially Henrich, U ber die Einheit der Subjektivitat, Philosophische Rundschau
3(1955), pp. 2869; English version, On the Unity of Subjectivity, tr. G. Zoeller, in R.
Velkley (ed.), The Unity of Reason (Cambridge: Harvard University Press, 1994), pp. 1754.
11 See Kant and the Fate of Autonomy, p. 37, which cites Lectures on Metaphysics/ Immanuel
Kant, p. 111.
12 See my Zu Kants Argumentation am Anfang des Dritten Abschnitts der Grundlegung, in
Systematische Ethik mit Kant, ed. Hans-Ulrich Baumgarten and Carsten Held (Freiburg:
Alber, 2001), pp. 2454; and Pure Reason of Itself Alone Suffices to Determine the Will, in
O. Hoffe (ed.) Immanuel Kant: Kritik der praktischen Vernunft (Berlin: Akademie Verlag,
2001), pp. 99114.
13 See my On Schneewind and Kants Method in Ethics, Ideas y Valores 102 (1996), pp. 28
53; as well as John Hare, Review of Christine Korsgaard, Creating the Kingdom of Ends,
and The Sources of Normativity, Faith and Philosophy 17 (2000), pp. 37183, and Kant on
Recognizing our Duties as Gods Commands, ibid., pp. 45978; and Patrick Kain, Kants
Moral Constructivism and his Conception of Legislation, in The Paideia Project Online:
Proceedings of the Twentieth World Congress of Philosophy. 1999= http://www.bu.edu/wcp/
Papers/Teth/TethKan.htm; and cf. Clemens Schwaiger, Kategorische und andere Impera-
tive: zur Entwicklung von Kants praktischer Philosophie bis 1785 (Stuttgart: Frommann-
Holzboog, 1999).
14 This tendency is effectively criticized in several works by Larmore and also in Donald H.
Regan, The Value of Rational Nature, Ethics 112 (2002), pp. 26791.
15 Hence I agree with much of Regans analysis but not his presumption that the Kantian agent
must either choose on the basis of empirical desire or else must launch herself arbitrarily
(ibid., p. 281). This perspective too quickly forecloses the possibility that a rational being
can have an understandable attachment to its own rational nature that is neither arbitrary nor
rooted in a prior empirical interest.
16 I am indebted to Ido Geiger for drawing my attention to the striking use of this term in the
play.
17 For this reason and many other considerations given in the material cited above, note 13, I
believe that even these passages fit a Kantian value theory that is consistent with taking
moral ends to be fully real, although not in the absurd sense of stretching to a realm that is
beyond reason altogether. Here again it is important to keep in mind the point that Kants
transcendental idealism is specifically a doctrine of what is ideal because of space and time,
and it in no way implies a non-realism about values. On Kantian value realism and respect
as the acknowledgement of objective value, cf. Allen Wood, Kants Ethical Thought
(Cambridge: Cambridge University Press, 1999), pp. 46f. and 157f.
18 I believe Kant does have an answer here about how reason can appropriately provide an
interest in acting morally, but I am not so sure he has a non-question-begging answer to the
question of why we must choose always to let the voice of reason override that of nature. See
my Kant on the Good Will, in O. Hoffe (ed.), Grundlegung zur Metaphysik der Sitten
(Frankfurt: Klostermann, 1989), pp. 4565; and Kant and the Fate of Autonomy, Ch. 7. The
deepest problem here may be that the general capacity to set ends may not be of clear moral
worth (since all the ends that a free being might actually set could be evil and compatible
292 Karl Ameriks

with an externally harmonious world of agents), whereas the capacity to set good ends may
turn out to have value only on question-begging grounds or insofar as something external to
rational nature is given moral priority after all.
19 I wish to express my indebtedness to my co-symposiasts very helpful essays as well as their
seminal earlier work on related themes, especially Breazeales Fichtes Aenesidemus
Review and the Transformation of German Idealism, Review of Metaphysics 34 (1981),
pp. 54568; and Between Kant and Fichte: Karl Leonhard Reinholds Elementary
Philosophy , Review of Metaphysics 35 (1982), pp. 785821; and Larmores The Morals of
Modernity (Cambridge: Cambridge University Press, 1996) and his recent philosophy
reviews in The New Republic. I am also indebted to discussants at two APA symposia on the
volume, and especially to Eric Watkins, Marcelo Stamm, Fred Beiser, Patrick Kain, and Paul
Franks.

Received 14 August 2002

Karl Ameriks, Department of Philosophy, University of Notre Dame, 100 Malloy Hall, Notre
Dame, IN 46556, USA. E-mail: Karl.P.Ameriks.2@nd.edu